(2016) Diffusion of Ict and Sme Performance

E-Services Adoption: Processes by Firms in Developing
Nations
Diffusion of ICT and SME Performance
Md Shah Azam
Article information:
To cite this document: Md Shah Azam . "Diffusion of ICT and SME Performance" In
E-Services Adoption: Processes by Firms in Developing Nations. Published online: 07
Oct 2015; 7-290.
Downloaded by National University of Singapore At 22:47 30 January 2016 (PT)
Permanent link to this document:

http://dx.doi.org/10.1108/S1069-096420150000023005
Downloaded on: 30 January 2016, At: 22:47 (PT)
References: this document contains references to 0 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 71 times since NaN*
Users who downloaded this article also downloaded:
Darlene K. Mutalemwa, (2015),"Does globalisation impact SME development in
Africa?", African Journal of Economic and Management Studies, Vol. 6 Iss 2 pp.
164-182 http://dx.doi.org/10.1108/AJEMS-01-2015-0012
S. Agut, R. Grau, J.M. Peiró, (2003),"Individual and contextual influences on
managerial competency needs", Journal of Management Development, Vol. 22 Iss 10
pp. 906-918 http://dx.doi.org/10.1108/02621710310505494
David C. Roach, (2011),"The impact of product management on SME firm
performance", Journal of Research in Marketing and Entrepreneurship, Vol. 13 Iss 1 pp.
85-104 http://dx.doi.org/10.1108/14715201111147969
Access to this document was granted through an Emerald subscription provided by

emerald-srm:397875 []
For Authors
If you would like to write for this, or any other Emerald publication, then please
use our Emerald for Authors service information about how to choose which
publication to write for and submission guidelines are available for all. Please visit
www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society.
The company manages a portfolio of more than 290 journals and over 2,350 books
and book series volumes, as well as providing an extensive range of online products
and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner
of the Committee on Publication Ethics (COPE) and also works with Portico and the
LOCKSS initiative for digital archive preservation.
time of download.
*Related content and download information correct at
DIFFUSION OF ICT AND SME
PERFORMANCE
Md Shah Azam
ABSTRACT
Information and communications technology (ICT) offers enormous

opportunities for individuals, businesses and society. The application of
ICT is equally important to economic and non-economic activities.
Researchers have increasingly focused on the adoption and use of ICT by
small and medium enterprises (SMEs) as the economic development of a
country is largely dependent on them. Following the success of ICT
utilisation in SMEs in developed countries, many developing countries
are looking to utilise the potential of the technology to develop SMEs.
Past studies have shown that the contribution of ICT to the performance
of SMEs is not clear and certain. Thus, it is crucial to determine the
effectiveness of ICT in generating firm performance since this has impli-
cations for SMEs’ expenditure on the technology. This research exam-
ines the diffusion of ICT among SMEs with respect to the typical stages
from innovation adoption to post-adoption, by analysing the actual usage
of ICT and value creation. The mediating effects of integration and utili-
sation on SME performance are also studied. Grounded in the innovation
diffusion literature, institutional theory and resource-based theory, this
study has developed a comprehensive integrated research model focused
E-Services Adoption: Processes by Firms in Developing Nations

Advances in Business Marketing & Purchasing, Volume 23A, 7290
Copyright r 2015 by Emerald Group Publishing Limited
All rights of reproduction in any form reserved
ISSN: 1069-0964/doi:10.1108/S1069-096420150000023005
7
8 MD SHAH AZAM
on the research objectives. Following a positivist research paradigm, this

study employs a mixed-method research approach. A preliminary con-
ceptual framework is developed through an extensive literature review
and is refined by results from an in-depth field study. During the field
study, a total of 11 SME owners or decision-makers were interviewed.
The recorded interviews were transcribed and analysed using NVivo 10
to refine the model to develop the research hypotheses. The final research
model is composed of 30 first-order and five higher-order constructs
which involve both reflective and formative measures. Partial least

squares-based structural equation modelling (PLS-SEM) is employed to
test the theoretical model with a cross-sectional data set of 282 SMEs in
Bangladesh. Survey data were collected using a structured questionnaire
issued to SMEs selected by applying a stratified random sampling
technique. The structural equation modelling utilises a two-step proce-
dure of data analysis. Prior to estimating the structural model, the mea-
surement model is examined for construct validity of the study variables
(i.e. convergent and discriminant validity).
The estimates show cognitive evaluation as an important antecedent for
expectation which is shaped primarily by the entrepreneurs’ beliefs (per-
ception) and also influenced by the owners’ innovativeness and culture.
Culture further influences expectation. The study finds that facilitating
condition, environmental pressure and country readiness are important
antecedents of expectation and ICT use. The results also reveal that
integration and the degree of ICT utilisation significantly affect SMEs’
performance. Surprisingly, the findings do not reveal any significant
impact of ICT usage on performance which apparently suggests the pos-
sibility of the ICT productivity paradox. However, the analysis finally
proves the non-existence of the paradox by demonstrating the mediating
role of ICT integration and degree of utilisation explain the influence of
information technology (IT) usage on firm performance which is consis-
tent with the resource-based theory. The results suggest that the use of
ICT can enhance SMEs’ performance if the technology is integrated and
properly utilised. SME owners or managers, interested stakeholders and
policy makers may follow the study’s outcomes and focus on ICT
integration and degree of utilisation with a view to attaining superior
organisational performance.
This study urges concerned business enterprises and government to look
at the environmental and cultural factors with a view to achieving ICT
usage success in terms of enhanced firm performance. In particular,
Diffusion of ICT and SME Performance 9
improving organisational practices and procedures by eliminating the tra-

ditional power distance inside organisations and implementing necessary
rules and regulations are important actions for managing environmental
and cultural uncertainties. The application of a Bengali user interface
may help to ensure the productivity of ICT use by SMEs in Bangladesh.
Establishing a favourable national technology infrastructure and legal
environment may contribute positively to improving the overall situation.
This study also suggests some changes and modifications in the country’s
existing policies and strategies. The government and policy makers

should undertake mass promotional programs to disseminate information
about the various uses of computers and their contribution in developing
better organisational performance. Organising specialised training pro-
grams for SME capacity building may succeed in attaining the motiva-
tion for SMEs to use ICT. Ensuring easy access to the technology by
providing loans, grants and subsidies is important. Various stakeholders,
partners and related organisations should come forward to support
government policies and priorities in order to ensure the productive use
of ICT among SMEs which finally will help to foster Bangladesh’s
economic development.
Keywords: Diffusion; ICT; SME; organisational performance; culture;
environmental pressures
1. INTRODUCTION
1.1. Overview
Information and communications technology (ICT) has become an integral

part of human life in the 21st century. The rapid expansion of ICT and its
application into various economic activities have opened new opportunities
for individuals, businesses and society. In particular, the introduction of
various ICT devices has increased productivity in business organisations
and has assisted them to manage intra- and inter-organisational affairs.
The technology also provides customers with the ability to conduct perso-
nal communications, business transactions and banking operations in a
more flexible and efficient manner.
Organisations are integrating ICT into business operations to enhance
productivity.1 The wide-ranging applications of ICT, particularly
Internet-based digital technology, have reshaped the ways in which
10 MD SHAH AZAM
communication occurs and have also made changes to the systems, proce-
dures and processes of relevant services. Consequently, this affects the ways
in which customers, suppliers, regulatory bodies and other external parties
deal with business organisations.
The past decades have produced much research investigating the use of
information technology (IT) by both individuals and organisations (see
Agarwal & Prashad, 1997, 1998, 1999; Davis, 1993; Davis, Bagozzi, &
Warshaw, 1989; Kendall, Tung, Chua, Ng, & Tan, 2001; Mathieson, 1991;
Moore & Benbasat, 1991; Premkumar & Potter, 1995; Sathye & Beal, 2001;
Tan & Teo, 2000; Taylor & Todd, 1995b; Venkatesh, Brown, Maruping, &
Bala, 2008; Venkatesh & Davis, 2000; Venkatesh & Morris, 2000;
Venkatesh, Morris, Davis, & Davis, 2003). A primary focus of these
research studies is the identification of the factors that determine adoption
and diffusion of ICT.
Prior studies follow the notion that the use of ICT would enhance orga-
nisational performance. In the past few years, the growth in ICT usage has
increased tremendously across industries around the world. Business enter-
prises’ investment in ICT has also increased dramatically which poses the
valid question about whether organisational spending on ICT results in
improved organisational performance. Zhu and Kraemer (2005) referred to
the wave of debate over the new ‘IT value paradox’ which was triggered by
Carr (2003) in the article ‘IT Doesn’t Matter’.
Prior studies reveal ambiguous results about the contribution of ICT to
business performance, such as having a negative effect (Warner, 1987);
zero effect (Venkatraman & Zaheer, 1990); a contingent positive effect
(Powell & Dent-Micallef, 1997; Tippins & Sohi, 2003; Wu, Yeniyurt,
Kim, & Cavusgil, 2006) and a direct positive effect (Bharadwaj, 2000;
Zhu & Kraemer, 2005). What is clear is that the contribution of ICT to
organisational performance is not straightforward and certain.
Organisational sustainability is largely dependent on the productive use
of ICT. Thus, answers to this renewed paradox will have important impli-
cations for the way businesses approach IT investment and management.
Zhu and Kraemer (2005) revealed that ICT use (e-commerce use) generates
organisational performance. By applying the notion of the resource-based
view, they reported that the integration of ICT with front-office functional-
ities and back-end databases can create unique ICT capabilities which
cannot be easily imitated and thus have the potential to create improved
business performance (Bharadwaj, 2000; Zhu & Kraemer, 2002). This
aspect of ICT use, named, ‘ICT integration’, has not been widely documen-
ted in the contemporary research literature and, hence, its effect on
business performance has not been clearly identified. Similarly, another

aspect of ICT use may correlate with firm performance which is ‘ICT utili-
sation’, that is, whether the ICT is used properly and effectively.
This phenomenon of ICT use is particularly related to the developing
country environment. Anandarajan, Igbaria, and Anakwe (2002) have
stated that many information systems (IS) in the less developed countries
are under-utilised and, thus, do not make a significant contribution to
improving the performance of organisations that are using them (Forster &
Cornford, 1992; Ordedra, Lawrie, Bennett, & Goodman, 1993). This state-
ment is indicative of the fact that proper utilisation of ICT may result in
improved firm performance; that is, utilisation can make a bridge between
ICT use and firm performance. Integration may also act in the same way.
Contemporary research initiatives have not produced adequate facts and
figures about integration, utilisation and their effects, particularly the medi-
ating effects, on firm performance which may be considered as the agenda
for a new research study.
The perspective of the study, that is, the developed versus developing
country perspective, has become another research motivation. The past few
decades have witnessed an unprecedented increase in the trend of ICT
usage to accomplish a wide range of functions at individual, organisational
and society levels. The phenomenal usage growth has inspired organisa-
tions to become ICT-dependent in accomplishing various internal and
external functions. The trend for organisations to become ICT-dependent
is significantly higher in developed countries than for organisations from
the developing world. Past studies have therefore mainly focused on the
developed country context. Consequently, most of the theories in ICT have
been formulated from the perspective of developed countries, particularly
the American perspective (Zhu & Kraemer, 2005).
In reality, the theories formulated in the developed country environment
may not be appropriate in addressing similar phenomena in developing
countries as there remains a wide digital divide between the two types
of countries. Utilising the advantages of high Internet penetration, developed
countries have become ICT-dependent to accomplish various functions such
as governance, business, education and utilities (for individuals and organi-
sations) with this gradually trickling down to the developing world.
The success of various ICT applications in the developed world has a
significant implication for technology adoption and usage behaviour in
developing countries. Recent statistics have suggested significantly higher
Internet usage growth (between 2000 and 2013) in the developing part of
the world than in developed countries which has created the grounds for
12 MD SHAH AZAM
diverse use and utilisation of ICT in developing countries. The changing

pattern of the world’s ICT usage growth in recent years may attract
researchers’ attention towards developing and least developed countries
(LDCs).2
Bangladesh is a member of the LDCs. The country is characterised by
high population, low income and inadequate infrastructural supports such
as inefficient technology, inadequate power supply, low teledensity, very
low Internet penetration, absence of effective legal systems, and financial
and banking mechanisms which are not supportive for facilitating commu-
nication and transactions via the Internet. Despite poor teledensity and
Internet penetration, the government has planned to make Bangladesh a
technology-dependent society by the 50th anniversary of the birth of the
country in 2021 with this adopted in the national development plan, Vision
2021. The development of ICT and its utilisation in the country’s economic
development were included in the election manifesto and post-election
agendas of the present government which have later attracted the attention
of many researchers, policy makers, practitioners and the general public as
a movement called Digital Bangladesh.
The growth indices, development initiatives and infrastructure of
Bangladesh are identical with those of a typical developing country. The
characteristics of the country and its move towards establishing an ICT-
based society are the rationale behind undertaking this survey-based
research on the diffusion of ICT and SME performance seeking reliable
and valid research outcomes that are representative of developing
countries.
1.2. Research Questions
The success of the use of ICT as a tool for increasing productivity in

various sectors is the main driver for new adopters. The uniqueness of the
technology viz., cost, convenience and efficiency (speed and accuracy) moti-
vates users in their new and continued use of technology for personal and
business use. The productivity implications of ICT have received the bulk
of researchers’ attention around the world in their studies of its adoption
and diffusion phenomena. However, the majority of the previous ICT
adoption-diffusion studies have investigated intention (e.g. Gefen &
Straub, 2000; Kendall et al., 2001; Lal, 1999; Pavlov & Chai, 2002; Teo,
Wei, & Benbasat, 2003) and actual usage behaviour (e.g. Anandarajan
et al., 2002; Johnson & Hignite, 2000; Thatcher, Foster, & Zhu, 2006;
Venkatesh et al., 2008). Thus, the question about whether the adoption or
use of technology induces improved organisational performance remains
relatively under-researched, especially in the case of SMEs. This may be
considered as a vital research gap which has therefore provided the motiva-
tion to link technology adoption-diffusion behaviour with organisational
performance in studying the effects of ICT on SME performance.
The innovation adoption theories explain the adoption and diffusion
phenomena mainly through intention (Ajzen, 1985; Davis, 1986;
Fishbein & Ajzen, 1975; Rogers, 1983; Venkatesh et al., 2003). A criticism
of the existing theories is that intention is a reflection of the adopter’s
internal schema of beliefs which may fail to adequately explain actual
behaviour in a situation under incomplete volitional control. Intention,
furthermore, may not be reflected in the decision if a time gap exists
between intention and actual behaviour.
A recent technology adoption research study has examined the role of
behavioural expectation in order to avoid the inabilities of the cognitive
factor, behavioural intention. The new construct, behavioural expectation,
is not a cognitive factor thus it is able to address the roles of various inter-
nal and external factors (Venkatesh et al., 2008). The measurement of
behavioural expectation as well as its contribution to actual ICT usage
behaviour from an organisational perspective has not been well examined.
The adoption and diffusion phenomena of technological innovation
have been analysed by applying a number of theoretical models in the last
few decades. The parsimony of the research model has been an important
concern for researchers seeking accuracy in outcomes and ease of estima-
tions.3 To maintain parsimony in the analytical model, a number of
external factors, such as various aspects of culture and environment, have
remained outside of the study due to the scope of the research model pre-
viously employed to examine the diffusion of a technological innovation
(viz., Ajzen, 1985; Davis, 1986; Fishbein & Ajzen, 1975; Rogers, 1983;
Venkatesh et al., 2003, 2008). In some cases, although these external
variables have attracted researchers’ attention, the effects have not been
comprehensively analysed.
Generally, the external variables have been considered either separately
or a part of the context has been taken into consideration. For example,
the effects of culture or environmental pressures have been examined sepa-
rately to address the ICT adoption and usage behaviour in many past
studies (see Teo et al., 2003 for environmental pressure and Erumban &
Jong, 2006 for culture). Some other studies have analysed a part of culture
or environment such as competitive pressure, in some cases normative
14 MD SHAH AZAM
pressure has been analysed (with other antecedent factors) to study the
effects of environmental pressures on the diffusion of a technological inno-
vation (see Zhu, Dong, Xu, & Kraemer, 2006; Zhu & Kraemer, 2005; Zhu,
Kraemer, Xu, & Dedrick, 2004). Inclusion of an adequate number of
variables on the other hand can improve model efficiency for reliable and
valid research outcomes.4
In recent years, significant progress has been made in the process of sta-
tistical analysis. In particular, the introduction of PLS-SEM) (Chin, 2010)
has offered robust estimates and an increased ability to handle a large

number of latent variables in a single platform. It can also handle the com-
pound or higher-order constructs by estimating higher-order hierarchical
modelling (see Wetzels, Odekerken-Schröder, & van Oppen, 2009). The
construction of a comprehensive research model with a range of variables,
such as individual, organisational, cultural and environmental variables,
and a robust analysis may offer a detailed and clear understanding about
the organisational technology diffusion behaviour.
Contemporary researchers have applied additional variables with the
existing innovation diffusion theoretical frameworks (see Venkatesh et al.,
2003, 2008; Zhu & Kraemer, 2005; Zhu, Dong, et al., 2006; Zhu et al.,
2004). For example, Venkatesh et al. (2003) included perceived perfor-
mance expectancy, effort expectancy, social influence and facilitating condi-
tion as exogenous variables in the model. Facilitating condition is defined
as the extent to which the organisation is able to use the technology.
Employee skill and knowledge, compatibility and the organisational
resource base constitute the facilitating condition. Past literature has
supported the view that the higher the facilitating conditions, the higher
the usage of a technological innovation in an organisation (Gupta,
Dasgupta, & Gupta, 2008; Venkatesh et al., 2008).
ICT, particularly Internet-based digital technologies, functions over a
public network. Thus, the technology infrastructure of a country, govern-
ment telecommunication policy, and the availability and speed of the
Internet may matter in organisational ICT usage behaviour. Like facilitat-
ing condition, country readiness may play a significant role in ICT diffu-
sion behaviour. Recent studies have investigated country readiness a little.
For example, government policy and the regulatory environment have
been studied in some previous studies (see Zhu & Kraemer, 2005; Zhu,
Kraemer, & Xu, 2006; Zhu et al., 2004). The effects of country readiness
on organisational IT usage behaviour, although addressed conceptually
(see Molla & Licker, 2005), have not been documented widely in the exist-
ing literature. Besides this, previous research studies have mostly favoured
the view that organisations adopt technology that is useful and that
provides them with some economic benefit while, although important,
non-economic factors such as cultural and environmental factors have been
overlooked (Thatcher et al., 2006).
The economic development of a country is largely dependent on SMEs
in the present market-based global competitive environment. Although
equally important in terms of economic significance, SMEs in developed
countries have continued to contribute substantially to their country’s
growth process while those from developing countries have not fared well.
The rapid expansion of IT and its application to almost every sphere of
economic activity have initiated the process of encouraging SMEs to make
the most cost-effective use of new technologies in production, marketing
and networking (Mandal, 2007). Thus, how SMEs can be tapped into new
opportunities and produce enhanced performance in economic develop-
ment is considered to be a worthwhile research study.
Taking into consideration the notion of ICT’s performance implications,
recent high ICT usage growths in the developing world and SMEs’ eco-
nomic significance, this study has opted to examine how ICT affects eco-
nomic development through improved organisational performance in the
developing country context. Considering SMEs as a population, this study,
thus, has attempted to answer the main research question: ‘How does diffu-
sion of ICT correlate with SME performance in Bangladesh’? The study
has also raised the following primary questions:
RQ-1. What is the contribution of internal and external factors in beha-
vioural expectation in the context of Bangladesh?
RQ-2. How does behavioural expectation correlate with ICT use by

SMEs in Bangladesh?
RQ-3. What is the role of ICT use, integration and utilisation in deter-
mining SME performance?
1.3. Objectives
The main objective of this study is to formulate a theoretical framework

for examining the contribution of the use and utilisation of ICT to SME
performance. The specific objectives of the study are to:
RO-1. Explore the cognitive factors influencing ICT adoption by SMEs.
16 MD SHAH AZAM
RO-2. Delineate the effects of culture and environmental pressure on the

adoption and usage of ICT by SMEs.
RO-3. Measure the effects of facilitating condition and country readiness

on ICT adoption and use by SMEs.
RO-4. Examine the role of behavioural expectation in explaining ICT

usage by SMEs.
RO-5. Examine the contribution of ICT usage to SME performance.
RO-6. Examine the mediating effects of ICT integration and utilisation

in explaining the contribution of ICT usage to SME performance.
1.4. Research Background
ICT, particularly the Internet, underpins almost every single activity under-
taken in the modern world and affects everyone on the planet even those
who do not themselves have first-hand access to ICT (ITU, 2010). Good
examples include food distribution, power networks, water supplies or
mass transportation, all of which are controlled and managed today by
ICT networks and applications.
According to the World Telecommunication Report 2010, released to
review the mid-term status and achievement between the World Summit on
the Information Society (WSIS) 2005 and the Millennium Development
Goals (MDG) 2015, tremendous progress has been made over the past
decade, with almost two billion people throughout the world now having
access to the Internet.
Although significant progress has been evident in the world’s Internet
penetration, household Internet penetration levels vary substantially
between countries and regions. At the end of 2008, one out of four house-
holds in the world had access to the Internet but only one out of eight
households in developing countries was connected, compared to three out
of five in developed countries.
While by the end of 2008, 58.1% of households in Europe had Internet
access, only 16.8% of households in Asia-Pacific countries were connected
to the Internet. The Internet penetration of Asia-Pacific countries remains
at a lower level in comparison to Europe, America, the Commonwealth of
Independent States (CIS) and the Arab States. The Internet penetration
of Bangladesh is significantly lower (below 1%) than that of other
Asia-Pacific countries, such as Japan, Malaysia, Korea, Singapore and
Australia.
According to the World Bank (WB) (World Bank, 2010) the Internet
penetration in various countries was estimated as 75.9% (USA), 76%
(UK), 70.8% (Australia), 69.6% (Singapore) and 55.8% (Malaysia), while
Bangladesh’s Internet penetration was only 0.347% in 2008 (Azam, 2013).
The Internet penetration of those countries was more recently calculated as

78.1% (USA), 83.6% (UK), 88.8% (Australia), 75.0% (Singapore), 60.7%
(Malaysia) and 5.0% (Bangladesh) at 30 June 2012.5 The statistics show an
increasing trend in Internet penetration in all of the above countries;
however, the volume of Internet penetration in Bangladesh is very small in
comparison to the other countries. While many developed and developing
countries have achieved significant advantages through the computerisation
of government departments, business firms and educational institutions,
the digital initiatives of Bangladesh still remain at risk due to the poor
digital participation of the citizenry.
The Internet began operating commercially in Bangladesh in the mid-
1990s. Bangladesh has a poor teledensity in comparison to other developed
and developing countries around the world (ITU, 2013). Like many other
developing countries, Bangladesh suffers from inadequate infrastructure,
outdated telephone systems, limited access to telephones and computers,
poor service quality with high prices, lack of qualified personnel and a low
level of literacy and IT skills, as well as cultural and language barriers
(Azam, 2007). The main obstacle to using the Internet in Bangladesh is its
distribution. The Internet is still an urban privilege in Bangladesh as
the facilities are more concentrated in urban areas, especially being
Dhaka6-based. The population living outside urban areas is mostly
deprived of gaining the benefit of the Internet although they have the
potential.7 The noticeable digital divide is observed in Bangladesh in the
following areas (Rahman, 2003):
• Urban and rural populations

• Dhaka (the capital) and rest of the country
• Different educational streams
• Rich population and poor population
• Male and female
• Mainstream and tribal populations
• Government and private organisations.
18 MD SHAH AZAM
Despite suffering from multidimensional problems in terms of Internet

access, Bangladesh has the potential of achieving success in its use of the
Internet in the education, development, business and other services sectors.
The Government of Bangladesh considers the ICT sector as the thrust
sector in the fifth five-year plan of the People’s Republic of Bangladesh. In
order to enjoy high bandwidth at a lower usage cost, the country has been
connected with the information super highway through submarine optic
fibre networks with a 16-country consortium through the SEA-ME-WE-4
project.8 The Government of Bangladesh inaugurated this connectivity on
21 May 2006 (Daily Star, 2006; Prothom Alo, 2006). The landing station
has been established at Cox’s Bazar, the southern city, near the Bay of
Bengal.
Despite Bangladesh’s poor Internet penetration, the present
Government of Bangladesh has given the highest priority to ICT and has
initiated diverse policies and programs to achieve the digital goal provi-
sioned in the 2009 national election and post-election agenda. The
country’s yearly national budgets (in the last few years) have allocated
a substantial amount of resources for ICT development thus reiterating
the expansion of ICT networks to rural communities to achieve
government, citizen and business interactions and exchanges through the
Internet. The government has also initiated some modifications to
the country’s national ICT policy in 2009 which reiterate the necessity of
establishing e-government, e-services and e-commerce environments in
order to gain economic potential. In addition, the government has empha-
sised the formulation of appropriate policies and strategies to facilitate
Internet-related communication, e-commerce operation and e-governance.
In order to achieve the potential of ICT, the government is dedicated to
utilising the Internet in the education and services sectors (Azam &
Quaddus, 2009a, 2009b, 2009c). Although numerous policy initiatives have
been adopted to utilise the potential of ICT in the economic development
of the country, the success of digitisation or computerisation is still
doubtful.
Bangladesh has initiated steps to possibly fight against the hurdles and
hindrances of ICT adoption, such as, limited accessibility to the Internet,
poor teledensity, poor electricity network. However, poor Internet penetra-
tion is still considered as the main issue in establishing an e-based transpar-
ent society along with other issues such as the limited affordability of
computers and limited knowledge; inadequate legal and regulatory support;
inefficient and traditional systems of banking operation; poor financial sup-
port and traditional payment mechanisms; lack of human resources; and
high Internet usage cost as well as security concerns (Azam, 2005, 2006a,
2006b; Azam & Lubna, 2008a, 2008b; Azam & Quaddus, 2009a; Hossain,
2000; ITRC, 2000; Rahman, 2002).
Bangladesh is basically an agricultural-based country. Its recently
developed industrial base, particularly the ready-made garments (RMG)
industry, has emerged as the main vehicle for the country’s economic
development. The overall culture of the country is characterised by a high
population, low incomes and quite a large number of unemployed people
with labour cheap and available as a result. Like many other Asian
states, Bangladesh’s culture has also been characterised by high power dis-
tance, collectivism and low uncertainty avoidance (House, Hanges,
Javidan, Dorfman, & Gupta, 2004). The power is concentrated at the top
of Bangladeshi society.
The Internet usage statistics shown in the preceding section are inspiring
Asian countries to utilise the potential of ICT in their economic develop-
ment. The Internet penetration in Bangladesh likewise is also growing.
The rapidly increasing trend of ICT usage, particularly Internet use, in
business provides a motivation for large organisations as well as for smaller
organisations in developing countries to adopt this technology. This usage
encompasses the management of organisational internal communication,
external communication, shop floor management, inventory control and
customer integration as well as online order processing and transactions to
acquire increased and competitive organisational performance.
Small and medium-sized enterprises (SMEs) have played an important
role in the development of all countries. SMEs in developed countries con-
tribute substantially to those countries’ growth processes. Although lagging
behind, SMEs in developing countries are also contributing positively.
The prospects and contribution of Bangladesh’s SMEs in its economic
development are enormous. SMEs account for about 45% of the manufac-
turing value-add in Bangladesh. They account for about 80% of industrial
employment, about 90% of total industrial units and about 25% of the
total labour force. Their total contribution to export earnings varies
between 75% and 80% (Azam & Quaddus, 2009c, 2009d).
According to the Bangladesh Bureau of Statistics (BBS), SMEs provide
about 44% of the country’s employment. The 2003 Private Sector Survey
estimated that about six million micro-, small- and medium-sized enter-
prises, defined as enterprises with fewer than 100 employees, contributed
around 2025% of gross domestic product (GDP) (The New Nation,
2008). The number of SMEs and their significant contribution to the
national economy in terms of employment generation, GDP contribution
and export earnings thus create a significant research opportunity exploring
the adoption and diffusion of ICT.
20 MD SHAH AZAM
1.5. Definition of Terms
Behavioural expectation: Refers to an individual’s self-reported subjective

probability of his or her performing a specified behaviour, based on his
or her cognitive appraisal of volitional or non-volitional behavioural
Antecedents. It is the immediate antecedent of actual behaviour which is
stronger than intention and is able to include the effects of some contextual
and other external factors (Venkatesh et al., 2008).
Cognitive evaluation: Refers to an individual’s overall evaluation towards

an innovation. The evaluation process involves functional as well as
psychological consequences. These cognitive and affective evaluations form
one’s attitude, that is, negative or positive feelings about performing a
behaviour: this has been widely researched in the consumer, marketing and
IS research. This study modelled cognitive evaluation to address attitude
which is reflected by salient beliefs and respective evaluations (Ajzen, 1985,
1991).
Country readiness: Refers to an individual’s perceptions about the coun-
try’s preparedness or resources base which may support one’s (an indivi-
dual or organisation) decision to use a technological innovation. Country
readiness comprises technology infrastructure, financial infrastructure, legal
infrastructure, and Government policy and supports.
Culture: Refers to the collective programming of the mind that distin-
guishes the members of one group or category of people from another
(Hofstede, 2001).
Degree of utilisation: Refers to the degree through which the proper and
actual use of technology is ensured.
Environmental pressure: Refers to the pressure arising from the
institutional environment. For example, environmental pressure includes
various encouragement, advice or pressures which a firm receives from
its customers, suppliers, the regulatory authority and other stakeholders.
Environmental pressure comprises normative, mimetic and coercive
pressures.
Facilitating condition: Refers to an individual’s perceptions of the avail-
ability of technological and/or organisational resources (i.e. knowledge,
resources and opportunities) that can remove barriers to using a system
(Venkatesh et al., 2003).
ICT: Refers to ICT which may range from a simple digital phone or
computer operation to highly sophisticated computer-driven and Internet-
driven automated equipment. In this study, ICT is used to refer to the com-
puter and computer-driven Internet and networking technologies which
include various levels of ICT applications such as: (i) basic computing
(computer and Internet); (ii) computing with a homepage operation which
entails product cataloguing systems; (iii) computing with a homepage
operation which offers online order receiving and processing systems;
(iv) computing with interactive homepage operations which involve online
order processing with online transaction processing systems and (v) com-
puting with complete internal and external digital communication (enter-
prise resource planning (ERP)).
ICT usage: Refers to an individual’s self-reported subjective assessment

about rate of usage of various ICT applications. This study applies terms
‘ICT usage’ and ‘ICT use’ interchangeably to indicate SMEs’ ICT usage
behaviour.
Organisational performance: Refers to the outcome of organisational
processes in a given time. Profit growth, sales growth, market share, pro-
ductivity growth and firm competitiveness are the dimensions of organisa-
tional performance.
SMEs: Refer to small and medium-sized enterprises. In this study,
SMEs are defined according to the Industrial Policy 2010 of the People’s
Republic of Bangladesh as:
• A manufacturing firm with 100250 employees or having fixed assets
from 10 Crore (1 Crore = 10m) Bangladeshi Taka (BDT) to 30 Crore
BDT (excluding the value of land and factory) is considered a medium
enterprise while any firm with 2599 employees or having fixed assets
from 55 Lakh (1 Lakh = 10k) BDT to 10 Crore BDT (excluding the
value of land and factory) is considered a small enterprise.
• Also, a firm in the services industry with 50100 employees or having fixed
assets from 1 Crore BDT to 15 Crore BDT (excluding the value of land
and factory) is considered a medium enterprise while a firm with 1049
employees or having fixed assets from 5 Lakh BDT to 1 Crore BDT
(excluding the value of land and factory) is considered a small enterprise.
1.6. Research Significance
This research has been dedicated to the extensive study of the ICT diffusion
process and its effects on SME performance. In its design, the study has
extended beyond the traditional innovation diffusion theories and has
combined this with the resource-based view to examine SME performance.
The study has employed a mixed-method research approach to overcome
22 MD SHAH AZAM
the limitations of a mono-method application, that is, either a qualitative

or quantitative method. In this research, the qualitative study has led to the
presentation of a fine-tuned and contextualised comprehensive research
model. A number of new constructs and measurement procedures have
also been identified during the field study. The quantitative survey and
analyses have validated the anticipated model. However, the significance
of this research lies in its contribution to theory and its contribution to
practice.
1.6.1. Contribution to Theory

To examine organisational performance, this study has reviewed several
existing innovation diffusion theories and has combined them with
the institutional theory and resource-based views. More specifically, the
following theories have been reviewed: theory of reasoned action (TRA)
(Fishbein & Ajzen, 1975); diffusion of innovation (DOI) theory (Rogers,
1983); theory of planned behaviour (TPB) (Ajzen, 1985); technology accep-
tance model (TAM) (Davis, 1986); unified theory of acceptance and use of
technology (UTAUT) (Venkatesh et al., 2003); technology, organisation
and environment (TOE) framework (Tornatzky & Fleischer, 1990); institu-
tional theory (DiMaggio & Powell, 1983) and resource-based theory
(Barney, 1991).
A comprehensive research model has been constructed by synpapering
existing theories: the model focuses on the diffusion of ICT process and its
impacts on organisational performance. As most of the previous innova-
tion diffusion studies have focused on the developed country perspective,
the primary research model has been fine-tuned and contextualised through
a qualitative study. The rigour in the qualitative field study analysis has
resulted in a comprehensive model which includes a range of variables
from individual, organisational, socio-cultural and environmental levels.
The qualitative research has also provided an extensive understanding
about this field of research through the inclusion of some new variables
and the examination of their relationship with other study variables. For
example, the study has included integration and utilisation as new con-
structs and has anticipated their mediating roles in examining the effects of
ICT usage on SME performance. The field study has also provided a valu-
able contribution by presenting the measurement procedures for the newly
introduced variables. By adding some higher-order compound variables
which cover a wide range of variables as each higher-order variable has
two or more manifest variables, the framework has become a holistic
research model.
The theoretical framework now offers an opportunity to examine the

whole process of innovation diffusion and its effect on organisational
performance in a comprehensive model which is an important theoretical
contribution. The robustness of estimates, furthermore, suggests the
suitability of the comprehensive model for analysing the diffusion of
ICT by SMEs and its effects on organisational performance. This frame-
work is potentially suitable for testing similar phenomena in the large
organisation environment from both a developing and developed country
perspective.
1.6.2. Contribution to Practice

This study provides a clear picture of how internal and external factors
affect ICT usage by SMEs in a developing country and its consequent
effects on organisational performance. In analysing the data collected from
SMEs using different levels of ICT applications, it was revealed that SMEs
with a positive cognitive evaluation towards ICT, and that had adequate
technological resources and received positive pressures from their institu-
tional environment were more likely to use ICT while the overall culture
and country infrastructure were not complementary to the use of ICT.
The study has provided a unique contribution in addressing the ICT
paradox in a developing country context. This study has found that ICT
use does not increase the performance growth of SMEs straightaway.
Organisations should acquire the integration and proper utilisation of the
technology to achieve performance growth. This is consistent with the
notion of the resource-based view of ICT which reiterates that combining
ICT into different organisational functionalities (i.e. integration with
front-end functionalities or back-end integration) can create IT capabilities
that are rare, un-imitable, valuable and sustainable, thereby contributing
to value generation. Managers, interested stakeholders and policy
makers may follow the outcome of this study and focus on ICT integration
and the degree of utilisation with the view of attaining superior firm
performance.
This study has also explored the patterns of ICT usage by SMEs and
forecasts the degree and magnitude of the effects of various individual,
organisational and environmental factors on the adoption and usage of
different levels of ICT applications. This study suggests some changes and
modifications to Bangladesh’s existing policies and strategies to promote
ICT adoption and use by SMEs for economic development. Government
policy and supports focusing on the integrated and proper use of technol-
ogy could contribute positively to SME performance.
24 MD SHAH AZAM
1.7. Organisation of the Chapter
The chapter is structured as eight sections followed by references and

appendices. The first section has presented an introduction and overview of
the study including the research background with an overview of the over-
all structure of the research. The section started with the grounding of the
research questions and has drawn general and specific research objectives.
The significance of the research has also been discussed.
The second section deals with the review of the literature. The literature
review includes reviewing some important theories in the domain of inno-
vation diffusion along with institutional theory and resource-based theory.
The empirical studies are also reviewed to develop a conceptual framework
for the study. A preliminary research model is proposed which provides the
basis for the field study and for the quantitative data collection survey. The
third section deals with research method. This section discusses the research
methodology and design incorporated in the study. It first introduces the
research paradigm of the study. An interpretive (qualitative) approach
and a positivist (quantitative) approach are discussed. A mixed-method
approach incorporating qualitative approach into a quantitative frame-
work is adopted.
The fourth section deals with the field study and the development of the
final research model. This section presents the detailed results of the quali-
tative data analyses. It presents a brief sample profile for the field study
and a wide discussion of the factors and variables identified during the
interviews. The section also illustrates the final model of the study by incor-
porating the field study results and the factors identified from the literature
review. The fifth section deals with the hypotheses developed and the ques-
tionnaire construction. This section develops and describes the hypotheses
for the study and also describes the instrument with its origins and sources.
A brief description of the pre-testing of the survey instrument is also
presented at the end of the section.
The sixth section deals with analysis of the quantitative data. This sec-
tion presents the analyses of the quantitative data in detail. The rationale
for sample size determination is firstly presented followed by the assessment
of the non-response bias and outer model estimation. The outer model
estimation deals with the assessment of the psychometric properties of the
model which includes assessment of the reliability and validity of the
measures. The composite reliability, content validity, construct validity and
discriminant validity of the constructs and measures are discussed in detail.

Finally, the results of the inner model estimates are presented. A summary
of the entire analyses with the results of hypotheses testing is presented at
the end of the section.
The seventh section discusses the findings of the PLS results in the light
of the major research questions and the hypotheses proposed in this study.
The theoretical and practical implications from these results are provided
in this section. The final section (Section 8) presents the conclusion and
future research directions. This section provides an overview of the study

and presents its theoretical and practical contributions. The section also
discusses the limitations and weaknesses of this study and concludes with a
brief discussion of possible future research directions in the subject area of
this study.
1.8. Summary
This section provides background information and an overview of the

research approach documented in this chapter. The review of the litera-
ture pertaining to the research topic has shown where key aspects have
not been explored, and thus has led to the research questions posed
here. This section has formulated the research questions which provide
the basis for determining the objectives of the study. The significance of
the research and definitions of related terms have been discussed.
Finally, the organisation of the chapter structure was outlined. The next
section presents the literature review and outlines a primary research
model.
2. LITERATURE REVIEW
2.1. Introduction
This section focuses on reviewing the literature, both the theories

and empirical studies, to address the ICT diffusion process and firm per-
formance from an SME perspective. Various theories relating to ICT
adoption and organisational performance are critically reviewed to build
26 MD SHAH AZAM
a comprehensive model which finally employs individual, organisational,

socio-cultural and environmental factors to look at SMEs’ ICT usage
behaviour and its consequential effects on firm performance. The following
sections of the literature review serve a number of purposes. Section 2.2
describes the diffusion process while the theoretical frameworks that
address ICT adoption-diffusion, institutional factors and firm performance
are discussed in Section 2.3. Section 2.4 reviews empirical studies and
Section 2.5 addresses the Antecedents of ICT use, while Section 2.6 dis-
cusses the Antecedents of firm performance. Finally, a testable research

model (Fig. 1) is developed based on the discussion and literature review
described in the sections below.
2.2. Diffusion Process
The adoption of a new product is a decision process that moves through

different stages of time. Diffusion refers to the process by which an innova-
tion is communicated through certain channels over time among the
members of a social system. According to Rogers (1995), the decision
process begins with knowledge about a new innovation and ends with
implementation and confirmation that the innovation is being used. The
process involves the five different phases, namely, knowledge, persuasion,
decision, implementation and confirmation, through which the adoption
and diffusion of an innovation occur.
A range of various characteristics, variables and factors affect the diffu-
sion process in different phases (Rogers, 1983). For example, the socio-
economic characteristics, personality traits and communication behaviour
of the decision-making unit influence the knowledge stage while decision-
makers’ perceptions about the innovation affect the intention or decision to
adopt an innovation. However, the implementation and continued use of
the innovation are related to several other variables.
Researchers face two fundamental questions when studying the adoption
and diffusion of an innovation. These questions are related to the differ-
ences between adoption and diffusion, and the patterns of adoption-
diffusion of an innovation by an individual and organisation. The answers
are not straightforward and vary as the process (discussed above) involves
several stages of diffusion, namely knowledge, persuasion, decision, imple-
mentation and confirmation, through which the adoption and diffusion
are projected into an integrated environment. However, according to the

innovation diffusion research, the organisation first has to make a
decision on the adoption of the innovation (e.g. new technology).
Information about the innovation is collected. This information then
leads to the formation of perceptions about the innovation (Xu &
Quaddus, 2005). In a simplistic view, adoption may be defined as the
mental state which accepts or rejects an innovation for use while the
process by which an innovation becomes popular (implementation and
confirmation) is called diffusion.

The second question is crucial for adoption researchers as there are big
differences between the individual and the organisational adoption and dif-
fusion processes. The individual adoption-diffusion process mainly relates
to an individual’s knowledge, personality, perceptions and motivation while
the organisational adoption-diffusion process depends on the organisa-
tional rules and procedures and the people involved with the decision.
Normally, a decision-making unit (in an organisation) comprises three or
more people who are primarily guided by the company policy, rules,
regulations and customs. Thus, the nature, complexity and the process of
organisational adoption and diffusion appear differently to that of the pro-
cess and nature of individual adoption and diffusion. This study explored
the organisational adoption-diffusion process from the perspective of an
SME which is similar to the individual adoption-diffusion process. In
SMEs, the decision about ICT adoption is more likely to be made by the
owner (Doukidis, Lybereas, & Galliers, 1996; Kendall et al., 2001; Matlay &
Addis, 2003; Quaddus & Hofmeyer, 2007) with often little concern about the
importance of ICT strategy and planning within their business (Beckett, 2003;
Nejadirani, Behravesh, & Rasouli, 2011).
An innovation also passes through several stages in its life cycle which
starts from the introduction stage and passes through growth, maturity and
decline stages. People from different socio-cultural, economic and financial
positions adopt an innovation in different stages. The risk-takers who
come first and try an innovation promptly (during the introduction stage)
are called innovators. Other types of adopters such as early adopters, early
majority, late majority and laggards use an innovation during the growth,
maturity and decline stages accordingly. Interestingly, while the laggards
are the users of the innovation when an innovation reaches its decline stage,
the innovators have already shifted to a new innovation which has just
started its journey.
28 MD SHAH AZAM
An innovation comes to the market to satisfy specific needs of indivi-

duals, groups or society. Thus, potential adopters use an innovation with
the motivation of satisfying some specific needs. Similarly, SMEs use an
innovation with the motivation of increasing organisational productivity
to enhance organisational performance, such as profit increase, revenue
increase or cost saving (Beard, Madden, & Azam, 2014; Madden,
Azam, & Beard, 2013). Numerous past studies have investigated the
adoption and diffusion of ICT based on the notion that ICT adoption,
diffusion or use generates improved organisational performance.

However, the question of whether the adoption and usage of the technol-
ogy have succeeded in addressing the motivation or reasons for the
adoption and usage is not well understood. This study thus has been
motivated to look at the impact of the diffusion of ICT on organisational
performance with particular attention given to the mediating effects of
integration and utilisation. An integrated framework has then been devel-
oped by extending the scope of existing adoption-diffusion theories to
address organisational performance. The following section discusses rele-
vant theories of innovation, diffusion and organisational performance and
outlines the theoretical framework for studying the ICT diffusion process
and organisational performance.
2.3. Theoretical Framework
This research has investigated SMEs’ ICT diffusion behaviour by extending

the scope of existing diffusion theories through the link with firm perfor-
mance. The theoretical grounding for the constructs under study was devel-
oped by combining the notions and principles of various theoretical
frameworks applicable to the adoption and diffusion of ICT. The theories
reviewed were the DOI theory (Rogers, 1983, 1995); TRA (Fishbein &
Ajzen, 1975); TPB (Ajzen, 1985); TAM (Davis, 1986); TOE framework
(Tornatzky & Fleischer, 1990); institutional theory (DiMaggio & Powell,
1983); UTAUT (Venkatesh et al., 2003) and resource-based theory
(Barney, 1991).
2.3.1. DOI Theory (Rogers, 1983)9

The DOI theory (Rogers, 1983) explains the diffusion process of an innova-
tion and highlights various factors that affect different phases of the
process. According to Rogers (1983), the diffusion process begins with the
knowledge of the existence of the innovation and matures through persua-
sion, decision, implementation and confirmation stages. During the knowl-
edge stage, the consumer is exposed to the innovation’s existence and gains
some understanding of how it functions. The persuasion stage refers to that
period in which the consumer forms a favourable or unfavourable attitude
towards the innovation. The persuasion stage is followed by a decision
phase. If the decision is in favour of adoption, an implementation process
starts. An innovation may be confirmed and incorporated into the culture

of the user population, may undergo changes (re-invention) or may be
rejected during the implementation phase.
The persuasion stage is considered to be the most important stage in
terms of the relative importance of the five stages explained in the DOI
theory since potential adopters gather information from various sources
and attempt to determine the utility of the innovation during this stage
(Rogers, 1995). Potential adopters’ adoption intention or willingness is
formed during the persuasion stage. Rogers (1995) highlighted five attri-
butes of innovation as perceived by the individual or organisation
perceived innovation characteristics which determine adopter willingness or
rate of adoption. These five innovation attributes are relative advantage,
compatibility, complexity, trialability and observability. These innovation
characteristics may explain 4987% of the variance in the rate of adoption
of the innovation (Rogers, 1995). The DOI theory (Rogers, 1983), although
focused on the individual’s innovation adoption process, is also effective in
examining organisational innovation adoption (Attewell, 1992; Azam &
Quaddus, 2009b; Brancheau & Wetherbe, 1990; Kendall et al., 2001; Tan &
Teo, 2000).
Numerous past studies have shown the significant impacts of perceived
innovation characteristics, as highlighted by Rogers (1983, 1995, 2003); on
the rate of adoption of an innovation. Table 1 shows that the innovation
characteristics highlighted by Rogers (1983, 1995, 2003) are supported by a
number of empirical studies.
The impacts of relative advantage, compatibility, complexity, trialability
and observability have been empirically proven in different countries
around the world. Past studies have also revealed that Rogers’ (1983, 1995,
2003) model is suitable for examining the adoption rate of a technological
innovation from the organisational perspective, particularly the SME
perspective.
30 MD SHAH AZAM
Table 1. Empirical Evidence in Favour of Rogers’ (1983) Model.

Attributes of References
Innovation
Relative advantage Kuan and Chau (2001), Wymer and Regan (2005), Jeon, Han, and Lee
(2006), Kendall et al. (2001), Tan and Teo (2000, 2004), Lim and
Speece (2002), Sathye and Beal (2001), Holak and Lehman (1990),
Azam and Quaddus (2009b), Scupola (2003b), Premkumar,
Ramamurthy, and Nilakanta (1994)
Compatibility Teo and Ranganathan (2004), Kendall et al. (2001), Jeon et al. (2006),
Azam and Quaddus (2009b), Tan and Teo (2000), Hoppe, Newman,
and Mugera (2001), Cooper and Zmund (1990)
Complexity Jeon et al. (2006), Hoppe et al. (2001), Tan and Teo (2000), Lederer,
Mirchadani, and Sims (1997), Cockburn and Wilson (1996), Azam and
Quaddus (2009b)
Trialability Kendall et al. (2001), Azam and Quaddus (2009b)
Observability Azam (2007, 2009b)
2.3.2. TRA (Fishbein & Ajzen, 1975)

The TRA (Ajzen & Fishbein, 1980; Fishbein & Ajzen, 1975) is one of the
well-researched theories for measuring behavioural intention: it explains
the causes of behavioural intent and illustrates the structures of the
relationships.
The TRA has two unique factors which are the attitude towards
the behaviour (ATT) and subjective norms (SNs) that contribute to
behavioural intention (BI) which finally explains the actual behaviour. The
TRA is guided by the underlying basic assumption that humans are quite
rational and make use of all available information, both personal and
social, before they act (Crawley & Coe, 1990).
The TRA has been widely used in previous years to examine consumers’
BI as well as to address the causes of actual consumer behaviour in various
consumption-related issues (Bang, Ellinger, Hadjimarcou, & Traichal,
2000; Chang, 1998; Crawley, 1988; Crawley & Coe, 1990). The TRA has
also been used in the IT field as the basis for testing several technologies
and has spanned a variety of subject areas, for example, word processing
(Davis et al., 1989); MS Windows (Karahanna, Straub, & Chervany, 1999);
e-commerce (Vijayasarathy, 2004); Internet information management
(Celuch, Taylor, & Goodwin, 2004) and e-banking (Shih & Fang, 2004,
2006). A particularly helpful aspect of TRA from an IS perspective is that
attitude and SNs are theorised to mediate the effect of external variables
on the intention to use new IT (Davis et al., 1989).
Many studies have shown the applicability of the TRA for studying indi-
viduals’ BI and actual usage behaviour with regard to an innovation by
revealing the significant effects of attitude towards intention (Chang, 1998;
Davis et al., 1989; Lu, Yu, Liu, & Yao, 2003; Mathieson, 1991; Ramayah,
Jamaludin, & Azam, 2007; Rhodes & Courneya, 2003; Taylor & Todd,
1995b) and SNs (Crawley & Coe, 1990; Ramayah et al., 2007; Taylor &
Todd, 1995b; Venkatesh & Davis, 2000) on BI.
The TRA was subsequently reviewed and modified to comprehensively
explain usage behaviour by adding more variables as antecedents of BI.

For example, the TPB is an extension of the TRA which includes a new
variable, perceived behavioural control, within the TRA framework.
2.3.3. TPB (Ajzen, 1985)

The TPB was proposed by Ajzen (1985, 1991) and, as explained in the pre-
vious section, is an extension of the TRA (Ajzen & Fishbein, 1980). The
TPB was developed to address the original model’s limitations in dealing
with behaviours over which people have incomplete volitional control
(Ajzen, 1985, 1991). Thus, it overcomes the problematic predictive validity
of the TRA to explain the behaviour under study which is not under full
volitional control. Ajzen (1985) made the extension by including one addi-
tional construct, perceived behavioural control, within the TRA framework
to predict BI and behaviour. Perceived behavioural control refers to
‘people’s perception of [the] ease or difficulty of performing the behaviour
of interest’ (Ajzen, 1991). A number of external factors (such as environ-
mental or organisational factors) can make a given behaviour easier or
harder to perform.
The TPB holds that human action is guided by three kinds of considera-
tions: beliefs about the likely outcomes of the behaviour and the evalua-
tions of these outcomes (behavioural beliefs); beliefs about the normative
expectations of others and motivation to comply with these expectations
(normative beliefs) and beliefs about the presence of factors that may facili-
tate or hinder performance of the behaviour and the perceived power of
these factors (control beliefs). The TPB also explains that certain factors or
constructs, known as control beliefs, may facilitate and impede people’s
behaviour; thus, they can influence a person’s adoption intention or their
purchase of a product or service (Ajzen & Madden, 1986).
Many studies have shown the effectiveness and applicability of the TPB
for examining individuals’ behaviour towards an innovation by revealing
the significant effects of attitude towards intention (Chang, 1998; Davis
et al., 1989; George, 2002, 2004; Lu et al., 2003; Mathieson, 1991;
32 MD SHAH AZAM
Ramayah, Ignatius, & Aafaqi, 2005; Ramayah et al., 2007; Ramayah,

Jantan, Noor, Razak, & Ling, 2003; Ramayah, Noor, Nasurdin, & Sin,
2004; Rhodes & Courneya, 2003; Shih & Fang, 2004; Taylor & Todd,
1995a); SNs (Ramayah et al., 2003, 2004; Taylor & Todd, 1995b;
Venkatesh & Davis, 2000); and perceived behavioural control (Cheung,
Chang, & Lai, 2000; Jiang, Hus, Klien, & Lin, 2000; Jones, Sundaram, &
Chin, 2002; Taylor & Todd, 1995a) on BI. Ajzen (1991) showed the direct
link between perceived behavioural control and actual behaviour although
it has an indirect effect on actual behaviour through significantly contribut-

ing to BI.
2.3.4. TAM (Davis, 1986)

The TAM was developed by Davis (1986) to explain IT usage behaviour. It
is an adaptation of the TRA and states that BI to use a technology is
directly determined by two key beliefs: perceived usefulness and perceived
ease of use. Perceived usefulness assesses the extrinsic characteristics of IT,
that is, task-oriented outcomes such as ‘the prospective user’s subjective
probability that using a specific application will increase his or her job per-
formance within an organisational context’. On the other hand, perceived
ease of use examines the intrinsic characteristics of IT, that is, its ease of
use (how easy it is to use), ease of learning, flexibility and clarity of the
interface. Perceived ease of use is stated as ‘the degree to which the pro-
spective users expect the target system to be free of effort’ (Davis et al.,
1989).
The earlier version of the TAM included SNs with perceived ease of use
and usefulness as antecedents of BI which was omitted from the later
model. One key benefit of using the TAM to understand system usage
behaviour is that it provides a framework for examining the influence of
external factors on system usage (Hong, Thong, Wong, & Tam, 1999).
Various external variables such as computer self-efficacy, social influ-
ence, experience, voluntariness, diversity of technology, trust, culture and
relevance have been added in the context of the TAM in different settings
in previous initiatives to gain more insight into technology acceptance
(Agarwal & Prashad, 1999; Davis et al., 1989; Hong et al., 1999; Shih,
2004; Taylor & Todd, 1995a; Venkatesh & Davis, 2000; Venkatesh &
Morris, 2000; Wang, Wang, Lin, & Tang, 2003; Yoon, 2009).
The TAM (Davis, 1986) has been one of the most frequently used
research models for examining systems’ usage behaviour over the past two
decades. Many previous studies have revealed that the fundamental compo-
nents of TAM, perceived ease of use and perceived usefulness, were found
to have strong significant effects on BI (e.g. Chau, 1997; Chau & Hu, 2002;
Davis, 1989; Davis et al., 1989; Lu et al., 2003; Mathieson, 1991;
Subramanian, 1994; Szajna, 1996; Taylor & Todd, 1995a; Venkatesh &
Davis, 2000; Yoon, 2009).
2.3.5. TOE Framework (Tornatzky & Fleischer, 1990)

In addition to these individual adoption-diffusion theories, Tornatzky and
Fleischer (1990) proposed a TOE framework to look at organisational
aspects of technology diffusion. The TOE framework identifies the follow-

ing three aspects of a firm’s context that influence the process by which it
adopts, implements and uses technological innovations:
(i) Technological context is concerned with existing technologies as well
as new technologies relevant to the firm.
(ii) Organisational context addresses descriptive measures about the orga-
nisation such as scope, size and the amount of slack resources avail-
able internally.
(iii) Environmental context refers to the aspects of how a firm conducts its
business, responds to its industry, customers and competitors, and
deals with government.
This framework has received more attention and acceptance from
diverse fields of study as it is consistent with the classical DOI theory
(Rogers, 1983). Rogers emphasised technological characteristics, and both
internal and external characteristics of the organisation, as drivers for tech-
nology diffusion.
Many previous studies have utilised the TOE framework to examine
organisational technology usage behaviour and have analysed the effects of
technological, organisational and environmental factors (e.g. Marques,
Oliveira, Dias, & Martins, 2011; Oliveira & Martins, 2010; Zhu &
Kraemer, 2005; Zhu, Kraemer, & Xu, 2003; Zhu, Kraemer, et al., 2006;
Zhu et al., 2004).
2.3.6. Institutional Theory (DiMaggio & Powell, 1983)

The institutional theory posits that organisations face pressures to conform
to these shared notions of appropriate forms of behaviours, since violating
them may call into question the organisation’s legitimacy and thus affect its
ability to secure resources and social support (DiMaggio & Powell, 1983;
Tolbert, 1985).
DiMaggio and Powell (1983) distinguished between three types of iso-
morphic pressures normative, mimetic and coercive and suggested that
34 MD SHAH AZAM
coercive and normative pressures normally operate through interconnected

relationships while mimetic pressures act through structural equivalence.
Ajzen and Fishbein (1980) explored the role of SNs and studied how it
affects individual behaviour. SNs, in other words, pressures from friends
and family, play a vital role in the formation of the intention to use an
innovation. When applied to organisations, a focal organisation can learn
about an innovation and its associated benefits and costs from other user
organisations to which it is directly or indirectly tied, and is likely to be per-
suaded to behave in a similar way (Burt, 1982). Many studies have consid-
ered normative pressure as an antecedent of organisational innovation
adoption (Kuan & Chau, 2001; Teo et al., 2003).
Mimetic pressures are the influences of other structurally equivalent
organisations that have initiated some innovations and have become suc-
cessful. These pressures may cause an organisation to change over time to
become more like the other organisations in its environment (DiMaggio &
Powell, 1983). Many past studies have included mimetic pressures when
looking at organisational ICT adoption behaviour (Premkumar &
Ramamurthy, 1995; Teo et al., 2003).
Coercive pressures address various kinds of powers or influences, infor-
mal or formal, exercised by other organisations upon which an organisa-
tion is dependent. A dominant customer, supplier or parent organisation
sometimes exercises their power or coercively influences the organisation to
act in a certain way where the dependent organisation has no option other
than to comply with the requirements. Coercive pressures are significant
when studying ICT diffusion behaviour (Quaddus & Hofmeyer, 2007;
Teo et al., 2003).
2.3.7. UTAUT (Venkatesh et al., 2003)

Venkatesh et al. (2003) proposed a new theoretical framework to explain
user intentions and subsequent usage behaviour with regard to the use of
IS. This theory applies a theoretical structure similar to the TRA or TAM.
That is, the intention explains the actual usage of ICT while it (BI) receives
various influences from the antecedent factors performance expectancy,
effort expectancy, social influence and facilitating condition. Gender, age,
experience and voluntariness of use are anticipated as possible moderators
of the four key constructs on usage intention and behaviour. Venkatesh
et al. (2003) developed the theory through a review and consolidation of
the constructs of eight models that earlier research had employed to explain
IS usage behaviour, namely, the TRA, TAM, motivational model (MM),
TPB, a combined technology acceptance model and theory of planned
behaviour (C-TAM-TPB), model of PC utilisation (MPCU), DOI theory

and social cognitive theory (SCT). Subsequent validation of UTAUT in a
longitudinal study found that it accounted for 70% of the variance in usage
intention.
Many past studies have shown the applicability of the UTAUT model
for examining usage behaviour of a technological innovation in both devel-
oped and developing countries (e.g. Gupta et al., 2008; Venkatesh &
Zhang, 2010).
2.3.8. Resource-Based View (RBV) (Barney, 1991)

The resourced-based theory, popularly known as the RBV of the firm
(Barney, 1991), has been widely used to examine organisational competitive
advantage. The theory has received attention from numerous researchers
who have been willing to investigate firm performance. The RBV is a
promising contemporary theory that combines strategic insights on compe-
titive advantage and organisational insights on firm existence. According
to Barney (1991), valuable, rare, imperfectly imitable and imperfectly
substitutable resources could generate sustainable competitive advantage
for the firm with the prerequisite of heterogeneity and imperfect mobility of
resources among competing firms. Peteraf (1993) focused on heterogeneity,
ex-post limits to competition, imperfect mobility and ex-ante limits to
competition as the characteristics for strategic resources in the generation
of sustainable competitive advantage for the firm.
Prior to the RBV framework (Barney, 1991; Peteraf, 1993), Porter
(1985) indicated that competitive advantage and its constituents could mea-
sure the firm’s success relative to its competitors. With the view to assess
competitive advantage, numerous researchers have applied the RBV to
investigate firms’ performance (Bharadwaj, 2000; Powell & Dent-Micallef,
1997).
The RBV may also look at the performance impact of ICT taking into
consideration that ICT resources and capabilities are intangible, unable to
be imitated and unique. How IT resources and capabilities can generate
firm performance or competitive advantage is the main focus of applying
RBV within IS research. The traditional RBV of ICT is successfully applied
to look at the impact of ICT on organisational performance (Bharadwaj,
2000; Powell & Dent-Micallef, 1997).
2.3.9. Review of the Existing Theories

The theoretical frameworks that address ICT diffusion can be categorised
into three groups based on their focus, scope and structures. The DOI
36 MD SHAH AZAM
theory (Rogers, 1983), TRA (Fishbein & Ajzen, 1975), TPB (Ajzen, 1985),
TAM (Davis, 1986) and UTAUT (Venkatesh et al., 2003) all focus on the
prospective users’ BI to adopt an innovation. These theories primarily
anticipate a positive link between users’ perceptions about the innovation’s
characteristics and their BI. Adopters’ perceptions about the innovation’s
characteristics are reflected in the perceived innovation characteristics in
DOI theory (Rogers, 1983); attitude in TRA (Fishbein & Ajzen, 1975) and
TPB (Ajzen, 1985); perceived usefulness and perceived ease of use in TAM
(Davis, 1986); and perceived performance expectancy and perceived effort

expectancy in the UTAUT (Venkatesh et al., 2003) model. Rogers (1983)
stated that user intention is formed during the primary stage of the diffu-
sion process, the persuasion stage, where perceived innovation characteris-
tics play a vital role in explaining the intention. Innovation characteristics
generally explain 4987% of the variation in an innovation’s adoption.
Furthermore, Rogers (1983) indicated that the diffusion process starts from
the knowledge stage, that is, when adopters are exposed to an innovation.
Adopter demographics, risk-taking behaviour and innovativeness play a
vital role in the early adoption of an innovation. Table 2 summarises the
existing theoretical frameworks.
The TRA, TPB, TAM and UTAUT show a similar structural relation-
ship among the various internal and external factors, BI and actual beha-
viour, which is different from the DOI theory. All four of these theoretical
frameworks anticipate that BI is the only antecedent of actual usage beha-
viour. User perceptions about an innovation influence BI which ultimately
explains actual usage behaviour. For example, the impact of attitude on
actual behaviour is mediated through BI under the TRA and TPB frame-
works. Likewise, the TAM and UTAUT model are designed to study the
impacts of perceived ease of use and perceived usefulness (TAM) as well as
performance expectancy and effort expectancy (UTAUT) on actual usage
behaviour mediated through BI. In addition to the internal variables, var-
ious external variables are used as antecedents of BI in the existing theories.
For example, the TRA includes SNs whether friends, family and peer
groups like or dislike the performance of a certain behaviour with
attitude. The TPB includes SNs and perceived behavioural control the
ability to perform a certain behaviour. The UTAUT includes SNs and
facilitating condition the availability of the required technological and
human resources to use a technological innovation with perceived per-
formance expectancy, effort expectancy and social influence.
The TOE framework (Tornatzky & Fleischer, 1990) anticipates a posi-
tive link between external and internal variables and organisational
Table 2. Review of the Theoretical Frameworks.
Endogenous Exogenous Variables
Variables
DOI TRA TPB TAM UTAUT Institutional Systems Usea RBV Proposed
Theory Theoretical
Constructs
Individual Owner/CEO
characteristics/ characteristics
innovativenessb
Relative advantage Attitude Attitude Perceived Performance Attitude
usefulness expectancy
Compatibility
Behavioural Complexity Perceived ease Effort
intention of use expectancy
Trialability
Observability
Subjective Subjective Social influence Normative
norms norms pressure
Perceived Facilitating Facilitating
behavioural condition condition
control (organisational
resources)
Competitive Environmental
pressure pressure
(industry
environment)
Coercive Country readiness
pressure (country
infrastructure &
resources)
Culture (societal
environment)

Table 2. (Continued )
Endogenous Exogenous Variables
Variables
DOI TRA TPB TAM UTAUT Institutional Systems Usea RBV Proposed
Theory Theoretical
Constructs
Behavioural Behavioural Behavioural

expectation intention intention
Facilitating
condition

Use Intention Intention Intention Intention Behavioural Behavioural
intention expectation
Behavioural
expectation
Facilitating Facilitating
condition condition

Performance Use and utilisation ICT use
of organisational Degree of
resources utilisation
a
A theoretical framework employed to study system usage behaviour (Venkatesh et al., 2008).
b
Personal demographics and innovativeness are the main influencers in the knowledge stage when adopters are exposed to an innovation. Early adopters intend to and adopt an innovation
very promptly.
technology usage behaviour. This framework divides all external and inter-
nal antecedent factors into three different categories, namely:
(i) Technological context the existing technologies as well as new tech-

nologies relevant to the firm
(ii) Organisational context the descriptive measures about the organisa-
tion viz., size, and the amount of slack resources available internally
(iii) Environmental context the aspects of how a firm conducts its busi-
ness, responds to its industry, customers and competitors, and deals

with government, and the process by which the organisation chooses,
adopts, implements and uses a technological innovation.
By definition, the environmental context includes various aspects

of environmental factors related to the adoption of an innovation. The
concerns arising from the organisation’s responses to its customers, compe-
titors and regulatory authorities are environmental factors. Institutional
theory (DiMaggio & Powell, 1983) posits that organisations face pressures
to conform to the shared notions of appropriate forms of behaviour in an
institutional environment. Firms working in an environment are influenced
by their customers, suppliers, parent organisations and other similar firms.
Institutional theory describes these influences by categorising three types of
isomorphic pressures coercive, mimetic and normative and suggests
that coercive and normative pressures normally operate through intercon-
nected relationships while mimetic pressures act through structural equiva-
lence. Institutional theory and the TOE framework explain the effects of
environmental factors on the process by which an organisation adopts,
uses and implements an innovation; however, they do not explain how
these factors affect the formation of intention or entrepreneurs’ perceptions
while forming the intention to adopt an innovation. The TOE framework,
although it considers the organisational and technological context in
conjunction with environmental factors, is silent about any influence from
perceived innovation characteristics.
Furthermore, how organisations deal with government policies, relevant
legislation and relevant infrastructural supports is another aspect of envir-
onmental factors which is related to country readiness. The issues related
to how an organisation conducts its business, responds to society and pro-
ceeds to choose or use an innovation are again different from the other two
types of environmental issues, namely, environmental pressures and coun-
try readiness. These phenomena are environmental concerns that are more
culture-specific.
40 MD SHAH AZAM
The above-mentioned theories explain the influence of diverse factors on

BI or actual behaviour (technology usage); however, none of them address
any aspect of organisational performance. The notion that a satisfied user
initiates repeat use and becomes a loyal user is a driver for the inclusion of
organisational performance with technology diffusion research. Satisfied
users also act as a reference for others who are willing to use the same inno-
vation. Thus, the final outcome of an innovation adoption viz., enhanced
organisational performance, has become an important concern in present
IT research. The RBV (Barney, 1991) posits that resources which are valu-
able, rare, imperfectly imitable and imperfectly substitutable could generate
sustainable competitive advantage. ICT usage is valuable and rare for the
firm. However, combining the resources with organisational processes
may make it imperfectly imitable and substitutable. Thus, the use and
degree of utilisation of ICT in an SME play a vital role in enhanced firm
performance.
The above-mentioned discussion spells out the fact that a single model
doesn’t comprehensively cover all relevant issues and also provides a
platform to synpaper the theories (the TRA, TAM, TPB, DOI theory,
UTAUT, TOE framework, institutional theory and RBV) and build a
comprehensive theoretical framework to look at the innovation diffusion
process of ICT in the SME sector and its resultant outcome in terms of
organisational performance. The comprehensive theoretical framework
combines relevant external and internal factors as antecedents of BI which
influence ICT usage behaviour with a view to addressing the impact of ICT
usage on organisational performance.
2.4. Review of Empirical Studies
This section highlights some important empirical findings which explore

the research gaps in the existing literature and provides a logical grounding
of various internal and external factors to be synpapered to develop an
integrated theoretical framework for examining the diffusion of ICT and
SME performance. When undertaking the critical review of the relevant
literature, as shown in Table 3, the motivation for the research as well as
the research direction were addressed.
The literature review revealed that studying the impacts of ICT usage on
organisational performance is an important research issue which has not
been widely documented.10 Although the diffusion of ICT has been a much
researched area of study in the past two decades, there remains a wide
Table 3. Review of Related Empirical Studies.

Author (Year Published) Research Method Independent Variables Results Comments
Zhu and Kraemer (2005) Quantitative survey E-business use: E-business use: The study addressed e-
Activity analysed: Structural equation modelling Technology competence Technology competence business diffusion and
E-business usage and (PLS) was employed with a Org. size Org. size value creation while other
value (post-adoption data set of 624 retail Org. international scope Org. financial commitment aspects of ICT
variation) industry firms across 10 Org. financial commitment Competitive pressure applications, such as
countries. Competitive pressure Regulatory support general purpose ICT, ERP,
The integrated research model Regulatory support E-business value: were overlooked.
was grounded in the E-business value: Front-end functionality The study, although featuring
innovation diffusion Front-end functionality Back-end integration technological,
literature and resource- Back-end integration E-business use organisational and
based theory which E-business use The study reported significant environmental factors, did
featured technological, differences in the firms’ not include cognitive
organisational and e-business use and value evaluation, culture and
environmental factors used between developed and country readiness in the
to examine e-business developing countries. model.
usage and value creation. The study investigated the
phenomena from the
developed country
perspective.
Zhu et al. (2003) Quantitative survey Technology competence Technology competence Only the adoption decision
Activity analysed: Data source was ECaTT, a Firm scope Firm scope was examined which is not
Electronic business (EB) data set developed by Firm size Firm size adequate for inferring the
adoption Empirica, a research Consumer readiness Consumer readiness implementation process or
institute based in Bonn, Lack of trading partner Lack of trading partner its impact on firm
Germany. readiness readiness performance.
The Logit model was used in a Competitive pressure Competitive pressure The study was based on a
data set of 3,100 businesses Industry dummies Low EB intensity countries European countries’ data
and 7,500 consumers from Country dummies were similar to full sample set which poses the
eight European countries while high EB intensity question as to whether the
countries differed. result would apply to
developing or newly
41
industrialised countries.
42
Zhu et al. (2006) Quantitative survey E-business use: Initiation: The study focused on
Activity analysed: Structural equation modelling Technology readiness Technology readiness adoption and diffusion of
Assimilation of Internet- (AMOS) was employed for Technology integration Technology integration e-business.
based e-business the estimations with a data Firm size Managerial obstacles (-) From this study, it was not
(initiation, adoption set of 1,857 firms in the Global scope Competition intensity possible to predict if the
and routinisation) retail industry across 10 Managerial obstacles Regulatory environment diffusion of e-business had
countries. Competition intensity Adoption: any positive impact on firm
The model featured Regulatory environment Technology readiness performance.
technological, Technology integration
organisational and Firm size (-)
environmental factors to Competition intensity
examine the three stages of Routinisation:
assimilation. Technology readiness
The study also compared e- Technology integration
business assimilation Firm size (-)
between developed and Managerial obstacles (-)
developing countries to Competition intensity (-)
investigate whether there Regulatory environment
was any economic effect.
Molla and Licker (2005) Quantitative survey Perceived org. e-readiness: Organisational factors, The study focused on e-
Activity analysed: Multiple discriminant function Awareness especially the human, commerce adoption and
E-commerce adoption and analysis was conducted Resources business and technological institutionalisation while
institutionalisation with a data set collected in Commitment resources, and awareness other applications of ICT
(developing country South Africa. Governance were more influential than were not included in the
MD SHAH AZAM
perspective) Perceived external e-readiness: environmental factors in model. The study looked at
Government e-readiness the initial adoption of the phenomena from the
Market forces e-readiness e-commerce. developing country
Support industries e-readiness Environmental factors, perspective. However, it
together with commitment lacked the construction of a
and the governance model comprehensive model to
that organisations analyse the phenomena.
installed, affected For example, cognitive
e-commerce evaluation, institutional
institutionalisation. pressures, culture and
country infrastructure were
not included in the model.
Hong and Zhu (2006) Quantitative survey Technology integration E-business adoption: The study focused on e-
Activity analysed: The theoretical framework Web spending Technology integration commerce adoption and
E-commerce adoption and was developed grounded in Web functionalities Web spending migration to e-commerce.
migration to technology diffusion Electronic data interchange Web functionalities The study was silent about the
e-commerce theory. (EDI) use Partner usage impact of e-commerce on
Multinomial logistic regression Partner use The adoption rate varied by firm performance.
was used with a data set of Perceived obstacles firm size in the case of the The study dealt with the
1,036 firms in a broad Control variables: overall sample and non- phenomena from the
range of industries. Firm size adopters versus adopters developed country
Industry type sub-sample while it had no perspective.
effect on potential adopters
versus adopters sub-
sample.
Migration to E-business:
Web spending
Web functionalities
EDI (-)
Partner usage (-)
Perceived obstacle (-)
Madden et al. (2013) Quantitative survey The study included strategic Firms that enter into the This study basically proved
Dependent variable: Trivariate Probit model reasons efficiency, online market for market that ICT application (online
Small firms online market estimations were conducted market expansion, expansion are successful market entry) in SMEs
entry with a data set of 1,001 introducing new goods, while firms that targeted improved firms’ market
small and medium-sized responding to customer cost reduction were performance (market
businesses in Australia. requests, supplier disappointed. expansion and profit). This
requirements, anticipating The study had no evidence study included
competition for entry that blended firms enjoyed environmental pressures
into online business and the any important advantage such as customer requests,
effect on market over their virtual supplier requirements and
performance with various competitors. anticipating competition;
43
firm characteristics, however, it did not explain
44
industry characteristics and the effects of other

web investment. technological,
organisational and
environmental factors on
ICT use or subsequent firm
performance.
Jehangir and Downe Quantitative survey Business performance: E-commerce capability: IT infrastructure and IT
(2011) Linear regression was used IT infrastructure IT infrastructure human resources can create
Activity analysed: with a data set of 243 IT human resources IT human resources e-commerce capability
Business performance Malaysian manufacturing E-commerce capability Business performance: which finally affects
firms. E-commerce business performance.
Konings and Roodhooft Quantitative survey IT infrastructure Use of e-business was E-business contributed
(2002) The final sample included IT human resources substantially higher among positively to the
Activity analysed: 836 firms. The survey was a E-commerce large firms than in small productivity, that is the
E-business effect on firm postal survey in which firms which however varied performance of the firm.
productivity and cost survey instruments were in different industries.
efficiency sent to 5,718 firms. Large firms engaged in e-
business have higher total
factor productivity than
large firms that are not.
E-business has no effect on
factor productivity of small
firms.
MD SHAH AZAM
Oliveira and Martins Quantitative survey The research model included: Firms with high levels of TOE The TOE framework, in
(2010) The final sample included Technology readiness factors also had an general, is applicable to the
Activity analysed: 6,694 firms belonging to Firm size enhanced level of e- study of European firms’ e-
E-business adoption by the EU27 members, Expected benefits and barriers business. business adoption.
European firms excluding Malta and of e-business High levels of competitive Competitive pressures and
Bulgaria. Computer-aided Improved products or services pressure led to high levels industry-specific
telephone interview (CATI) or internal processes of e-business adoption. characteristics are related
technology was used with Internet penetration Industry-specific to high levels of e-business
Competitive pressure characteristics, not adoption.
the randomly selected Industry country-specific
samples. characteristics, better
Factor analysis and cluster explained e-business
analysis were conducted adoption.
for estimates.
Powell and Dent-Micallef Empirical study Human resources Human complementary The findings showed that IT
(1997) Quantitative survey Business resources resources accounted for alone did not produce
Activity analysed: Technology resources significant overall sustainable performance
IT and firm performance Dependent variable performance variance advantages in the retail
IT performance with the human resources industry, but some firms
Overall performance set yielding a large positive gained advantages by using
Profitability coefficient for all IT to leverage intangible,
Sales growth performance measures. complementary human and
Technology resources did not business resources such as
account for any significant flexible culture, strategic
firm performance variance: planningIT integration,
they only influenced IT and supplier relationships.
performance. The results also helped to
explain why some firms
outperform others using
the same IT, and why
successful IT users often
fail to sustain IT-based
competitive advantages.
Gupta et al. (2008) Quantitative survey Intention The study found that This framework can be used to
Activity analysed: Data were collected from 102 Performance expectancy performance expectancy, look into the technology
ICT adoption in a employees of a government Effort expectancy effort expectancy and social diffusion phenomena in
government organisation in India. Social influence influence were significantly other organisations
organisation in a PLS-based SEM was Usage associated with intention to (government or private)
developing country employed for data analysis. Facilitating condition use ICT while facilitating from a developing country
Behavioural intention condition was found to perspective.
have a significant positive An interesting outcome of this
effect on ICT-use study was that the
behaviour. intention variable was not
45
relevant in contexts where
46
The study did not find any the technology was already
significant relationship being used. The finding
between behavioural may suggest possible
intention and actual changes to the existing
Internet usage. Intention to theoretical frameworks;
use is relevant in situations thus, it should be re-
where the technology is examined in different
new and the users have not contexts.
used it which was not the
case here.
Venkatesh and Zhang Quantitative survey Intention Intention The framework which
(2010) A longitudinal survey was Performance expectancy Performance expectancy included performance
Activity analysed: conducted of the employees Effort expectancy Gender and age positively expectancy, effort
Use of IT at in an organisation which Social influence moderate the effects of expectancy, social
organisational level in operated in USA and Usage performance expectancy, influence, behavioural
USA and China China. Facilitating condition while the interaction of intention and facilitating
In all, 300 employees were Behavioural intention gender, age and experience condition, is worth
surveyed in each country Gender, age, experience and negatively moderates effort working with in different
from one business unit. voluntariness of use were expectancy (for both countries.
PLS-based SEM was used as included as moderators in countries); however, the Although this framework
the analytical tool. various relationships. interaction of gender, age, provided an illustration of
voluntariness and the factors influencing ICT
experience on social usage behaviour in a
influence has a negative developed as well as in a
MD SHAH AZAM
effect in the United States. developing country, it did
Usage not indicate whether the
Behavioural intention use of technology could
Interaction of age and generate organisational
experience positively affects performance.
facilitating condition in Importantly, the study
both countries. indicated that PLS-based
structural equation
modelling (SEM) was
efficient in analysing the
usage of IT.
Venkatesh et al. (2008) Quantitative survey Behavioural expectation Behavioural intention and Behavioural expectation
Activity analysed: In all, 321 users of new Behavioural intention behavioural expectation emerged as a strong
Use of new information information systems were Facilitating condition played significant roles only indicator for predicting the
systems surveyed in a longitudinal Usage in direct effects analysis system usage behaviour
field study. Behavioural intention while behavioural addressing some of the key
Behavioural expectation expectation emerged as the limitations of behavioural
Facilitating condition only significant factor in intention and facilitating
Gender, age and experience direct and interaction condition and providing a
were used as moderators in effects analysis for better understanding of
various situations. frequency and intensity of systems use.
use; however, behavioural Behavioural expectation could
intention affected the be included as a new
duration of use. construct to study IT usage
behaviour.
Thong (1999) Quantitative survey Decision-maker CEO innovativeness and IS PLS-based structural equation
Activity analysed: Data analysis was carried out characteristics included knowledge, relative modelling (SEM) was
Adoption of information with a data set of 166 small innovativeness and IS advantage and functional in the analysis
systems (IS) in small businesses. knowledge. compatibility, complexity, of IS usage behaviour.
businesses To collect data, a mail order IS characteristics included business size and CEO innovativeness and
survey was conducted with relative advantage, employees’ IS knowledge knowledge, relative
1,200 small businesses in compatibility and were significantly related to advantage/compatibility,
Singapore: while 294 complexity. the likelihood of IS business size and
responses were returned, Organisational characteristics adoption while business employees’ knowledge
only 166 were complete included business size, size, employees’ IS (skilled human resources)
responses. employees’ IS knowledge knowledge and information emerged as key constructs
and information intensity. intensity played a for examining the
Environmental characteristics significant role in explaining likelihood of IS adoption
particularly included the extent of IS adoption. by small firms.
competition. Although CEO innovativeness
and environmental
pressures were not
47
observed as significant in
48
explaining the extent of

information about IS, these
factors should be looked
into in different contexts.
Yap, Soh, and Raman Quantitative survey Consultant effectiveness Consultant effectiveness The study confirmed the
(1992) Data analysis was carried out Vendor support Vendor support significant effect of
Activity analysed: with a data set of 96 small CBIS experience CBIS experience consultant effectiveness,
Success of computer- businesses in Singapore. Financial resources Financial resources vendor support, CBIS
based information CEO support CEO support experience, financial
systems (CBIS) in Level of user participation Level of user participation resources, CEO support
small businesses Number of administrative and level of user
applications participation on CBIS
Presence of systems analyst success while administrative
applications and the
presence of a systems
analyst had no significant
effects.
Thong and Yap (1995) Quantitative survey Individual characteristics CEO characteristics With business size, CEO
Activity analysed: Data analysis was carried out included CEO Business size characteristics played a
IT adoption in small with a data set of 166 small innovativeness, CEO significant role in IT
businesses businesses attitude towards adoption adoption by small
of IT and CEO IT businesses. Clearly, small
knowledge. businesses were more likely
Organisational characteristics to adopt IT if CEOs were
MD SHAH AZAM
included business size, more innovative, had a
competitiveness of positive attitude towards
environment and the adoption of IT, and had
information intensity. greater IT knowledge.
Kendall et al. (2001) Quantitative survey Perceived innovation Relative advantage Relative advantage,
Activity analysed: characteristics included Compatibility compatibility and
E-commerce adoption by Data analysis was carried out relative advantage, Trialability trialability played
SMEs in Singapore with a data set of 58 SMEs compatibility, complexity, significant roles in the
in Singapore. trialability and adoption of e-commerce by
observability Singaporean SMEs while
complexity and
observability were not
found to have any
significant effect.
Teo et al. (2003) Quantitative survey Coercive pressures Coercive pressures Adoption of
Activity analysed: Data analysis was carried out Mimetic pressures Mimetic pressures interorganisational systems
Adoption intention of with a data set of 492 Normative pressures Normative pressures was largely influenced by
financial electronic individuals from 222 non- the environmental pressures
data interchange (EDI) adopting organisations in arising from institutional
Singapore. environments.
Dada (2006) Case study Initial adoption and The study found that The results may be used as
Activity analysed: The study modified the institutionalisation: organisational factors were validation of UTAUT
Group support systems framework of Molla and Organisational factors influential in the adoption factors particularly
(GSS) in two Licker (2005) reflecting the (factors leading to usage and institutionalisation of perceived effort expectancy
developing countries notion of the UTAUT behaviour from UTAUT GSS while country-level and performance
Tanzania and South framework (Venkatesh framework) factors did not have any expectancy, facilitating
Africa et al., 2003). Environmental factors influence on the intention condition and subjective
(e-readiness measure) and likelihood of using norms. The study, however,
GSS. was important as it looked
at the effects of country-
level e-readiness, finally
nullifying its effect which
should be further looked
into with a quantitative
survey.
Marques et al. (2011) Quantitative survey Technology (equipment and Technology readiness, country Country-level factors such as
Activity analysed: The study analysed the data processes), organisation wealth and education level country wealth (GDP per
Adoption of medical set comprising 448 (size, location, managerial were significantly associated inhabitant, % of
record system hospitals in Europe (data structure), human context with medical record systems households connected to
source: e-Business Watch (user involvement) and adoption in European the Internet, total spending
49
environmental context hospitals. on R&D as a % of GDP)
50
2006 decision-maker (cultural environment of emerged as antecedents for

survey). the country and regulatory the adoption of a new
The study added influence) information system.The
environmental aspects into Three factors (country wealth, study was conducted in the
the human, organisation competition and context of Europe; thus, it
and technology fit (HOT- technology readiness) were is not sure how this variable
fit) framework and built a finally explored through would behave in other
new framework which it principal component countries particularly in
called the HOTE (human, analysis. developing countries.
organisation, technology
and environment)
framework.
Erumban and Jong (2006) Secondary sources of data Power distance Without country dummy and The results of the study
Activity analysed: were used. This study used Uncertainty avoidance education: suggested that ICT
ICT adoption rate data about cultural Individualism Uncertainty avoidance adoption rate in a country
differences across countries Masculinity Individualism was closely related to its
based on Hofstede’s Long-term orientation With country dummy: national culture. The study
dimensions. The proportion Country dummy Power distance revealed that power
of ICT expenditure across Education Uncertainty avoidance distance and uncertainty
42 countries and per capita Masculinity avoidance were the most
computers across 49 Country dummy important dimensions
countries (from 1991 to Education explaining ICT adoption
2001) were used to establish rate in a country.
MD SHAH AZAM
the ICT adoption rate. The
study used ICT expenditure
and per capita computer
data for those countries for
which the corresponding
Hofstede indices were
available.
Gibbs and Kraemer Quantitative survey External pressure The study found significant The study confirmed the
(2004) Data collected from 2,139 Perceived benefits effects of perceived benefits, importance of strategic
Activity analysed: establishments from three Government promotion external pressure, benefits, external pressure,
Scope of usage of industries across 10 Org. compatibility government policy, technology and financial
e-commerce countries Legislation barriers legislation barriers, resources, and policy

Technology resources technology resources, environment on the range
Financial resources financial resources and all of e-commerce use. US
Size country dummies, except firms enjoyed significantly
Industry and country dummy for Denmark, on the range higher range of use than
Organisational of e-commerce usage. firms from other countries
IT sophistication; Top which also confirmed the
management support; significant country effects
Firm size on the range of e-commerce
use.
Thatcher et al. (2006) The study adopted both Industry All factors under This study confirmed that
Activity analysed: qualitative and interpretive Importance of responsiveness; organisational, industry, cultural dimensions were
Adoption of e-commerce methodology. The four Importance of cost cutting; government and culture associated with the
systems in electronics phases of the methodology Multinational companies; were significantly associated adoption of B2B e-
and textile industries were data collection, data Trend-setting companies with B2B adoption in the commerce in both
reduction, data display and Government Taiwanese electronics Taiwanese electronics and
drawing conclusions. Policies promoting B2B industry. However, the textile industries.
(business to business); effects varied with the Interestingly, the effects
Subsidies promoting B2B adoption of B2B e- were positive in the
Culture commerce in the textile electronics industry while
Power distance; Uncertainty industry. More specifically, the cultural dimensions
avoidance; Individualism IT sophistication, were negatively related to
versus collectivism; importance of B2B e-commerce adoption
Masculinity versus responsiveness and in the textile industry.
femininity; Time government subsidy
orientation; High context promoting ICT had positive
versus low context effects: firm size and being a
trend-setting company had
no effects while all other
factors were negatively
associated.
51
52 MD SHAH AZAM
research gap between the perspectives of developed and developing coun-

tries on ICT diffusion behaviour. The applicability of existing diffusion
theories to developing countries is also not clear. Contemporary researchers
who wish to learn about the diffusion of an innovation from a holistic
approach seek to analyse the important aspects of the ICT diffusion
phenomena under an integrated and comprehensive model.
Previous studies have analysed various issues of the diffusion of an inno-
vation by applying a wide range of theoretical frameworks. Most of them
have investigated one aspect at a time of the contextual and environmental

factors with the view of avoiding analytical complexity by creating a parsi-
monious research model. It is known that the inclusion of a large number
of important and relevant variables improves the goodness of model fit and
enhances the variation of the model explained. It is not unlikely that vari-
ables (endogenous or exogenous) have different effects in different contexts
as well as with the number of different variables included in the model.
The literature review (see Table 3) and discussion in Sections 2.5 and 2.6
have built the foundation of an integrated and comprehensive research model.
The primary model based on the literature review is presented in Fig. 1.
2.5. Antecedents of ICT Use
2.5.1. Cognitive Evaluation and Owner Innovativeness

Beliefs, evaluations, intention and actual behaviours are modelled in the
TRA with SNs. Fishbein and Ajzen (1975) suggested that an overall
affective evaluation, that is, one’s beliefs (cognitive) towards performing a
behaviour, and the respective evaluations (affective) heavily influence an
individual’s intention of performing the behaviour. The evaluation is made
based on one’s cognitive evaluation of the consequences of that behaviour
(Ajzen & Fishbein, 1980; Sheppard, Hartwick, & Warshaw, 1988). The
evaluation process involves functional consequences, that is, outcomes
which are immediate, direct and tangible, as well as psychological conse-
quences, that is, the consequences are internal, personal and abstract in
nature. These cognitive and affective evaluations form one’s attitude, that
is, negative or positive feelings about performing a behaviour with this
having been widely researched in the consumer, marketing and IS literature
(Ajzen, 1985; Cho, 2004; Fishbein & Ajzen, 1975; Gehrt & Carter, 1992;
Liao & Cheung, 2001; Mehta & Sivadas, 1995). Prior studies have shown
the significant effect of attitude, that is, cognitive evaluation towards inten-
tion (Chang, 1998; Davis et al., 1989; George, 2002, 2004; Lu et al., 2003;
Mathieson, 1991; Ramayah et al., 2003, 2004, 2005, 2007; Rhodes &
Courneya, 2003; Shih & Fang, 2004; Taylor & Todd, 1995a).
Personal characteristics play a significant role in individual decision-
making or choice behaviour. Individuals affiliated with different demo-
graphic groups show different brand preferences. For example, variations
in product preference and choice behaviour appear in educated versus
non-educated, rich versus poor, urban versus rural and between individuals
affiliated with different professional positions. Individuals are variously
classified according to their receptivity and adoption of an innovation.

Rogers (1995) categorised five different types of adopters according to the
time of adoption as innovators, early adopters, early majority, late majority
and laggards (Weber & Kauffman, 2011). Innovators are adopters who are
innovative in nature and who are prepared to bear the risks associated with
the early adoption of an innovation. The adoption of an innovation in a
small firm is highly influenced by the personal characteristics and prefer-
ences of the owner or CEO with most small businesses being managed by
the owner who also acts as the CEO (Solomon, 1986; Steinhoff & Burgess,
1986; Yap et al., 1992). Thus, innovative SME owners adopt innovation at
the early stage. Prior studies have shown that the innovativeness of the
CEO or owner of an SME significantly affects the adoption (Thong, 1999;
Thong & Yap, 1995) and the implementation of IS (Thong, 1999).
2.5.2. BI and Behavioural Expectation

Intention is one of the strong predictors of the adoption of an innovation.
From the earliest theories of adoption-diffusion, intention has been used to
address the adoption of innovation. The TRA predicts intention as a func-
tion of attitude towards behaviour and SNs (Fishbein & Ajzen, 1975). The
intention is measured through the function of five innovation characteris-
tics, namely, relative advantage, perceived compatibility, perceived com-
plexity, perceived trialability and perceived observability in the DOI model
(Rogers, 1983). Perceived risk (Davila, Gupta, & Palmer, 2002; Lim &
Speece, 2002; Tan & Teo, 2000) and adopters’ characteristics (Rogers, 1983;
Thong & Yap, 1995) are also correlated with intention. Again, perceived
behavioural control has a direct correlation with BI (Ajzen, 1985). Intention
has also been elaborately presented as the function of attitude towards beha-
viour and SNs while attitude towards behaviour is analysed as a function of
perceived ease of use and perceived usefulness (Davis, 1986).
Despite significant improvements of the past innovation adoption the-
ories from the TRA (Fishbein & Ajzen, 1975) to the TAM (Davis, 1986),
54 MD SHAH AZAM
some major drawbacks still prevail. Diversified uses of technology, from

individual level to different trading situations, have involved some modifi-
cations of findings from previous studies in order to adequately address
technology adoption-diffusion behaviour. Some new dimensional con-
structs have therefore been included within the existing theories as is
evident in the literature. Venkatesh et al. (2003) proposed some modifica-
tions in the existing behavioural theory by having their UTAUT model
address performance expectancy, effort expectancy and SNs as antecedents
of BI to explain actual IT usage behaviour. Although intention, a reflection

of the adopter’s internal schema of beliefs, had great influence in earlier beha-
vioural models, it cannot predict the situation when the adopter is not under
full volitional control or the situation where a time gap exists between inten-
tion and actual behaviour (Venkatesh et al., 2008). However, a stronger
predictor of actual behaviour may incorporate some other external (social
and environmental) factors which need to be considered in the existing model.
Venkatesh et al. (2008) further modified their UTAUT model (Venkatesh
et al., 2003) and included behavioural expectation as a new construct which is
a stronger predictor of behaviour than BI. Taking into consideration the
scope and limitations of various previous theories, this study has included
behavioural expectation as a possible predictor of actual ICT use.
2.5.3. Facilitating Condition and Country Readiness

The studies utilising the traditional RBV have proposed that if firms can
combine IT-related resources to create a unique IT capability, this can
result in superior firm performance (Bharadwaj, 2000; Ravichandran &
Lertwongsatien, 2005). IT capability is addressed as the facilitating
condition in ICT diffusion studies (Venkatesh et al., 2008). The facilitating
condition is defined as individuals’ perceptions of the availability of techno-
logical and/or organisational resources (i.e. knowledge, resources and
opportunities) that can remove barriers to using a system (Venkatesh et al.,
2003). Thus, network connectivity, Internet speed, computer hardware
resources, software and skilled manpower produce favourable facilitating
conditions which foster the use of ICT. This phenomenon is also evident in
the previous classical behavioural theory as behavioural control in the TPB
(Ajzen, 1985, 1991) and self-efficacy in SCT (Bandura, 1986).
The organisational ICT RBV and facilitating condition need to be stated
differently as, although the organisation may have enough resources, this
does not necessarily mean that these resources facilitate the achievement of
better performance. The facilitating condition’s effect on behavioural
expectation and ICT use is not limited only to organisational resources.
The TOE framework (Tornatzky & Fleischer, 1990) includes environ-

mental factors with technological and organisational factors for studying
the adoption of ICT at firm level. Previous studies have considered various
aspects of external environmental factors, such as institutional environmen-
tal concerns (Oliveira & Martins, 2010; Teo et al., 2003); country infra-
structure and governance concern (Gibbs & Kraemer, 2004; Marques et al.,
2011; Xu, Zhu, & Gibbs, 2004; Zhang, Cui, Huang, & Zhang, 2007; Zhu,
Kraemer, et al., 2006) and cultural concerns (Burn, 1995; Erumban &
Jong, 2006; Gefen & Straub, 1997; Straub, 1994; Thatcher et al., 2006).
The effects of country infrastructure, governance-related external factors
and organisational facilitating conditions are alike. In reality, these aspects
are related to country readiness (Marques et al., 2011) in the facilitation
of ICT diffusion at individual, organisational and government level. The
availability and geographical coverage of the Internet, speed and price of
Internet service, ICT resources (hardware and software), regulatory frame-
work, market conditions, delivery systems, government policy and support
have direct effects on ICT adoption. The condition of the favourableness
or unfavourableness of the infrastructural factors accelerates or inhibits
ICT use at firm level. Thus, this phenomenon (country readiness) should be
considered as a new aspect of the facilitating condition which is different to
the facilitating condition (Venkatesh et al., 2008; Venkatesh & Zhang,
2010) explained in previous studies. This study thus considers both of these
aspects which are logically added as separate predictors of behavioural
expectation as the facilitating condition (firm-level resources) and country
readiness (national level infrastructure and governance).
2.5.4. Culture and Environmental Pressure

Culture is a broad spectrum of behavioural study which has been defined
differently in different studies in the literature. In the broadest sense,
culture may be defined as the sum total of shared learned beliefs, values,
norms and customs which guide individual or group behaviour in a society.
Hofstede (2001) treated culture as the collective programming of the mind
that distinguishes the members of one group or category of people from
another. Although the effects of culture on leadership and organisational
processes as well as on individual behaviour have been explored in many
studies (House et al., 2004), its effects when studied in ICT adoption
research are inconclusive.
The importance of culture in determining individual innovativeness relative
to various personal and perceptual factors during adoption of an innovative
or new product is a vital research issue. This aspect of cultural study is also
56 MD SHAH AZAM
neglected in consumer research. Parker and Sarvary (1996) suggest that

culture has no significant impact on innovativeness at the individual level.
There is controversy about the role of culture in the use and adoption of
IT. Some researchers favour the view that organisations adopt technology
that is useful and provides them with some economic benefit with culture
having no significant role. However, other researchers hold the alternative
view that culture plays an important role in determining not only whether
organisations in a particular country adopt a certain technology but it also
impacts on the degree to which that technology is accepted and the ways in
which it is used (Thatcher et al., 2006).
Although they are limited in number, some previous studies have
reported a significant link between cultural dimensions and different facets
of IT use (Bertolotti, 1984; Burn, 1995; Erez & Early, 1993; Gefen &
Straub, 1997; Harris & Davison, 1999; Hill, Loch, Straub, & El-Sheshai,
1998; Ho, Raman, & Watson, 1989; Straub, 1994). Erumban and Jong
(2006) found that the national culture and the ICT adoption rate of a
country are closely related. They further reported that most of Hofstede’s
dimensions are important in influencing adoption with power distance
and uncertainty avoidance dimensions seeming to be the most influential.
Thatcher et al. (2006) supported this outcome in ICT adoption, particu-
larly B2B e-commerce adoption, in the Taiwanese electronics industry
context.
Organisations that operate within a cultural setting are believed to be
influenced by the shared learned beliefs, values, norms and customs of the
specific culture. Similar to national culture, organisational culture has been
defined as the collective programming of the mind that distinguishes the
members of one organisation from those of another.
In the ‘Global Leadership and Organisational Behaviour Effectiveness’
(GLOBE) study (House et al., 2004), culture has been examined in two
ways: as the way of carrying out behaviour (practice) and the way through
which behaviour should be carried out (values). House et al. (2004) used
nine different cultural dimensions to study the effects of the national
culture on the organisation and its leadership, namely, power distance,
uncertainty avoidance, in-group collectivism, institutional collectivism,
humane orientation, assertiveness, gender egalitarianism, future orientation
and performance orientation. On the other hand, Hofstede (1984) used five
dimensions to study culture, namely, power distance, individualism versus
collectivism, masculinity versus femininity, uncertainty avoidance and
long-term versus short-term orientation. Although House et al. (2004) have
defined culture by further elaboration of Hofstede’s (1984) concepts, both
theoretical frameworks have been used in multidisciplinary fields of study

when addressing culture.
After reviewing Hofstede’s (1984) concepts and House et al.’s (2004)
dimensions as well as the context of Bangladesh, the following dimensions
have been included in the proposed research model: power distance (refers
to the inequality of the distribution of power in a country); in-group collec-
tivism (the degree to which individuals express pride, loyalty and cohesive-
ness in their organisations or families); uncertainty avoidance (the extent to
which a society, organisation or group relies on social norms, rules and

procedures to alleviate the unpredictability of future events) and gender
egalitarianism (the degree to which a collective minimises gender inequal-
ity). The measurement items for national culture and organisational culture
have been adapted from House et al. (2004).
In addition to country readiness and national culture, the usage of ICT
at the firm level may be influenced by some other factors arising from the
institutional environment (see Teo et al., 2003).
The impact of these aspects of environmental factors and their charac-
teristics has been well addressed in institutional theory. This needs to be
synpapered with the adoption-diffusion theories so as to reach the external
variables reflected in Rogers’ (1983) DOI theory and the subsequent TOE
framework and other innovation diffusion frameworks.
Institutional theories have posited that organisations face pressures to
conform to these shared notions of appropriate forms of behaviours, since
violating them may call into question the organisation’s legitimacy and
thus affect its ability to secure resources and social support (DiMaggio &
Powell, 1983; Tolbert, 1985).
DiMaggio and Powell (1983) distinguished between three types of iso-
morphic pressures coercive, mimetic and normative and suggested
that coercive and normative pressures normally operate through inter-
connected relationships while mimetic pressures act through structural
equivalence.
Ajzen and Fishbein (1980) explored the role of SNs and studied how
they affect individual behaviour. SNs, in other words, pressures from
friends and family play a vital role in the formation of intention to use an
innovation. In organisational aspects, a focal organisation is able to learn
organisations that are directly or indirectly tied to it, and is likely to be per-
suaded to behave similarly (Burt, 1982). Many studies have considered nor-
mative pressure as an antecedent of organisational innovation adoption
phenomena (Kuan & Chau, 2001; Teo et al., 2003).
58 MD SHAH AZAM
Mimetic pressures are the influences of other structurally equivalent

organisations that have initiated some innovations and have become suc-
cessful. This pressure may cause an organisation to change over time to
become more like other organisations in its environment (DiMaggio &
Powell, 1983). Many past studies have included mimetic pressures
in exploring organisational ICT adoption behaviour (Premkumar &
Ramamurthy, 1995; Teo et al., 2003).
Coercive pressures address various kinds of powers or influences, infor-

do a certain thing with the dependent organisation having no option other
than to comply with the requirements. Coercive pressures are of great
importance when studying the influences on ICT diffusion behaviour
(Quaddus & Hofmeyer, 2007; Teo et al., 2003).
2.6. Antecedents of Organisational Performance
2.6.1. ICT Use, Integration and Degree of Utilisation

A large number of empirical studies have predicted IS usage behaviour by
applying BI as the final dependent variable. The proposition ‘positive inten-
tion leads to positive behaviour’ has provided the rationale for such studies.
Some studies have also examined actual usage as a dependent variable for
predicting actual behaviour. Different indicators of systems use, such as
duration of use, frequency of use and intensity of use, have been used to
reach actual usage behaviour (Venkatesh et al., 2008). Organisational ICT
use has also been examined in different stages of assimilation such as initial
adoption (the firm’s initial evaluation of an ICT-based operation at the pre-
adoption stage); adoption (formal adoption of an ICT-based operation)
and routinisation (full-scale deployment of ICT at the post-adoption stage
in which the ICT-based operation becomes an integral part of the value-
chain activities) (Fichman, 2000; Zhu, Kraemer, et al., 2006). Use of ICT
has also been applied as the independent variable in examining the impact
of ICT on organisational performance (Bharadwaj, 2000; Zhu, Dong,
et al., 2006; Zhu & Kraemer, 2005). The results of such studies have
revealed that greater ICT usage leads to greater impact on business
performance.
Zhu and Kraemer (2005) explored beyond the dichotomous ‘adoption
versus non-adoption’ phenomenon and examined the impact of actual ICT
usage on organisational ICT value creation, that is, organisational perfor-

mance. In analysing 624 firms across 10 countries in the retail industry, their
study revealed that front-end capabilities and back-end integration of ICT
had significant impact on organisational value creation while the impact of
back-end integration appeared stronger than that of front-end functional-
ities. This outcome was consistent with the RBV as back-end integration
may create ICT resources which are firm-specific and difficult to imitate
thus providing enhanced firm performance. Integration thus becomes an
important dimension of the post-adoption stages of ICT usage behaviour.

Anandarajan et al. (2002) reported that the mere adoption of IT by
organisations does not necessarily confer benefits as these would only be
achieved by its effective usage. The ICT adoption phenomena in developing
countries as well as in less developed countries is assumed to be similar as
other studies have shown that many IS in LDCs are under-utilised and
hence do not make a significant contribution to improving the performance
of the organisations that are using them (Forster & Cornford, 1992;
Ordedra et al., 1993). In support of the previous studies, Song and
Mueller-Falcke (2006) stated that SMEs are confronted with a number of
challenges in adopting and using ICT with the result that they often under-
utilise the available technologies (see UNCTAD, 2006).
The above discussions reveal that the adoption and post-adoption stages
of ICT usage comprise various strategic uses of ICT which start from the
initial stage of ICT usage through to the highly integrated ICT-based
environment. Proper utilisation of ICT is an important issue for the pro-
ductive use of ICT in developing countries. The various uses of ICT may
be categorised as ICT usage (Venkatesh et al., 2008) (depth of use which
deals with frequency and intensity of ICT use); integration (Powell & Dent-
Micallef, 1997) (breadth of ICT use which is related to the strategic use of
the technology, that is, to what extent the technology is integrated with the
organisational goals, plans and processes) and utilisation (how appropri-
ately the ICT is utilised and how much of the facilities of the technology
are used in different functional areas in the organisation, that is, frequency
of use and its strategic applications), all of which have implications for
2.6.2. Organisational Performance

Defining and measuring business performance has been of interest to
researchers for centuries (Pham & Jordan, 2007). Despite the controversy
over the business value of computer investments, known as the ‘productiv-
ity paradox’, which still continues even in the face of successful evidence
60 MD SHAH AZAM
about pay-offs from IT (Brynjolfsson, 1993; Brynjolfsson & Hitt, 1996),

numerous studies have found a positive link between IT use and organisa-
tional performance (Baldwin & Sabourin, 2001; Cragg, King, & Hussin,
2002; Dvir, Segev, & Shenhar, 1993; Gretton, Gali, & Parham, 2004;
Pilat & Wyckoff, 2005). However, the measurement of firm performance is
not straightforward. Past researchers have used different conceptualisations
to measure business performance. Szymanski, Bharadwaj, and Varadarajan
(1993) suggested studying business performance based on market share
and profitability while Voss and Voss (2000) viewed performance as the
outcome of the company business process.
Organisational performance, an important indicator of economic devel-
opment, has been conceptualised in different ways in different studies. It
can be measured objectively based on historical data (Bharadwaj, 2000;
Sanders & Premus, 2005) or measured subjectively based on respondents’
perceptions of organisational performance in relation to their expectations
and goals, or in comparison with the company’s competitors (Powell &
Dent-Micallef, 1997; Ravichandran & Lertwongsatien, 2005).
In previous studies, various dimensions of organisational performance
have been considered for linkages with ICT such as profit growth, market
share, productivity growth, improved performance and the firm’s competi-
tiveness. In this study, organisational performance was measured using
the subjective approach by evaluating the organisation’s performance in
different functional areas according to the plans and goals of the organisa-
tion that had been previously determined.
2.6.3. Mediation of Integration and Utilisation

Researchers and practitioners involved in studying ICT use and utilisation
in the organisational context are interested in discovering the answer to the
question of how and to what extent ICT usage is associated with organisa-
tional performance. The answer is equally important for the CEOs or own-
ers of organisations that are currently using ICT or that intend to use ICT
in different functional areas. The DOI theory contends that the impact of a
new technology depends on the extent to which it is used in key value-chain
activities. Previous studies have found different results when examining the
impact of ICT usage on organisational performance, such as negative effect
(Warner, 1987); no effect (Sager, 1988; Venkatraman & Zaheer, 1990);
mixed effects (Powell & Dent-Micallef, 1997; Tippins & Sohi, 2003; Wu
et al., 2006) and direct positive effects (Bharadwaj, 2000; Clemons & Row,
1991; Clemons & Weber, 1990; Feeny & Ives, 1990; Kettinger, Grover,
Guha, & Segars, 1994; Li & Ye, 1999; Schwarzer, 1995). Zhu and Kraemer
(2005) suggested that merely examining the initial adoption might not
reveal the variations in IT value, because IT creates business value in
sequential stages. They again stated that despite its theoretical importance,
usage has been under-studied in empirical research which was therefore
their motivation for analysing the linkage between usage and performance
impact. Zhu and Kraemer (2005) found a positive link between ICT use
(e-commerce) and value creation which was consistent with empirical find-
ings of the importance of usage in different contexts such as electronic data
interchange (EDI) (Mukhopadhyay, Kekre, & Kalathur, 1995) and deci-

sion support systems (DSSs) (Devaraj & Kohli, 2003). Their findings also
complied with the notion that ‘without broad and deep use of e-business
along the value chain, it would be impossible for e-business to generate any
impact on firm performance in terms of sales, procurement, or internal
operations’ (Zhu & Kraemer, 2005, p. 70). Although not statistically exam-
ined, the above-mentioned notion and interpretation of study results
significant positive effects of ICT usage, front-end functionalities and ICT
integration into back-office databases on firm performance imply the
mediating roles of ICT integration on firm performance.
Kim, Cavusgil, and Calantone (2006) revealed a mediation role of inter-
firm systems integration to explain supply chain communication systems
(SCCS) innovation on firm supply chain performance. This current
study focuses on studying the effects of ICT innovation on firm perfor-
mance; thus, it is logically anticipated that ICT integration may have a
mediation role to explain the effects of ICT usage on organisational
performance.
The degree of utilisation, that is, whether ICT is used properly and effec-
tively is another aspect of ICT use which may also correlate with firm
performance. Despite its practical importance for developing countries, this
aspect of usage, that is, the appropriate usage of ICT, has been greatly
under-studied. Anandarajan et al. (2002) stated that many IS in less
developed countries are under-utilised and, thus, do not make a significant
contribution to improving the performance of the organisations that have
implemented them (Forster & Cornford, 1992; Ordedra et al., 1993). A
conceptual linkage between ICT usage, degree of utilisation and organisa-
tional performance may be forecast from those statements which imply a
mediating role for the degree of utilisation. There is a paucity of research
studies that have looked at the inter-relationship of ICT usage, integra-
tion, degree of utilisation and organisational performance. This study has
anticipated that ICT usage would impact on organisational performance
with this impact mediated though integration and the degree of utilisation.
62 MD SHAH AZAM
2.7. Preliminary Research Model
Based on the above literature review, a preliminary research model (Fig. 1)

has been constructed to attain the research objectives. The model posits that
the diffusion process of ICT starts with SMEs’ intention to adopt the technol-
ogy. The BI is largely influenced by the characteristics of the owner as Rogers
(1983) stated that adoption of an innovation is a decision process in which an
innovation passes through different states of adoption and is finally implemen-
ted or has its use confirmed. He further stated that the adoption starts from
the knowledge stage where the adopter is first exposed to an innovation and
passes through the persuasion, decision and confirmation stages. A variety of
variables play different roles in various stages of an innovation adoption.
Rogers (1983) affirmed that the personal characteristics of an adopter such
as demographics, income, risk-taking behaviour, innovativeness largely
influence the knowledge stage which is very important for early adopters.
‘Cognitive evaluation’ towards an innovation is an important factor
which influences BI. The TRA (Fishbein & Ajzen, 1975) and the TPB
(Ajzen, 1985) explain the influence of cognitive evaluation on BI.
Environmental pressures that have arisen from the institutional environ-
ment are believed to have a positive influence on BI. The TRA (Fishbein &
Ajzen, 1975) and TPB (Ajzen, 1985) explain these influences as SNs with
this also supported by a number of ensuing theoretical frameworks (such
as UTAUT, Venkatesh et al., 2003). These theories focus on the individual
adoption process thus the influence of family, friends and peer groups
become significant. The organisational adoption behaviour is also
Owner/CEO
Cognitive Innovativeness
Evaluation
Environmental Behavioural Integration

Pressure Intention
Behavioural Organisational
Culture ICT Usage
Expectation Performance
Country
Degree of
Readiness
Utilisation
Facilitating
Condition
Fig. 1. Primary Research Model Based on Literature Review.

influenced by similar variables which arise from the influence of close

associates of the firms, partners and peer organisational groups. These phe-
nomena have been addressed as environmental pressures (DiMaggio &
Powell, 1983; Kuan & Chau, 2001; Teo et al., 2003) and have a significant
effect on organisational adoption intention (Teo et al., 2003).
Venkatesh et al. (2008) explained BI as an important variable in the
innovation diffusion process. They distinguished between intention and
expectation stating that intention is a reflection of the adopter’s internal
schema of beliefs; thus, it cannot predict the situation which is not under
the adopter’s full volitional control. Furthermore, intention may also fail
to adequately explain the situation where a time gap exists between inten-
tion and actual behaviour. Behavioural expectation is anticipated as a
stronger predictor of actual behaviour (in comparison to intention) which
may incorporate the influences of external variables arising from the socio-
cultural environment. Venkatesh et al. (2008) also posited that behavioural
expectation is highly influenced by BI, culture, country readiness and facili-
tating factors (or conditions).
As illustrated in Fig. 1, the preliminary model shows that behavioural
expectation plays a significant role in predicting ICT use. The model incor-
porates two aspects of ICT usage behaviour. The depth of ICT usage, that
is, the frequency or intensity of use, is indicated as ICT usage while the
breadth of ICT use involves strategic uses of ICT. The breadth of ICT use
is further categorised into two strategic variables, namely, integration
which indicates the extent to which ICT is integrated into various func-
tional areas, and utilisation which indicates how appropriately ICT is used
in the organisation. Organisational performance is explained by the impact
of ICT usage, integration and the degree of utilisation. The primary model
also anticipates a mediating role for integration and utilisation to explain
the impact of ICT usage on organisational performance.
2.8. Summary
This section has presented the literature review which addressed various
aspects of the research. The conceptual grounding of the study has been
postulated through reviews of relevant theories such as the DOI theory,
TRA, TPB, TAM, UTAUT, institutional theory and the RBV. Relying on
the structure developed by analysing existing theoretical frameworks, this
section also reviewed relevant contemporary empirical studies. Finally, a
preliminary research model was constructed (which was later fine-tuned
and contextualised by a field study) to attain the research objectives.
64 MD SHAH AZAM
3. RESEARCH METHODOLOGY
3.1. Introduction
This section presents the research design, research methods, and analytical
tools and techniques employed to attain the objectives of the study. The
study employed a mixed-method research approach which involved an
exploratory qualitative field study followed by a comprehensive quantita-
tive survey. The following sections discuss the entire research methodology
employed in the study which includes the research paradigm, the qualitative
field study design process of transcription and qualitative interview
analysis and quantitative research design survey instrument design,
data collection, data screening, data validation and data analysis.
3.2. Research Paradigm
A paradigm is a set of assumptions or beliefs that provides a conceptual

framework which, in turn, guides us in looking at and interpreting the
world around us (Suppe, 1977; cited in Deshpande, 1983). Guba and
Lincoln (1994) considered a paradigm to be a set of basic beliefs (or meta-
physics) that deals with first principles through which the nature of the
world, the individual’s place in it, and the range of possible relationships to
the world and its component parts are defined. In reality, a paradigm com-
prises the principles or philosophies which guide a person, a professional or
a researcher in selecting challenging issues, developing models and theories,
establishing criteria for methodology, instrument design and data collec-
tion, and also providing principles, procedures and methods appropriate
for looking at similar phenomena (Filstead, 1979; cited in Deshpande,
1983). Based on the epistemological, ontological, procedural and methodo-
logical concerns, several paradigms or dichotomies have shaped research
approaches in the fields of behavioural studies, business, IS and social
sciences, such as positivism versus interpretivism, quantitative versus quali-
tative, induction versus deduction and exploratory versus confirmatory
(Fitzgerald & Howcroft, 1998).
In general, two broad scientific paradigms, namely, the positivist and inter-
pretivist research paradigms, have guided most research in business, beha-
vioural studies, social sciences and IS (Hudson & Ozanne, 1988; Marsden &
Littler, 1996). However, Myers and Klein (2011) emphasise critical research
as an emerging stream in IS research and support Orlikowski and Baroudi’s
(1991) classification of research paradigm. Orlikowski and Baroudi (1991)
suggest three research paradigms: positivist, interpretive and critical.11 The

recent IS literature is broadly guided by these philosophies.
The positivist research paradigm attempts to apply the methods and prin-
ciples of the natural science model to explain behaviour or phenomena
through causal relationships while the interpretivist approach attempts to
interpret the inter-subjective meanings where a phenomenon is explained
through multiple explanations or realities rather than by one causal relation-
ship or one theory (Creswell, 2003; Denzin & Lincoln, 1994; Neuman, 2006).
In critical research, a critical stance is taken towards taken-for-granted

assumptions about organisations and IS, and where the aim is to critique to
status quo ‘through the exposure of what are believed to be deep-seated,
structural contradictions within social systems’ (Myers & Klein, 2011, p. 19).
Although a mixed-method approach was undertaken, especially during
the field study based on which the initial research model was examined and
fine-tuned, this study is primarily guided by the positivist research para-
digm by administering extensive use of quantitative methods to objectively
measure the variables and determine the causal relationships among the
constructs under study.
3.3. Research Method
The positivist and interpretivist paradigms have been supported by numer-

ous researchers in their separate explorations of reality. As the positivist
paradigm depends on the methods and principles of natural science, it has
no flexibility when it comes to interpreting phenomena from multiple reali-
ties which is, in fact, what the interpretivist research paradigm employs. In
reality, the interpretation or analyses of various situations, events and phe-
nomena in business, social sciences or behavioural sciences are not always
straightforward when it comes to employing the principles of natural
science in a single reality. On the other hand, the interpretivist paradigm
has the flexibility to look at the event from multiple realities and can find
the source of the problem while it does have a limitation regarding the gen-
eralisability of the results. Although both paradigms have their strengths
and are successfully utilised in multidisciplinary fields of study, such as
marketing, IS, organisational theory, social sciences and behavioural
science, in fact, each method, either qualitative or quantitative, has its own
limitations (Greene, Caracelli, & Graham, 1989). A single-method study
could lead to inaccurate and inadequate results.
Thus, a combination of the two paradigms, the mixed-method approach,
has been proposed as a third paradigm view based on pragmatism
66 MD SHAH AZAM
(Tashakkori & Teddlie, 1998). A combination of qualitative and quantita-

tive approaches within different phases of the research process is employed
under this research paradigm. The mixed-method research approach is
developed based on the notion that the combination of qualitative and
quantitative methods would compensate for their mutual and overlapping
weaknesses (Greene et al., 1989) as well as providing cohesive and coherent
outcomes (Hohental, 2006).
Although purely qualitative researchers are not receptive to the idea of
implementing a combination of two research-guiding philosophies in an

integrated platform, a growing body of researchers are employing the
mixed-method research approach in the field of IS and marketing.
In order to receive more insights from a mixed-method research
approach, different types of research design are employed that may be clas-
sified as the triangulation design, the embedded design, the explanatory
design and the exploratory design (Creswell, 2003). The triangulation
design refers to the collection and comparison of the data from both quali-
tative and quantitative methods with the view to validate or expand the
quantitative estimates using the qualitative data. The embedded design sug-
gests the collection of both qualitative and quantitative data while either of
the data types plays a supplementary role within the overall design. The
explanatory design leads to the collection and analysis of quantitative data
followed by the subsequent collection and analysis of qualitative data.
Unlike the explanatory design, the exploratory design starts with qualita-
tive data to explore a phenomenon and then estimates quantitative data.
In determining the appropriate mixed method for the current research, it
was essential to again reflect upon the objectives. As discussed in Section 1
(see Section 1.3), the main aim of this research was to explore the internal
and external factors that affect SMEs’ intention and actual usage of ICT
and its resultant outcome in terms of organisational performance. The
study, in its design, required the synpaper of various theories of innovation
diffusion and the creation of a link with the theories and conceptual frame-
works from different fields, such as, resource-based theory, institutional
theory and cultural study (Hofstede, 1984; House et al., 2004).
Based on previous theoretical frameworks and empirical studies, the
initial model (see Fig. 1 in Section 2) was proposed. The model needed to
be tested in terms of its applicability and validity to provide sufficient com-
prehensiveness to explain such behaviour. Thus, a field study through semi-
structured interviews was employed. Finally, a survey was carried out to test
the comprehensive model to ensure its generalisability and improve its expla-
natory power (the details of the process are in the next section). Based on the
description of the process and the research objectives, a triangulation design
was employed. This design allowed the data from both qualitative and
quantitative methods to be compared and merged during the analysis to
increase the reliability and validity of the research. Furthermore, triangula-
tion has been acknowledged as the most common mixed-method approach
(Creswell, 2003).
3.4. Research Process

This study investigated the factors that affect SMEs’ intention to use and
their actual usage of ICT and also examined how the different levels of
ICT application can influence organisational performance. This study
employed a mixed-method research approach which involved qualitative
and quantitative methodologies. Moreover, the study executed different
tasks in different phases to conclude the research: these are discussed in the
following sections.
3.4.1. Step 1: Literature Review

The first phase of the study employed an extensive review of the literature
related to the theories of innovation diffusion such as the TRA, TPB, DOI
theory and the TAM. Since the study was related to SMEs, the theories
related to organisational technology diffusion were also reviewed. Thus,
the TOE framework and institutional theory have been reviewed. The RBV
was reviewed within the proposed framework when looking at the contri-
bution of organisational ICT usage to organisational performance. The
study also examined the effects of culture and environmental pressures on
the diffusion of ICT by SMEs in Bangladesh. Thus, contemporary cultural
studies such as Hofstede (1984) and House et al. (2004) (GLOBE
study) were reviewed. Most recent related studies, namely, Venkatesh
et al. (2003, 2008), have been duly inspected to develop a conceptual frame-
work for the study. The cross-evaluation of relevant theories has resulted
in the discovery of some strong points in each theory which could contri-
bute to conceptualising the key constructs of BI, actual usage of ICT and
organisational performance. In addition to the theories and conceptual
studies, a good number of contemporary empirical studies were also
reviewed to augment the theoretical model in terms of its scope and
meaning and to establish the inter-relationship between the constructs.
3.4.2. Step 2: Preliminary Research Model

The systematic and rigorous literature review, both of the theories and
empirical studies, resulted in the development of a comprehensive model
68 MD SHAH AZAM
for examining the effects of different variables and their interrelationships.

A theoretically grounded, comprehensive initial research model was devel-
oped which gathered together the potential constructs within the domain of
the study (see Fig. 1 in Section 2).
3.4.3. Step 3: Qualitative Field Study

The initial model developed through the literature review could be
suitable for adequately explaining ICT diffusion by SMEs in Bangladesh,
measuring organisational performance and exploring the appropriate ante-

cedents of performance. The model was then evaluated for its consistency
and suitability within the context and for its adequacy in explaining issues
according to the study’s objectives. Next, a field study was conducted by
interviewing 11 owners or decision-makers of SMEs who were selected
through a convenience sampling technique.
The objectives of the interview were to: (1) seek out and identify con-
cepts and procedures that might not be reported or recognised in the litera-
ture review, and (2) evaluate the worthiness of the concepts identified in the
literature review. Each of the interviews was recorded and later transcribed
by the researcher with the view of transforming it into text. Each of the
transcripts was analysed by NVivo 10.
3.4.4. Step 4: Model Refinement

The preliminary model was refined based on the exhaustive related litera-
ture review as well as the findings from the field study. Where necessary,
the model was augmented with new items and/or constructs, or simplified
at that stage through elimination of duplicate constructs and items. The
primary model was refined, fine-tuned and finalised through this process.
3.4.5. Step 5: Hypotheses Construction

Basing on the refined model, a number of hypotheses were constructed at
this stage with these hypotheses also supported by prior theories and rele-
vant research. The resource-based theory (RBT) and institutional theory
were used together with innovation diffusion theories to guide the develop-
ment of the hypotheses.
3.4.6. Step 6: Questionnaire Design

Based on the hypotheses developed in the previous stage, a tentative ques-
tionnaire was designed. Each of the questions was used as a source of data
and was adapted from previous relevant research with slight modifications
to adjust to the context and to ensure clarity in expression for the
interviewees. Some new measurements were also developed through the

qualitative field study with these later supported by relevant literature. A
total of 84 items and 16 constructs were included in the questionnaire to
investigate the research problems. The questionnaire was validated through
pre-testing and a pilot survey before conducting the countrywide main
survey in Bangladesh.
3.4.7. Step 7: Pre-testing of the Questionnaire

As mentioned in the previous stage, the questionnaire was pre-tested to

ensure that it was ready for wide circulation for the final data collection.
ICT experts, owners of SMEs or their representatives, and some academics
were invited to pre-test the questionnaire. The objective of pre-testing was
to contextualise the instruments and also to ensure reliability of the ques-
tions. During the questionnaire pre-testing phase, different professionals
and experts scrutinised the questionnaire from different perspectives and
provided their feedback which enhanced the content validity of the
questionnaire.
3.4.8. Step 8: Questionnaire Refinement

The questionnaire was modified in accordance with feedback received from
the expert respondents during pre-testing. The tentative instruments were
refined through a careful revision of the questionnaire. The changes, adjust-
ments and modifications adopted during this stage resulted in the final
questionnaire for the study.
3.4.9. Step 9: Data Collection

The main data collection process entailed distributing the final structured
questionnaires to the owners or decision-makers of different SMEs in
Bangladesh seeking their responses. In Bangladesh, SMEs are divided into
two classes, namely, the manufacturing industry and the services industry.
In considering the nature of businesses and their potential to contribute to
the national economy, the manufacturing industry in Bangladesh and,
particularly, the RMG industry was considered as the subject of this
study due to its high potential in terms of its contribution to the GDP,
employment generation and export earnings. A total of 300 SMEs were
targeted for data collection: they were selected through a stratified
random sampling from a list of companies developed by using the data-
bases of Bangladesh Garment Manufacturers and Exporters Association
(BGMEA), Leather and Footwear Manufacturers and Exporters Associa-
tion of Bangladesh (LLMEB), databases of the Bangladesh Export Promotion
70 MD SHAH AZAM
Bureau, SMEs Foundation of Bangladesh and the Yellow Pages. From a

comprehensive list of SMEs, a sample was selected using the principle of
stratified random sampling. The target subjects were either the owner of
the firm or their representative.
A total of 282 responses were gathered proportionately from the RMG
(225) and leather industries (57) which was considered adequate for
PLS-based data analysis. Section 3.6.5 describes this step in detail.
3.4.10. Step 10: Data Analysis

A PLS-SEM (Barclay, Higgins, & Thompson, 1995; Chin, 2010) technique
was employed to analyse the data collected through the comprehensive sur-
vey. A two-step procedure, involving assessment of the measurement model
by examining its psychometric properties and estimation of the structural
model, was followed for the quantitative data analysis.
3.4.11. Step 11: Result Interpretation and Report Writing

After successful completion of the entire data analysis (both qualitative and
quantitative), the results were discussed and interpreted in accordance with
the previously developed research questions and objectives. The write-up
of the paper, however, was a continuous process which, in fact, began dur-
ing the literature search: the final write-up started with the reporting of the
results and their interpretations. The outcomes of different phases of
the research presented in various sections were articulated and integrated
during this stage. The process ended with a complete paper which includes
the research questions, study objectives, literature review, research metho-
dology, qualitative field study analysis and final model specification,
hypotheses development, data analysis, results interpretation, conclusion
and references.
3.5. Qualitative Field Study Method
This study has adopted a mixed-method research approach to attain its

objectives. In accordance with the exploratory research design, this study
has employed a qualitative field study followed by a quantitative survey.
The study was dedicated to looking at the diffusion of ICT by SMEs in
Bangladesh and the resultant outcome in terms of the organisational per-
formance achieved through integration and proper utilisation of ICT. The
study undertook an extensive literature review to develop the conceptual
framework and primary model. Most of the literature reviewed for this
study focused on developed countries’ perspectives, and mainly the

American perspective. As this was an initiative to look at the phenomena
from a developing country perspective and, more clearly, from the
Bangladesh perspective, the research instruments and the constructs may
not have been well suited to this perspective. Thus, a qualitative field study
was appropriate in order to contextualise the model and any modification
or extension. To validate and enhance the factors and variables identified
through the comprehensive literature review, the study, during this phase,
explored the phenomenon of ICT usage by SMEs in Bangladesh and its

effects on organisational performance by interviews with the owners of
SMEs and other stakeholders. Researchers have argued that the goal of
understanding a phenomenon from the point of view of participants and its
particular social and institutional context was difficult to achieve when
textual data were quantified (Kaplan & Maxwell, 1994). Therefore, it was
believed that a ‘pseudo case study’ that involved a qualitative study of a
small number of participants would meet the objectives of this phase of the
study.
As such, a field study approach (Patton, 1999; Zikmund, 2000) was
adopted as the research method for the qualitative phase. Qualitative meth-
ods permit the evaluator to study selected issues in depth and detail.
Approaching fieldwork without being constrained by predetermined knowl-
edge contributes to the depth, openness and detail of qualitative inquiry
(Patton, 1999). The field study enabled the researcher to be involved in
investigating the factors influencing diffusion of ICT and its effects on
organisational performance in the participating SMEs.
3.5.1. Sample Selection

The study adopted the convenience sampling technique and selected
11 SMEs for qualitative interviews. This qualitative study began with the
notion of an open-ended number of cases and stopped interviewing after
the 11th case after taking into consideration informational redundancy and
theoretical saturation (Denzin & Lincoln, 2005; Lincoln & Guba, 1985;
Strauss & Corbin, 1998). The sample comprised 11 owners of different
SMEs or their representatives who were selected by adopting a convenience
sampling method. This is one of the most commonly used sampling meth-
ods in business research (Zikmund, 2000) as it relies on available subjects
who are close at hand or easily accessible (Berg, 2004). The convenience or
judgment sampling procedure is also effective for qualitative studies: appro-
priate and resourceful people who can easily be identified and included in
the sample may provide a valuable contribution to the research.
72 MD SHAH AZAM
As diffusion of ICT among SMEs and organisational performance were

the focus, the study considered organisations that were using any kind of
ICT in various functional areas such as in shop floor management, account
management, inventory control, for qualitative interviews. The convenience
sampling technique also allowed the researcher to include SMEs who were
functioning at different user levels such as basic computing, e-commerce
users and users of enterprise resource systems.
A wide range of SMEs operate in different sectors in Bangladesh. These
sectors are broadly classified as the services industry and the manufacturing
industry. SMEs play a substantial role in the economic development of a
country. Despite the fact that the contribution of SMEs in economic
development is similar, the definitions of SMEs are diverse in different
country contexts. The Industrial Policy 2010 of the People’s Republic of
Bangladesh defines an SME as:
• A manufacturing firm having 100250 employees or having fixed assets
from 10 Crore BDT to 30 Crore BDT (excluding the value of land and
factory) is considered a medium enterprise while any firm having 2599
employees or having fixed assets from 55 Lakh BDT to 10 Crore BDT
(excluding the value of land and factory) is considered a small enterprise.
• Also, a firm in the services industry having 50100 employees or having
fixed assets from 1 Crore BDT to 15 Crore BDT (excluding the value of
land and factory) is considered a medium enterprise while a firm having
1049 employees or having fixed assets from 5 Lakh BDT to 1 Crore
BDT (excluding the value of land and factory) is considered a small
enterprise.
Feasible organisations for the qualitative research were selected from the
Yellow Pages directory. Personal connections and the nature of the organi-
sation as well as the owner’s availability and interest in participating in the
interview were the criteria for sample selection.
3.5.2. Data Collection

The data collection and analysis method used in qualitative studies is differ-
ent from that used in quantitative techniques (Lincoln & Guba, 1985).
Interviews have been accepted as one of the major data collection techni-
ques for qualitative studies (Maykut & Morehouse, 1994) and are therefore
employed as a method of qualitative data collection in various fields of
study. The main three different categories of interviews employed in data
collection are: standardised (formal or structured); unstandardised (infor-
mal or non-directive) and semi-standardised (guided semi-structured or
focused) interviews (Babbie, 2001; Merriam, 1998; Nieswiadomy, 2011).

The overall objective of the study, the nature of the problem and the
research perspective help in selecting the type of interview appropriate for
the study. It was mentioned earlier that the purpose of this study was to
explore and examine the antecedents of ICT adoption and usage by SMEs
and their effects on organisational performance. Utilising the outcomes of
previous research and theories, this study developed a primary model in
order to inquire about the phenomena. Although much research has been
conducted in this field in the past few years, these studies have mainly
focused on developed country perspectives and, particularly, on the
American perspective (Watson, Kelly, Galliers, & Brancheau, 1997; Zhu
et al., 2004). The theories or the empirical evidence, although relevant,
should be contextualised before their execution in addressing the phenom-
ena from a developing country perspective.
This study, thus, has adopted a semi-structured interview for collecting
relevant qualitative data to explore various variables and factors in order
to refine the model to address SMEs’ ICT diffusion and organisational per-
formance from the Bangladesh perspective. By design, the semi-structured
interview ensures that the list of themes, issues to be addressed and
questions to be asked are identified and pre-defined by the researcher.
Even though its nature is to have a significant understanding about the
context and nature of the variables associated with the study and their
interrelationships, the semi-structured interview ensures that information is
captured from respondents’ perspectives rather than being imposed by the
researcher (Burns & Bush, 2000).
The semi-structured questions and probing guide were developed
through the comprehensive literature review. The primary guiding semi-
structured questions were also pre-tested by interviewing some SME owners
so that any adjustments or modifications could be made thus ensuring that
the guide was useful for the interviews and also confirming the data from
respondents’ perspectives. The semi-structured interview questions focused
on the following areas of information that were needed from the field study:
• General perceptions, beliefs towards ICT and intended ICT use

• The main factors that influenced SMEs’ intentions to use ICT
• The usage of different applications of ICT
• Conceptualisation of the national culture and its role in the diffusion of
ICT by SMEs
• Conceptualisation and the role of environmental pressures in shaping
SMEs’ intentions to use ICT
74 MD SHAH AZAM
• Conceptualisation and the role of country readiness in intention to use

and usage of ICT by SMEs
• Conceptualisation and the factors associated with organisational
performance
• The role of ICT use in organisational performance.
Before the interviews, the selected interviewees were personally con-
tacted and an appointment was made for the interview. An outline of the
interview along with a cover letter specifying several important perspectives
of the research such as the objective, scope and perspective of the study,
process of the interview and the role of the interviewee were also sent
beforehand so interviewees would be comfortable with the formalities and
able to express their opinions freely. The interviews were recorded in their
entirety by a micro-audio recorder with permission duly given by intervie-
wees. Applying the rules of transcription, the recorded interviews were
transformed into text immediately after the interview, generally, on the
following day.
3.5.3. Analyses of Qualitative Data

This study employed the content analysis technique to analyse the qualita-
tive data (Siltaoja, 2006). Qualitative data analysis involves both inductive
and deductive approaches to compare the qualitative data with the theore-
tical framework and also to develop a causal relationship between the
constructs under study (Berg, 2001).
The themes, sub-themes and concepts explaining the variables and
factors were explored by the inductive process. This process also explored
the measurement scales of some variables or factors. The variables and
factors explored by the induction process were further induced into a
common framework. The ultimate objective of the field study was to pre-
pare a comprehensive conceptual framework by comparing the induced
framework with the initial research model developed from the literature
review. This framework was therefore compared with the initial research
framework.
In this study, NVivo (Version) 10 software was used to analyse the qua-
litative interviews. NVivo (registered under QSR International) is designed
to facilitate common qualitative techniques for organising, analysing and
sharing data. It can analyse unstructured or semi-structured data like inter-
views, surveys, field notes, web pages and journal articles from a range of
sectors which includes social sciences, education, health care and business.
The development of NVivo has provided greater convenience through
using the programme and has also enhanced the capacity of the researcher
to undertake interpretive analysis (Bazeley, 2007).
With the intention being to analyse the qualitative data by NVivo, the
recorded interviews were transformed into interview transcripts. The
NVivo 10 software first imported all interview transcripts and provided
the opportunity for analysis by nodes, modelling and query support.
Through these analyses, the qualitative study’s findings assisted in devel-
oping a comprehensive model with this then used as a basis for the
construction of hypotheses. The outcome of the field interviews and the

extensive literature review provided the rationale for the construction of
hypotheses. The qualitative study finally refined the study model and devel-
oped hypotheses that formed the basis for the quantitative study, the main
concern of this research.
3.6. Quantitative Study Method
Although the theoretical framework of the study was developed through an

extensive literature review which was contextualised and fine-tuned through
an in-depth field study, the design and execution of quantitative research
was a vital part of this study as the study stood under the positivist
research paradigm. The quantitative survey explored the degree and magni-
tude of Hypothesised relationships between the variables and constructs
under study. Based on the theoretical framework discussed in the preceding
section, a number of hypotheses were developed for possible rejection or
acceptance which would help to infer the situation and policy implications
(Anderson, 1973). As the study was seeking to empirically test some
assumptions or hypotheses and the methods used in this phase were
designed to be detached and independent of the specific situation under
study, a quantitative method was considered most appropriate. The size
and geographical dispersion of the sample units also provided the rationale
for adopting a survey method and, in particular, a questionnaire-based
survey.
3.6.1. Developing the Questionnaire

A questionnaire was the instrument used for data collection. Often, a ques-
tionnaire is called an interview form or schedule. Malhotra (2004) described
a questionnaire as a formalised set of questions for obtaining information
from respondents. A questionnaire is a very important research instrument
for the researcher engaged in research based on quantitative primary data.
76 MD SHAH AZAM
A researcher must prepare a questionnaire to collect primary data. The

notion of translating the information needed into a question format is the
principle behind the development of a questionnaire. Researchers should
pay heed to carefully developing the questionnaire so that it is effective for
collecting the needed information. The style, language and symbols used in
the questions should match prospective respondents’ status, culture and
ability to make them (respondents) comfortable and motivated to respond.
A well-developed questionnaire encourages respondents to provide accu-
rate information which not only ensures accurate information but also
minimises response errors.
The questionnaire was designed to collect SMEs’ owners’ or decision-
makers’ perceptions or evaluations of SMEs’ intention to use or usage of
ICT and how they were influenced by different factors and variables. The
questionnaire also endeavoured to collect respondents’ subjective assess-
ments or evaluations about SMEs’ performance and how performance was
influenced by ICT utilisation and integration. The study adopted subjective
or indirect measurements rather than direct or objective measures. As with
many research debates, researchers have differences of opinion between
applying subjective or objective measures. Some researchers believe that
the responses to subjective measures are likely to be influenced by selection
bias. In addition, it is also true that, in many cases, collection of quantita-
tive data by observed measures is very hard and sometimes erroneous. For
instance, small firms in developing countries are not formalised and they
do not keep their business records formally. Sometimes, they hide some
information due to reductions in tax and worry with regard to their compe-
titors. In this situation, data from using objective measures would not be
suitable or effective. Thus, although there was the possibility of selection
bias, the use of subjective measures in this research was appropriate. The
literature supported the general reliability of self-reported and subjective
measures (Dess & Robinson, 1984). Nevertheless, care was taken to select
measurement items that previous research had shown to be valid and
reliable.
In general, the scaling technique is widely used to assess the variation in
data collected by subjective measures. The scaling technique yields the
highest level of information feasible in a given situation which permits
the use of a great variety of statistical analyses. This study mainly used the
Likert rating scale with the help of the semantic differential scale. The
Likert rating scale has been widely used in psychology, sociology and busi-
ness research as it is easy to construct and administer. A Likert scale pro-
vides the respondents with a complete picture of a phenomenon so they can
easily indicate the degree of their agreement or disagreement with a variety

of statements related to the phenomenon (Aaker, Kumar, & Day, 2004).
Respondents readily understand how to use the scale, making it
suitable for mail or telephone surveys or personal interviews. The main dis-
advantage of the Likert scale is that it takes longer to complete than other
itemised rating scales because respondents have to read each statement.
The points on the rating scale could vary between two to any higher num-
ber, but most researchers prefer rating scales of between five and nine
points (Cox, 1980; Reynolds & Neter, 1982). Although the literature shows
that the correlation coefficient decreases with a reduction in the number of
scale categories, affecting all statistical analyses based on the correlation
coefficient (Givon & Shapira, 1984), the convenience and ease of execution
of the questionnaire and its ability to encourage respondents to accurately
assess their feelings during the survey should be considered equally. By
including rating scales with too many points such as a 9-point rating scale,
respondents may be confused in assessing their feelings between so many
points. It may also create a situation in which the respondent evaluates his/
her position deeply again and again when indicating his/her feelings. This
may lead to a longer response time while, at the same time, spontaneous
perceptions or responses may not be provided which would in the end
cause a biased response. Thus, the rating scale should include a fair number
of points that, on one hand, can produce scope for variations in perception
while, on the other hand, being easily perceivable and executable by the
respondent. In showing the researcher’s understanding of this situation,
this study has adopted a 5-point rating scale.
3.6.2. Questionnaire Translation

The study has endeavoured to investigate the diffusion of ICT and the
effect of its usage on SMEs’ performance by collecting primary data mainly
from the owners of SMEs in Bangladesh. The native language of the
study area is Bengali and a significant number of SME owners, particularly
the owners of small industries in Bangladesh, are Bengali-dependent.
Questionnaires in English were not suitable for data collection in this study
context. The English version of the approved questionnaire was thus trans-
lated into Bengali before conducting the survey to achieve greater conveni-
ence and ease of operation. As the respondents easily understood the
Bengali version of the questionnaire and could easily respond with the
most appropriate answer, the questionnaire became very useful for data
collection. Although the Bengali version of the questionnaire was easy to
use and very useful for data collection, the translation of the questionnaire
78 MD SHAH AZAM
and creating equivalence between the Bengali version and the English ver-
sion was not an easy task.
Four different procedures namely, one-way translation (direct transla-
tion), back-translation (double translation), translation by committee and
decentring are used for translation of instruments. Among the above
four alternatives, direct translation is the most frequently used translation
procedure where a bilingual translator translates the questionnaire directly
from a base language to the respondent’s language. This procedure is con-
sidered as less expensive and less time-consuming than other translation

procedures. It is important that the translator is fluent in both the lan-
guages and familiar with both the cultures otherwise direct translation of
certain words and phrases may not project similar meanings and sentiments
which may cause erroneous responses. However, back-translation is a
two-way translation procedure which can overcome the errors of one-way
(direct) translation procedure.
In back-translation, a researcher prepares material in one language and
asks a bilingual translator to translate it into another (target) language. A
second bilingual translator independently translates the material back into
the original language. The researcher then has two original language forms
to examine and even if he/she does not know the target language can
make a sound judgment about the quality of the translation. However,
several repeat translations and back-translations may be necessary to
develop equivalent questionnaires, and this process can be expensive and
time-consuming.
In the committee approach, a group of bilingual translators translate
from the base language to the target language. The mistakes of one mem-
ber can be caught by others on the committee. The weakness of this
method is that committee members may not criticise one another and may
even unify against the researcher. Werner and Campbell (1970) proposed
decentring as a way to develop instruments that would be culturally appro-
priate when cross-cultural research is conducted. In the decentring process,
the original language instrument is not considered final until the entire
translation process is completed. Therefore, if a translator believes that a
grammatical structure or word or tense must be changed to appropriately
fit the cultural group under study, the original instrument should also be
changed to reflect these linguistic and cultural characteristics. There is a
constant comparison of the two instruments, and modifications are made
to the first to account for limitations of the target language. Although the
instrument may be more accurate culturally and linguistically, this may
also be more costly in terms of time and resources, and the length of the
instrument may be increased.
The literature describes back-translation as one of the most adequate
translation procedures (Marin & Marin, 1991). The researcher has there-
fore adopted the back-translation approach for translating the question-
naire and ensuring equivalence of the two versions. To explain in more
detail, the researcher first translated the original English questionnaire,
which had been approved by the university’s Human Research Ethics
Committee, into Bengali. The first version of the Bengali questionnaire was
thoroughly checked by a university academic in Australia. The refined ver-
sion of the translated questionnaire was further reviewed by two university
academics from Bangladesh who were involved in research in Australia.
The final version of the Bengali questionnaire was again translated back
into English. A third person, a researcher from Bangladesh, performed the
back-translation of the questionnaire into English. The back-translated ver-
sion of the questionnaire provided the opportunity to check whether the
translated version of the items was projecting a similar meaning and
approach as in the original version. Interestingly, although some words
were found to be different, all items in both versions of the questionnaire
were observed as being similar in meaning which finally ensured equiva-
lence of the two versions of the questionnaire.
3.6.3. Pre-Testing of the Questionnaire and Pilot Test

The initial questionnaire was refined through pre-testing prior to adminis-
tering the actual survey. Five people in Australia who had direct experience
in running an SME in Bangladesh were invited to pre-test the question-
naire. The objectives of pre-testing were to contextualise the instrument
and to ensure reliability of the questions. During the questionnaire pre-
testing phase, different respondents scrutinised the questionnaire from
different perspectives and gave their feedback which finally enhanced the
content validity of the questionnaire. In addition to respondents’ feedback
on the content of the questionnaire, the overall length of time taken and
difficulties that arose during completion of the questionnaire were also
inspected. A follow-up interview was conducted with each participant to
identify any weaknesses in the instrument. Basing on the experience of the
pre-testing and respondents’ feedback on several points, the questionnaire
was finalised after initiating the necessary corrections and modifications.
A pilot test was also undertaken administering a mini sample survey
using the final version of the questionnaire. A total of 35 SMEs were
80 MD SHAH AZAM
selected through a convenience sampling technique and surveyed. The ques-

tionnaire was further refined through the pilot test. During the pilot test,
numerous errors in the questionnaire were detected and a number of
suggestions were received from respondents. As the questions had been
generated to assess respondents’ perceptions or evaluation of different
statements by rating them on a Likert scale and, in some cases, on a seman-
tic differential scale, respondents perceived that many of the questions were
very close to each other in their meaning. In addition, many respondents
faced trouble in differentiating their exact evaluation using the given five
points. More specifically, respondents faced difficulties judging the intensity
between ‘4’ and ‘5’ as well as between ‘1’ and ‘2’ on the 5-point rating scale.
In addition to respondents’ comments, these errors were also detected in
the analysis of the data from the pilot study. As the respondents could not
differentiate between their feelings in their evaluations, they mostly chose
the extreme value. Thus the factor loading of the different items of some
constructs tended to reach value ‘1’. To overcome this error, a statement
about each point of the scale was given in the questionnaire which helped
respondents to select the right point. The questionnaire was again revised
to incorporate respondents’ suggestions (which the researcher felt were logi-
cal) to ensure a valid and reliable questionnaire. The final version of the
questionnaire was used to collect data from a national survey.
3.6.4. Sample Selection

Sample selection is one of the major tasks in a research project and, parti-
cularly, in survey research. Although this study was designed to draw the
conceptual framework through the literature survey and to contextualise
and fine-tune the framework through the findings from a qualitative field
study, the main outcome was produced by analysing quantitative data (col-
lected through a sample survey) to reject the null hypotheses and draw
some policy implications. The results drawn from a survey are said to be
representative if the sample contains characteristics that are similar to the
population. The literature has suggested that the application of probability
principles in sample selection may help to produce a representative subset
of the population. The results of data collected from a representative subset
of the population strengthen the confidence in anticipating any situation.
Researchers select probability sampling by applying different strategies
which include simple random sampling, proportionate stratified random
sampling, disproportionate stratified random sampling and cluster sam-
pling. Stratified sampling involves the extraction of proportionate represen-
tation of multiple groups of firms considered in the research (Reynolds,
Simintiras, & Diamantopoulos, 2003) where units within a group are simi-
lar and between groups that bear dissimilar characteristics. Each stratum
includes the same characteristic of interest that facilitated the systematic
random sampling procedure (Churchill, 1991; Malhotra, 2002). Contrary
to random sampling, numerous researchers have adopted the convenience
or judgment sampling technique. However, this study has adopted the dis-
proportionate stratified random sampling technique for sample selection.
In addition to the sampling procedure, the size of the sample is also
another important factor to ensure the representativeness of the sample as

well as its suitability for executing the appropriate statistical tools. For
example, in a qualitative field study, 48 cases are sufficient for analysis.
However, the application of basic statistical tools in quantitative data ana-
lysis requires normal distribution. The rule of thumb suggests that more
than 30 cases can justify the normality of data and may be used in any
basic statistical analysis. However, requirements of sample size may vary in
different types of statistical analysis and a variety of opinions was also
observed in the literature even when applying the same tools (Hair,
Anderson, Tatham, & Black, 1998; Tabachnick & Fidell, 1996). For exam-
ple, the standard and sophisticated statistical analysis including structural
equation modelling (SEM) recommends sampling of 200 as fair and 300 as
good (Tabachnick & Fidell, 1996). Similar to that, Hair et al. (1998) recom-
mended a sample size of 200 to test a model using SEM, because 200 is a
‘critical sample size’ (Hoelter, 1983) that can be used in any common
estimation procedure for valid results.
Aaker et al. (2004) suggested an alternative approach which recom-
mends using a comparable and similar extant study’s sample size with a
satisfactory level of reliability as a guide. A number of closely related inno-
vation diffusion studies (Bayo-Morionesa & Lera-López, 2007; Twati &
Gammack, 2006; Venkatesh et al., 2008) used a sample size ranging from
200 to 300. Based on the above-mentioned examples and notions from
previous studies, this study has adopted a sample size close to 300 as SEM
was the main tool being used for data analysis.
The SMEs in Bangladesh formed the population for this study while the
owner of the firm or their designated representative engaged in the firm’s
decision-making were the study’s subjects. There are approximately seven
million micro-industries and SMEs in Bangladesh. Micro-industries were
outside the scope of the study’s objectives and therefore were excluded.
According to Bangladesh’s Industrial Policy 2010, the country’s SMEs are
categorised into two different industries the manufacturing industry
and the services industry. This study executed a multistage process to
82 MD SHAH AZAM
achieve a representative sample selection which started by categorising all

industries in Bangladesh into the formal and informal sectors. Formal firms
were considered for sample selection while informal firms were excluded. In
a consecutive step, formal firms were categorised into large-scale enterprises
and SMEs. In accordance with the objective of the study, large-scale firms
were excluded and SMEs were considered for further treatment. Basing on
their environment, SMEs were divided into two categories rural firms
and urban firms. Rural business firms were excluded from the sample frame
as they lacked the infrastructure for ICT-dependent procedures in their

rural environment while urban firms were given full attention. Urban firms
were again classified into four distinct categories based on their location,
namely, firms located in Khulna, Rajshahi, Chittagong and Dhaka.
Khulna, Rajshahi, Chittagong and Dhaka are four major cities in
Bangladesh. Dhaka is the capital of the People’s Republic of Bangladesh
and 80% of Internet services are Dhaka-centric. One important locational
benefit held by Dhaka is that almost every firm has its head office located
there. Based on the number of organisations in their industry and their
concentration and IT penetration, the firms located in Dhaka were consid-
ered for the study. Those firms were further divided into two groups based
on the nature of the industry, whether it was the manufacturing industry or
the services industry. Based on the export contribution and economic sig-
nificance, the manufacturing industry was logically considered for sampling
and the services industry was excluded. The manufacturing industry com-
prises various industrial units. Although each of these industrial units con-
tributes positively to the national economy, their performance varies.
Based on their contribution and its potential in the economic development
of the country, some industries have been announced as the thrust sector
for the economy. Owing to their contribution and economic potential, the
RMG industry and leather and leather product industry have attracted the
attention of the Government of Bangladesh, are considered as the thrust
sector and receive preferential treatment and trade benefits.
In the manufacturing industry sector, the RMG industry is the leading
industry contributing around 7080% of export earnings and is therefore
considered to be the goose that laid the golden egg in Bangladesh while the
leather and leather product industry is an emerging industry. The study has
logically considered both the leading industry and an emerging industry to
build a sample frame. The proportion of the contribution of the RMG and
leather industries in Bangladesh is roughly 80:20. The sample frame, thus,
was generated by taking sample units from these two industries according
to the proportion of their contribution. Understanding the nature of the
study and the statistical tools to be used for analysis, this study decided to
select a sample of 300 cases by incorporating 240 (80%) from the RMG
industry and 60 (20%) from the leather industry. The sample units were
selected on a random basis from a list of related industries. In accordance
with the rough contribution of the two industries being 80:20, the sampling
procedure was justified as proportionate stratified random sampling.
Interestingly, if the number of firms engaged in the sectors had been consid-
ered, the selection of the sample may be fallen into disproportionate strati-
fied random sampling as a very large number of firms are engaged in the
RMG industry while the number of firms engaged in the leather industry is
small.
Blalock (1960) suggested that the disproportionate stratified random
sampling method is better than the proportionate stratified method in stu-
dies in which the differences in numbers among groups are large. Thus, the
process and method adopted for selecting firms from the RMG and leather
industries by the 80:20 ratio were still justified as disproportionate stratified
random sampling. Churchill (1991) supported disproportionate stratified
sampling as being a method by which to increase precision without increas-
ing cost and to reduce the sample variation. This technique offers the
opportunity to reduce sampling error to achieve an increased level of
confidence by the representation of two different categories of pertinent
sample characteristics that accurately reflect the population (Davis, 2000;
Zikmund, 2000). A flow chart of the sample selection process has been pro-
vided in Appendix A.
3.6.5. Quantitative Data Collection

A list of companies was prepared from the BGMEA members’ directory
20102011 which contained 2,917 firms and the Bangladesh Knitwear
Manufacturers and Exporters Association (BKMEA) members’ directory
20102011 which contained 1,987 firms. A list of firms engaged in leather
and leather products manufacturing and exporting was developed from the
Leather and Footwear Manufacturers and Exporters Association of
Bangladesh (LLMEB) members’ list 20102011 which contained 58 firms,
a list from the Bangladesh exporters’ category on the Export India web
page which contained 28 firms, and names and addresses of another 45
firms were collected from the records of the Ministry of Trade and
Commerce. With the aim of collecting responses from 300 firms (or a num-
ber close to that figure), a total of 1,320 questionnaires are sent out to
respondents: 1,200 questionnaires to the RMG industry and 120 question-
naires to the leather industry.
84 MD SHAH AZAM
Before sending the questionnaires, respondents were contacted by tele-

phone. The reason for the telephone calls before sending the questionnaires
was to take the opportunity to judge the category of the firms selected as
primary sample units through screening questions and also to receive prior
consent from respondents. Upon receiving respondents’ consent, the survey
instrument together with a covering letter explaining the purpose and
instructions for the survey were sent directly to the person contacted which
was believed to be an efficient way to receive a satisfactory response.
The first round of survey packages was sent out in the third week of
September 2011. Surprisingly, no survey responses were received until
November 2011 even after follow-up calls. Respondents were contacted
through email and telephone calls. However, sending emails was deemed to
be not effective in this context. The telephone follow-up, although calls
reached the respondents, was not efficient for pursuing respondents to get
them to complete and return the questionnaire. After realising the situation
and having an extensive meeting with some experts including the supervisor
of this research project, three educated surveyors were employed to follow-
up and collect the completed questionnaires. Extensive training was pro-
vided to the surveyors to acquaint them with the research and survey. The
surveyors also received training about remaining neutral during the follow-
up process and the collection of responses. It was important to note that,
at the preliminary stage, the researcher had planned to use an online survey
procedure through Survey Monkey software in conjunction with a mail-out
survey. A version of the questionnaire was also posted to the web for data
collection. It was also surprising that the online data collection procedure
was not successful. However, the final approach worked well, that is, using
a combination of telephone follow-up and door-to-door physical visits
resulted in a number of good responses. The first round of the study was
conducted during the period from September 2011 to March 2012. A total
of 150 completed responses were received during this period.
The second round of the study was conducted during the period from
April 2012 to August 2012. During this phase, the researcher himself made
repeated telephone calls to respondents. The surveyors also were more effi-
cient in facilitating the procedure and were successful in boosting the num-
ber of responses. During the second phase, another package consisting of
reminder letters, a copy of the questionnaire and a souvenir was once again
sent to the contact persons. Thus, a more efficient result was achieved and
a total of 132 responses were received during this phase.
Finally, 225 usable responses from the RMG industry and 57 useable
responses from the leather industry were received. Finally, the data set
containing 282 records was used for the entire data analysis of the study
which included SEM. The final response rates were calculated as 22% from
the RMG industry and 55% from the leather industry. As the size of the
leather industry in Bangladesh is small, the researcher put in hard labour
and paid extensive attention to collecting a suitable number of responses
close to the targeted number of 60 which ensured a good response rate.
However, administering a successful survey in a country like Bangladesh is
a hard job as many of the popular modes for surveys are not suitable
for Bangladesh. Previous research experience has admitted to receiving a

lower rate of responses when surveying organisations in Bangladesh
with a response rate of 17.5% considered as satisfactory (Saleh, 2006;
Shamsuddoha, 2004). Thus, the response rate of 22% was valid for draw-
ing inferences representative to the population. The sample size was also
adequate for applying SEM.
3.6.6. Quantitative Data Analysis

This study employs SEM for quantitative data analysis. SEM, a second
generation of the statistical analytical tool, incorporates a two-step proce-
dure in data analyses which administers the assessment of the measurement
model followed by undertaking the structural model estimation. The
measurement model involves the assessment of the construct validity, con-
vergent validity and discriminant validity of the reflective constructs.
Convergent validity ensures that the items explaining a construct converge
well through examining whether the items in each construct are highly cor-
related and reliable; while discriminant validity ensures that the reflective
constructs are different from each other by estimating the average com-
munalities, construct correlation and cross-loading matrix.
SEM was appropriate for analysing the data in accordance with the
proposed conceptual framework. A growing number of researchers are
adopting causal or SEM as it allows the analysis of complex networks
of constructs, each construct typically measured by multiple variables.
Covariance structure analysis, as implemented by LISREL, EQS or
AMOS, is the well-known approach to causal modelling. However, PLS is
a complementary approach with features that are well suited to the domain
of technology research (Barclay et al., 1995).
Although a combination of software, including MS Excel (for data
management); SPSS (for data manipulation and descriptive statistics);
LIMDEP (for cross-correlation) and AMOS (for estimating some of the
SEM components for cross-checking) was used for the treatment of
the data, checking and rechecking of reliability and validity of the
86 MD SHAH AZAM
measurements, and facilitating data analyses in various stages of the

research, the main part of the quantitative data analysis was performed by
using PLS-Graph Version 3 software.
The decision by any researcher when selecting an appropriate analytical
tool for reliable estimates is vital. Past research initiatives have employed
both covariance-based structural equation modelling and correlation-based
structural equation modelling to examine similar phenomena while these
two different methods of SEM have some specialties and also some
constraints. Chin (1995) reported that covariance-based SEM is superior

on mathematical grounds while correlation-based SEM had superiority on
practical grounds. The covariance-based structural equation modelling
(CBSEM) software, such as AMOS, LISREL, considers reflective items
when analysing both measurement and structural models while component-
based structural equation modelling software, such as PLS-Graph, can
handle both reflective and formative measures in estimating the measure-
ment model and structural model.
The theoretical framework of the study involved a complex structure of
hierarchical latent constructs comprised of both reflective and formative
items. The nature of the study constructs and their measurement items thus
justified the application of PLS-SEM (Barclay et al., 1995).
With the arguments as stated above, the PLS technique was considered
to be the most appropriate data analysis tool for the quantitative study.
This study attempted to use the PLS technique to establish the relationship
between the different model constructs, thus testing the hypotheses. As
such, the data collected in this study were analysed using the PLS technique
utilising the PLS-Graph Version 3.0 computer software developed by Chin
(2002) (www.plsgraph.com). In this regard, PLS path estimates were stan-
dardised regression coefficients, and the loadings of items on the constructs
could be construed as factor loadings (Barclay et al., 1995). The PLS
technique also produced R-squared (R2) values for all endogenous con-
structs which could be interpreted in the same manner as R-squared (R2)
values produced by regression analyses (Igbaria, Guimaraes, & Davis,
1995).
3.6.7. PLS Procedures

In the preceding section, it was stated that SEM incorporates a two-step
procedure in data analyses which administers the assessment of the mea-
surement model followed by the structural model estimation. The measure-
ment model involves the assessment of the construct validity, convergent
validity and discriminant validity of the reflective constructs. This study,
thus, followed the two-step sequential procedure by employing PLS-SEM

(Barclay et al., 1995).
3.6.7.1. Step 1: Assessment of the Measurement Model. The measurement

model dealt with the relationships between the observed variables and the
constructs. Items which represented the observed variables measured the
constructs. The analysis of the measurement model led to the calculations
of loadings providing the researcher with an indication of the strength of
the measures.
3.6.7.2. Step 2: Assessment of the Structural Model. The structural model

dealt with the relationships between the constructs in a structured relational
framework or path diagram. Through the analyses, the PLS technique pre-
sented a clear picture of the magnitude and degree of contribution of various
constructs under a common and structured path diagram which helped
the researcher to test the hypotheses according to the theoretical framework.
3.6.8. Assessment of Measurement Model

During the measurement model assessment phase, the relationships
between indicators and their corresponding constructs were examined by
assessing construct validity which consisted of convergent validity and dis-
criminant validity. Individual item reliability and internal consistency were
the measures for convergent validity which evaluated how closely the items
in a single construct correlated with each other (Barclay et al., 1995;
Santosa, Wei, & Chan, 2005). The discriminant validity of the study con-
structs was also assessed at this phase. Discriminant validity refers to the
degree to which the study constructs differ from each other (Barclay et al.,
1995). Adequate convergent and discriminant validity item reliability,
internal consistency and discriminant validity (Barclay et al., 1995;
Hulland, 1999; Santosa et al., 2005) build confidence and accuracy in the
structural model estimation.
3.6.8.1. Convergent Validity. The assessment of convergent validity was a

fundamental part of assessing the measurement model. The psychometric
properties of the measurement model were assessed by evaluating the relia-
bility, convergent validity and discriminant validity (Fornell & Larcker,
1981). The reliability of the constructs was assessed by considering compo-
site reliability and Cronbach’s alpha. Inter-item correlations were the
strong measurement of convergent validity as it (convergent validity)
assessed consistency across multiple operationalisations. The magnitude
88 MD SHAH AZAM
and significance of standard path loadings were considered to examine con-

struct validity.
3.6.8.1.1. Item Reliability Item reliability assessed the loadings for each
individual item. The loadings indicated the correlation of the items with
their respective constructs. Therefore, maintaining low loading items would
decrease the correlation between the items in the construct (Nunnally,
1994). Item reliability also measured the level of random error for each
construct; the lower the item loading, the higher the level of random error.
Therefore, this procedure could identify and eliminate the items in a parti-
cular construct that could increase the construct’s level of random error
(Fornell & Larcker, 1981).
High item loadings indicated the reliability of the measures of the latent
variable; however, the previous literature supported some agreed level of
item loadings as thresholds for fairly reliable measures (Hair et al., 1998;
Igbaria et al., 1995). Igbaria et al. (1995) deemed 0.4 as an acceptable mini-
mum loading. Hair et al. (1998) suggested that loadings above 0.3 were sig-
nificant, above 0.4 were more significant and above 0.5 were very significant.
Chin (1998a) believed that item loadings should be above 0.5. Carmines and
Zeller (1979) maintained 0.7 as the reliability limit whilst Barclay et al.
(1995) specified 0.707 as the minimum limit. However, Nunnally (1994)
argued that in the case of strong theoretical support, further reviews of low
loading items were warranted. This would be especially pertinent if the low
loading items added to the explanatory power of the model.
Taking into account all the recommendations in the literature and to
maximise the measurement model’s ability to fulfil the requirements of con-
vergent validity, 0.5 was determined as the minimum value.
3.6.8.1.2. Internal Consistency Internal consistency was measured through

calculating composite reliability (Fornell & Larcker, 1981). Composite
reliability is considered to be superior to traditional measures of consis-
tency (such as Cronbach’s alpha) because it is not influenced by the number
of indicators (Hanlon, 2001). Eq. (1) is the formula for how internal consis-
tency was calculated:
P 2
λyi
α = P 2 P ð1Þ
λyi þ λVarðɛ i Þ
where α = internal consistency, λ = component loading to an indicator,

Y = construct, i = item, Varðɛ i Þ = 1 − λ2yi .
Constructs with a coefficient value of 0.70 and above in the estimates

of composite reliability were accepted as reliable for further analysis (as
suggested by Barclay et al., 1995; Igbaria, Zinatelli, Cragg, & Cavaye,
1997).
3.6.8.1.3. Average Variance Extracted (AVE) Fornell and Larcker (1981)

specified that average variance extracted (AVE) should be at least 0.5 for
convergent validity to be satisfied. Thus AVE, although it is not a usual
measure of convergent validity, was rationally computed for robustness of
the statistical analysis. The AVE was calculated using Eq. (2):
P2
λyi
AVE = P 2 P ð2Þ
λyi þ λVarðɛ i Þ
where λ = component loading to an indicator, Y = construct, i = item,

Varðɛi Þ = 1 − λ2yi .
3.6.8.2. Discriminant Validity. The measurement model also involved

evaluating the discriminant validity which was the extent to which differ-
ent constructs diverged from one another. Discriminant validity, a proof
of construct validity, is defined as the degree to which any given
construct is different from any other (Barclay et al., 1995). The square
root of the AVE and cross-loading matrix is a widely used measure
for discriminant validity (Barclay et al., 1995; Igbaria et al., 1995).
According to Igbaria et al. (1995), a model is assessed as having
acceptable discriminant validity if the square root of the AVE of a con-
struct is larger than its correlation with other constructs. On the other
hand, the constructs may be considered as discriminant if the loading of
items within a construct (shown in columns in a cross-loading matrix)
are greater than the loading of any other item within the same column
(Barclay et al., 1995). Both approaches for assessing discriminant validity
were important in adequately proving discriminant validity at construct
level as well as at item level.
3.6.8.2.1. Discriminant Validity at Construct Level Fornell and Larcker’s

(1981) suggestion involving estimating the AVE is often used as an effec-
tive criterion to prove discriminant validity among reflective constructs.
According to this criterion, discriminant validity is assessed by compar-
ing the square root of AVE with the inter-construct correlations. The
90 MD SHAH AZAM
square root of AVE should be greater than the inter-construct correla-

tions when the constructs are considered to have adequate discriminant
validity.
The inter-construct correlations are presented in the off-diagonals while
the square root of AVE is placed in the main diagonal in bold font on the
matrix. In order to prove discriminant validity among the reflective con-
structs, the off-diagonal elements (correlation of latent variables) must be
less than or equal to the bolded, diagonal elements (square root of AVE)
in the corresponding rows and columns (Barclay et al., 1995; Gefen,

Straub, & Boudreau, 2000; Igbaria et al., 1997).
3.6.8.2.2. Discriminant Validity at Item Level Barclay et al. (1995) sug-

gested another way for assessing discriminant validity by using a cross-
loading matrix which considers loadings and cross-loadings of measures
to test discriminant validity at the item level. The cross-loading matrix
displays the constructs in the columns and the measurement items in the
rows which enables the researcher to check the itemconstruct correla-
tion at any point. Thus, the correlation matrix provides an opportunity
to compare the constructitem correlation for discriminant validity. The
loading of items within a construct (shown in columns) should be greater
than the loading of any other item within the same column in order
to prove discriminant validity among the constructs (see Table 4 for
threshold values and assessment procedures for convergent and discrimi-
nant validity).
Table 4. Threshold Values for Reliability and Validity.

Measurement Assessment Procedure and Threshold Values
1. Convergent validity
(a) Item reliability Item loading ≥ 0:7
(b) Internal consistency
i. Composite reliability Calculated value ≥ 0:7
ii. Average variance extracted (AVE) Calculated value ≥ 0:5
2. Discriminant validity pffiffiffiffiffiffiffiffiffiffi
(a) Construct level AVE > correlation between the constructs
(b) Item level Item loadings of construct > all other cross-item
loadings of the construct
3. Nomological validity R2 ≥ 0:10
3.6.9. Structural Model Estimation and Nomological Validity

The structural model was estimated to examine the degree and magnitude
of the relationships between endogenous and exogenous variables. More
specifically, the structural model, using the PLS technique, estimated path
coefficients, t-statistics, standard errors and R2 to examine the Hypothesised
relationships. The path coefficients indicated the strengths and direction of
the relationships, and t-statistics and standard errors indicated the signifi-
cance of the influence, while the R2 value indicated the amount of variance
explained. The variances associated with the endogenous variables deter-

mined the explanatory power of the proposed model. The nomological
validity of the endogenous variables of the model was examined by the R2
values (Santosa et al., 2005). As proposed by Falk and Miller (1992), this
study used 0.10 as the cut-off value for nomological validity of the endo-
genous variables under the theoretical framework.
This study adopted a bootstrap resampling procedure to generate
t-statistics and standard errors (Chin, 1998b; Gefen et al., 2000). The tech-
nique which bootstrapping employs for calculating the t-statistic is similar to
the traditional t-test that is also used to interpret the significance of the paths
between study constructs (Barclay et al., 1995). By utilising a confidence esti-
mation procedure other than the normal approximation, the bootstrap pro-
cedure reproduced samples with replacements from the original sample
set and continued to sample until it reached the specified number as required
for the analysis. This study used 500 resamples for the bootstrap.
3.7. Summary
The methods employed and processes undertaken to attain the purpose of

this study were the key focus of this section. As the study adopted a mixed-
method research approach, the methods and procedures employed for
executing the qualitative field study and quantitative survey were stated
separately in different sections. The data collection process carried out
through field interviews and the methods for analysing the interviews were
described. The design and structure of the questionnaire, the tests of
reliability and validity of the measures, and the methods of data analysis
and hypotheses testing were also discussed in detail. The following section
deals with the analysis of the field interviews and the fine-tuning or
re-constructing of the research model.
92 MD SHAH AZAM
4. FIELD STUDY ANALYSIS AND COMPREHENSIVE

RESEARCH MODEL
4.1. Introduction
This section addresses the various stages of the qualitative research which
was undertaken to provide a foundation for the comprehensive quantitative
study. The field study was administered to compare the conceptual frame-
work developed by the literature review with the context of the study. The
main purpose of the qualitative research was to contextualise and fine-tune
the research model. The field study also explored new constructs and their
measurement items. The entire process of the qualitative research, from
designing the field study questionnaire to the final outcomes, is reported
in this section which concludes with a refined research model (Fig. 2) for
this study.
4.2. Operation of the Field Study
4.2.1. Sample
Researchers have applied different principles to determine the sample
size for qualitative data collection. Some researchers have agreed on an
open-ended number of cases while other researchers have supported the
idea of selecting a pre-defined range of cases. Eisenhardt (1989) and
Perry (1998) favoured a restricted range of cases and suggested that 48
cases were appropriate for qualitative research while Sandelowski (1995)
stressed that it was hardly possible to determine in advance the minimum
number necessary to ensure an adequate sample size in qualitative
research. The adequacy of the sample for credible research findings there-
fore becomes relative. Thus, the point of focus should be whether the
sample strategy should be large or small to achieve the intended objec-
tives of the investigation instead of judging a sample too small or too
large. However, the sample size may be reasonably considered as ade-
quate for analysis if either informational redundancy (Denzin & Lincoln,
2005; Lincoln & Guba, 1985) or theoretical saturation (Strauss &
Corbin, 1998) is reached.
This study applied the notion of an open-ended number of cases
and stopped interviewing after the 11th case taking into consideration
informational redundancy and theoretical saturation. The sample com-

prised 11 owners of different SMEs or their representatives selected by
adopting a convenience sampling method. Table 5 shows the demo-
graphic profiles of the interviewees who participated in the qualitative
survey.
4.2.2. Demographic Profiles of Study Respondents

A convenience sampling method was used to select the sample respondents.
The sample comprised owners or their representatives from eight small and
three medium-sized organisations. The study also included five organisa-
tions from the manufacturing industry and six organisations from the
services industry. Personal judgment was applied to ensure diversity among
the sample units. The market competitive position of six firms was good;
two firms possessed a very good market position while only one firm pos-
sessed an extremely good competitive position. On the other hand, one
organisation declared that they were experiencing bad market conditions
while one organisation explained that they had neither a good nor a bad
market position.
Again, 10 organisations had experienced revenue increases in the last
few years: among them, five organisations showed highly increased revenue
and one organisation had secured substantially increased revenue in the
last few years. The sample also included an organisation which had no
change in its revenue in that period of time. It is important to note that all
Table 5. Sample Profile for Qualitative Research.

Firm ID SME Type Employees Size Market Position Revenue
A S 35 Small Good Increase

B M 95 Small Good Increase
C S 21 Small Extremely good High increase
D S 47 Medium Good High increase
E S 9 Small Good Increase
F S 5 Small Good Substantial increase
G M 21 Small Good High increase
H S 92 Medium Bad Steady
I M 35 Small Neither good nor bad Increase
J M 35 Small Very good High increase
K M 89 Medium Very good High increase
94 MD SHAH AZAM
of the organisations included in the sample for the qualitative survey had
experience of ICT use in their organisation.
4.2.3. Data Collection

The study employed direct and face-to-face interviews with the persons
selected for the qualitative survey. The interviews were organised and
recorded with the consent of the interviewee. The individuals participating
in the field study were provided with a consent form which clearly outlined
the purpose and ethical issues related to the study. It was also mentioned
that his/her participation in the interview was completely voluntary and
thus they could withdraw themselves from the study at any time. A semi-
structured interview technique was used as the primary tool to collect data
(the data collection procedure has been discussed in detail in Section 3).
A complete English version of the semi-structured interview guide has been
provided in Appendix B.
4.2.4. Data Analysis

The content analysis technique was employed to analyse the qualitative
data (Siltaoja, 2006). During the qualitative data analysis, both inductive
and deductive methods were utilised to compare the qualitative data with
the theoretical framework and also to develop a causal relationship
between the constructs under study (Berg, 2001).
The themes, sub-themes and concepts explaining variables and factors
were explored by using the inductive process. The process also explored
measurement scales of some variables or factors. The variables and factors
explored by the induction process were further induced into a common fra-
mework. The ultimate objective of the field study was to prepare a compre-
hensive conceptual framework by comparing the induced framework with
the initial research model which had been developed from the literature
review. This common framework was then compared with the initial
research framework presented in Section 2.
In the past, content analysis using Berg’s (2001) guidelines was carried
out through a completely manual process. In the manual process, the tran-
scripts were read word by word, a process through which quite common
phrases were identified. Similar patterns or relationships that existed were
also marked and notes were taken of any section that was quite similar
or common from one transcript to another. These phrases, similarities
or patterns were analytically coded and finally categorised into various
subsections known as factors and variables. The process for these analyses
became easier and more sophisticated with the introduction of text analysis
software, particularly NVivo.
NVivo provides some user-friendly tools which can be used to identify,
record and analyse various themes, sub-themes and concepts, and the
causal relationships among various concepts. It also provides options
for inducing all of the concepts, variables or factors that were explored
through the inductive process and provides options for model building.
It becomes quite easy to explore categories and manage the themes, sub-
themes and concepts using the free and tree nodes.
This study used Nvivo 10 to analyse the qualitative interviews. In order
to analyse the interviews using NVivo, the recorded interviews were trans-
formed into interview transcripts. NVivo first imported all of the interview
transcripts. The researcher created numerous free nodes by labelling each
segment of the data. Each ‘free node’ summarised and accounted for each
concept in the data which provided the basis for developing various tree
nodes comprising similar free nodes. The tree nodes were developed from
the free nodes that had been developed in the immediate earlier stage.
The tree nodes comprised a number of relevant free nodes with a similar
concept which might become a construct. For example, 99 variables (free
nodes) were explored while analysing the interviews for organisational
performance. Further analysis was undertaken to combine similar variables
into one which resulted in nine distinct variables. Later, those nine vari-
ables were grouped into three tree nodes named competitiveness, internal
operation productivity and financial performance. These three distilled
factors were again grouped into a distinct broader construct of interest
which was called performance.
4.3. Findings (1st Stage: Inductive Analysis)
This section presents the findings from the field study analysis based on the
first stage of content analysis. Factors and variables explored by the con-
tent analysis are firstly presented. Relationships among the study constructs
and construction of the final combined model are presented in subsequent
sections. A summary of the findings of the 1st stage inductive analysis
comprising a high-level list of the factors and variables with subsequent
frequencies is presented in Table 6.
Table 6. Factors and Items from Field Study.
96
Factors/Variables Participants
A B C D E F G H I J K
Owner innovativeness
Courageous ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Knowledge of ICT ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Receptive to new ideas ✓ ✓ ✓ ✓ ✓ ✓ ✓
Cognitive evaluation
Usefulness
Useful in the organisation ✓ ✓ ✓
Increases productivity ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Enables performing tasks more quickly ✓ ✓ ✓ ✓
Helps increase chances of getting a raise ✓ ✓ ✓
Ease of use
The system is easy to use ✓ ✓ ✓ ✓ ✓
Easy to become skilful using the system ✓ ✓
We find the technology easy to use ✓ ✓
Learning to operate the technology is easy for us ✓ ✓ ✓ ✓
Behavioural intention and expectation
I don’t find any differences between our intention to use ICT ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
and what we are going to do in evaluating our ability or the
facilities or constraints arising from the external environment
Environmental pressure
MD SHAH AZAM
Coercive pressure
Conformity with parent corporation’s practice ✓ ✓ ✓
Dominance of customer adopters ✓ ✓
Our major suppliers demand we use ICT
Mimetic pressure
Competitors use ICT ✓ ✓ ✓
Competitors are benefited by using ICT ✓ ✓ ✓
Normative pressure
Customers use ICT ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Suppliers use ICT ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓ ✓
Stakeholders (important business partners) use ICT

Regulatory bodies use ICT ✓ ✓ ✓ ✓ ✓
Global pressure (time pressure)
Pressure from globalisation ✓ ✓ ✓ ✓ ✓
Hard to sustain without ICT at present time ✓ ✓ ✓ ✓ ✓
Facilitating condition
Existing system and process compatibility ✓ ✓ ✓ ✓ ✓ ✓
Resources necessary to use the system ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Training and maintenance facility ✓ ✓ ✓ ✓
Country readiness
Technology infrastructure
Internet service availability ✓ ✓ ✓ ✓ ✓
Speed and sophistication of Internet ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Hardware and accessories ✓ ✓ ✓ ✓ ✓ ✓
Uninterruptible power supply ✓ ✓ ✓
Human infrastructure
Individual’s knowledge of and skill in IT and Internet-based ✓ ✓ ✓ ✓ ✓ ✓
business
Knowledge and skill in IT and Internet-based business at ✓ ✓ ✓
institutional level
Institutional support for computer education (availability and ✓ ✓ ✓
affordability)
Legal infrastructure
Law relating to online communication ✓ ✓
Law relating to online consumer protection ✓ ✓
Financial infrastructure
Online banking facility (adequate) ✓ ✓ ✓
Debit and credit card service ✓ ✓ ✓
97
98
Factors/Variables Participants
Policy and subsidy

Government tax and customs policy ✓ ✓ ✓
Government’s motivational programme ✓ ✓ ✓ ✓ ✓ ✓
Government grants ✓ ✓ ✓ ✓
Government subsidies ✓ ✓ ✓
Culture
In-group Collectivism
Group members take pride in the individual accomplishments of ✓ ✓ ✓ ✓
their group managers
Group managers take pride in the individual accomplishments ✓ ✓ ✓ ✓ ✓
of their group members
Managers encourage group loyalty even if individual goals ✓ ✓ ✓ ✓ ✓
suffer
Power distance
Power is highly concentrated at the top of society ✓ ✓ ✓ ✓
People blindly follow the leader ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Uncertainty avoidance
Orderliness and consistencies are not apparent ✓ ✓ ✓ ✓ ✓ ✓ ✓
Lack of standard operational rules and procedures ✓ ✓ ✓ ✓
Bengali values
MD SHAH AZAM
Face-to-face communication is vital in daily lives ✓ ✓ ✓ ✓
People respect and prefer to communicate through the Bengali ✓ ✓ ✓ ✓ ✓
language
Strong and close social bonds ✓ ✓ ✓ ✓
Ethical culture
Bribery and corruption ✓ ✓ ✓ ✓
Nepotism and politicisation ✓ ✓
ICT use
Basic computing and Internet ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Have own homepage ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
✓ ✓ ✓ ✓ ✓
Have interactive homepage which supports product cataloguing

and order processing (online order receiving and processing)
Interactive homepage which supports online transactions and ✓ ✓ ✓
account management (e-business)
Online customer service and complete digitisation ✓ ✓ ✓ ✓
Degree of utilisation
Diversified use of the same technology ✓ ✓
Proper utilisation ✓ ✓ ✓ ✓ ✓
Use full capacity ✓ ✓
ICT integration
Entire organisation is under single system ✓ ✓ ✓
External organisations are under integrated systems ✓ ✓
Front-office functions are digitally integrated with back-office ✓ ✓ ✓ ✓ ✓ ✓
functions and databases
Extent of utilisation for external communication ✓ ✓ ✓ ✓
Performance
Competitiveness (competitive performance)
Sales area has been widened ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Interaction with customers has been increased ✓ ✓ ✓
Competitive position has been improved ✓ ✓ ✓ ✓ ✓
Internal operation productivity (internal performance)
The internal operation of the organisation has become ✓ ✓ ✓
transparent
Productivity of the employees has been improved ✓ ✓ ✓ ✓
The internal operation of the organisation has become ✓ ✓ ✓
structured
Financial performance
Sales of the company have been increased ✓ ✓ ✓ ✓ ✓ ✓
Profitability of the company has been increased ✓ ✓ ✓ ✓ ✓ ✓
99
Overall performance is increased ✓ ✓ ✓ ✓ ✓ ✓ ✓
100 MD SHAH AZAM
4.3.1. Factors and Variables

4.3.1.1. Owner Innovativeness. The field study explored owner innovative-
ness as one of the important factors influencing ICT adoption by SMEs in
Bangladesh. Six of the 11 respondents thought that the adoption of ICT
was related to the risk-bearing ability of the owner of a firm. The respon-
dents who noticed that the owners’ or CEOs’ risk-taking propensity or
ability greatly impacted on ICT adoption by SMEs divulged the fact by
saying; ‘In my opinion, it is a big task [to implement ICT] and our owners
have shown their courage by using such a high-cost and sophisticated

software’ (Firm B), or ‘… the owner of the company should be such a person
who can take a high risk on his shoulder’ (Firm J), or ‘[h]e loves to take high
risks in various business situations’ (Firm C).
Seven respondents were also unanimous in recognising owners’ or
CEOs’ innovativeness in the adoption and diffusion of ICT by SMEs in
Bangladesh, saying that ‘… our owners are knowledgeable and educated and
once they felt the necessity of the technology in the organisation they decided
to use it’ (Firm K), or ‘[o]ur owner is highly educated and an innovative
man’ (Firm C), or ‘[t]hey [owner/CEO] have adequate knowledge on the
use and utilitisation of ICT’ (Firm B).
Seven other respondents expressed their feelings differently in terms
of supporting the influence of owners’ or CEOs’ innovativeness in the
adoption or use of ICT by SMEs as: ‘I can mention the special quality of the
owner of this organisation which is that they can easily receive new things
particularly newly innovated technology … They [owners] should come
forward and receive new ideas in relation to the use of ICT in the company’
(Firm B); ‘[w]ithout innovativeness, nobody can run a venture. Our owner is
an innovative person’ (Firm F); ‘[o]ur owner has given utmost importance
with computerisation. Innovative [receptive to new ideas] and risk-taking
behaviour are the fundamental qualities of the owner’ (Firm I).
4.3.1.2. Cognitive evaluation. The field study indicated that respondents’

cognitive evaluation of technology mattered in regard to the intentions
towards ICT usage by SMEs in Bangladesh. Firm G admitted the influence
by saying: ‘[i]t (ICT) is entertaining. It is providing more scope for docu-
mentary evidence than telephone and other media. It is interactive, reasonable
and easy …’ and ‘… customers send some documents which are easy to check,
send or forward to the appropriate departments. It helps the employee to
become confident …’ Firm K and Firm E also felt similarly. For example,
Firm K stated that:
[i]t [application of ICT] makes the operation easier. Suppose we have many depart-
ments. To run a business, I have to communicate with other departments and write a
letter to the head of another department: maybe he is situated on the 8th floor while I
am on the ground floor. If I have to send it physically to the department, it will take a
lot of time and effort. (Firm K)
As with Firms G, K and E, most respondents indicated their evaluation

of technology was driving them towards using it. The field study analysis
revealed that cognitive evaluation resulted from user perceptions of the
various benefits and barriers with regard to innovation. Various benefits
and utilities were assessed from the field study which were associated with
two perceptions perceived usefulness and ease of use. The following sub-
sections address users’ perceptions about ICT through which a user or a
prospective user can develop cognitive evaluation, that is, favourableness
or unfavourableness towards the technology.
4.3.1.2.1. Perceived Usefulness The field study revealed that almost all
respondents felt that organisations’ perceptions about the usefulness of the
technology influenced the adoption of ICT by SMEs in Bangladesh. For
example, Firm H stated: ‘[w]e get information at the same time from every
corner of the country. If we don’t use ICT, we cannot get information at the
same time from everywhere. In that case, some districts can get it earlier;
some districts may get it later which may create a disparity among different
units of the organisation. We receive information at the same time and very
promptly and timely [currently] through ICT’.
Similarly, Firm D stated: ‘[i]t saves time, reduces the cost and increases
efficiency as I can get access to the report from anywhere and can send my
response also from anywhere. So it provides better opportunity to overcome
the place and time barriers of the operations and business’.
The usefulness of the technology was also explored in other firms’ state-
ments. Firm C indicated that ICT was a tool for organisational efficiency
and productivity while Firm I mentioned ICT’s capability for creating
effective and efficient communication in a flexible manner. For example,
Firm I said: ‘[c]ommunication through ICT reduces the time [prompt], it
produces evidence of communication. Cost reduction: place and time don’t
matter with it’.
4.3.1.2.2. Perceived Ease of Use As with perceived usefulness, the field
study also revealed that ‘perceived ease of use’ influenced the adoption and
use of ICT by SMEs in Bangladesh. Nine of the 11 respondents felt that
the ease of use of the technology mattered in their adoption intention
or usage of ICT with this explored from various statements. As Firm
102 MD SHAH AZAM
A said: ‘[n]ow I am quite comfortable and friendly with the diverse use of
ICT.’ Firm H agreed with the notion and stated: ‘[i]t [use of ICT] is inter-
active, reasonable and easy (to use) … when a customer sends some docu-
ments which are easy to check, send or forward …’ Firm F also supported
this view by perceiving that international communication became easier
through the use of ICT. Firm F stated: ‘ICT mainly makes communication
from our country with other countries easier’ (Firm F).
The analysis of the field interviews revealed that cognitive evaluation
mattered in SMEs’ adoption and usage of ICT. The field study again
explored two constituents of cognitive evaluation which were perceived use-
fulness and perceived ease of use. By applying an interpretive research
approach, the field study thus anticipated that, by its nature, the concep-
tualisation cognitive evaluation would be reflected through perceived
ease of use and perceived usefulness of the technology.
4.3.1.3. Culture. The field study revealed that all respondents felt that cul-
ture influenced the adoption of ICT by SMEs in Bangladesh. For example,
Firm B stated: ‘[i]n terms of using ICT, a formal and disciplined work envir-
onment and capable and educated manpower are required … however, it
[importance of culture] has been stated in several counts in the previous
discussion. To me, national culture is one of the strong constructs affecting
individuals’ or groups’ ICT use behaviour’.
Similarly, Firm F stated: ‘[t]he cultural aspects are related to ICT adop-
tion [intention or usage decision] by SMEs in various counts’.
Firm J also felt the same and said: ‘… as [we feel] the societal norms and
practice deter [on many counts] automated and non-personal exchanges, we
are facing trouble in getting upgraded and wide use of the technology’.
The above-mentioned statements and quotes inferred that the national
culture affected cognitive evaluation and intention to use ICT by SMEs in
Bangladesh. The field study also explored various dimensions of culture
that were related to SMEs’ ICT usage intention.
The interpretive research revealed that ‘power distance’, ‘in-group collec-
tivism’, ‘uncertainty avoidance’, ‘ethical culture’ and ‘Bengali values’ were
dominant in the formation of the Bengali culture which may also have
some relationship to ICT adoption by SMEs in Bangladesh.
4.3.1.3.1. Power Distance All respondents mentioned that huge power

distance was evident in the social or managerial hierarchy which charac-
terised Bangladeshi national culture. For example, Firm B stated: ‘[i]n our
country, the general people blindly follow their leaders although they do not
accept them in their mind’
Firm G expressed a similar assessment about the power distance dimen-
sion of national culture by saying: ‘I don’t expect that the people would obey
and follow their leader without any question [but, in fact,] people cannot
challenge their leader’s or superior’s decision and reflect their own opinion or
decision’.
Some interviewees reported that obeying the leader was a norm in the
country’s culture and, in reality, subordinates did not have any scope to
question their leader. For example, Firm J stated: ‘[t]hey [the common peo-
ple] should not follow their leaders without question. In the present situation,
in practice, there is no scope for the follower to question their leader’.
4.3.1.3.2. In-Group Collectivism The field study also explored in-group
collectivism as one of the strong cultural dimensions that influence ICT
adoption by SMEs in Bangladesh. For example, Firm H said: ‘I would feel
pride with my affiliation to the organisation. In our organisation, workers
take pride in the individual accomplishment of their managers and the man-
agers also take pride in the individual accomplishment of their subordinates.
I wish the organisation would be run collectively. I don’t prefer the autocratic
environment’.
Past studies have suggested that the cultures of Asian countries are
collective cultures while the cultures of Western countries are mostly indivi-
dualistic (Hofstede, 2001; House et al., 2004). The view that the culture of
Bangladesh was also considered to be collective was supported by many
statements. For example, Firm G said: ‘I like transparency in our organisa-
tion and for any problems that arise to be settled with discussion among all
members concerned. The members of our organisation are not obliged to do
everything but they do so as expected ’.
4.3.1.3.3. Uncertainty Avoidance One respondent (Firm F) considered
uncertainty avoidance as a strong part of culture by saying: ‘[o]ur organi-
sation is not a disciplined or organised entity. The formal approach is not evi-
dent everywhere in the organisation. Maybe there exist some rules but those
are for a few people. In our organisation, the needs of the organisation and
what is the responsibility of the general workers are not clearly explained and
recorded ’.
Similarly, Firm G said: ‘like our country, orderliness and consistencies are
not stressed in different functional areas in the organisation. Although many
needs of the organisation are conveyed to the general workers by word of
mouth, who is responsible for what and what the organisation expects from
104 MD SHAH AZAM
individuals are not clearly stated … many inconsistencies are evident in var-
ious stages in society as in our organisation. If people do anything wrong,
they may be forgiven as the requirements and their responsibilities are not
clearly stated’.
4.3.1.3.4. Ethical Culture Ethical issues are crucial for the development of
Bangladesh. Unethical approaches and misdeeds are becoming an integral
part of day-to-day life and corporate culture in the country.12 Politicians
and bureaucrats are surrounded by nepotism, unfairness and corruption in

their working environment. The field study explored the view that ethical
culture was one of the strongest actors in the decision-making process of
SMEs’ ICT usage. For example, Firm A said: ‘[i]n my personal experience,
it is evident that inefficient workers or candidates are offered jobs in different
areas in many organisations in Bangladesh. So they play bad roles to cover
their inefficiencies … The selection procedure should be transparent. In
Bangladesh, the selection is full of nepotism’.
Similarly, Firm J stated about politicisation and the bribery culture:
‘[i]n our country, politicisation is quite common, and bribes have become a
common phenomenon. I feel the individuals’ qualifications and performance
should be evaluated rather than their affiliation with politics, groups or
economic strength’.
Firm E emphasised the unethical issues of Bengali culture by saying:
‘[i]n our country, we have championed in corruption several times around the
world. Besides, corruption has become an integral part of our culture. We
have achieved the championship in corruption for three consecutive years
around the world’.
4.3.1.3.5. Bengali Values The culture of Bangladesh is constituted with

some values and norms called Bengali values that are different from the
cultural dimensions stated by Hofstede (2001) and House et al. (2004).
These values were explored in statements by Firms G, K and B. Firm G
said: ‘[i]n our country, people do exchange through [the language] Bengali.
But ICT is not compatible with Bengali and there is no Bengali user interfaces
for various ICT applications which is a vital problem for the mass use of
technology’.
Firm K emphasised people’s preference for communicating in the
Bengali language. For example, Firm K said: ‘[w]e have pride in our lan-
guage movement and independence. In honour of the language martyrs and
their sacrifice in language movement and in respect of the Bengali language,
the government of the country declared that all work and communication
should be in Bengali which creates very easy communication among people

with the same language’.
Furthermore, the people of Bangladesh live in a society that maintains
very close bonds among community members. This aspect of culture was
explored in a statement from Firm B: ‘face-to-face communication and
personal social exchanges are vital in the day-to-day lives of the country’s
population’.
4.3.1.4. Country Readiness. Every respondent expressed their concern

about infrastructure and other country readiness indicators with regard
to establishing an ICT-based working and market environment. For
example, Firm D said: ‘[t]o develop ICT and SMEs, the government has
prime responsibilities. Nobody can use ICT without government approval as
the government has some policy and rules. The infrastructure for ICT cannot
be developed by individuals. The government should provide the developed
infrastructure for using ICT such as the submarine fibre optic cable network,
telephone systems, etc.’ The above-mentioned statement implied that
country readiness was composed of various infrastructures. The field
study revealed that country readiness comprised technology infrastructure,
human infrastructure, legal infrastructure, financial infrastructure, policy
and supports. The following subsections present respondents’ assessments
and predictions on different aspects of country readiness.
4.3.1.4.1. Technology Infrastructure Technology infrastructure was consid-

ered the most important aspect of country readiness which would enable an
SME to implement ICT-based functionalities. Firm C indicated that
Internet connectivity and Internet speed-related infrastructure were essen-
tial for developing an ICT-based business or society. Firm C addressed the
issue as: ‘[i]n our country, we are connected online through the submarine
fibre optic cable network under the sea which connects our country to the east
Asian countries and other countries around the world. Sometimes, this cable
network is disrupted due to environmental uncertainty or man-made hostile
activity and the whole operation stagnates’. Firm C again mentioned the
discriminatory Internet speed and sophistication of Internet services across
the country by saying: ‘although we are not facing any difficulties in Internet
speed as we are directly linked with DSE [Dhaka Stock Exchange], we
sometimes receive complaints from our customers who are not receiving
adequate speed’. The discriminatory Internet services were also proved by
another statement from the same firm: ‘[w]e have 24 hour Internet
106 MD SHAH AZAM
connectivity. We use Ravi Internet. Those who are using laptops, they use
Grameen Internet. We don’t have much problem with Internet speed’.
Although power supply was not relevant to ICT operation particularly
in developed countries, it was an important part of country readiness with
regard to establishing an ICT-dependent working environment. This aspect
of technology infrastructure was explored with the statement from Firm B:
‘[w]e have a crisis with power supply in our country. Where it will go, we
don’t know. We also have our back-up plan. We will set up our own genera-
tor. We are also willing to establish a 1 (one) megawatt electricity generation

plant’.
Similarly, Firm F said: ‘firstly, we feel that electricity should be available.
Computer accessories should be available. What information we use, there
should be a back-up facility. UPS [uninterruptible power supply] should be
available. And the UPS back-up time should be enhanced’.
4.3.1.4.2. Human Infrastructure Human infrastructure was considered to
be another important part of country readiness in facilitating ICT in the
workplace. An ICT-based operation requires knowledgeable and expert
human resources for its operation and maintenance. Analysing statements
from the various SMEs, the field study explored the view that human infra-
structure was an important component of country readiness. For example,
Firm C stated: ‘[w]e face some difficulties and time lapses in getting
adequate support according to our needs. If we get an educated and technolo-
gically competent employee at first [entry level], then we could save some
effort and time which we spend in educating our new employees. Another
important concern in terms of the use of ICT is the quality and skill of
employees.’.
Firm E expressed concern about employee skills at entry level by saying:
‘[n]ew employees who come to work lack skills [ICT skills]’.
On the other hand, Firm I highlighted the development aspects of
human resources which are required for establishing an ICT-based working
environment. Firm I stated: ‘[I]n our time, there was no computer education
at school and college level, now it is included at school and college. More
comprehensive computer education should be given as part of a general educa-
tion to provide fundamental computer skills’.
4.3.1.4.3. Legal Infrastructure Legal infrastructure has implications for
ICT-based operations. Communication and transactions via the Internet,
in particular, require significant legislative supports. Organisations partici-
pating in online transactions and exchanges feel the necessity of having
effective laws and practices to solve any dispute that occurs during an
online transaction, communication or exchange. Firm K stated the impor-

tance of legal infrastructure: ‘[w]e don’t have adequate rules and regulations
to solve any dispute of conflict arising from online communication, exchanges
and transactions’.
Firm I described the importance of legal support in regard to accepting
digital signatures which was essential for any digital contract by saying:
‘[t]o authenticate a digital contract, a digital signature should be accepted as
valid in various contracts’.
4.3.1.4.4. Financial Infrastructure ICT has provided enormous opportu-

nities for individuals, businesses and society. In particular, the introduction
of Internet-based business operations has become a revolution in managing
businesses and transactions leading to changes in the ways in which these
were traditionally carried out. These changes are also reflected in banking
operations. Success in online business and transactions requires an efficient
online banking mechanism. Developed countries are now almost online
banking-dependent. Developing countries although lagging behind are also
seeking to develop an efficient online banking mechanism. The field study
explored the importance of financial infrastructure in the successful imple-
mentation of online business operations and transactions. For example,
Firm K expressed their concern about online banking facilities by saying:
‘[w]e don’t have adequate rules and regulations … Our banking mechanism
for online transactions and operations is also not satisfactory’.
Firm C mentioned the importance of online banking in regard to custo-
mer compatibility with the mechanism by saying: ‘[n]ormally most of our
customers are not habituated with online money transfers’. Firm I raised a
deep concern by saying: ‘[i]n our country, we don’t access bank accounts
from home’.
4.3.1.4.5. Policy and Supports Government policy and supports play a
pivotal role in the implementation of ICT into different functional areas in
a country. The field study explored the importance of government policy
and supports in promoting ICT usage in Bangladesh and, in particular, for
achieving organisational productivity and efficiency in the business sector.
For example, Firm K stated: ‘[a]t government level, they can reduce the tax
and custom duties on ICT-related products. Once government has withdrawn
all taxes and duties from ICT-related products, they can provide their other
supports such as grants and subsidies to also encourage ICT usage at the
organisational level and country level’.
Firm C referred to the importance of government policy and supports.
The present government’s move towards establishing an ICT-based society
108 MD SHAH AZAM
is an example of policy and supports to launch an ICT-based business

environment. The importance of policy and supports was also explored in
this statement by Firm C: ‘Besides this, the present government is very much
encouraging ICT utilisation in the country and building a digital Bangladesh
by 2021 so we hope that, within a few years, we will see some improvement at
our national level and we will also receive some supports and grants from
government in this regard’.
4.3.1.4.6. Environmental Pressure The field study explored the view that
environmental pressure has direct and significant influence on SMEs’ adop-

tion and usage of ICT. For example, Firm H said: ‘[w]e have our own dis-
tribution system. We have some suppliers and other collaborators for
ingredient supply. They are in a developed country and they have highly
sophisticated ICT usage. On the other hand, those who are taking our product
as middlemen are not IT-based’.
Firm C mentioned the influence of environmental pressure on SMEs’
adoption and usage of ICT focusing on the effects on competitors and
suppliers: ‘[w]e are encouraged to use better technology and upgrade our
existing systems when we see that our competitors and suppliers are using
better ICT. We also feel deterred from using and upgrading our technological
standards when we see our customers do not have an adequate technological
fit, and they are not receptive to or compatible with ICT and our systems’.
From the above-mentioned statement, it can be easily assumed that
environmental pressure has diverse effects. The field study explored differ-
ent dimensionalities of environmental pressure, such as coercive pressure,
mimetic pressure, normative pressure and global pressure which affect the
adoption and diffusion of ICT by SMEs in Bangladesh.
4.3.1.4.7. Coercive Pressure Many SMEs in Bangladesh have adopted
ICT-based operations or have ensured their online presence in conducting
or supporting a business in response to pressures from the regulatory
authority or dominant customers. For example, Firm I stated: ‘[o]ur custo-
mers and suppliers do all their work online. We have to comply with them.
Besides the amount that we claim for the service should be claimed online and
the money is transferred online. Accessories and cloth from different coun-
tries: India, Sri Lanka, Indonesia, China, the UK. A very limited portion we
collect from domestic sources. They are more competent online. They require
everything ordered online and require payment online. They have high band-
width online, they have no power interruption, they have high mobile online
access, so they can access from anywhere at any time. They don’t have band-
width fluctuations’.
Firm E also mentioned coercive pressure focusing on the demands of a

powerful customer: ‘our big customers encourage us to use sophisticated
technology like them to make us compatible with their systems’.
4.3.1.4.8. Mimetic Pressure Pressures from competitors influenced the

adoption and diffusion of ICT by SMEs in Bangladesh. The study explored
mimetic pressure as an important aspect of environmental pressure. For
example, Firm B said that: ‘[o]ur competitors are also using sophisticated
technology. In fact, two of our competitors are using this software and they
started [to do so] prior to our use. So, buyers can adequately know about
their orders and their progress so he/she can understand the situation and the
stage that it has reached, etc. at any time from their premises. They have
given the customers these privileges and the opportunity to track their order
so they are getting bigger advantages than us. Thus, they are achieving better
performance and business than us. We are now at the initial stage. We have
the intention to extend our technological support to the buyers’ premises’.
Firm G also stated: ‘[i]n the competitive market, a business, like the
garments trade, is impossible to run without ICT although it was possible in
earlier days’.
4.3.1.4.9. Normative Pressure The field study explored normative pressure

as a dimension of environmental pressure which influenced the adoption
and diffusion of ICT by SMEs in Bangladesh. For example, Firm B stated:
‘[o]ur customers are high-profile ICT users. They have their own server and
own homepage, as well as an automated customer and supplier service envir-
onment. They use specialised software. They have provided us with some scan-
ners. During the procedure, we are connected with them’.
Similarly, Firm J said: ‘[o]ur all customers are also communicating with
us through emails and online communication. Our customers are also suggest-
ing that we use ICT communication’.
Firm G indicated the importance of normative pressure with regard to
accessing suppliers, competitors and customers by saying: ‘[o]ur suppliers,
customers, and competitors use the Internet, email and computer which is the
same as what we use’.
4.3.1.4.10. Global Pressure In addition to coercive, mimetic and normative

pressures, some pressures have been affecting SMEs’ ICT adoption and
usage which differed from these existing three categories. These pressures,
named as global pressure, were related to ongoing technological develop-
ment and global market pressures. For example, Firm D stated: ‘[i]n the
age of globalisation, without adopting ICT, we cannot think of a business. At
110 MD SHAH AZAM
what stage would we be if we didn’t use ICT or the Internet facility was
closed? We would simply be paralysed/collapsed’.
Firm K agreed with Firm D by saying: ‘I can’t work without ICT in the
age of globalisation’.
Firm K also said: ‘[i]t [ICT usage] is the same as the technology usage
behaviour of the individual. Say I am using a mobile phone. After some days,
I find that a new phone has arrived on the market. Then I go for the new
phone to get the new features, style and benefits or model. It is actually the
effect of innovation. The old technology was obsolete and new technology
emerges and we are all following these cycles. In other ways, I can tell that
we are each affected by the hype of using technology’.
Firm J mentioned the importance of global pressure in the adoption and
diffusion of ICT by saying that: ‘[i]f ICT is seized from our company, our
company will collapse. We can’t go one step without ICT. We are seriously
ICT-dependent. We can’t even sustain communication through letters, phone
or fax. It is not possible to run without ICT. It is the demands of time and
our surroundings’.
4.3.1.5. Facilitating Condition. The field study explored the view that some
resources and competencies were essential for using ICT-based technology
or systems: these were computer hardware, software, skilled manpower and
a compatible working environment. These were resource-based although
essential for ICT-based operations, and their adequacy and diversity would
accelerate or deter firms’ ICT usage. For example, Firm I expressed this as:
‘[g]ood quality computer, good bandwidth, computer literacy is required for
ICT use’.
Firm J mentioned the importance and effectiveness of a facilitating con-
dition that focused on human resources by stating that: ‘[w]e have skilled
human resources and technology-competent employees. We have adequate
hardware and software resources’.
Similarly, Firm I said: ‘We have Internet connectivity, computer hard-
ware, skilled manpower, and other necessary resources [human resource and
technology resources]’.
Firm A commented about the procedure they followed which reflected
the necessity of having skilled manpower in the adoption and diffusion of
ICT by SMEs in Bangladesh. Firm A stated: ‘[o]ur working operations are
directed and controlled by different departments and their head or person-in-
charge. We have an MIS (Management Information Systems) department.
The head of MIS deals with and controls all ICT-related tasks. In fact, they
are the custodian of ICT and its utilisation. We have that skilled manpower.
We also have computer hardware and software resources. We have the

connectivity’.
Many organisations mentioned the adequacy of resources which
make them habituated to and almost dependent on ICT. For example,
Firm I said: ‘[i]t is not only a demand from our clients; it has become a part
of our everyday work culture’. Firm B supported this expression by saying:
‘[n]ow I [as a leader of the organisation] am quite compatible and friendly
with the diverse use of ICT’ while Firm I approached this differently
as: ‘[w]e have the problem that we are still depending on paper-based
documentation’.
The above-mentioned analyses indicated that the existing systems and
working procedure may foster or deter technology usage by firms. This
aspect referred to the facilitating condition of technology usage which was
clearly reflected in the response of Firm J: ‘[i]n each and every organisation,
you will find some strengths and weaknesses; we have them also. But what I
would like to state is that we have more strengths in our organisation in terms
of using ICT. Our other organisational systems are ICT-compatible’.
4.3.1.6. Intention or Expectation of ICT Use. The field study explored

whether the respondents evaluated their move towards the adoption of ICT
in two phases, namely, intention and expectation. They had their plan and
most had the firm’s willingness to move forward. Their plan was rational
and did not fluctuate in considering different environmental factors or
situations. For example, Firm C expressed the view that: ‘[w]e can provide
that support, as some of our competitors have already established this kind of
set-up to support their customers. To make our organisation competitive, we
will be able to establish the systems to integrate our customers and other
stakeholders in the online environment’.
Organisations rationally drew up their intention. The following state-
ments also suggested that firms’ intentions were not simply a plan:
The supports and surrounding facilities screen the intention. We are not in any doubt
about whether we could do this according to our intention. We will upgrade the existing
technology. (Firm A)
I think it is possible to implement our intention. (Firm C)
The above-mentioned analyses indicated that the intention of the organi-

sations was well-founded and rational and was not easily changed by the
effects of external or environmental factors. Thus, this intention should
actually be called ‘expectation’.
112 MD SHAH AZAM
4.3.1.7. ICT Use. The field study explored the situation that, to some
extent, all respondents were using ICT in their firms. The results from the
field study analysis categorised all responding firms into six groups based
on their level of ICT applications usage, namely, basic computing with
Internet; static homepage; interactive homepage which supported product
cataloguing; interactive homepage which supported online transactions,
that is, e-commerce; and digitisation among different functional areas; and
ERP. For example, Firm A said: ‘Yes, we use ICT. We use Internet. To support
our Internet operation, we use a computer, printer, scanner, etc. Besides this,
we have installed a cc [closed circuit] camera for security purposes’.
Similarly Firm F said: ‘Yes, we are using ICT. We are using the Internet,
only the Internet. We have our own homepage’.
Firm D mentioned the sophisticated operation of ICT in their organisa-
tion by saying: ‘We also take orders through the Internet. We sell to some of
our valued dealers and our branches. But the sales relate to any kinds of
declarations and reports that we send through the Internet and also receive
through the Internet. We collect dealers’ requirements through mobile phone:
although for those who have the Internet facility, we get their requirements
through the Internet’.
Firm G also mentioned e-commerce participation and said: ‘[t]he custo-
mer places an order online in detail. We send this to the merchandise dept.
and scrutinise it in that section then send it to the sample section. We select
the price and then bid the price. Send the price to the customer once … Most
of the works are [Inter]net-based’ (Firm G).
The field study explored digitisation or automation as an application of
ICT usage. More specifically, Firm B said: ‘[b]esides, we have special soft-
ware by which we analyse our production and other related works. Finally, I
find that ICT is a very useful technology. Now I am quite compatible and
friendly with the diverse uses of ICT. Our firm is also computerised’.
4.3.1.8. ICT Integration. The field study explored organisational concerns

about the integration of the variety of technological facilities used in an
organisation. The field study supported the view that the integration of
diversified technologies was dependent on the level of ICT usage by SMEs
in Bangladesh. Integration was also seen to affect organisational perfor-
mance. For example, Firm K said: ‘[a]fter that [introduction of ICT use],
the scope of ICT use has widened and strengthened. It seems that a new tech-
nology is just introduced in some department and gradually it will spread over
the organisation and work under the common system’.
Firm B mentioned integration with other firms within the industry by

saying: ‘[w]e have accumulated some of our closely related similar group of
companies in very close contact for those who are working with us in the same
network environment through our server. We are doing the above-mentioned
works through ICT’.
Similarly Firm K said: ‘[i]f the external organisations are also integrated
or communicated with entirely through ICT, the performance [organisational
performance] will certainly be increased’.
4.3.1.9. Degree of Utilisation. The field study explored the view that the
degree of utilisation was an important factor explaining organisational per-
formance, as was ICT integration. The factor degree of utilisation was also
developed from the firm’s usage of different levels of ICT applications. For
example, Firm B said: ‘[w]e can earn productivity and efficiency [organisa-
tional performance] through proper utilisation of ICT particularly through
some software by which we can track, control and initiate alternative
strategy … those companies from whom we are getting benefits and gradually
developing, they are actually developing due to the proper utilisation of the
technology [ICT]. Some big companies from the world perspective are giving
their employees Blackberries so, from anywhere around the world, they can
mail and be connected with the network’.
Similarly, Firm H said: ‘[w]e are trying to develop the IT-literate compe-
tent employee day by day. We hope we shall be capable of the appropriate
and comprehensive use of ICT’. Some firms emphasised the importance of
the appropriate use of technology for customer satisfaction as well as for
organisational performance. For example, Firm I said: ‘[i]f the ICT is not
appropriate, we will not be able to satisfy the customer or target group’. Firm
J also said similarly: ‘The performance depends on how well we will use ICT’.
Proper and appropriate use of ICT was considered crucial for organisa-
tional performance. By utilising ICT properly, organisations could attain
improved organisational performance. However, ICT may cause decreased
productivity or damaging output if not utilised properly. For example,
Firm K stated: ‘ICT enhances the employees’ performance. ICT normally
reduces the errors. But it depends on the person who operates the ICT. It
depends on their skill. I have a doubt because if the ICT is not properly
handled, it may produce bad and damaging output’.
4.3.1.10. Organisational Performance. The field study explored the view

that the use of different levels of ICT applications may foster the perfor-
mance of the organisation. ICT integration and the degree of utilisation
114 MD SHAH AZAM
also had an effect on organisational performance. For example, Firm B

said: ‘[a]fter that [as we were not getting reasonable performance], I con-
sulted with some external experts and learned that our existing technological
set-up was good enough to do that job. We actually had a lack of operational
knowledge about our set-up [which was not properly used] thus we were
getting less productivity [organisational performance]’.
The field study also explored different factors under the conceptualisa-
tion of organisational performance such as competitiveness, internal
operation productivity and financial performance. For example, Firm D

said: ‘[i]t [use of ICT] also enhances the sales growth in respect to our
competitors’.
4.3.1.10.1. Competitiveness The field study explored competitiveness as

a dimension of organisational performance. For example, Firm J said:
‘[i]t [ICT use] enhances productivity. It enhances competitive position. It
increases sales’. Firm I also supported Firm J by saying: ‘It is the tool which
creates competitiveness’.
Similarly, Firm H said: ‘[t]he ICT helps to exchange information related
to sales, marketing, HR, different policy strategy, market conditions, etc.
which makes the company competitive … To stay competitive, gaining the
competitive advantage through IT [ICT] is essential’.
4.3.1.10.2. Internal Operation Productivity The field study explored the

view that ICT use influenced productivity in the internal operation of the
organisation. Firm D said: ‘[w]e work in this process [by using ICT] trans-
parently and promptly and in the age of globalisation we are able to provide
international standard products and related facilities and opportunities in our
country promptly’. Firm G supported the view that the use of ICT enhanced
transparency and productivity in internal operations by saying: ‘[s]o the
organisations, those that are operating and starting operations in this field,
are using the Internet and ICT which make them transparent to each other.
[Inter]net connection is obvious. … When we use ICT, we save time and are
able to make the relationship transparent between all parties involved in the
work’.
4.3.1.10.3. Financial Performance Firms (B, C, D, E, G, I and K)

expressed their views agreeing with the concept that ICT use facilitated
organisational performance. For example, Firm G said: ‘[i]t [ICT use] is
entertaining. It [ICT use] provides more scope of documentary evidence than
telephone and other media. It [ICT use] is interactive, reasonable and easy.
It [ICT use] is not only saving time, but also providing some documentary
evidence. It [ICT use] is easy to control, and to have a conversation with a
customer through ICT is also easier. It [ICT use] reduces errors. It [ICT
use] helps increase performance … When we use efficient ICT, it enhances
performance. It increases our company’s service sales’.
Firm K also supported Firm G by saying: ‘[i]t [ICT use] enhances the
productivity of the company. The financial performance, that is, the sales,
profitability and overall performance of the company are enhanced by the use
of ICT’
The field study explored nine different free nodes to address organisa-
tional performance (see Table 6). The interpretive analysis further cate-
gorised the nine free nodes into three sub-themes (first-order factors) of
organisational performance. Further analyses included all of these three
sub-themes (first-order factors) in a broader higher-order complex concep-
tualisation which is organisational performance. Thus, the field study
explored organisational performance as a higher-order conceptualisation
formed by three different aspects of organisational performance competi-
tiveness, internal operation productivity and financial performance that
were also measured by nine items.
4.3.2. Linkages among the Factors

The objective of the field study was to explore the variables and their mea-
sures. However, the development of the relationships between the factors is
also considered to be an important task during a qualitative data analysis
(Xu & Quaddus, 2005). Table 7 indicates the relationships between the
factors that were explored in the qualitative analysis.
Table 7 is also an outcome of the analysis of field interviews presented in
the previous section (Section 4.3.1). The table presents all of the explored
factors and establishes a link (a causal link) between them. For example,
the notion of CE → EXP represents the influence of cognitive evaluation
on intention to adopt ICT by SMEs. It was observed that all firms except
Firms E, G and H found that cognitive evaluation had a direct influence
on their intention to adopt ICT at their own firm. The relationship between
the two factors was explored and justified as Firm B said: ‘… so I think
mind set-up and attitude [cognitive evaluation] are largely related to ICT
use. If I don’t have a positive attitude, how can I accept [intention to adopt]
the new thing? I wouldn’t accept and use ICT … I think they [similar SMEs]
will change their attitude [cognitive evaluation] and would be interested to
utilise ICT’s potential in their organisation’. Firm D expressed a similar
116 MD SHAH AZAM
Table 7. Causal Links Explored from the Qualitative Survey.

Construct Participants
Cognitive evaluation
CE & EXP ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
OI & CE ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
OI & EXP ✓ ✓ ✓ ✓ ✓ ✓ ✓
OI & USE ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Behavioural expectation
EXP & USE ✓ ✓ ✓ ✓ ✓ ✓
Environmental pressure
EP & EXP ✓ ✓ ✓ ✓ ✓ ✓
EP & USE ✓ ✓ ✓ ✓ ✓
FC & EXP ✓ ✓ ✓ ✓ ✓ ✓
FC & USE ✓ ✓ ✓ ✓ ✓ ✓
Country readiness
CR & EXP ✓ ✓ ✓ ✓ ✓
CR & USE ✓ ✓ ✓
Culture
CUL & CE ✓ ✓ ✓ ✓ ✓
CUL & EXP ✓ ✓ ✓ ✓ ✓
ICT use
USE & PERF ✓ ✓ ✓ ✓ ✓ ✓
USE & INT ✓ ✓ ✓ ✓ ✓ ✓
USE & UTL ✓ ✓ ✓ ✓ ✓ ✓
UTL & PERF ✓ ✓ ✓
Integration
INT & PERF ✓ ✓
Performance
USE → INT → PERF ✓ ✓ ✓
USE → UTL→ PERF ✓ ✓
USE → UTL
INT
PERF ✓ ✓
CE = Cognitive evaluation, OI = Owner innovativeness, EXP = Expectation, EP =

Environmental pressure, CUL = Culture, CR = Country readiness, FC = Facilitating
condition, USE = Actual usage, INT = Integration, UTL = Utilisation, PERF = SME
performance.
feeling indicating as follows that cognitive evaluation created a positive

intention to adopt ICT:
… what types of work are being done in Bangladesh? If ICT is not used in that area,
the nation will stay backward. So attitude plays a positive role in ICT-related decisions.
It creates [a] strong intention to use ICT.
The relationships between constructs were developed from the direct

statements. In some cases, it was not possible to find the relationship from
a direct statement. In those cases, detailed data analysis was performed and
interpretive research philosophy was applied to interpret the underlying
relationship between the factors. For example, Firm C stated that: ‘[t]o
operate ICT in an organisation, some resources are required. We should
have competent employees who have adequate knowledge and capacity to
understand and use ICT … The monetary resources are required to buy the
ICT resources, that is, the hardware, computer, printer, scanner, Internet
connectivity, server, etc. and the software support which will manage the
operation of ICT according to the company’s needs in an integrated
environment’.
From the above-mentioned statement, it was inferred that some
resources were necessary for ICT use. The resources indicated the facilitat-
ing condition. The statement might also be used to develop the causal link
between the facilitating condition and ICT use by the organisation.
Although the statement was not directly expressing any distinct relation-
ship, by analysis utilising the interpretive philosophy, a positive relation-
ship may be indicated between the facilitating condition and ICT use.
4.3.3. Comparison between Field Study Findings and the Initial Model
(2nd Stage: Deductive Model)
Most of the variables in the field study were supported by the literature
which was discussed earlier in Section 4. This section discusses the factors
which either evolved from the field study or were different from the existing
literature.
The initial model developed through the literature review projected that
ICT use, integration and degree of utilisation influenced SME performance.
Similar to the literature review, the field study revealed that ICT usage in
conjunction with integration and utilisation strongly affected SME perfor-
mance. The field study also supported the mediating role of ICT integra-
tion and degree of utilisation to explain the influence of ICT usage on
SMEs’ performance.
118 MD SHAH AZAM
The existing literature suggested that BI towards the adoption of an

innovation leads to behavioural expectation which may be affected by
many external factors like facilitating condition, country factors, culture,
environment, etc. The primary model thus included two factors to indicate
the firms’ two different levels of willingness to adopt ICT. The field study,
analysing the interviews, explored this situation which was different to what
had been revealed in the existing literature. The field study justified one fac-
tor for addressing firms’ willingness to adopt an innovation. The field study
explored that the firm’s intention was not only cognitive, it was rational
with this screened through external factors and forces; that is, intention
acted as expectation in the context of SMEs’ ICT adoption behaviour in
Bangladesh. Thus, BI has been screened out and only one factor, beha-
vioural expectation, was justified for testing the Hypothesised relationship.
The field study also revealed two new constructs Bengali values and
ethical culture for addressing national culture. The field study distilled the
cultural dimensions explained in the existing literature (Hofstede, 1984;
House et al., 2004) and discovered that three among the five cultural dimen-
sions, power distance, uncertainty avoidance and in-group collectivism, may
have some influence on ICT adoption or usage by SMEs in Bangladesh.
The field study explored various country-level factors that may have
some influence on ICT adoption and usage by SMEs. The field study also
revealed a new dimension of environmental pressure. The qualitative analy-
sis explored the view that ‘pressure from globalisation’ was an important
factor which may have some influence (in conjunction with coercive
pressure, mimetic pressure and normative pressure) on the adoption of ICT
by SMEs in Bangladesh.
The existing literature suggested that positive attitude led to positive inten-
tion which was formed through favourable or unfavourable perceptions
about the innovation. Similar to the primary model, the field study explored
this factor as a state of cognitive evaluation which was formed by SME
owners’ perceptions about the usefulness and ease of use of the innovation.
The field study justified the formation of the constructs: cognitive evaluation,
culture, environmental pressure, country readiness and organisational perfor-
mance as a higher-order complex hierarchical conceptualisation.
4.3.4. Justification of the Findings in the Literature Review

With support from the literature, this section has provided the justification
for the selected constructs and dimensions that were developed from the
field study. It has emphasised that the factors and dimensions that were
derived from the field study, on the basis of commonality and consistency,
were also supported by the existing literature. Therefore, this justification
has established the competency and adequacy of each construct and dimen-
sion in the existing literature. Table 8 presents the factors and the dimen-
sions that have been finalised and the relevant support from the literature.
4.3.5. Justification of Combining BI and Expectation

BI plays a vital role as a strong and immediate antecedent in predicting
actual behaviour. The past theories in behavioural studies have recognised
the strong and important role of intention in predicting actual behaviour
such as the TRA (Fishbein & Ajzen, 1975); TPB (Ajzen, 1985); TAM
(Davis, 1986) and DOI theory (Rogers, 1983). The pattern and formation
of these theoretical models have demonstrated the mediating role of
Table 8. Justification of the Field Study Variables by the Literature.

Constructs Subconstructs Reference
Cognitive evaluation Ease of use Venkatesh et al. (2003)

Usefulness Venkatesh et al. (2003)
Owner characteristics Owner’s innovativeness Thong and Yap (1995), Thong
(1999)
Environmental Coercive pressure Quaddus and Hofmeyer (2007)
pressure Mimetic pressure Teo et al. (2003)
Normative pressure Teo et al. (2003)
Global pressure Field study
Facilitating condition Venkatesh et al. (2008)
Country readiness Technology infrastructure Field study
Human infrastructure Field study
Legal infrastructure Zhu and Kraemer (2005)
Financial infrastructure Zhu and Kraemer (2005)
Government policy and Roessner (1988), Goldsmith (1990)
supports
Culture Power distance House et al. (2004), Hofstede (1984)
Uncertainty avoidance House et al. (2004), Hofstede (1984)
In-group collectivism House et al. (2004)
Bengali values Field study
Ethical culture Field study
Behavioural Venkatesh et al. (2008)
expectation
ICT use Davis (1989), Venkatesh et al.
(2008)
Integration Zhu and Kraemer (2005)
Degree of utilisation Field study
Performance Competitiveness Zhu and Kraemer (2005)
Internal operation productivity Zhu and Kraemer (2005)
Financial performance Zhu and Kraemer (2005)
120 MD SHAH AZAM
intention which has been reflected in successive research initiatives around the
world which have explored innovation diffusion in multidisciplinary fields of
study such as in psychology, business, social sciences and IS research.
Taking into consideration the role of BI, many diffusion researchers
have studied innovation diffusion behaviour involving intention as the final
dependent variable (e.g. Gefen & Straub, 2000; Kendall et al., 2001; Lal,
1999; Pavlov & Chai, 2002). On the other hand, some researchers who have
investigated actual usage behaviour (e.g. Venkatesh et al., 2008) have also
included intention as an strong antecedent reflected by a number of cogni-

tive, individual, social and environmental factors.
The innovation adoption theories have explained the adoption and dif-
fusion phenomena mainly through intention (Ajzen, 1985; Davis, 1986;
Fishbein & Ajzen, 1975; Rogers, 1983; Venkatesh et al., 2003). The existing
theories have been criticised as intention is a reflection of the adopter’s
internal schema of beliefs which may fail to adequately explain actual
behaviour in a situation in which the adopter is under incomplete volitional
control. Intention, furthermore, may not be reflected in the decision if a
time gap exists between intention and actual behaviour. A stronger antece-
dent, termed as behavioural expectation, was therefore proposed to explain
technology adoption behaviour which was not a cognitive factor which
would thus be able to address the roles of all external and internal antece-
dents (Venkatesh et al., 2008).
Warshaw and Davis (1985) explained BI as ‘the degree to which a
person has formulated conscious plans to perform or not perform some
specified further behaviour’.
Warshaw and Davis (1984) explained behavioural expectation as ‘an
individual’s self-reported subjective probability of his or her cognitive
appraisal of volitional and non-volitional behavioural Antecedents’.
Venkatesh et al. (2008) argued that the role of intention logically
involved a new construct, behavioural expectation, to address the reflec-
tions of numerous internal as well as external influences with the view
to predicting actual behaviour. The limitations of intention in predicting
actual usage behaviour were also reflected in subsequent research.
Venkatesh et al. (2003) suggested that the conclusion of previously studied
conceptualisations of systems use such as duration, frequency and intensity
of use (these conceptualisations were used in many studies rooted back to
Davis et al., 1989) was true only within an intentionality framework, where
external factors were taken into account via facilitating conditions (FC),
defined as ‘the degree to which an individual believes that an organisational
and technical infrastructure exists to support use of the systems’
(Venkatesh et al., 2008, p. 484). To overcome the limitations of BI and the
facilitating condition in predicting actual behaviour, a new construct,
behavioural expectation, was proposed (Warshaw & Davis, 1984). Venkatesh

et al. (2008) recognised the rationale and importance of behavioural expecta-
tion in explaining the limitations of BI and the strengths of behavioural
expectation which was stronger than intention in predicting actual behaviour
from a systems usage perspective in the organisational context. They proved
it through a cross-sectional empirical study in a different time period.
The innovations and changes in models and theories have encouraged
the investigation of existing issues in a developed and comprehensive
setting of conceptualisations and underlying relationships among the study

constructs. The chronological arguments and developments in IS research
to address actual systems usage behaviour (e.g. Venkatesh et al., 2003,
2008; Warshaw & Davis, 1984, 1985) have suggested the inclusion of BI
and behavioural expectation as predictors of actual behaviour.
Venkatesh et al. (2008) proved the strength of behavioural expectation
and BI in an organisational setting by looking at the behaviour of newly
innovated systems where the recipients or adopters had no experience dur-
ing the first phase of the research. In supporting the outcome and recom-
mending that the conclusion of the research was to introduce behavioural
expectation as a new construct to predict actual behaviour, we doubted the
applicability of behavioural expectation in a situation where the adopters
had some experience with the innovations. In addition, would it perform
similarly in a setting where adopters were carrying out rational behaviours?
Prior experience was found to be an important determinant of behaviour
(Ajzen & Fishbein, 1980; Bagozzi, 1981; Fishbein & Ajzen, 1975; Triandis,
1979). Specifically, it has been suggested that knowledge gained from
past behaviour will help to shape intention (Eagly & Chaiken, 1993;
Fishbein & Ajzen, 1975). This is, in part, because experience makes knowl-
edge more accessible in memory (Fazio & Zanna, 1978; Regan & Fazio,
1977) and also because past experience may make low probability events
more salient, ensuring that they are accounted for in the formation of
intentions (Ajzen & Fishbein, 1980). Thus, direct experience will result in a
stronger, more stable BIbehaviour relationship (Ajzen & Fishbein, 1980).
This stable BI was reflected in behavioural expectation as conceptualised
by Warshaw and Davis (1984) and Venkatesh et al. (2008).
Where the adopters have proceeded with logical steps and realistically
evaluated every situation, the BI was formed through the reflections of var-
ious realistic and possible benefits, the favourableness or unfavourableness
of the innovations as well as the expected outcomes which were similar to
the concept ‘behavioural expectation’. Furthermore, respondents of this
study were SME owners or their representatives who were responsible for
organisational prospects and prosperity; thus, they behaved rationally in
any decision involving the running of their businesses. They thought,
122 MD SHAH AZAM
planned and had realistic intentions to stay competitive in the ongoing

changing globalised market environment. In this situation, we may con-
clude that the intention used in this study, in fact, indicates the rational
intention, that is, expectation. Thus the final model has combined intention
and expectation (these were proposed in the initial model separately) and
has used rational intention (expectation) as an immediate antecedent of
SMEs’ technology usage behaviour.
4.3.6. The Comprehensive Research Model

As discussed earlier, a comparison was made between the initial model and
the findings of the field study. Justifications of the selected constructs and
dimensions were then made. As a result, this section proposes a comprehensive
model for the current research. Fig. 2 illustrates this comprehensive model.
The comprehensive model (cognitive evaluation, expectation, use and
organisational performance) states that cognitive evaluation and entrepre-
neurs’ innovativeness are the primary antecedents of BI. The model argues
that in the situation of having prior experience, the outcome and
OI
INT
COMP
PEU
CE EXP USE PERF INTOP
PU
FPERF
CUL FC UTL
PD ING UN BV EC
CR EP
TI HI LI FI PS CP MP NP GP
Fig. 2. The Comprehensive Research Model. Notes: CE = Cognitive evaluation,

OI = Owner innovativeness, EXP = Expectation, EP = Environmental pressure,
CUL = Culture, CR = Country readiness, FC = Facilitating condition, USE =
Actual usage, INT = Integration, UTL = Utilisation, PERF = SME performance.
(First-order constructs, PU = Perceived usefulness, PEU = Perceived ease of use,
PD = Power distance, ING = In-group collectivism, UN = Uncertainty avoidance,
BV = Bengali value, EC = Ethical culture, TI = Technology infrastructure, HI =
Human infrastructure, LI = Legal infrastructure, FI = Financial infrastructure, PS =
Government policy and supports, CP = Coercive pressure, MP = Mimetic pressure,
NP = Normative pressure, GP = Global pressure, COMP = Competitiveness,
INTOP = Internal operation productivity, FPERF = Financial performance.
predictions are almost known to the adopters thus the intention becomes
rational which is similar to expectation. Thus, the model uses expectation
instead of intention to explain the actual technology usage behaviour.
Expectation is again logically used in the model to adequately receive
the influences of external forces culture, environmental pressure and coun-
try readiness along with the internal and organisational factors cognitive
evaluation, entrepreneurs’ innovativeness and facilitating condition.
The model has extended one further step in comparison to the TRA,
TPB, TAM, the DOI theory or other traditional innovation diffusion fra-
meworks as it includes performance as the final outcome which is explained
by the usage of the innovation. The model emphasises that straightforward
use of the technology may not be fruitful for enhanced organisational
performance. The usage of the innovation will generate organisational per-
formance growth if it is properly utilised within an integrated framework.
Thus two new constructs, ICT integration and utilisation, have been
included as immediate antecedents of performance: they also act as media-
tors on the effects of ICT usage on organisational performance.
The model applies higher-order modelling to conceptualise cognitive
evaluation, culture, environmental pressure, country readiness and organi-
sational performance. That is, each of those constructs is conceptualised as
a higher-order construct of two or more latent variables. The performance
construct has been conceptualised as a higher-order construct of competi-
tiveness, internal operation productivity and financial performance.
4.4. Summary
This section has presented the findings of the field study and has proposed
a research model. Qualitative data were generated from 11 interviews con-
ducted in Bangladesh to assess the experience, expertise and perceptions of
SME owners or decision-makers on the adoption and diffusion of ICT at
firm level. The main objective of this field study was to test the applicability
of the proposed initial model based on the prior literature, and to explore
the dimensionality of related constructs. Overall, the content analysis tech-
nique consisting of inductive and deductive phases was employed to ana-
lyse the data. Moreover, theoretical as well as lateral replication was used
in the deductive phase. Factors and variables, as well as some measures,
have been explored with these being further scrutinised in light of the relevant
literature. Furthermore, relationships among factors have been established.
Based on the analysis, a combined model (integrating all factors and variables
from each interview) has been developed. Later, this model was compared
with the initial model (derived from the literature review) to propose the
124 MD SHAH AZAM
comprehensive research model. The final model has demonstrated a wide and
comprehensive process of ICT diffusion and its resultant outcome as organi-
sational performance in the context of SMEs in Bangladesh. On the basis of
the comprehensive model, several hypotheses have been constructed which
are reported in Section 5. The following section (Section 6) reports the data
analysis and findings of the quantitative survey.
5. HYPOTHESES AND QUESTIONNAIRE

DEVELOPMENT
5.1. Introduction
This section presents the development of the hypotheses and quantitative

research instruments based on the comprehensive research model (see
Fig. 2). The model was primarily developed from a rigorous literature
review process which was then contextualised and modified by the qualita-
tive field study analysis. The hypotheses are presented in different subsec-
tions under Section 5.2. Following the presentation of the hypotheses,
Section 5.3 discusses the development of the research instrument.
5.2. Hypotheses Development
5.2.1. Hypothesis Related to Cognitive Evaluation

The TRA (Fishbein & Ajzen, 1975) suggests that a structured model be
used to examine the relationships between one’s beliefs, evaluations, inten-
tion and actual behaviour. Fishbein and Ajzen (1975) suggested that an
overall affective evaluation, that is, one’s beliefs (cognitive) with regard to
performing a behaviour and the respective evaluations (affective) heavily
influence an individual’s intention to perform the behaviour. The evalua-
tion is made based on one’s cognitive evaluation of the consequences of
that behaviour (Ajzen & Fishbein, 1980; Sheppard et al., 1988). The evalua-
tion process involves functional consequences, that is, the outcomes which
are immediate, direct and tangible, as well as the psychological conse-
quences, that is, the consequences which are internal, personal and abstract
in nature. These cognitive and affective evaluations form one’s attitude,
that is, negative or positive feelings about performing a behaviour with this
having been widely researched in the consumer, marketing and IS research
(Ajzen, 1985; Fishbein & Ajzen, 1975; Gehrt & Carter, 1992; Mehta &
Sivadas, 1995; Reynolds, 1974 (for catalogue shopping); Schlosser,
Shavitt, & Kanfer, 1999 (for Internet advertising); Liao & Cheung, 2001 (for
e-shopping); Cho, 2004 (for intended online transactions). In this study, the
affective and cognitive aspects were associated with SMEs’ intention to use
various applications of ICT. The owners’ or managers’ feelings about using
ICT (favourable or unfavourable) are formed through their beliefs towards
using ICT, which was termed as ‘attitude’ towards ICT in many previous
research studies (Liao & Cheung, 2001; Mehta & Sivadas, 1995; Reynolds,
1974; Schlosser et al., 1999) as a higher conceptualisation of perceived ease of
use and usefulness of the technology. The construct was measured by the
owners’ or managers’ evaluation of the usefulness of the technology in their
firms’ current settings as well as the extent to which the technology was
considered easy to use by their respective employees or operators. Thus, this
evaluation, although similar to attitude, is termed ‘cognitive evaluation’.
The outcome of the field study was supportive of previous studies with
the ‘cognitive evaluation’ being positively related to the SMEs’ intention to
use ICT. All participants agreed upon the relationship between attitude and
intention. The participants expressed the view that the owners’ or managers’
positive evaluations about the usefulness as well as ease of use of the technol-
ogy led to the positive intention to use ICT. Unlike the formation of consu-
mers’ attitudes towards an innovation, the SMEs’ owners had prior
knowledge about the innovation and tried to make a rational decision by jus-
tifying the prospective gain and sacrifice in using the technology. To make
the firm profitable, the owners or entrepreneurs avoided acting on impulse
or making an emotional decision both of which are common in the consumer
decision process. The intention of SMEs’ owners or managers regarding ICT
was not simply a dream or aspiration but rather a rational and justified
intention which was at least screened through basic financial and managerial
concerns. Thus, the intention was termed as an expectation which is a more
expressed state of will than intention and can explain the effects of both
internal and external factors (Venkatesh et al., 2008; Warshaw & Davis,
1984). The implication was that a positive ‘cognitive evaluation’ was an
important indicator that would influence SMEs’ expectation to use ICT.
Based on the above discussions, the following hypothesis is proposed:
H1. Cognitive evaluation has direct and positive influence on SMEs’
expectation to use ICT.
5.2.2. Hypotheses Related to Owner Innovativeness

The adoption of a new product is a diffusion process that moves through
different stages over time. Moreover, diffusion is defined as the process by
which an innovation is communicated through certain channels over time
among the members of a social system (Kendall et al., 2001). According to
126 MD SHAH AZAM
the DOI theory (Rogers, 1983, 1995, 2003), the diffusion process begins
with knowledge of the existence of the innovation which is followed by per-
suasion, decision, implementation and confirmation stages. Rogers (1995)
indicated that the decision-maker’s characteristics, such as, socio-economic
characteristics, personality variables and communication behaviours, play
a vital role during the knowledge stage which also provides the basis for
intention formation. Individuals’ innovativeness has a significant influence
on the formation of adoption intention of an innovation although it is
formed during the persuasion stage.

The adoption of an innovation in a small firm is highly influenced by
the personal characteristics and preferences of the owner or CEO of
the organisation as most small businesses are managed by the owner who
also acts as the CEO (Solomon, 1986; Steinhoff & Burgess, 1986; Yap
et al., 1992). Interestingly, the SMEs have a flat organisational structure
(Raymond & Magnenat-Thalmann, 1982; Steinhoff & Burgess, 1986). The
CEO in an SME usually performs more than one function in running the
business operations, makes most decisions, and has full control of organisa-
tional resources (Quaddus & Hofmeyer, 2007; Yap et al., 1992). Yap et al.
(1992) explained how a CEO could play supportive roles in computer-based
information systems (CBIS) in a small business by establishing the firm’s
appropriate CBIS goals, identifying critical business information needs,
allocating sufficient financial resources to achieve CBIS goals and making
decisions relating to CBIS implementation and control.
Rogers (1995) also categorised five different adopters according to the
time of adoption as being innovators, early adopters, early majority, late
majority and laggards (Weber & Kauffman, 2011). The innovators are the
adopters who are innovative in nature and bear some risks associated with
the early adoption of an innovation. The innovativeness of the CEO or
owner of an SME has a significant effect on the adoption and implementa-
tion of IS (Thong, 1999; Thong & Yap, 1995). Therefore, based on the
above discussion, it is hypothesised that:
H2a. Owner innovativeness has direct and significant effects on cognitive

evaluation.
H2b. Owner innovativeness has direct and significant effects on SMEs’

H2c. Owner innovativeness has direct and significant effects on SMEs’

use of ICT.
5.2.3. Hypotheses Related to Environmental Pressure

The TOE framework (Tornatzky & Fleischer, 1990) posits the effects of
some external or contextual variables such as organisational and environ-
mental factors on the organisational aspects of technology diffusion.
The TOE framework identifies three aspects of a firm’s context that
influence the process by which it adopts, implements and uses technological
innovations. The environmental context is an important context among
them referring to the aspects of how a firm conducts its business, responds
to its industry, customers and competitors, and deals with government.

This framework has received more attention and acceptance from diverse
fields of study as it is consistent with the classical DOI theory (Rogers,
1983). Rogers emphasised the technological characteristics, and both inter-
nal and external characteristics of the organisation, as drivers for technol-
ogy diffusion.
The impact of environmental factors and their characteristics are well
addressed in institutional theory which posits that organisations face pres-
sure to conform to these shared notions of appropriate forms of beha-
viours, since violating them may call into question the organisation’s
legitimacy and thus affect its ability to secure resources and social support
(DiMaggio & Powell, 1983; Tolbert, 1985).
DiMaggio and Powell (1983) distinguished between three types of
isomorphic pressure coercive, mimetic and normative and suggested that
coercive and normative pressure normally operate through interconnected
relationships while mimetic pressure acts through structural equivalence.
Ajzen and Fishbein (1980) explored the role of SNs and studied how
they affect individual behaviour. SNs, in other words, pressure from friends
and family plays a vital role in the formation of intention to use an innova-
tion. In the organisational aspect, a focal organisation is able to learn
organisations with whom it is directly or indirectly tied, and is likely to be
persuaded to behave similarly (Burt, 1982). Many studies consider norma-
tive pressure as an antecedent of organisational innovation adoption
phenomena (Kuan & Chau, 2001; Teo et al., 2003).
Mimetic pressure is the influence of other structurally equivalent organi-
sations that have initiated some innovations and have become successful.
This pressure may cause an organisation to change over time to become
more like other organisations in its environment (DiMaggio & Powell,
1983). Many past studies have included mimetic pressure when looking at
organisational ICT adoption behaviour (Premkumar & Ramamurthy,
1995; Teo et al., 2003).
128 MD SHAH AZAM
Coercive pressure addresses various kinds of power or influence, infor-

do a certain thing with the dependent organisation having no option other
than to comply with the requirement. Coercive pressure has had great
importance in previous studies on ICT diffusion behaviour (Quaddus &
Hofmeyer, 2007; Teo et al., 2003). This study has applied a hybrid concept
of environmental pressure combining all aspects of isomorphic pressure in

a higher-order construct. Thus, the effect of environmental pressure in the
ICT adoption/diffusion process at firm level is hypothesised as:
H3a. Environmental pressure has a significant influence on SMEs’
H3b. Environmental pressure has a significant influence on SMEs’ actual

use of ICT.
5.2.4. Hypotheses Related to Facilitating Condition

Within the environmental context, the TOE framework (Tornatzky &
Fleischer, 1990) also posits the influence of technological context and orga-
nisational context. The technological context is concerned with the existing
technologies as well as new technologies relevant to the firm while the organi-
sational context addresses descriptive measures of the organisation such as its
scope, size and the amount of slack resources available internally. These
factors are particularly important for IS’ diffusion at organisational level as
some resources such as technological resources and human resources are
required for the operation of ICT. The presence of such resources may facili-
tate the adoption and use of the technology (Venkatesh et al., 2003).
The facilitating condition the existence of technological or organisa-
tional support infrastructure is grounded in the TPB (Ajzen, 1985). This
construct is also included in other successive models such as the decomposed
theory of planned behaviour (DTPB) (Taylor & Todd, 1995c), combined
TAM and TPB (C-TAM-TPB) (Taylor & Todd, 1995a) and the MPCU
(Thompson, Higgins, & Howell, 1991) which have looked at its effects on
intention and actual usage of technological innovations. The construct facili-
tating condition (or PBC, perceived behavioural control) is included as an
antecedent of intention (Ajzen, 1985; Taylor & Todd, 1995c; Thompson
et al., 1991) and actual use (Ajzen, 1985; Tornatzky & Fleischer, 1990). Thus,
based on the above-mentioned discussion, it is hypothesised that:
H4a. Facilitating condition has a significant influence on SMEs’ expecta-

tion to use ICT.
H4b. Facilitating condition has a significant influence on SMEs’ actual

use of ICT.
5.2.5. Hypotheses Related to Country Readiness

Country readiness, an important external factor, has been discussed in var-
ious past studies that have examined the adequacy and access costs of basic
information infrastructure (Kraemer & Dedrick, 1994; Shih, Dedrick, &
Kraemer, 2005); government policy and legislation (Lee & Shim, 2007;
Teo, Tan, & Buk, 1998); regulatory supports (Zhu & Kraemer, 2005; Zhu,
Kraemer, et al., 2006); the rule of law, political openness and property
rights protection (Caselli & Coleman, 2001; Oxley & Yeung, 2001; Shih
et al., 2005); and education levels (Caselli & Coleman, 2001). Dewan and
Kraemer (2000) reported that developed and developing countries differed
in terms of the level of IT use and the factors shaping that use. In referring
to Rogers (1983), Tornatzky and Fleischer (1990), and Caselli and
Coleman (2001), Zhu and Kraemer (2005) mentioned that, at the general
level, technology diffusion studies presented evidence of an unevenly occur-
ring diffusion across countries with different environments which indicated
that country-specific characteristics mattered when it came to technology dif-
fusion. Likewise, a variety of economic, social and political factors, including
income, education, technology policies, cultural norms and access to formal
and informal communication networks impact on the extent of diffusion.
Iacovou, Benbasat, and Dexter (1995) and Kuan and Chau (2001) con-
sidered readiness primarily as the availability of resources (financial and
technical) as well as strategic readiness. Country readiness has become an
important factor to be investigated in the area of organisational technology
diffusion studies as, even though an organisation has enough resources,
this does not necessarily mean that they facilitate the situation towards
better performance. Thus, the facilitating condition’s effect on behavioural
expectation and ICT use (Venkatesh et al., 2003, 2008) is not limited solely
to organisational resources. Internet availability and facilities at national
level, Internet speed, the cost of Internet access and ICT resources, regula-
tory framework, market conditions, delivery systems, and government pol-
icy and supports also have direct effects on ICT adoption producing a
regulatory environment that accelerates or hinders ICT use at firm level.
The field study also supported the inclusion of country readiness as an
antecedent of expectation and actual ICT usage behaviour. Participants in
130 MD SHAH AZAM
the field study indicated that country-level technology infrastructure,

human infrastructure, legal infrastructure, financial infrastructure and gov-
ernment policy and supports were important concerns that created country
readiness for rapid adoption and diffusion of ICT. The above-mentioned
four dimensions of country readiness should not be considered in isolation
from each other, but should be treated in a collective and mutually reinfor-
cing manner. Hence, the country readiness construct represents an integra-
tive measure of the level of readiness along these four dimensions. A
second-order factor modelling approach can capture correlations among

the four first-order factors and explain them using a higher-order construct
that is an integrative latent representation of country readiness. Previous
studies have noted that this operational approach represents a theoretically
strong basis for capturing complex measures (Segars & Grover, 1998;
Sethi & King, 1994; Stewart & Segars, 2002; Venkatraman, 1990; Zhu &
Kraemer, 2002; Zhu et al., 2003). Country readiness is not quite evenly
reflected in the four dimensions.
This factor is considered as a new aspect of the facilitating condition
which is different from the facilitating condition (organisational-level
resources and capabilities) explained in previous studies (Venkatesh et al.,
2003, 2008) and is termed ‘country readiness’. This study thus logically
added country readiness as a separate predictor of behavioural expectation
and actual use behaviour. Based on the above discussion, it is hypothesised
that:
H5a. Country readiness has a significant influence on SMEs’ expectation
to use ICT.
H5b. Country readiness has a significant influence on SMEs’ use of ICT.
5.2.6. Hypotheses Related to Culture

The existing theories have admitted the effects of various external environ-
mental factors on the adoption and diffusion of an innovation in the
organisational context (Rogers, 1995; Tornatzky & Fleischer, 1990). The
theoretical frameworks developed by Hofstede (2001) and House et al.
(2004) stressed the effects of culture on organisational leadership. Hofstede
(2001) treated culture as the collective programming of the mind that
distinguishes the members of one group or category of people from another
and studied culture in five dimensions power distance, collectivism,
uncertainty avoidance, masculinity versus femininity and long-term versus
short-term orientation.
House et al. (2004) addressed culture differently with nine dimensions
performance orientation, assertiveness, future orientation, humane orientation,
institutional collectivism, in-group collectivism, gender egalitarianism,

power distance and uncertainty avoidance. House et al. (2004) indicated
that national culture has effects on leadership and organisational processes.
Thus, the planning, execution and controlling aspects of an organisation
are highly influenced by culture. By utilising Hofstede et al.’s (2001) and
House et al.’s (2004) frameworks, many previous studies have examined the
impact of culture on organisational technology adoption. Thatcher et al.
(2006) reported that culture plays an important role in the adoption of a
certain technology by organisations in a particular country and it also

impacts on the degree to which the technology is accepted and the ways in
which it is used. Past studies have also supported the link between cultural
dimension and different facets of IT use (Bertolotti, 1984; Burn, 1995;
Erez & Early, 1993; Gefen & Straub, 1997; Harris & Davison, 1999; Hill
et al., 1998; Ho et al., 1989; Straub, 1994). Erumban and Jong (2006) found
that the national culture and ICT adoption rate of a country are closely
related. They further reported that most of Hofstede’s dimensions were
important in influencing adoption with power distance and uncertainty
avoidance dimensions seeming to be the most influential.
This study has undertaken a qualitative research study to explore and
justify the effects of various cultural dimensions in the diffusion of ICT
among SMEs in Bangladesh. The analysis of the interview transcripts
resulted in anticipation of the effects of five dimensions of culture, namely,
power distance, uncertainty avoidance, in-group collectivism, ethical cul-
ture and Bengali values on the intention to use ICT.
The exploratory search also anticipated negative structural relationships
between cultural dimensions and SMEs’ intention to use ICT. Based on the
above discussion, it is hypothesised that:
H6a. Culture has a significant influence on cognitive evaluation.
H6b. Culture has a significant influence on SMEs’ expectation to use

ICT.
5.2.7. Hypothesis Related to Expectation

The majority of the innovation diffusion theories included intention as a
strong predictor of actual usage behaviour (Ajzen, 1985; Davis, 1986;
Fishbein & Ajzen, 1975; Rogers, 1983; Venkatesh et al., 2003). As an
immediate antecedent of actual innovation usage behaviour, BI can explain
the effects of various internal or external factors in predicting actual
behaviour. For example, intention is formed through the adopter’s charac-
teristics (Rogers, 1983; Thong & Yap, 1995); PBC (Ajzen, 1985); SNs
(Fishbein & Ajzen, 1975); perceived ease of use and perceived usefulness
132 MD SHAH AZAM
(Davis, 1986); and performance expectancy and effort expectancy

(Venkatesh et al., 2003). Although intention, a reflection of the adopter’s
internal schema of beliefs, had a significant influence in earlier behavioural
models, it cannot predict the situation which is not under full volitional
control of the adopter or the situation where a time gap exists between
intention and actual behaviour (Venkatesh et al., 2008). However, a stron-
ger predictor of actual behaviour could incorporate some other external
(social and environmental) dimensions which would need to be considered
in the existing model. Venkatesh et al. (2008) further modified their

UTAUT model (Venkatesh et al., 2003) and included behavioural expecta-
tion as a new construct which is a stronger predictor of behaviour than
intention.
The field study did not find any differences between intention and expec-
tation. Furthermore, it proved that the SME owner who intended to adopt
a new technology did not simply dream about it nor was it unachievable.
The intention was formed by justifying the feeling through taking into
consideration several business and resource factors. For example, Firm A
indicated that: ‘[t]he supports and surrounding facilities screen the intention.
We are not in any doubt whether we could do so according to our intention.
We will upgrade the existing technology’.
Based on the previous literature (Venkatesh et al., 2008) and the field
study results, this study has hypothesised that behavioural expectation is a
possible predictor of actual ICT use. Thus,
H7. Expectation has a significant influence on actual ICT usage.
5.2.8. Hypotheses Related to ICT Use

ICT use is included as the final dependent variable in most of the theoreti-
cal frameworks which involve IT innovation diffusion behaviour (Davis,
1986; Taylor & Todd, 1995a; Venkatesh et al., 2003, 2008). Past studies
have employed ICT use as the final dependent variable based on the notion
that ICT use would generate performance. Researchers have applied differ-
ent aspects of IS use such as duration of use, frequency of use and intensity
of use to measure the actual behaviour in relation to any IT usage
(Venkatesh et al., 2008).
Some studies have employed different forms of IT usage such as
ICT/system use (Anandarajan et al., 2002; Davis, 1986; Taylor & Todd,
1995a; Zhu, Dong, et al., 2006; Zhu & Kraemer, 2005); EDI integration
(Iacovou et al., 1995); e-business initiation, adoption and routinisation
(Zhu, Kraemer, et al., 2006); and e-commerce migration (Hong & Zhu,
2006). Although the investigation of ICT use was found to be the focus of
a number of ICT innovation diffusion studies in the past few years, ICT’s
impact on performance growth has become a vital issue in recent years.
The literature has supported a direct and positive relationship between IT
usage by SMEs and firm performance (Zhu, Dong, et al., 2006; Zhu &
Kraemer, 2005).
Anandarajan et al. (2002) indicated a concern relating to the productivity
paradox and stated that the mere adoption of IT by organisations does not
necessarily confer on them the benefits that could only result from its effec-
tive usage. Past literature has enunciated that many IS in LDCs are under-
utilised and thus do not make a significant contribution to improving the
performance of the organisations that are using them (Forster & Cornford,
1992; Ordedra et al., 1993). In support of the previous studies, Song and
Mueller-Falcke (2006) stated that SMEs are confronted with a number of
challenges in adopting and using ICT and they often end up under-utilising
the available technologies (see UNCTAD, 2006). IS usage implies the
role that IS utilisation plays in generating organisational performance.
Zhu and Kraemer (2005) also indicated the positive and effective role of
ICT integration on organisational performance. Thus, the motivation for
organisational ICT adoption or usage is to integrate or properly utilise the
technology for performance growth.
The field study supported the view that the integration of diversified
technologies is dependent on the level of ICT usage by SMEs in
Bangladesh. For example, Firm K stated: ‘[a]fter that [introduction of ICT
use], the scope of ICT use has widened and strengthened. It seems that a new
technology is just introduced in some department and gradually it will spread
over the organisation and work under a common system’.
The field study further explored the view that the degree of ICT utilisa-
tion is also developed from the firm’s usage of different levels of ICT appli-
cations. For example, Firm B stated: ‘[we]e can earn productivity and
efficiency (organisational performance) through proper utilisation of ICT
particularly through some software by which we can track, control and initi-
ate alternative strategy … those companies [from whom] we are getting
benefits and gradually developing, they are actually developing due to the
proper utilisation of the technology (ICT). For example, in [the] interna-
tional context, some big companies are giving their employees ‘Blackberries’
so they (employees) can mail and be connected with the network from
anywhere around the world’.
134 MD SHAH AZAM
Based on the above discussion, it is hypothesised that:

H8a. ICT usage has a significant influence on integration.
H8b. ICT usage has a significant influence on utilisation.
H8c. ICT usage has a significant influence on performance.
5.2.9. Hypothesis Related to ICT Integration

The RBV (Barney, 1991) emphasises combining ICT resources to build

unique IT capabilities which would generate organisational performance.
The IT capabilities are dependent on how ICT is integrated and utilised
among different functional areas. Recent studies have shown that ICT
integration facilitates transactional efficiencies and expansion of existing
channels. Zhu and Kraemer (2002, 2005) revealed that integration (of the
front-end functionalities of e-business) helps firms to provide real-time
product information to customers, offers customisation capability and
facilitates self-service via online account management, thereby improving
transactional efficiencies and expanding the existing channels: ICT integra-
tion (back-end integration) also enables information sharing within the
firm and along the value chain.
The field study has justified the effects of ICT integration on organisa-
tional performance. For example, Firm B stated that: ‘[w]e have accumu-
lated some of our similar closely related group of companies into very close
contact between those are working with us in the same network environment
through our server. We are performing the above-mentioned work through
ICT (as we have got good results)’.
It (ICT integration) makes our communication and interaction with customers
easy.’(Firm J). Firm J also indicated the positive effect of ICT integration on organisa-
tional performance by saying, ‘[w]e find the ICT is working better when we see that [ICT
by integration] makes the communication with our supplier easy.
Based on past studies and the results of the field study, it is hypothesised
that:
H9. ICT integration has a significant influence on performance.
5.2.10. Hypothesis Related to ICT Utilisation

As with ICT integration, appropriate utilisation of ICT can contribute to
organisational ICT capabilities. Thus, according to the RBV, organisa-
tional performance also logically depends on the utilisation of ICT. Past
studies have revealed that many IS in LDCs are under-utilised and hence
do not make a significant contribution to improving the performance of
the organisations that are using them (Forster & Cornford, 1992; Ordedra
et al., 1993).
In analysing the field study interviews, it was anticipated that utilisa-
tion of ICT would influence organisational performance as the field study
had explored whether ICT utilisation has a direct positive link with orga-
nisational performance. For example, Firm H stated that: ‘[w]e are
trying to develop the IT-literate competent employee day by day. We hope

we shall be capable of appropriate and comprehensive use of ICT’. Firm I
expressed the importance of ICT utilisation on organisational perfor-
mance in terms of ensuring customer satisfaction when it stated that:
‘[i]f the ICT is not appropriate [appropriately utilised], we will not be
able to satisfy the customer or target group’. Firm J, agreeing with Firms
H and I, also stated the importance of ICT utilisation by saying: ‘[t]he
performance is dependent on how well [how appropriate] we will use [uti-
lise] ICT’.
The field interview analysis revealed that most of the organisations, such
as Firms H, I and J recognised the importance of ICT utilisation to secure
improved organisational performance. However, Firm K expressed their
strong concern about the erroneous process or bad organisational perfor-
mance which may happen if ICT is not properly utilised. The importance
of ICT utilisation on organisational performance is clearly reflected in the
following statement:
ICT enhances the employees’ performance. ICT normally reduces the error. But it var-
ies from person to person operating the ICT. It depends on their skill. I have a doubt if
the ICT is not properly handled [properly utilised], it may produce bad and damaging
output. (Firm K)
Based on the above discussion, it is thus hypothesised that:

H10. ICT utilisation has a significant influence on performance.
5.2.11. Hypothesis Related to the Mediation Effects of Integration and

Utilisation
Numerous studies have found a positive link between IT use and organisa-
tional performance (Baldwin & Sabourin, 2001; Cragg et al. 2002; Dvir
et al., 1993; Gretton et al., 2004; Pilat & Wyckoff, 2005). Although
achieving enhanced organisational performance is the motivation for orga-
nisations using ICT, past studies have also revealed differing results such as
research by Warner (1987) who found a negative effect; Venkatraman
136 MD SHAH AZAM
and Zaheer (1990) and Sager (1988) who found no effect; and Powell and
Dent-Micallef (1997) and Tippins and Sohi (2003) who found mixed effects;
while Bharadwaj (2000) and Zhu and Kraemer (2005) found direct positive
effects. Previous sections (Sections 5.2.8, 5.2.9 and 5.2.10) have analysed
ICT usage, integration and utilisation and have also hypothesised relation-
ships between them. Hypotheses H8a, H8b and H8c proposed that ICT
usage would have positive effects on integration, utilisation and organisa-
tional performance. On the other hand, Hypotheses H9 and H10 proposed
that integration and utilisation would have positive effects on organisa-

tional performance. The above-mentioned relationships may imply that
integration and utilisation have mediational effects which has not been
widely investigated in previous studies.
Kim et al. (2006) revealed the mediation role of inter-firm systems
integration in explaining the effects of SCCS innovation on firm supply
chain performance. Although the mediation effects of integration have
been empirically justified, there remains a paucity of empirical findings in
support of the role of utilisation. This study thus looked at the inter-
relationship of ICT usage, integration, degree of utilisation and organisa-
tional performance through the use of the field study analysis in which the
mediation effect of utilisation was found. For example, Firm I indicated
that: ‘[i]f the ICT is not appropriate [appropriately utilised], we will not be
able to satisfy the customer or target group’ and Firm J also stated the rela-
tionship by saying: ‘[t]he performance is dependent on how well [how appro-
priate] we will use [utilise] ICT’. The field study also justified the
mediating roles of both integration and utilisation. For example, Firm K
stated that: ‘… if the technology is used [utilised] completely in different
working areas in the entire organisation as well as if the external organisa-
tions are also integrated or communicated with entirely through ICT, the per-
formance [organisational performance] will certainly be increased’. Based on
the above discussion, it is hypothesised that:
H11a. Integration plays a significant mediation role in explaining the
effects of ICT usage on SME performance.
H11b. Utilisation plays a significant mediation role in explaining the

H11c. Integration and utilisation jointly play a significant mediation role

in explaining the effects of ICT usage on SME performance.
5.3. Questionnaire Development
The previous section has discussed the various constructs under study and
has anticipated possible relationships between them with the result that 10
sets of hypotheses (see Fig. 3) have been formulated. These hypotheses
were tested by the survey data. The data were collected by a questionnaire
survey in which the questionnaire contained a set of questions which speci-
fically examined different hypotheses developed for this study. The follow-
ing sections describe the questionnaire development.
5.3.1. Overview of the Questionnaire

This study collected data in two phases. The first phase was confined to col-
lecting qualitative data by interviews which were conducted using a semi-
structured questionnaire. The final phase involved data collection in order
to test the hypotheses. An extensive survey was administered through a
structured questionnaire which was designed to test the research hypotheses
according to the comprehensive model as shown in Fig. 2.
OI
INT
COMP
H2a H2c H8a
H2b H9
PEU
PU H1 H7 H8c
H10
H6a H6b H4a H8b FPERF
H4b
CUL FC H3b UTL
H5a
H5b H3a
PD ING UN BV EC
CR EP
CE = Cognitive evaluation, OI = Owner innovativeness, EXP = Expectation, EP = Environmental pressure, CUL = Culture,
CR = Country readiness, FC = Facilitating condition, USE = Actual usage, INT = Integration, UTL = Utilisation,
PERF = SME performance.
(First-order constructs, PU = Perceived usefulness, PEU = Perceived ease of use, PD = Power distance, ING = In-group
collectivism, UN = Uncertainty avoidance, BV = Bengali value, EC = Ethical culture, TI = Technology infrastructure,
HI = Human infrastructure, LI = Legal infrastructure, FI = Financial infrastructure, PS =Government policy and supports,
CP = Coercive pressure, MP = Mimetic pressure, NP= Normative pressure, GP = Global pressure,
COMP = Competitiveness, INTOP = Internal operation productivity, FPERF = Financial performance)
Fig. 3. The Comprehensive Research Model and Hypotheses.

138 MD SHAH AZAM
The questionnaire was developed from the existing relevant literature

and field study. The questionnaire included a wide range of questions speci-
fic to the study hypotheses. A variety of forms, patterns and scales were
used in different questions which included, where applicable, open-ended,
closed-ended, dichotomous, numerical, categorical questions and questions
with answers rated on a Likert-type scale. In order to assess respondents’
perceptions, both semantic differential scales and Likert scales were used.
The questionnaire was broadly divided into three main sections: questions
related to respondents’ demographic details; questions related to different

factors of ICT use and questions related to organisational performance.
The next section is divided into the following subsections: questions related
to cognitive evaluation; owner innovativeness; expectation; environmental
pressure; country readiness; culture; ICT use; integration and utilisation.
5.3.2. Measurement Instrument

This section discusses the development of the measurement instrument
which includes the form of the instrument and its sources. The main
sources of the instrument were the existing literature and the field study.
5.3.2.1. Cognitive Evaluation. In accordance with the theoretical frame-

work developed for this study (see Section 4), cognitive evaluation reflects
the adopter’s favourableness or unfavourableness towards an innovation
with this view formed through various attitudinal beliefs and respective
evaluations. Fishbein and Ajzen (1975) addressed various attitudinal beliefs
and their evaluations as Antecedents of attitude towards an innovation.
Many past studies have measured adopter favourableness or unfavourable-
ness (attitude) towards an innovation by a few reflective items such as
‘using the innovation is good/bad idea’, ‘the innovation makes work more
interesting’, ‘working with the innovation is fun’ and ‘I like/dislike working
with the innovation’ (Davis et al., 1989; Venkatesh et al., 2003). In measur-
ing the TPB constructs, Taylor and Todd (1995b) combined all belief items
with the evaluative components using the expectancy-value
P approach sug-
gested in the TPB
P (Ajzen, 1985, 1991, for example,
P i=1 b i e i for measuring
the attitude, j = 1 nbj mcj for the SN and k = 1 cbk pfk for PBC). Shih and
Fang (2004) followed this approach in predicting customers’ intention to
adopt Internet banking.
With reference to Ajzen (1991), Taylor and Todd (1995b) stated that the
relationship between the belief structures and the Antecedents of intention
(A [attitude], SN and PBC (or facilitating condition)) is particularly not
well understood due to two factors. Firstly, the belief structures are
P P
combined
P into unidimensional constructs (i.e. i = 1 bi e i , j = 1 nbj mcj ,
k=1 cb pf
k k ) in the theory. Such monolithic belief sets may not be consis-
tently related to attitude, SN or PBC (Bagozzi, 1981, 1982; Miniard &
Cohen, 1979, 1981, 1983; Shimp & Kavas, 1984). Secondly, the belief sets,
especially those relating to attitude, are idiosyncratic to the empirical
setting, making it difficult to operationalise the theory. In contrast to the
TPB, the TAM (Davis et al., 1989) proposed a belief set, consisting of ease
of use and usefulness that was consistent and generalisable across different
settings. Based on the above discussion, Taylor and Todd (1995b) recom-
mended a set of stable, decomposed belief structures for the TPB model.
The field study explored eight different items which measured SME
owner favourableness or unfavourableness towards ICT. Interestingly, all
eight measurement items belonged to two distinct factors. The past litera-
ture has recognised these two factors as perceived usefulness and perceived ease
of use (Davis, 1986, 1989; Taylor & Todd, 1995b; Venkatesh et al., 2003).
Guided by the field study and following the suggestion by Taylor and Todd
(1995b), this study combined all eight reflective items to measure cognitive
evaluation. Applying the principle of hierarchical modelling molecular
model (Chin & Gopal, 1995) perceived usefulness and perceived ease of
use were used as reflective first-order constructs of a hybrid conceptualisa-
tion called ‘cognitive evaluation’. The measurement items of the first-order
constructs, although developed from the field study, were also supported
by the existing literature. Items for the first-order manifest variables of the
higher-order cognitive evaluation and their sources are stated in Table 9.
5.3.2.2. Owner Innovativeness and Facilitating Condition. Owner innova-

tiveness was modelled as an antecedent of ICT adoption by SMEs. Thong
and Yap (1995) adopted the risk-taking ability of the owner and the own-
er’s ‘ability to produce original ideas’ to face the competitive environment
as measurement items for owner innovativeness. The field study explored
two items for measuring owner innovativeness which were similar to the
measurement items adopted by Thong and Yap (1995) (see Table 10).
This study modelled facilitating condition as an antecedent factor of
expectation as well as ICT usage. Past studies have adopted a few items
related to the adopter’s assessment of their firm’s ability and resources,
such as, assessing whether the adopter has the resources necessary to use
the innovation, if the adopter has the knowledge necessary to use the inno-
vation, if the innovation is incompatible with other systems, and if a speci-
fic person (or group) is available for assistance with any operational
difficulties (Venkatesh et al., 2003; Venkatesh & Zhang, 2010).
140 MD SHAH AZAM
Table 9. Cognitive Evaluation.

Items Statement Sources
Ease of use
PEU1 My interaction with the technology would be Field study, Venkatesh et al. (2003)
clear and understandable
PEU2 It would be easy for me to become skilful at Field study, Venkatesh et al. (2003)
using the technology
PEU3 I would find the technology easy to use Field study, Venkatesh et al. (2003)
PEU4 Learning to operate the technology is easy for me Field study, Venkatesh et al. (2003)
Usefulness
PU1 I would find ICT useful in my job Field study, Venkatesh et al. (2003)
PU2 Using the technology enables me to accomplish Field study, Venkatesh et al. (2003)
tasks more quickly
PU3 Using the technology increases my productivity Field study, Venkatesh et al. (2003)
PU4 If I use the technology, I will increase my Field study, Venkatesh et al. (2003)
chances of getting a raise
The field study explored the items relating to the existing systems or pro-
cesses of the organisation (i.e. whether the systems or processes were com-
patible with ICT usage), the resources of the organisation (whether the
organisation had the resources necessary to use the systems) and the skill
of the employees (whether employees had the knowledge and skills neces-
sary to use the systems). The measurement items for facilitating condition
explored through the field study were similar to the measurement scales
adopted by Venkatesh et al. (2003). Measurement items for owner innova-
tiveness and facilitating condition along with their sources are presented in
Table 10.
5.3.2.3. Environmental Pressure. This study modelled environmental pres-

sure as an antecedent of behavioural expectation and usage of ICT by
SMEs with a view to examining the effects of ICT on organisational perfor-
mance. The institutional theory (DiMaggio & Powell, 1983) suggests that
organisations face three dimensions of environmental pressure, namely,
coercive pressure, mimetic pressure and normative pressure. Past studies
have examined the effects of one particular dimension or the joint effects of
these three dimensions to estimate the impacts of environmental pressure
on organisational ICT adoption behaviour. For example, Teo et al. (2003)
studied coercive pressures, mimetic pressures and normative pressures, Zhu
et al. (2003) studied competitive pressures which are known as mimetic
pressures, and Quaddus and Hofmeyer (2007) studied coercive pressures.
Table 10. Items for Owner Innovativeness and Facilitating Condition.

OI1 He often risks doing things differently Field study, Thong and Yap (1995)
OI2 He has original ideas Field study, Thong and Yap (1995)
FC1 The ICT system is compatible with other existing Field study, Venkatesh et al. (2008)
systems that we use
FC3 Our employees have the knowledge and skill Field study, Venkatesh et al. (2008)
necessary to use the system
FC4 We have the resources necessary to use the system Field study, Venkatesh et al. (2008)
Past studies have employed various scale items to measure different

dimensions of environmental pressure. For example, Teo et al. (2003) used
seven subconstructs to measure the three dimensions of environmental
pressure. Specifically, they adopted the subconstructs, extent of adoption
among competitors and perceived success of competitor adopter for
mimetic pressures; perceived dominance of suppliers, perceived dominance
of adopter customers, conformity with parent corporation’s practices for
coercive pressures and extent of adoption among suppliers, extent of adop-
tion among customers and participation in industry, business and trade
associations for normative pressures. As they were also latent variables,
these subconstructs were measured by multiple indicators (Teo et al., 2003).
In contrast to predictive estimates, the dimensions of environmental pres-
sure were directly measured by multiple items such as coercive pressure
being measured by the adopter’s assessment on whether the parent com-
pany liked the adopting firm’s use of ICT, whether major customers
demanded firms’ ICT use and whether major suppliers demanded firms’
ICT use (Quaddus & Hofmeyer, 2007); and competitive pressures were
measured by the adopter’s assessment about the degree affected by compe-
titors in the local market and the degree affected by competitors in the
national market (Zhu & Kraemer, 2005).
The field study explored 11 items which belonged to four distinct groups.
Three of the four constructs, namely, coercive pressure, mimetic pressure
and normative pressure (Teo et al., 2003), were similar to the dimensions
adopted in previous studies. The field study explored a new dimension of
environmental pressure named global pressure which had arisen from the
ongoing initiatives of globalisation.
In accordance with the results of the field study analysis, coercive pres-
sure was measured by the extent to which dominant customers and parent
142 MD SHAH AZAM
organisations forced SMEs to use the technology. Mimetic pressure was

measured by items relating to the extent to which competitors used ICT
and to those competitors who had largely benefited from their use of
ICT. Normative pressure was measured by items relating to the extent of
ICT use by customers, by suppliers, by important stakeholders or business
partners and by regulatory bodies. The pressure relating to globalisation
was measured through items which enquired about the extent to which the
organisation used ICT to attain the business opportunities arising from
globalisation, and about the extent to which the organisation believed that
no business at present could be sustained without utilisation of ICT. All of
the above-mentioned measurement items were supported by the existing
literature except the items for global pressure which were explored from the
field study.
Importantly, the final research model included overall environmental
pressure as an antecedent of SMEs’ expectation to use ICT. In accordance
with the results of the field study, the effects of overall environmental pres-
sure were measured by the four dimensions: coercive pressure, mimetic
pressure, normative pressure and global pressure. These four dimensions
were used as index of the higher-order conceptualisation. Past studies used
composites of various dimensions to measure a compound variable; for
example, Taylor and Todd (1995b) used a composite of peer influences and
superior influences to measure normative pressure. Analysis of the field study
findings, and particularly the interpretation of the free node and tree node
structures, suggested the application of a molar model structure (Chin &
Gopal, 1995) to measure overall environmental pressure as a second-order
construct developed by these four constructs instead of through aggrega-
tion or averaging their effects. Table 11 presents contents and sources of all
measurement items for the first-order latent variables of the higher-order
environmental pressure.
5.3.2.4. Country Readiness. This study modelled country readiness as an

antecedent of behavioural expectation and usage of ICT by SMEs. Past
studies have addressed various environmental factors related to the prepa-
redness of the country which may impact on fostering the usage of ICT by
organisations, particularly SMEs. Important country readiness dimensions
explored in the existing literature included national infrastructure (Huy &
Filiatrault, 2006; Kurnia & Peng, 2010); regulatory environment (Zhu
et al., 2003); government supports (Jeon et al., 2006); technology vendors
supports (Doolin, McLeod, McQueen, & Watton, 2003) and technology
readiness, country wealth and level of education (Marques et al., 2011).
Table 11. Items for Environmental Pressure.

Coercive pressure
CP1 My parent company likes us to use ICT Field study, Quaddus and
Hofmeyer (2007)
CP2 Our major customer demanded our ICT use Field study, Quaddus and
Hofmeyer (2007)
CP3 Our major suppliers demanded our ICT use Field study, Quaddus and
Hofmeyer (2007)
Mimetic pressure
MP1 What is the extent of ICT adoption by your firm’s Field study, Teo et al. (2003)
competitors currently?
MP2 My main competitors that have adopted ICT have Teo et al. (2003)
benefited largely
Normative pressure
NP1 What is the extent of ICT adoption by your firms’ Field study, Teo et al. (2003)
customers currently?
NP2 What is the extent of ICT adoption by your suppliers Field study, Teo et al. (2003)
currently?
NP3 What is the extent of ICT adoption by your important Field study, Quaddus and
business partner currently? Hofmeyer (2007)
NP4 What is the extent of ICT adoption by your regulatory Field study
bodies?
Global pressure
GP1 I am encouraged to adopt ICT in accessing Field study
opportunities arising by globalisation
GP2 I feel this is the time when nobody can sustain [their Field study
business] without utilising ICT
The field study explored 16 items to measure the impact of country

readiness. During the analysis, those 16 items were further grouped into
five dimensions of country readiness, namely, technology infrastructure,
human infrastructure, legal infrastructure, financial infrastructure and gov-
ernment policy and supports (see Table 12).
The availability and coverage of Internet services in Bangladesh,
Internet speed, availability of necessary hardware and uninterrupted power
supply were the items explored by the field study to measure technology
infrastructure.
Items that measured human infrastructure related to the general people’s
level of education, knowledge of ICT, ICT skill and institutional supports
for ICT training and education.
144 MD SHAH AZAM
Table 12. Items for Country Readiness.

Technology infrastructure
TI1 Availability and coverage of Internet service in the country Field study
TI2 Speed and sophistication of Internet Field study
TI3 Availability of necessary hardware and accessories Field study
TI4 Uninterrupted power supply Field study
Human infrastructure
HI1 Level of education of the general people Field study

HI2 General people’s knowledge of ICT Field study
HI3 Computing skill of the general people Field study
HI4 Institutional support for computer education Field study
Legal infrastructure
LI1 Legal supports for digital communication and business Field study
through the Internet
LI2 Law relating to online consumer protection Zhu and Kraemer (2005)
Financial infrastructure
FI1 Online banking facility Field study
FI2 Credit and debit card usage trend Zhu and Kraemer (2005)
Government policy and supports
GOV1 Government tax and customs policy Field study, Zhu and
Kraemer (2005)
GOV2 Government’s motivational programme Field study
GOV3 Government grants Field study
GOV4 Government subsidies Field study
The measurement items for legal infrastructure were legal protection for
digital communication and online business, and online consumer protection.
Financial infrastructure was measured by items related to the extent to
which online banking services are available and trend of debit or credit
card use in the country.
Government policy and supports was one of the major constructs
explored through the qualitative field study. The measurement items were
government’s tax and customs policy, motivational programme, govern-
ment grants and supports.
By applying an interpretive research approach, a careful analysis of the
field study, particularly the structures of theme, sub-themes and nodes,
suggested that technology infrastructure, human infrastructure, legal infra-
structure, financial infrastructure, and government policy and supports were
subconstructs for measuring the construct of interest country readiness.
Past studies have used composites of various dimensions to measure a
compound variable, for example, Taylor and Todd (1995b) used a compo-
site of efficacy and facilitating condition technology and facilitating con-
dition resources to measure PBC.
As with the formation of the construct environmental pressure, the field
study suggested that country readiness be formed as a higher-order con-
struct by applying the molar model structure (Chin & Gopal, 1995).
Measurement items for all first-order constructs and their sources are
shown in Table 12.
5.3.2.5. Culture. Culture has been variously defined in the past literature.
Kroeber and Kluckhohn (1952) considered 160 different definitions of cul-
ture while attempting to develop an acceptable general definition of culture.
The variations in conceptualisation of the term ‘culture’ have also been
widened in the past 60 years.
Contemporary studies have followed Hofstede’s (2001) framework to
conceptualise national culture (Erumban & Jong, 2006). A more recent
framework for cultural study developed by the GLOBE cultural study
(House et al., 2004) has received numerous researchers’ attention with this
also having been developed based on Hofstede’s (2001) framework.
However, Hofstede’s (2001) and House et al.’s (2004) frameworks have
been the most recent developments in the study of culture.
In Hofstede’s (2001) model, culture is conceptualised by five dimensions:
power distance, collectivism, uncertainty avoidance, masculinity versus
femininity and long-term versus short-term orientation while House et al.
(2004) explained culture through nine dimensions: performance orientation,
assertiveness, future orientation, humane orientation, institutional collectivism,
in-group collectivism, gender egalitarianism, power distance and uncertainty
avoidance.
Numerous past studies, investigating the effects of culture in the field of
management, have considered a portion of Hofstede’s (2001) or House
et al.’s (2004) models as being relevant to their studies (Mustamil, 2010;
Wahab, 2010).
This field study explored 14 items to measure the impact of culture.
Through analysis, the field study grouped these 14 items into five dimen-
sions of culture. Three of the five dimensions of culture that may affect the
diffusion of IT among SMEs in Bangladesh and that were explored
through the field study were related to the dimensions explained by
Hofstede (2001) and House et al. (2004). The field study explored two new
dimensions of culture which were Bengali values and ethical behaviour-
related cultural issues. Although these aspects of culture were beyond the
146 MD SHAH AZAM
scope of Hofstede’s (2001) and House et al.’s (2004) frameworks, it was

considered that they may play an influential role in shaping SMEs’ ICT
diffusion behaviour. This study, thus, included five dimensions power
distance, collectivism, uncertainty avoidance, Bengali values and ethical
culture to measure the effects of culture.
As with environmental pressure and country readiness, culture was also
measured as a higher-order construct by applying the molar model struc-
ture (Chin & Gopal, 1995). Items measuring the first-order cultural dimen-
sions for culture and their sources are stated in Table 13.
5.3.2.6. Expectation and ICT Usage. This study modelled expectation and
ICT usage as antecedents of ICT usage and organisational performance,
respectively. Expectation was included in the research model to indicate the
state of rational intention towards adoption of ICT. Past studies have
applied a variety of measurement items for intention and expectation as
well as for actual usage behaviour (Davis, 1989; Kendall et al., 2001;
Venkatesh et al., 2003, 2008). For example, Davis (1989) applied three
items to measure intention, namely; I plan to use the system in the next
<n > month, I predict to use the system in the next <n > month and I am
going to use the system in the next <n > month. On the other hand,
Kendall et al. (2001) measured Singaporean SMEs’ e-commerce adoption
intention by assessing their intended state of various e-commerce applica-
tions. The intention was measured using a 6-point scale by categorising six
different intended states of e-commerce applications; 1 = Current year, 2 =
Intend to use within 1 year, 3 = Within 12 years, 4 = Within 25 years,
5 = Within 510 years and 6 = No intention. The order of these intended
states was applied to assess various levels of e-commerce. The construct
was estimated as a latent variable comprising multiple items, for example,
willingness to have a homepage in the company, willingness to conduct
sales through Internet, willingness to purchase the supplies through
Internet, etc.
Venkatesh et al. (2008) differentiated the expectation from the intention
and applied a relatively expressed state of intention for measuring expecta-
tion such as I will use the system in the next <n > month, I am likely to use
the system in the next <n > month and I am going to use the system in
the next <n > month. Interestingly, this current study did not find any
differences between intention and expectation (see the field study analysis
in Section 4). The field study analysis revealed that this state of willingness
was different from the ‘intention’ that was used in past studies. More
clearly, SMEs’ intention to use any level of ICT was not only a cognitive
Table 13. Items for Culture.

Power distance
PD1 In this society, followers are expected to obey their leader House et al. (2004)
without question/question their leader when in disagreement
PD2 In this society, power is concentrated at the top/shared House et al. (2004)
throughout the society
PD3 I believe that followers should obey their leader without House et al. (2004)
question/question their leader when in disagreement

PD4 I believe that power should be concentrated at the top/shared House et al. (2004)
throughout the society
UN1 In this society, orderliness and consistency are stressed, even at House et al. (2004)
the expense of experimentation and innovation
UN2 In this society, social requirements and instructions are spelled House et al. (2004)
out in detail so citizens know what they are expected to do
UN3 In this society, orderliness and consistency should be stressed, House et al. (2004)
even at the expense of experimentation and innovation
UN4 In this society, social requirements and instructions should be House et al. (2004)
spelled out in detail so citizens know what they are expected
to do
In-group collectivism
INGR1 Group members take pride in the individual accomplishments House et al. (2004)
of their group managers
INGR2 Group managers take pride in the individual accomplishments House et al. (2004)
of their group members
INGR2 Managers encourage group loyalty even if individual goals House et al. (2004)
suffer
Bengali values
BVAL1 Face-to-face communication is vital in our daily lives Field study
BVAL2 In honour of the language movement, the Bengali language is Field study
available everywhere
BVAL3 Very close and tight social bond Field study
Ethical culture
ECUL1 Bribery/corruption are seen as common phenomena everywhere Field study
in the main institutions of society
ECUL2 Politicisation/nepotism are seen as common phenomena Field study
everywhere in the main institutions of society
wish or simply a plan or a dream, it was definite and rational. Based on the
field study results, this study used ‘expectation’ as an logical immediate
antecedent of actual usage behaviour: expectation was directly influenced
by various individual, organisational and environmental factors.
148 MD SHAH AZAM
Furthermore, the previous innovation diffusion studies have mostly

focused on a particular technology or systems usage such as Internet, email,
personal computers, e-commerce and EDI, ; thus assessment using items
like I plan …, I will … or I predict … may be logical grounds from which
to measure potential users’ intentions. This study has focused on the diffu-
sion of ICT in general which has comprised a range of ICT applications.
The field study explored various levels of ICT applications that were being
currently used by different SMEs in Bangladesh. To overcome possible
ambiguity in construct measurement, these ICT applications were cate-

gorised into five distinct levels of ICT operation: basic computing which
includes computer and basic Internet operation such as email; homepage
operation (static homepage); interactive homepage which supports product
cataloguing and order processing (e-commerce); interactive homepage
which supports online transactions and account management (e-business)
and complete digital communication and exchanges within and outside the
organisation (ERP). Applying the measurement approach adopted by
Kendall et al. (2001), respondents’ assessments about their expected use of
those ICT applications were used as measures for expectation. Various
states of expectation were assessed through a differential scale: 1 = current
user; 2 = expect within 6 months; 3 = within 12 months; 4 = within 13
years and 5 = expectation is not specified.
As with intention, actual usage was measured variously in previous
studies. Venkatesh et al. (2003) measured actual usage behaviour as the
duration of use via system logs; Venkatesh et al. (2008) used observed scale
items including duration of use, frequency of use and intensity of systems
use to measure IS use; and Zhu and Kraemer (2005) measured e-business
usage as a latent variable of multiple observed variables such as percentage
of consumer sales conducted online, percentage of B2B sales conducted
online and percentage of goods for resale ordered online. For assessing
actual ICT usage, this study used a similar set of questions for expectation.
While measuring actual ICT usage, respondents were asked about how
often they used those five different levels of ICT applications. A semantic
differential scale was used to collect details of respondents’ actual ICT
usage with ‘5’ representing ‘use quite often’ and ‘1’ denoting ‘not used at
all’. The measures for expectation and actual ICT usage and their sources
are stated in Table 14.
5.3.2.7. Integration and Utilisation. This study modelled integration and

utilisation as antecedents of organisational performance. Integration and
utilisation were also predicted to be intervening variables between ICT
Table 14. Items for Expectation and ICT Usage.

Expectation
Scale: 1 = current user; 2 = expect within 6 months; 3 = within 12 months;
4 = within 13 years, 5 = expectation is not specified.
EXP1 ICT basic operation (computer, email, web surfing, etc.) Field study
EXP2 Internet with own homepage (static homepage) Field study
EXP3 Interactive homepage which supports product cataloguing and order Field study
processing
EXP4 Interactive homepage which supports online transactions and account Field study
management
EXP5 Complete digital/electronic communication and exchanges within and Field study
outside the organisation
ICT use
Scale: 5 = use quite often; 1 = not used at all.
USAGE1 ICT basic operation (computer, email, web surfing, etc.) Field study
USAGE2 Internet with own homepage (static homepage) Field study
USAGE3 Interactive homepage which supports product cataloguing and order Field study
processing
USAGE4 Interactive homepage which supports online transactions and account Field study
management
USAGE5 Complete digital/electronic communication and exchanges within and Field study
outside the organisation
usage and organisational performance. Past studies used integration as a

latent variable which was measured through multiple items, for example, to
what extent company web applications are electronically integrated with
back-office systems and databases and to what extent company databases
are electronically integrated with suppliers and partners (Zhu & Kraemer,
2005).
Although a few studies stated the importance and consequences of ICT
utilisation (particularly under-utilisation) in generating enhanced organisa-
tional performance from a developing country perspective, there was a pau-
city of empirical studies which quantified the variable, degree of utilisation,
and measured its effects on organisational performance. The field study
supported the importance of ICT integration in enhancing SMEs’ perfor-
mance and also reiterated the importance of the degree of utilisation of
ICT in different functional areas such as production (or services), adminis-
tration and accounts, as well as marketing and sales.
Combining the field study results with the relevant literature, integration
was measured by the respondents’ assessment of: (i) ICT is utilised in entire
150 MD SHAH AZAM
working areas in the organisation; (ii) all departments and functional areas
of this organisation are integrated through a single ICT system; (iii) our
website is well developed for front-end functionality which supports
information, product catalogue, customer customisation and account
management; (iv) our web applications are electronically integrated with
back-office systems and databases and (v) company’s databases are electro-
nically integrated with suppliers and partners. The respondents’ assessment
was quantified using a 5-point Likert scale.
Due to the paucity of literature, measurement items for the degree of uti-
lisation were explored from the field study. The degree of utilisation was
measured by the respondents’ assessment of the extent of appropriate utili-
sation of ICT into various functional areas such as production, administra-
tion and accounts, as well as marketing and sales. A differential scale was
employed to measure utilisation with the measures being: 1 = ICT not at
all utilised; 2 = utilised to conduct 20% of functions; 3 = utilised to con-
duct 2050% of functions; 4 = utilised to conduct 5080% of functions
and 5 = utilised to conduct more than 80% of functions. Measurement
items for integration and utilisation as well as their sources are given in
Table 15.
5.3.2.8. Performance. This study has examined the diffusion of ICT by

SMEs in Bangladesh: it has extended the scope of the traditional
Table 15. Items for Integration and Utilisation.

Integration
INT1 ICT is utilised in entire working areas in the organisation Field study
INT2 All departments and functional areas of this organisation are Field study
integrated through a single ICT system
INT3 Our website is well developed for front-end functionality which Field study, Zhu and
supports information, product catalogue, customer Kraemer (2005)
customisation and account management
INT4 Our web applications are electronically integrated with back- Field study, Zhu and
office systems and databases Kraemer (2005)
INT5 Company’s databases are electronically integrated with Field study, Zhu and
suppliers and partners Kraemer (2005)
Utilisation
UTL1 Utilisation of ICT in production and service-related functions Field study
UTL2 Utilisation of ICT in admin and accounts-related functions Field study
UTL3 Utilisation of ICT in sales and marketing-related functions Field study
innovation diffusion model by including two stages of post-adoption phe-

nomena relating to ICT, namely, ICT use and value creation in terms of
organisational performance. Previous studies have measured organisational
performance in various ways (Beard et al., 2014; Bharadwaj, 2000; Madden
et al., 2013; Powell & Dent-Micallef, 1997; Zhu & Kraemer, 2005). The
measurement items for organisational performance used in this study have
been mainly adapted from prior studies such as those by Zhu and Kraemer
(2005) and Powell and Dent-Micallef (1997). In addition to the items
adapted from the existing literature, the field study also explored new items
for measuring organisational performance. The results of the field study
provided support for items previously used for organisational performance
measurement and also suggested a second-order latent variable measure-
ment structure.
The field study analysis revealed nine items for measuring organisational
performance which were further categorised into three distinct groups,
namely, competitiveness, internal operation productivity and financial
performance. The overall organisational performance was measured as
a higher-order conceptualisation comprising these three performance
dimensions. Prior research has supported the measurement of organisa-
tional performance as a second-order construct (Rai, Patnayakuni, & Seth,
2006; Zhu & Kraemer, 2005; Zhu et al., 2003). This study employed higher-
order modelling and measured organisational performance as a composite
of first-order constructs: competitiveness, internal operation productivity
and financial performance. All items for the first-order latent constructs
of organisational performance and their sources are stated in Table 16.
5.4. Questionnaire Translation
It was stated in previous sections that the main objective of the study was
to look at the diffusion of ICT and its resultant outcome as organisational
performance in the Bangladeshi context. The owners of different SMEs
were the subjects of the study: they were comfortable with and used to
communication and exchanges in the Bengali language. Thus, the English
version questionnaire was translated into Bengali before conducting
the survey to achieve greater convenience and ease of operation. A
complete English version of the survey questionnaire has been provided in
Appendix C.
A back-translation method (Marin & Marin, 1991) was employed to
create an accurate Bengali-version questionnaire and ensure equivalence of
152 MD SHAH AZAM
Table 16. Items for Performance.

Competitiveness
COMP1 Interaction with customer has been Field study, Zhu et al. (2004), Zhu and
increased Kraemer (2005)
COMP2 Our sales area has been widened Field study, Zhu et al. (2004)
COMP3 Competitive position has been improved Field study
Internal operation productivity
INTOP1 The internal process of the organisation Field study

became transparent
INTOP2 The internal process of the organisation Field study, Zhu et al. (2004)
became structured
INTOP3 Productivity of the employees has been Field study, Zhu et al. (2004)
improved
PERF1 Overall performance is increased Field study, Powell and Dent-Micallef
(1997)
PERF2 Profitability of the company has been Field study, Powell and Dent-Micallef
increased (1997)
PERF3 Sales of the company have been increased Field study, Zhu and Kraemer (2005),
Powell and Dent-Micallef (1997)
the two versions (see Section 3). The back-translation process underwent a
series of translations, checkrecheck and modifications. The researcher
first translated the original English questionnaire, which had been
approved by the university’s Human Research Ethics Committee, into
Bengali. The first version of the Bengali questionnaire was thoroughly
checked by a university academic in Australia who was a native
Bangladeshi. This refined version of the translated questionnaire was
further reviewed by two university academics from Bangladesh who were
involved in research in Australia. The final version of the Bengali question-
naire was again translated back into English. A third person, a researcher
from Bangladesh, performed the back-translation of the questionnaire into
English. This back-translated version of the questionnaire provided the
opportunity to check whether the translated version of the items projected
a similar meaning and approach to the original version. Interestingly,
although some words were found to be different, all items in both versions
of the questionnaire were observed to be similar in their meaning which
finally ensured equivalence of the two versions of the questionnaire. A com-
plete Bengali-translated version of the survey questionnaire has been
provided in Appendix D.
5.4.1. Questionnaire Refinement and Pilot Test

The initial questionnaire was refined through pre-testing prior to adminis-
tering the actual survey (as described in Section 3). The primary version of
the questionnaire was distributed to five respondents in Australia who had
direct experience in SMEs in Bangladesh. The feedback from these five
respondents was used to modify the questions for the final survey. The
questionnaire was translated into Bengali before being used to conduct the
final survey.
Although the questionnaire had been refined and modified through pre-
testing, a pilot test was also conducted by administering a mini sample sur-
vey with the final version of the questionnaire. A total of 35 SMEs were
surveyed after being selected through a convenience sampling technique.
Some errors and mistakes related to different questions were detected
and a few suggestions were received during the pilot test. The errors and
mistakes were mainly associated with: (i) ambiguous meaning for some
words; (ii) some questions being very close in their meaning creating confu-
sion for respondents; (iii) difficulties experienced in indicating their appro-
priate evaluation on the Likert scale and (iv) the question’s approach not
being simple and easy.
Despite the errors, the data collected through the pilot survey were
recorded and the study model was estimated. The analysis also indicated
some loopholes in the questionnaire. For example, the factor analysis of the
data collected through the pilot survey showed that the factor loadings of
some constructs were ‘1’. The suggested reason for this was that when
respondents were unable to differentiate the level of their evaluation in
answering the questions, they mostly chose the extreme value. Thus, the fac-
tor loading of different items of some constructs tended to reach a value of
‘1’. Based on feedback from respondents during the pilot survey and errors
detected through analysing the pilot survey data, the entire questionnaire
was revised to ensure that it was a valid and reliable survey questionnaire. It
is worth noting that the incorporation of all respondents’ suggestions and
the analysis results led to slight modification of the final version of the ques-
tionnaire by replacing some words and making changes in the expression
of some questions which mostly involved simplification. The final version
of the questionnaire was then used to collect data from a national survey.
5.5. Summary
This section has presented the construction of the hypotheses in accordance

with the final research model developed in Section 4. To analyse the
154 MD SHAH AZAM
comprehensive research model (see Fig. 2), 18 hypotheses were developed

to describe the relationships between the study variables (see Fig. 3).
Furthermore, three hypotheses were developed to analyse the mediating
roles of ICT integration and degree of utilisation to explain the influence of
ICT usage on organisational performance. The development of the mea-
surement scales and items was also presented. Next, the survey question-
naire was developed based on the existing literature and field study results.
The back-translation method was employed to translate the instrument for
the survey. Pre-testing and a pilot study were used to validate the question-
naire. The final questionnaire played the vital role in the data collection by
a national survey which is discussed in Section 6.
6. DATA ANALYSIS
6.1. Introduction
This section presents the quantitative data analysis procedures and reports
on the results. In particular, this section presents the procedure for and
results from estimating the structural model using the PLS-SEM) by
applying PLS-Graph Version 3, as well as reporting on the validation of
the psychometric properties of the measurement model. Factor loadings,
t-statistics and the composite reliability of the reflective constructs are pre-
sented as proof of convergent validity, while average communalities, the
correlation matrix and cross-loading matrix are also provided as proof of
discriminant validity among the reflective constructs. The relative impor-
tance of the formative items towards the formation of the related latent
construct is also appraised. The assessment of the structural model com-
menced once the measurement model had been evaluated and adjusted.
Estimates of path coefficients, t-statistics and R2 values are presented to
determine the degrees and magnitudes of the effects of exogenous variables
and explanatory power of the model. This section concludes by describing
the results of the hypotheses testing.
6.2. Overview of Survey and Data Examination
The study analysed 282 survey responses collected from different SMEs in
Bangladesh. The response rate was calculated as 22% for the samples from
the RMG industry and 55% for the leather and leather product industry.
(PLS-SEM) was used as the key analytical tool for the estimates. The repre-
sentativeness or generalisability of the research outcomes was largely
dependent on the sample size. A reasonable number of observations could
generate reliable results. The size of the sample was considered adequate
for executing SEM for the estimates as it satisfied different sample size spe-
cifications as suggested by Hair et al. (1998), Gefen et al. (2000), Barclay
et al. (1995) and Chin (1998a).13
The theoretical model of the study consisted of 25 first-order constructs

and five higher-order constructs among which six constructs were endogen-
ous. The sample size should be at least 60 (6 × 10 = 60) according to Gefen
et al. (2000): in the model, the number of indicators within the most com-
plex formative construct was six. Another recommendation was that it
should be at least 50 (5 × 10 = 50) according to Barclay et al. (1995) and
Chin (1998a) as the largest number of antecedent constructs leading to an
endogenous construct as predictors in a regression was five. The study was
finally run with 282 sample units for data analyses which could be consid-
ered sufficient for a robust PLS model.
6.2.1. Sample Profile

As the study utilised a disproportionate stratified probability sampling
technique, different types of SMEs from the manufacturing industry sector
were included in the sample which brought in logical grounds for generalis-
ing the inferred outcomes.
The study investigated the opinions and perceptions of the owner,
owner-manager, manager (decision-maker) or delegated representative of
SMEs located in or adjacent to the city of Dhaka. Dhaka was selected as
the sampling area after taking into consideration the fact that a high con-
centration of industry was evident in or near Dhaka. Internet penetration
was also high in Dhaka comprising approximately 80% of Bangladesh’s
total Internet users. The sample consisted of 79.2% from the RMG indus-
try and 20.2% from the leather industry (see Table 17).
Of the firms surveyed, 96.1% had their own homepage, 32.3% had
a product cataloguing-capable homepage, 17.7% have experience in
e-commerce and 0.4% had an ERP system. The study included 18.1% of
firms who enjoyed preferences as SMEs whereas only 7.1% had received
any grants or subsidies for ICT.
Furthermore, the survey comprised 85.1% male and 14.9% female
respondents (see Table 18). The sample also included respondents of
156 MD SHAH AZAM
Table 17. Survey Firm Characteristics.

Description F %
Sector
Ready-made garment industry 225 79.8
Leather industry 57 20.2
Size
Small business 174 61.7
Medium business 108 38.3
Level of ICT use

Homepage 271 96.1
Online cataloguing 91 32.3
E-commerce 50 17.7
ERP 1 0.4
SMEs’ preferences and grants
Received preferences as an SME 51 18.1
Received subsidies or ICT grants 20 7.1
Nature of customer
Geographically diverse customers 234 83.0
Customers not geographically diverse 48 17.0
Internet connectivity
Dial-up/DSL 29 10.3
Broadband (cable and mobile) 236 83.7
Mobile broadband 17 6.0
Operating offices or stores
Single office or store 11 3.9
Two offices or stores 238 88.4
More than two offices or stores 33 11.7
ICT experience
Started ICT before 2000 68 24.1
Started ICT since 2000 214 75.9
different ages, levels of education, levels of income and ICT experience

which enhanced the representativeness of the sample data.
As shown in Table 18, 42.6% of respondents were aged below 34 years,
37.2% were 3544 years and 20.2% were aged 45 years or above. In terms
of education, 35.8% of respondents had Masters’ degrees, 47.5% had
bachelor degrees and 5.7% had higher secondary certificates: 11% had
other academic qualifications.
The study also included 17% of respondents with monthly income below
20,000 BDT, 40.4% between 20,000 BDT and 30,000 BDT, 21.3% between
30,000 BDT and 50,000 BDT, 17.0% between 50,000 BDT and 100,000
BDT while 4.3% of respondents had monthly income over 100,000 BDT.
Overall, 95.7% of respondents had prior ICT experience.
Table 18. Survey Respondent Characteristics.

Description F %
Gender
Male 240 85.1
Female 42 14.9
Age
Age 34 or below 120 42.6
Age 3544 years 105 37.2
Age 4560 years 54 19.1

Age Over 60 years 3 1.1
Education
Master’s degree 101 35.8
Bachelor degree 134 47.5
Higher secondary certificate 16 5.7
Others 31 11.0
ICT experience
Have ICT experience 270 95.7
No ICT experience 12 2.3
Monthly income
Below 20,000 BDT 48 17.0
20,00030,000 BDT 114 40.4
30,00050,000 BDT 60 21.3
50,000100,000 BDT 48 17.0
Over 100,000 BDT 12 4.3
6.2.2. Data Examination

Prior to analysis, the data were first screened for outlier checking and
missing values. Alreck and Settle (1995) recommended an extensive data
clean-up process which involves the review of the data line by line to check
for any errors due to missing or irrational data. In conjunction with review-
ing the data line by line, this study adopted frequency tables and the
Mahalanobis distance test for outlier checking. During screening, some
missing values were found. As the number of missing cells was small in size
considering the number of variables as well as the number of records, a
different approach for missing value adjustment, involving a re-visit to the
respondents, was employed in the belief that precaution was better than
cure. However, there still remained some missing values even after
re-visiting the respondents. Finally, an ML (maximum likelihood) algo-
rithm was used for the imputation of the remaining missing values. The
careful review and screening of the raw data resulted in 282 complete and
usable data records for the final data analysis.
158 MD SHAH AZAM
The theoretical model was composed of 25 first-order constructs which

comprised 12 reflective and 13 formative constructs. The model also
included five second-order constructs which comprised two reflective
higher-order constructs and three formative higher-order complex concep-
tualisations. The theoretical model involved six endogenous latent variables
to illustrate the structural relationships among the 25 first-order and five
second-order latent constructs.
The model comprised 34 reflective items and 40 formative items which
were composed of 25 first-order and six second-order conceptualisations.

This was used to examine the effects of predictors for the endogenous
latent variables of the study namely, cognitive evaluation, expectation of
ICT usage, actual ICT usage, ICT integration, ICT utilisation and organi-
sational performance. The number of constructs employed in the theoreti-
cal model indicated that the sample size of the study (n = 282) was
adequate to proceed with PLS-SEM. The analysis also conformed to the
minimum threshold requirement of allocating at least two indicators per
construct for SEM (Kline, 2010; Rahim, Antonioni, & Psenicka, 2001).
6.2.3. Justification of Reflective and Formative Measures

The differentiation and appropriate use of formative and reflective con-
structs in estimates has been a recent advancement of SEM. IS researchers
initially modelled mostly reflective constructs due to many reasons. These
included the availability of software that was supportive for estimates of
formative constructs (Chin, 1998a; Gefen et al., 2000); conceptual criteria
for determining whether constructs should be specified as reflective or for-
mative (Diamantopoulos & Winklhofer, 2001; Edwards & Bagozzi, 2000);
lack of a consistent standard for assessing psychometric properties of mea-
sures (Bagozzi & Yi, 1998; Bollen, 1989); and lack of requisite knowledge
for the subsequent estimates (Jarvis, MacKenzie, & Podsakoff, 2003). The
introduction of PLS-SEM has provided the analytical tools suitable for
modelling reflective and formative constructs. The development of software
for component-based SEM has provided enormous opportunities for
researchers who are involved in modelling reflective and formative constructs.
However, it is difficult to anticipate the nature of an indicator, that is,
whether it is reflective or formative.
Researchers have primarily judged the nature of a latent variable by
applying the definitions of reflective and formative indicators as described
below.
By nature, reflective items are highly correlated as they (reflective indica-
tors) represent reflections, or manifestations, of a construct. Hence,
variation in a construct leads to variation in its indicators (Bollen, 1989).

For example, an individual change in the latent perceived usefulness con-
struct results in corresponding changes in each manifest indicator of
perceived usefulness. Thus, perceived usefulness has been identified as a
reflective construct.
Formative indicators, on the other hand, are entirely opposite to the
character of a reflective indicator. The formative items show direct causal
relationships from the item to the latent variable, that is, the items cause
the latent variable (Diamantopoulos & Winklhofer, 2001). In other words,

the formative constructs are formed by their respective measurement items.
Thus, the items are not correlated and measure different underlying dimen-
sions of the latent variable (Chin, 1998b). For example, country readiness is
measured by technology infrastructure, human infrastructure, legal infra-
structure, financial infrastructure, and government policy and supports. The
measurement indicators are not correlated and the variation in the latent
construct does not lead to variation in its indicators. More clearly, an indi-
vidual’s favourable assessment about country readiness does not necessarily
mean that all of its indicators are favourable for SMEs’ ICT adoption.
Thus, the country readiness construct has been identified as a formative
construct.
Modelling reflective or formative constructs requires theoretical justifica-
tion (Coltman, Devinney, Midgley, & Venaik, 2008; Diamantopoulos &
Siguaw, 2006; Jarvis et al., 2003). However, it may be difficult to explore
the theoretical interpretation of a construct, reflective versus formative.
Jarvis et al. (2003) developed a set of conceptual criteria which are used as
a guideline for justifying the nature of variables, reflective or formative,
modelled to measure a phenomenon. More clearly, a variable is modelled
as formative when the following decision rules hold; otherwise, it is reflec-
tive: the direction of causality is from indicators to constructs; the indica-
tors need not be interchangeable; covariation among indicators is not
necessary, and the nomological net of indicators can differ (Jarvis et al.,
2003; Petter, Straub, & Rai, 2007; Rai et al., 2006).
The screening process, which applied the above conceptualisations
and decision rules (see Appendix E for the decision rules in detail),
resulted in the identification of 34 reflective items and 40 formative
items for 12 reflective and 13 formative first-order constructs. Five
second-order constructs comprising two reflective higher-order constructs
and three formative higher-order complex conceptualisations were also
justified.
160 MD SHAH AZAM
6.2.4. Examination of Possible Biases

The distributions of the sample into various types of firms resulted in a
balance between the different categories of SMEs which were of different
type, size, scope of business and competitive position. A quarter of the
respondents in the sample were from small enterprises and three quarters
were from medium-sized enterprises. The survey instruments were supplied
to different SMEs in Bangladesh anonymously selected using a stratified
random sampling technique. Through the use of several follow-up calls and
reminders by telephone and by sending surveyors as the researcher’s repre-

sentatives, the survey responses were received in two phases. In the first
phase, a total of 150 responses were received. The researcher again initiated
the push technique to receive more responses. At this stage, two trained
surveyors were employed to contact respondents who had not yet sent the
completed survey instruments. Repeated follow-up calls and SMS, and
surveyors’ personal contacts resulted in the receipt of 132 more responses.
The study thus analysed the data set which comprised responses collected
at two different points in time.
To ascertain the quality of the data, the study examined whether any
systemic biases could exist due to there being two groups of respondents.
As the survey responses were received during two different time periods,
a question could be asked about whether respondents who answered
quickly may be in a more advantageous position in terms of their com-
pany’s positioning or their status in using advanced ICT: knowing the
situation better might thus motivate them to answer the questionnaire. On
the other hand, late respondents may be affiliated with the group of firms
that were running with a lower level of ICT use; thus, they could be less
motivated to respond. However, it was not unlikely that the advanced
group, in enjoying various facilities and the uniqueness of the technology,
might overrate the benefits, utility and usage of ICT and other aspects of
the study. To test this possible bias, the total sample was divided into two
groups: Wave-1 (advanced group: firms responded quickly) and Wave-2
(general group: firms responded in the second time period).
The Mann-Whitney U test was used to compare the ranked means of
some discrete variables between the two groups. A Kolmogorov-Smirnov
(K-S) test was further used to examine if the sample distribution of Wave-1
group was equal to that of the Wave-2 group (Boes, Graybill, & Mood,
1974, reported in Zhu & Kraemer, 2005). As shown in Table 19, the p value
of each variable was insignificant (p > 0.10). That is, all variables such
as firm characteristics (sector, size, global business and SME preference);
level of ICT use (homepage, online cataloguing and e-commerce); intention
Table 19. Test of Possible Biases.

Variable Mean Rank Mann-Whitney U Test KS Test
Wave-1 Wave-2 Mann-Whitney Z-stat p value Z score p value
Firm characteristics
Sector 142.14 140.77 9804.000 −0.202 0.840 0.081 1.000
Size 140.14 143.05 9696.000 −0.355 0.723 0.173 1.000
Global business 143.88 138.80 9543.000 −0.803 0.422 0.302 1.000
SME preference 140.44 142.70 9741.000 −0.349 0.727 0.135 1.000
Level of ICT use

Homepage 142.30 140.59 9780.000 −0.524 0.601 0.102 1.000
Online cataloguing 143.00 139.80 9675.000 −0.407 0.684 0.190 1.000
E-commerce 140.00 143.20 9675.000 −0.498 0.619 0.190 1.000
Intention to use IT
Online cataloguing 142.58 140.27 9738.000 −0.248 0.804 0.218 1.000
E-commerce 145.10 137.41 9360.000 −0.867 0.368 0.627 0.826
Subjective perception
Perceived usefulness 137.29 146.29 9268.000 −0.973 0.331 0.691 0.727
Perceived ease of use 142.86 139.96 9696.500 −0.321 0.748 0.355 1.000
Personal characteristics
Gender 141.18 141.86 9852.000 −0.114 0.909 0.041 1.000
Age 142.89 139.92 9691.000 −0.329 0.742 0.321 1.000
Education 136.40 149.34 8865.500 −1.647 0.100 0.716 0.684
Income 142.98 139.82 9678.000 −0.339 0.734 0.686 0.735
to use ICT (online cataloguing and e-commerce); subjective perception

(perceived usefulness and perceived ease of use) and individual character-
istics (gender, age, education and income) seemed to be equal between
the groups. Thus, the study concluded that the category of respondents
(Wave-1 or Wave-2) did not cause any survey bias.
6.3. Model Assessment
6.3.1. Assessment of the Measurement Model

The model consisted of 25 first-order constructs and five second-order con-
structs. Both first-order and second-order constructs contained reflective as
well as formative items or indicators. Thus, the assessment of the measure-
ment model involved several tasks that assessed the reliability and validity
of the first-order reflective and formative constructs, and the higher-order
reflective and formative constructs. The strength of the psychometric prop-
erties was assessed by examining the convergent validity and discriminant
162 MD SHAH AZAM
validity of the reflective items and manifest indicators (for second-

order reflective construct validation) while the formative constructs
were validated by composite measurement as formative items do not mea-
sure the same underlying dimension. The indicators’ weights, which mea-
sure the relative importance of the formative items in the formation of
the constructs, were used to make composite latent variables. The same
principle was applied in the validation of the higher-order formative
constructs.
6.3.2. Convergent Validity

The assessment of convergent validity was a fundamental part of assessing
the measurement model. The psychometric properties of the measurement
model were assessed by evaluating the reliability, convergent validity and
discriminant validity (Fornell & Larcker, 1981). The reliability of the con-
structs was assessed by considering composite reliability and Cronbach’s
alpha. The magnitude and significance of standard path loadings were
considered to examine the construct validity.
6.3.2.1. Item Reliability. Item reliability assessed the loadings for each
individual item. Table 20 presents the detailed item loadings. The loadings
indicate the correlation of the items with their respective constructs.
Therefore, maintaining low loading items would decrease the correlation
between the items in the construct (Nunnally, 1994). Item reliability also
measured the level of random error for each construct; the lower the item
loading, the higher the level of random error. Therefore, this procedure
could identify and eliminate the items in a particular construct that could
increase the construct’s level of random error (Fornell & Larcker, 1981).
High item loadings indicated the reliability of the measures of the latent
variable; moreover, the prior literature has supported some agreed levels of
item loadings as thresholds for fairly reliable measures. Igbaria et al. (1995)
deemed 0.4 as an acceptable minimum loading. Hair et al. (1998) suggested
that loadings above 0.3 were significant, above 0.4 were more significant
and above 0.5 were very significant. Chin (1998a) believed item loadings
should be above 0.5. Carmines and Zeller (1979) maintained 0.7 as the
reliability limit whilst Barclay et al. (1995) specified 0.707 as the minimum
limit. However, Nunnally (1994) argued that, in the case of strong theoreti-
cal support, further reviews of the low loading items were warranted. This
would be especially pertinent if the low loading items added to the explana-
tory power of the model.
Table 20. Psychometric Properties for First-Order Reflective Constructs.

Construct Items Loading t-statistics CR AVE
Perceived usefulness PU1 0.917 79.85 0.956 0.844

PU2 0.938 11.91
PU3 0.940 29.24
PU4 0.878 55.52
Perceived ease of use PEU1 0.948 50.00 0.967 0.879
PEU2 0.960 82.26
PEU3 0.963 9.22

PEU4 0.877 48.65
Facilitating condition FACICON1 0.769 30.92 0.875 0.701
FACICON2 0.853 33.31
FACICON3 0.895 76.24
Owner innovativeness OWNER1 0.765 3.50 0.800 0.669
OWNER2 0.674 1.99
Power distance PD1 0.963 89.29 0.964 0.930
PD4 0.965 90.74
In-group collectivism INGROUP1 0.746 10.78 0.900 0.752
INGROUP2 0.920 9.93
INGROUP3 0.924 6.53
Uncertainty avoidance UNAVOID1 0.980 63.54 0.974 0.951
UNAVOID2 0.971 40.51
Ethical culture ECUL1 0.975 200.04 0.967 0.936
ECUL2 0.960 130.29
Integration INTEGRA1 0.764 22.86 0.791 0.561
INTEGRA2 0.639 13.68
INTEGRA3 0.832 35.89
Internal operation productivity INTOP1 0.898 75.44 0.898 0.746
INTOP2 0.870 60.52
INTOP3 0.821 41.41
Competitiveness COMP1 0.732 11.63 0.731 0.478
COMP2 0.757 13.44
COMP3 0.572 6.08
Financial performance FPERF1 0.771 30.24 0.847 0.650
FPERF2 0.901 67.70
FPERF3 0.737 18.59
Taking into account all the recommendations in the literature and to

maximise the measurement model’s ability to fulfil the requirements of con-
vergent validity, the minimum value of 0.5 was determined. Hence, after
the first PLS run, four items with loadings below 0.5 were discarded, these
being INTEGRA4, INTEGRA5, PD2 and PD3. Although some of the
constructs had a small number of manifest items, each of the constructs
contained at least two indicators even after discarding the items with low
164 MD SHAH AZAM
loadings; thus, the final model conformed to the criterion suggested by

Kline (2010) and Rahim et al. (2001) that there should be a minimum of
two items in a construct.
The final run of the refined measurement model ensured high reliability
of the measures as all reflective items were found to have satisfactory load-
ings. All of the items were found to have loadings above 0.707 while only
one indicator had a loading of 0.572 which was acceptable as it satisfied the
threshold suggested by Hair et al. (1998) and Igbaria et al. (1995).
6.3.2.2. Internal Consistency. Internal consistency is measured through cal-

culating the composite reliability (CR) (Fornell & Larcker, 1981). CR is
considered to be superior to the traditional measures of consistency (such
as Cronbach’s alpha) because it is not influenced by the number of indica-
tors (Hanlon, 2001).
Constructs with a coefficient value of 0.70 or more in the estimates of
CR were accepted as reliable for further analysis (as suggested by Barclay
et al., 1995; Hair, Ringle, & Sarstedt, 2011; Igbaria et al., 1997). As shown
in Table 20, all constructs met this criterion. The reflective construct compe-
titiveness had the lowest internal consistency of 0.734 while uncertainty
avoidance had the highest of 0.974. Most of the study constructs had an
internal consistency of approximately 0.8 which was considered to be very
satisfactory.
6.3.2.3. Average Variance Extracted. Fornell and Larcker (1981) specified

that AVE should be at least 0.5 for convergent validity to be satisfied. Thus
the AVE scores, although not a usual measure of convergent validity, were
rationally computed for robustness of the statistical analysis.
As shown in Table 20, all of the reflective constructs satisfactorily met
this criterion for convergent validity while a little deviation was seen in the
AVE score for first-order reflective construct competitiveness which was
also used as a manifest latent indicator for the higher-order variable perfor-
mance. The variable competitiveness appeared with an AVE score of 0.478
which was not problematic as the score was very close to the cut-off value
as suggested by Fornell and Larcker (1981). Thus, the AVE results indi-
cated satisfactory convergent validity for all of the reflective constructs.
6.3.3. Discriminant Validity

The measurement model also involved the evaluation of the extent to
which different constructs diverged from one another which is termed
‘discriminant validity’. Discriminant validity, a proof of construct validity,
is defined as the degree to which any given construct is different from any
other (Barclay et al., 1995). The square root of the AVE and the cross-
loading matrix are widely used measures for discriminant validity (Barclay
et al., 1995; Igbaria et al., 1995). According to Igbaria et al. (1995), the
model is assessed as having acceptable discriminant validity if the square
root of the AVE of a construct is larger than its correlation with other
constructs (Hair et al., 2011). On the other hand, the constructs may be
considered as discriminant if the loading of items within a construct (shown
in the columns in a cross-loading matrix) are greater than the loading of

any other item within the same column (Barclay et al., 1995). Both the
approaches for assessing discriminant validity are important to adequately
prove discriminant validity at construct level as well as at item level.
6.3.3.1. Discriminant Validity at Construct Level. The square root of the

AVE and inter-construct correlations were compared to assess discriminant
validity at construct level. According to Fornell and Larcker (1981), con-
structs are considered to have adequate discriminant validity if the square
root of the AVE is greater than the inter-construct correlations.
Table 21 presents the square root of the AVE and the inter-construct
correlations. The inter-construct correlations are presented in the off-
diagonals while the values for the square root of the AVE are placed in the
main diagonal in the matrix in bold font. In order to prove discriminant
validity among the reflective constructs, the off-diagonal elements (correla-
tions of latent variables) must be less than or equal to the bolded, diagonal
pffiffiffiffiffiffiffiffiffi
elements ( AVE) in the corresponding rows and columns (Barclay et al.,
1995; Gefen et al., 2000; Hair et al., 2011; Igbaria et al., 1997). The mea-
surement model met the criterion for discriminant validity as the bolded
diagonal values were seen to be greater than the off-diagonal correlation
values in their corresponding rows and columns. Thus, the results proved
adequate discriminant validity among the study constructs allowing the
structural model estimation to proceed with no further item deletion from
the model.
6.3.3.2. Discriminant Validity at Item Level. The study also assessed the
discriminant validity at item level comparing the loadings and cross-
loadings of the measures. The loading of items within a construct (shown
in columns) should be greater than the loading of any other item in order
to prove discriminant validity among the constructs (Barclay et al., 1995).
Table 22 shows the loadings and cross-loadings of items and constructs
in the cross-loading matrix. The cross-loading matrix displayed high
166
Table 21. Correlation Matrix for First-Order Reflective Constructs.

FC OI INT PU PEU INTOP COMP FPERF PD UA ING ECUL
FC 0.835
OI 0.210 0.818
INT 0.442 0.068 0.749
PU 0.653 0.072 0.429 0.919
PEU 0.649 0.222 0.414 0.533 0.938
INTOP 0.557 0.051 0.521 0.638 0.337 0.864
COMP 0.324 −0.033 0.531 0.197 0.131 0.302 0.695
FPERF 0.389 0.033 0.586 0.500 0.321 0.730 0.314 0.806
PD −0.218 −0.144 −0.205 −0.130 −0.231 −0.155 −0.062 −0.145 0.964
UA 0.023 0.110 −0.100 0.022 0.109 −0.022 −0.102 0.029 −0.277 0.976
ING −0.191 −0.180 −0.100 −0.175 −0.203 −0.136 −0.023 −0.211 0.448 −0.368 0.867
ECUL −0.299 −0.082 −0.261 −0.267 −0.243 −0.393 −0.114 −0.347 0.390 −0.376 0.357 0.967
FC = Facilitating condition, OI = Owner innovativeness, INT = Integration, PU = Perceived usefulness, PEU = Perceived ease of use,
INTOP = Internal operation productivity, COMP = Competitiveness, FPERF = Financial performance, PD = Power distance, UA =
Uncertainty avoidance, ING = In-group collectivism, ECUL = Ethical culture.
MD SHAH AZAM
Table 22. Cross-Loading Matrix.

PU PEU OI FC PD ING UA ECUL INT INTOP COMP FPERF
PU1 0.9179 0.5182 0.0754 0.5828 −0.0767 −0.1167 0.0488 −0.2601 0.4686 0.6085 0.2474 0.5299
PU2 0.9372 0.4907 0.0539 0.6261 −0.0513 −0.1001 −0.0375 −0.2532 0.4338 0.6070 0.1920 0.4479
PU3 0.9392 0.4526 0.0981 0.5477 −0.0314 −0.0883 0.0017 −0.2137 0.3817 0.6203 0.1715 0.5262
PU4 0.8787 0.5017 0.0841 0.5549 −0.0622 −0.0997 0.0719 −0.2593 0.2777 0.5049 0.1213 0.3754
PEU1 0.4836 0.9400 0.1736 0.6457 −0.1848 −0.1710 0.1236 −0.2215 0.3537 0.3034 0.1219 0.2775
PEU2 0.5031 0.9518 0.1490 0.6269 −0.1920 −0.1942 0.0902 −0.2505 0.3905 0.3327 0.1496 0.3256
PEU3 0.5156 0.9554 0.1648 0.6275 −0.1637 −0.1889 0.0928 −0.2407 0.3593 0.3220 0.1495 0.2985
PEU4 0.4945 0.8984 0.2441 0.5226 −0.1414 −0.1420 0.1025 −0.1993 0.3343 0.3047 0.0928 0.3564
OWNER1 0.0406 0.2135 0.7370 0.2016 −0.1507 −0.1249 0.0647 −0.0625 0.0542 0.0215 −0.0407 −0.0173
OWNER2 0.0905 0.1416 0.9072 0.1496 −0.1069 −0.2190 0.1460 −0.0820 0.0651 0.0958 −0.0137 0.1223
FACICON1 0.7854 0.5153 0.1161 0.7462 −0.1557 −0.1627 −0.0007 −0.2819 0.4634 0.6478 0.3126 0.5109
FACICON2 0.4639 0.8794 0.1596 0.6543 −0.2131 −0.1769 0.0884 −0.2470 0.3362 0.3267 0.1156 0.2702
FACICON3 0.4545 0.8648 0.1838 0.6671 −0.2130 −0.1893 0.0895 −0.2697 0.3284 0.3423 0.1094 0.2834
PD1 −0.1187 −0.2211 −0.1325 −0.2121 0.8434 0.4126 −0.2643 0.3674 −0.1952 −0.1397 −0.0480 −0.1425
PD4 −0.1132 −0.2385 −0.1379 −0.2448 0.8419 0.4440 −0.2654 0.3943 −0.1675 −0.1517 −0.0013 −0.1152
INGROUP1 −0.0943 −0.1472 −0.1619 −0.1392 0.2669 0.9135 −0.3111 0.2711 −0.0255 −0.1098 0.0104 −0.1288
INGROUP2 −0.1207 −0.1900 −0.2295 −0.1260 0.3352 0.8649 −0.3804 0.3056 −0.0749 −0.0572 −0.0162 −0.1915
INGROUP3 −0.2109 −0.1847 −0.1537 −0.2011 0.3278 0.6823 −0.2901 0.3441 −0.1337 −0.1670 −0.0441 −0.2256
UNAVOID1 0.0439 0.1030 0.1269 0.0132 −0.2491 −0.3753 0.9771 −0.3560 −0.1153 −0.0214 −0.1171 0.0310
UNAVOID2 −0.0009 0.1117 0.1396 0.0352 −0.2470 −0.3740 0.9740 −0.3781 −0.0985 −0.0210 −0.0802 0.0480
ECUL1 −0.2880 −0.2463 −0.0582 −0.2953 0.2857 0.3013 −0.3636 0.9768 −0.2643 −0.3987 −0.1104 −0.3798
ECUL2 −0.2222 −0.2178 −0.1229 −0.2722 0.3117 0.3067 −0.3647 0.9575 −0.2369 −0.3581 −0.1191 −0.2951
INT1 0.3522 0.1620 −0.0275 0.2788 −0.0429 −0.0088 −0.1209 −0.1946 0.7851 0.4116 0.4646 0.4013
INT2 0.2805 0.5754 0.0879 0.4172 −0.0704 −0.0573 0.0470 −0.1663 0.5810 0.2714 0.2785 0.3336
INT3 0.3287 0.2592 0.1117 0.3026 −0.1562 −0.0500 −0.1214 −0.2223 0.8483 0.4652 0.4363 0.5557
INTOP1 0.5714 0.3221 0.0431 0.4592 −0.0664 −0.0821 0.0168 −0.3717 0.4683 0.8980 0.2193 0.7319
INTOP2 0.4566 0.1279 0.0588 0.3082 −0.0232 −0.0121 −0.0889 −0.3068 0.4974 0.8697 0.3045 0.6348
INTOP3 0.6343 0.4374 0.1106 0.6468 −0.1579 −0.1394 0.0176 −0.3417 0.3916 0.8209 0.2768 0.5378
167
COMP1 0.2482 0.1930 0.0510 0.3270 −0.0376 −0.0118 −0.0387 −0.0918 0.3215 0.2909 0.7318 0.2457
168
PU PEU OI FC PD ING UA ECUL INT INTOP COMP FPERF
COMP2 0.0794 0.0780 −0.0843 0.1674 0.0201 −0.0095 −0.1027 −0.1266 0.5042 0.2222 0.7563 0.2816
COMP3 0.0414 −0.0922 −0.0398 0.1396 0.0415 0.0537 −0.0827 0.0347 0.2483 0.0426 0.5725 0.0491
FPERF1 0.5701 0.4592 0.0768 0.4449 −0.1568 −0.1869 0.1863 −0.3491 0.4263 0.6872 0.2103 0.7711
FPERF2 0.3992 0.2733 0.1120 0.2973 −0.1142 −0.2026 0.0621 −0.3469 0.5212 0.6306 0.2437 0.9005
FPERF3 0.2516 0.0614 −0.0028 0.1663 0.0704 −0.0004 −0.1821 −0.1405 0.4737 0.4528 0.3239 0.7372
PU = Perceived usefulness, PEU = Perceived ease of use, FC = Facilitating condition, OI = Owner innovativeness, PD = Power distance,
ING = In-group collectivism, UA = Uncertainty avoidance, ECUL = Ethical culture, INT = Integration, INTOP = Internal operation
productivity, COMP = Competitiveness, FPERF = Financial performance.
MD SHAH AZAM
correlations between constructs and relevant measurement items. All reflec-

tive constructs of the model showed high correlation, with their respective
measurement items being also significantly higher than the items in the
same column measuring other constructs, the only exception being the cor-
relation between perceived ease of use and facilitating condition. Items 2 and
3, measuring the facilitating condition, demonstrated a very high correlation
with perceived ease of use (r = 0.88 and 0.86, respectively). This occurred
because facilitating condition is measured through questions that are, in a
sense, further proof of organisational technology competency making a sys-

tem easy to use and fostering the adoption rate of digital technology. This
measures whether the organisation has capable human resources who also
have the skills to do the work in an automated environment, whether the
organisation has adequate resources necessary to work in a digital environ-
ment and whether the existing systems are compatible with using digital
technology. Thus, high correlation between some measurement items of
facilitating condition and perceived ease of use was considered to be justified.
This high correlation was not considered problematic as the correlations of
perceived ease of use with its measurement items were observed as higher
than those with high correlation (r = 0.94, 0.95, 0.96 and 0.90, respectively).
Finally, it was concluded that the model now demonstrated both conver-
gent and discriminant validity.
6.3.4. Indicator Weights for Formative Constructs

The theoretical model comprised both reflective and formative constructs.
Among the 25 first-order and five second-order constructs of the model,
13 first-order constructs and four higher-order constructs were formative in
nature. By definition, reflective items measure the same underlying dimen-
sions and should be correlated, while formative indicators ‘cause’ the latent
construct, also called the emergent construct, which is explained as a func-
tion of the formative measures. The SEM involved a two-step procedure
which assessed the psychometric properties of the latent constructs by
examining reliability using Cronbach’s alpha or Fornell and Larcker’s
(1981) measure of CR; convergent validity by examining the loadings and
t-statistics; and discriminant validity by examining the AVE and inter-
construct correlations or loading and cross-loading matrix, before estimat-
ing the structural model. Although all of the tests for construct reliability
and validity were considered as fundamental requirements for reflective
constructs, these were not necessary requirements for formative constructs
(Jarvis et al., 2003; Rai et al., 2006) as the items were not correlated and
did not measure the same underlying dimension.
170 MD SHAH AZAM
In estimating the measurement model, researchers have generally pro-

ceeded by evaluating the construct validity and reliability for reflective con-
structs; however, these two forms of validity should not be conducted in
the same manner for formative measures, given the differences between
reflective and formative measures. Construct validity is typically assessed
by two aspects of validity: convergent validity which detects whether the
measures for a construct are more correlated with one another than with
measures of another construct, and discriminant validity which determines
if the measures are isolated as distinct constructs (Petter et al., 2007). The
methods of determining construct validity that focus on common variance,
although applicable to reflective constructs, logically do not apply well to
formative constructs as the correlations between indicators within a con-
struct do not need to be higher than the correlations between indicators of
different constructs (MacCallum & Browne, 1993) and there is no require-
ment that the measures of the construct be highly correlated (Rossiter,
2002). Unlike common factor analysis, where one examines the loadings, in
principal component analysis for formative constructs, the weights must be
examined.
Therefore, indicator weights, which provided information on the relative
importance of the formative items towards the formation of the corre-
sponding latent construct, were calculated. The weights for each of the
40 formative indicators are presented in Tables 23 and 24. The tables show
Table 23. Validity for First-Order Formative Constructs (Culture and

Environment).
Construct Items Weight t-statistics Tolerance VIF
Bengali values BVAL1 −0.751 4.78** 0.961 1.04

BVAL2 0.571 2.51** 0.980 1.02
BVAL3 0.048 0.19 0.960 1.04
Coercive pressure COERPRE1 −0.018 0.20 0.979 1.02
COERPRE2 0.682 10.31** 0.844 1.19
COERPRE3 0.512 7.19** 0.828 1.21
Mimetic pressure COMPRE1 0.424 1.97* 0.995 1.01
COMPRE2 0.877 7.68** 0.995 1.01
Normative pressure NORMPRE1 0.559 7.09** 0.910 1.10
NORMPRE2 0.583 7.09** 0.864 1.16
NORMPRE3 0.228 2.16** 0.964 1.04
NORMPRE4 0.110 0.99 0.979 1.02
Global pressure GLOPRE1 0.063 0.40 0.971 1.03
GLOPRE2 0.988 27.50** 0.971 1.03
**p < 0.01, *p < 0.05, VIF = variance inflation factor.

Table 24. Validity for First-Order Formative Constructs (Country

Readiness and ICT Use).
Construct Items Weight t-statistics Tolerance VIF
Technological infrastructure TECH1 0.203 0.68 0.870 1.15

TECH2 −0.386 1.21 0.721 1.39
TECH3 0.828 3.94** 0.740 1.35
TECH4 0.511 2.66** 0.969 1.03
Human infrastructure HINFRA1 0.612 1.46 0.354 2.82
HINFRA2 0.144 0.34 0.387 2.59

HINFRA3 0.367 1.09 0.608 1.65
Legal infrastructure LINFRA1 0.197 0.54 0.951 1.05
LINFRA2 0.938 4.18** 0.951 1.05
Financial infrastructure FINFRA1 0.897 2.81** 0.838 1.19
FINFRA2 0.210 0.54 0.838 1.19
Government policy and supports GOV1 0.808 3.26** 0.387 2.59
GOV2 0.862 3.39** 0.384 2.61
GOV3 −0.531 1.52 0.372 2.69
GOV4 −0.688 2.15* 0.355 2.82
Expectation EXP2 0.216 4.26** 0.951 1.05
EXP3 0.237 3.78** 0.425 2.35
EXP4 0.828 15.52** 0.452 2.21
EXP5 −0.062 1.08 0.69 1.45
Usage USAGE2 0.176 3.70** 0.981 1.02
USAGE3 0.491 6.48** 0.542 1.85
USAGE4 0.553 8.09** 0.543 1.84
USAGE5 0.052 2.25* 0.983 1.02
Utilisation UTL1 0.235 2.61* 0.919 1.09
UTL2 0.863 11.12** 0.689 1.45
UTL3 0.081 0.79 0.720 1.39
**p < 0.01, *p < 0.05, VIF = variance inflation factor.
that the weights for 26 formative items had significant t-values while this
was not the case for 14 items.
As there were a number of formative indicators with insignificant
weights, it was important to decide whether some indicators should be
discarded in order to have formative constructs with indicators that
would highly contribute to the emergent constructs. Diamantopoulos and
Winklhofer (2001) suggested that if any of the item weightings for forma-
tive measures were non-significant, it may be appropriate to remove the
non-significant indicators (one at a time) until all paths were significant
and a good fit was obtained (this may be applicable in CBSEM. Zhu,
Dong, et al. (2006) followed the guidelines of retaining measurement items
with significant weights and above the cut-off value of 0.3 (Chin, 1998b).
172 MD SHAH AZAM
Content validity is an issue that could have an adverse effect when adopt-
ing removal measures in formative constructs. Dropping the formative
indicators that had non-significant weights could pose the question whether
the construct was still measuring the entire domain and whether the content
validity was preserved (Bollen & Lennox, 1991).
The works of Fornell, Lorange, and Roos (1990), Santosa et al. (2005)
and Diamantopoulos and Winklhofer (2001) may be worth mentioning in
support of retaining all the indicators of a formative construct regardless of
their significance and the magnitude of their weight.

This study has developed 13 first-order formative constructs and four
higher-order constructs in analysing all possible aspects of conceptualising
the domain of the constructs. However, in its agreement with the proposi-
tion ‘omitting an indicator is omitting a part of the construct’, this study
showed that it suspected that the meaning or definition of the constructs
would be changed if any indicator was deleted. Thus, all the indicators of
the formative constructs were logically retained for further analyses.
In addition to the indicator weights, the test of multicollinearity was
conducted on the formative items. Unlike reflective indicators, where multi-
collinearity between construct items is desirable (illustrated by a high
Cronbach’s alpha or CR scores), excessive multicollinearity in formative
constructs can destabilise the model.
The existence of very high correlations among measures may suggest
that multiple indicators are tapping into the same aspect of the construct.
Thus, low collinearity among items is vital in ensuring the stability of the
estimates (Mathieson, Peacock, & Chin, 2001). Hence, this establishes
that each indicator has its own distinct influence. In general, multicollinear-
ity is examined by the variance inflation factor (VIF). Although a VIF
value greater than 10 is acceptable, different threshold values for VIF are
proposed as collinearity poses more problems in formative constructs.
Kleinbaum, Kupper, Muller, and Nizam (1998) recommended that the
threshold of VIF be 10 while Mathieson et al. (2001) and Hair et al. (2011)
suggested 5 as the maximum threshold for detecting multicollinearity
among indicators. However, Diamantopoulos and Siguaw (2006) stated
that a VIF statistic greater than 3.3 was problematic for formative mea-
sures. The SPSS 19 statistical package was utilised to run the regression
analysis with the PLS construct scores as the dependent variables and the
items as independent variables (Andreev, Maoz, Heart, & Pliskin, 2009).
As shown in Tables 23 and 24, the results depicted that all the VIF
scores were below 3.3 which was well below the recommended maximum
threshold recommended by Kleinbaum et al. (1998), Mathieson et al.
(2001) and Hair et al. (2011). Therefore, all the formative items were logi-
cally retained.
6.4. Validation of the Higher-Order Model
The estimation of hierarchically structured latent variables is a very useful

method under structural modelling which allows for more theoretical parsi-
mony and reduces model complexity (Edwards, 2001; Law, Wong, &
Mobley, 1998; MacKenzie, Podsakoff, & Jarvis, 2005). In the past, hier-
archical modelling was mainly analysed within the framework of CBSEM
and most of the cases focused on the second-order construct (Edwards,
2001; Edwards & Bagozzi, 2000; Jarvis et al., 2003; Law & Wong, 1999;
MacKenzie et al., 2005; Petter et al., 2007). These researchers defined
hierarchical constructs or multidimensional constructs as the constructs
involving more than one construct. Although still limited in number, recent
initiatives have explored a more advanced avenue in the analysis of
hierarchically structured latent variables, both on the extent of hierarchy as
well as in their underlying relationships.
Recent research has shown a third-order and fourth-order hierarchical
latent variable model design with varied underlying relationships among
the constructs, that is, formative or reflective, within a PLS-based path
modelling framework (Akter, D’Ambra, & Ray, 2010, 2011; Wetzels et al.,
2009). The higher-order construct may be defined as the latent variable
which is created through the indicators and latent variables in a multiple
hierarchy. For example, if a second-order latent variable consists of two
underlying first-order latent variables, each with four manifest variables,
the second-order manifest variables can be specified using all (eight) mani-
fest variables of the underlying first-order latent variables (Chin & Gopal,
1995). Consequently, the manifest variables are used twice: for the first-
order latent variable (‘primary’ loadings) and for the second-order latent
variable (‘secondary’ loadings).
With this established, the outer model (measurement model) and
the inner model (structural model) are thus specified accounting for the
hierarchical component of the model and representing the loadings of the
second-order latent variable on the first latent variables. Wetzels et al.
(2009) extended this approach to address the higher-order hierarchical
models.
Like first-order latent constructs, the second-order or higher-order con-
structs may also be composed of both reflective and formative indicators.
174 MD SHAH AZAM
The formation of two different types of higher-order constructs, the reflec-

tive higher-order construct and the formative higher-order construct, may
be explained by the following equations where Eqs. (3) and (4) explain first-
order and second-order reflective constructs, while Eqs. (5) and (6) state
the formation of first-order and second-order formative constructs,
respectively:
yi = Λy ⋅ηj þ ɛ i ð3Þ
where yi = manifest variables (e.g. items of perceived usefulness), Λy = load-

ings of first-order latent variable, ηj = first-order latent variable (e.g. per-
ceived usefulness), ɛi = measurement error.
ηj = Γ⋅ξk þ ζ j ð4Þ
where ηj = first-order factor (e.g. perceived usefulness), Γ= loadings of

second-order latent variable, ξk = second-order latent variable (e.g. cognitive
evaluation), ζ j = error of first-order factors.
η = γ 1 x1 þ γ 2 x2 þ … þ γ n xn þ ɛ ð5Þ
where η = first-order latent variable, xi = observed variables, γ i = expected

effect of xi on η, ɛ = a disturbance term, with COVðxi ; ɛÞ = 0 and EðɛÞ = 0.
Γ = φ1 η1 þ φ2 η2 þ … þ φn ηn þ ζ ð6Þ
where Γ = second-order latent variable, ηi = observed variables, φi = expected

effect of ηi on Γ, ɛ = a disturbance term with COVðni ; ζÞ = 0 and EðζÞ = 0.
This study has applied the above-mentioned functions to develop the
hierarchical conceptualisations of the study constructs through a multistage
process. The hierarchical conceptualisation can result in theoretical parsi-
mony and ease of analysis. The analytical procedures for hierarchical reflec-
tive models and formative models are depicted in Tables 25 and 26.
6.4.1. Validation of Second-Order Reflective Construct: Cognitive

Evaluation (ReflectiveReflective)
Cognitive evaluation was modelled as a second-order construct, capturing
two beliefs towards technological innovation perceived usefulness and
perceived ease of use, as validated through the field study. Resulting from
Table 25. Description of Second-Order Reflective Constructs.

Description Illustration
1 Construct the first-order latent variables

such as LV11 and LV12 and relate
them to their respective block of
manifest variables (such as LV11:
MV1-MV4; LV12:MV5-MV8) using
Mode A (reflective) in their outer
model. The loading represents the

first-order loadings (see Eq. (3)).
2 The second-order latent variable can
now be constructed by relating to the
block of the underlying first-order
latent variables and their constructs,
such as second-order construct LV21
by MV1-MV8, using Mode A
(reflective) in their outer model (the
indicators placed above the second-
order construct represent the
secondary loadings). The first-order
latent variables (LV11 and LV12) are
now related to the second-order
variable LV21 as reflective
dimensions. This inner model
represents the second-order loadings
(see Eq. (4)).
Researchers using PLS-SEM often refer to reflective measurement models (i.e. scales) as
Mode A, and formative measurement models (i.e. indices) are labelled Mode B (e.g. Rigdon,
Ringle, & Sarstedt, 2010).
the field study’s strong support for capturing complex measures, this
second-order approach viewed the two dimensions in a collective and
mutually reinforcing manner. Accordingly, cognitive evaluation was opera-
tionalised to be an integrative measure of beliefs towards technological
innovation (see Fig. 4).
The degree of explained variance of the second-order hierarchical cogni-
tive construct was reflected in its first-order components, perceived useful-
ness (77.5%) and perceived ease of use (75.8%). As shown in Table 27, the
path coefficients from second-order cognitive evaluation of the first-order
constructs, perceived usefulness (0.880) and perceived ease of use (0.870),
were greater than the suggested cut-off of 0.7 (Chin, 1998b; Zhu, Dong,
176 MD SHAH AZAM
Table 26. Description of Second-Order Formative Constructs.

Description Illustration
1 Construct the first-order latent

variables (LV11 and LV12) and
relate the manifest variables
(LV11:MV1-MV4; LV12:MV5-
MV8) to the formative constructs
using Mode B (formative) in their
outer model. The weight

represents the first-order weights.
Indicator weights are estimated to
validate the formative constructs
as well as calculate the composites
(see Eq. (5)).
2 The second-order latent variable can
now be constructed by relating the
block of the underlying first-order
latent variables (LV11 and LV12)
to the second-order formative
construct (LV21) using Mode B
(formative) in their outer model.
The first-order latent variables
(LV11 and LV12) are now related
to the second-order variable LV21
as formative dimensions and the
weight of these paths represent the
second-order weights (see Eq. (6)).
CE
0.880** 0.870**
PU PEU
PU1 PU2 PU3 PU4 PEU1 PEU2 PEU3 PEU4
**p < 0.01, CE = Cognitive evaluation, PU = Perceived usefulness,

PEU = Perceived ease of use
Fig. 4. Cognitive Evaluation.

et al., 2006) and were significant at p < 0.01. The CR (= 0.939) and AVE
(= 0.66) of the second-order construct were also well above the cut-off
values of 0.7 and 0.5, respectively (see Table 27) (Bagozzi & Yi, 1998;
Gefen et al., 2000). Thus, the conceptualisation of cognitive evaluation as a
higher-order, multidimensional construct seemed justified.
6.4.2. Validation of Second-Order Formative Constructs

The study involved four constructs as higher-order formative constructs,
namely, culture, environmental pressure, country readiness and organisa-

tional performance. As with the first-order formative construct validation,
the indicator weights and the test of multicollinearity were conducted to
assess the validity of second-order formative constructs.
6.4.2.1. Culture (Mixed). Culture was modelled as a second-order

composite variable of five first-order constructs power distance, in-group
collectivism, uncertainty avoidance, Bengali values and ethical culture. All
first-order constructs were reflective except Bengali values which was a
composite of three indicators (see Fig. 5). As shown in Table 28, the paths
from ethical culture and Bengali values were significant while the other three
paths were not found to be significant. Table 24 also showed that multi-
collinearity was non-existent among the five first-order constructs due to
their very small VIF values (Diamantopoulos & Siguaw, 2006). As stated
in Section 6.3.4, all formative indicators regardless of their magnitude
and level of significance were retained to ensure the measurement of the
entire domain and content validity (Bollen & Lennox, 1991) and thus the
hierarchical conceptualisation of culture was justified for the structural
model estimation.
Table 27. Second-Order Cognitive Evaluation Construct and Its

Association with First-Order Components.
Construct Component Items Loading t-value R2 CR AVE
Cognitive evaluation (second- Perceived 4 0.880** 72.57 0.775 0.939 0.66

order reflective) usefulness
Perceived 4 0.870** 47.73 0.758
ease of use
**p < 0.001.

178 MD SHAH AZAM
CUL
0.064 0.183 0.240 0.797** – 0.530**
PD ING UN ECUL BVAL
ING1 ING2 ING3

PD1 PD2 UN1 UN2 EC1 EC2 BV1 BV2 BV3

**p < 0.01, CUL = Culture, PD = Power distance, ING = In-group collectivism,
UN = Uncertainty avoidance, ECUL = Ethical culture, BVAL = Bengali values
Fig. 5. Culture.
Table 28. Second-Order Culture Construct and Its Association with First-
Order Components.
Construct Component Items Weight t-value Tolerance VIF
Culture (second-order Power distance 2 0.064 0.37 0.840 1.191

formative)
In-group 3 0.183 0.99 0.779 1.284
collectivism
Uncertainty 2 0.240 1.45 0.725 1.379
avoidance
Ethical culture 2 0.797** 6.18 0.788 1.270
Bengali values 3 −0.530** 3.68 0.932 1.073
**p < 0.01; VIF = variance inflation factor.
6.4.2.2. Environmental Pressure (FormativeFormative). Environmental

pressure was modelled as a second-order composite variable of four first-
order constructs coercive pressure, normative pressure, mimetic pressure
and global pressure. All of the first-order constructs used as manifest indica-
tors for the second-order formative construct, environmental pressure, were
also formative (see Fig. 6). As shown in Table 29, the paths from coercive
pressure, mimetic pressure and global pressure were significant while norma-
tive pressure did not produce any significant path weight. Multicollinearity
was non-existent among the four first-order constructs as the VIF value was
far below the conservative threshold level of 3.3 given by Diamantopoulos
and Siguaw (2006). Thus, the hierarchical conceptualisation of environmen-
tal pressure was justified for structural model estimation.
EP
0.231**
0.888** 0.139* – 0.040
CP MP NP GP
CP1 CP2 CP3 MP1 MP2 NP1 NP2 NP3 NP4 GP1 GP2
**p < 0.01, EP = Environmental pressure, CP = Coercive pressure, MP = Mimetic pressure,

NP = Normative pressure, GP = Global pressure
Fig. 6. Environmental Pressure.
Table 29. Second-Order Environmental Pressure Construct and Its

Construct Component Items Weight t-value Tolerance VIF
Environmental pressure Coercive 3 0.888** 13.19 0.451 2.216

(second-order formative) pressure
Normative 4 −0.040 0.48 0.475 2.107
pressure
Mimetic pressure 2 0.139* 1.98 0.935 1.069
Global pressure 2 0.231** 3.16 0.864 1.157
**p < 0.01, *p < 0.05; VIF = variance inflation factor.
6.4.2.3. Country Readiness (FormativeFormative). Country readiness

was modelled as a second-order composite variable of five first-order
constructs technology infrastructure, human infrastructure, legal infra-
structure, government policy and supports and financial infrastructure (see
Fig. 7). All first-order constructs of the higher-order model were reflective.
Reliability and validity of the first-order reflective constructs were justified
in Section 6.3.2.1. As shown in Table 30, the paths from technology infra-
structure ðγ = 0:543; t = 3:14; p < 0:01Þand government policy and supports
(γ = − 0:624, t = 3:72, p < 0:01) were significant while human infrastructure
(γ = 0:246, t = 1:10, p < 0:01), legal infrastructure (γ = 0:265, t = 1:62, p >
0.05) and financial infrastructure (γ = − 0:061, t = 0:22, p < 0:01) did not pro-
duce any significant effects. Multicollinearity was not existent among the
180 MD SHAH AZAM
CR
– 0.624**
0.246 0.265 – 0.061
0.543**
TI HI LI FI PS
TI1 TI2 TI3 TI4 HI1 HI2 HI3 HI4 LI1 LI2 FI1 FI2 PS1 PS2 PS3 PS4
**p < 0.01, CR = Country readiness, TI = Technology infrastructure, HI = Human

infrastructure, LI =Legal infrastructure, FI = Financial infrastructure, PS = Government

policy and supports
Fig. 7. Country Readiness.
Table 30. Second-Order Country Readiness Construct and Its

Construct Components Items Weight t-value Tolerance VIF
Country readiness Technology 5 0.543** 3.14 0.721 1.387

(second-order infrastructure
formative)
Human 4 0.246 1.10 0.569 1.757
infrastructure
Legal infrastructure 2 0.265 1.62 0.884 1.132
Government policy 4 −0.624** 3.72 0.994 1.006
and supports
Financial 2 −0.061 0.22 0.566 1.768
infrastructure
**p < 0.01; VIF = variation inflation factor.
five first-order constructs as the VIF values were between 1.006 and 1.768
which was far below the conservative threshold level. Thus, the hierarchical
conceptualisation of environmental pressure was justified for structural
model estimation.
6.4.2.4. Organisational Performance (ReflectiveFormative). The final

dependent variable of the study, organisational performance, was also mod-
elled as a second-order composite variable of three first-order reflective
constructs, competitiveness, internal operation productivity and financial per-
formance (see Fig. 8). As all first-order constructs were reflective, reliability,
convergent validity and discriminant validity were duly justified before
PERF
0.479**
0.566** 0.218**
COMP INTOP FPERF

CM1 CM2 CM3 IO1 IO2 IO3 FP1 FP2 FP3
**p < 0.01, PERF = Performance, COMP = Competitiveness, INTOP = Internal

operation productivity, FPERF = Financial performance
Fig. 8. Organisational Performance.
Table 31. Validity for Second-Order Performance Construct.

Construct Components Items Weight t-value Tolerance VIF
Performance (second- Competitiveness 3 0.566** 7.33 0.888 1.126

order formative)
Internal operation 3 0.218** 2.35 0.447 2.236
productivity
Financial 3 0.479** 4.33 0.444 2.251
performance
**p < 0.01; VIF = variation inflation factor.
estimating the composite variable (see Tables 2123). As shown in

Table 31, the paths from internal operation productivity (γ = 0:218, t = 2:35,
p < 0.05), competitiveness (γ = 0:566, t = 7:33, p < 0.01) and financial perfor-
mance ðγ = 0:479; t = 4:33; p < 0:01Þ were significant. The largest VIF value
of the constructs was 2.251 which confirmed that multicollinearity was
non-existent. Thus, the hierarchical conceptualisation of organisational
performance was valid for structural model estimation.
6.5. Structural Model
To assess nomological validity, a structural model was developed which

comprised five exogenous and six endogenous latent variables. The
182 MD SHAH AZAM
variances associated with the endogenous variables determined the explana-

tory power of the proposed model. The path coefficients and t-values were
calculated to address the effects of the constructs and their underlying rela-
tionships according to the proposed theoretical framework. The hypothe-
sised relationships between the constructs could be calculated by two
methods, namely, ‘bootstrap’ or ‘jackknife’ (Gefen et al., 2000). ‘Bootstrap’
is popularly used within the PLS framework as it produces both a t-value
and an R2 value.
The technique which bootstrapping employs for calculating the t-statis-

tic is similar to the traditional t-test that is also used to interpret the signifi-
cance of the paths between study constructs (Barclay et al., 1995).
Interpreted in a similar way in multiple regression analysis, the R2 value is
also used to indicate the explanatory power of exogenous variables within
a model. In other words, this value estimates the variance associated with
endogenous constructs; thus, the proposed overall model could be evalu-
ated. It is important to note that PLS had some advantages as it was ideal
for assessing the path loadings and structural relationships between the
study constructs which could handle both formative and reflective con-
structs (Chin & Newsted, 1999; Hanlon, 2001): it also did not require the
normal distribution of the data.
6.5.1. Nomological Validity

The nomological validity of the endogenous variables of the model was
examined by their R2 values (Santosa et al., 2005). Falk and Miller (1992)
proposed that the minimum R2 should be 0.10. As shown in Table 32, the
results of the SEM estimation employing a bootstrapping procedure indi-
cated that all the R2 values were above the minimum cut-off value proposed
by Falk and Miller (1992) which, in turn, ensured the nomological validity
of the model. The overall model explained 48.5% of the variance (R2 ) of
Table 32. Nomological Validity of the Endogenous Variables.

Endogenous Constructs R2
Cognitive evaluation 0.138

Expectation 0.585
Usage 0.765
Integration 0.416
Utilisation 0.239
Performance 0.485
The structural equation model estimation further explained that 13.8%

of variance in cognitive evaluation was accounted for by owner innovative-
ness and culture. The estimation also indicated that 58.5% of variance in
expectation was accounted for by cognitive evaluation, owner innovativeness,
culture, environmental pressure and country readiness. Expectation, owner
innovativeness, environmental pressure and country readiness also affected
actual usage with an R2 value of 0.765. The result indicated that 76.5% of
ICT usage behaviour was accounted for by the exogenous variables, expec-
tation, owner innovativeness, environmental pressure and country readiness.

The model finally indicated 41.6% of variance in integration and 23.9% of
variance in utilisation accounted for actual usage.
6.5.2. Tests of Hypotheses

As shown in Tables 32 and 33 (see also Fig. 9), the results depicted signifi-
cant effects of cognitive evaluation on the expectation of ICT usage by
SMEs (γ = 0:155, t = 2:74, R2 = 0:585). Thus, Hypothesis H1 was supported.
Table 33. Results of the Structural Model.

Loading t-value
CE → EXP 0.155** 2.74

OI → CE 0.163* 2.20
OI → EXP −0.067 0.97
OI → USE −0.038 0.85
EP → EXP 0.405** 8.51
EP → USE 0.191** 3.25
FC → EXP 0.294** 5.13
FC → USE 0.070* 1.94
CR → EXP −0.135** 2.83
CR → USE −0.005 0.03
CUL → EXP −0.048 0.71
CUL → CE −0.333** 5.81
EXP → USE 0.697** 12.58
USE → INT 0.643** 19.59
USE → UTL 0.488** 12.49
INT → PERF 0.541** 8.01
UTL → PERF 0.180* 2.74
USE → PERF 0.079 1.09
**p < 0.01, *p < 0.05.

CE = Cognitive evaluation, OI = Owner innovativeness, EXP = Expectation, EP =
Environmental pressure, FC = Facilitating condition, CR = Country readiness, CUL =
Culture, USE = Actual usage, INT = Integration, UTL = Utilisation, PERF = SME
performance.
184
OI
INT R2= 0.416

– 0.038
0.163* (0.85) COMP
(2.20) – 0.066 0.643** 0.541**
(0.97) (19.59) (8.01)
PEU
0.155** 0.697** 0.079
(2.74) (12.58) (1.09)
PU
R2= 0.138 R2=0.585 R2= 0.765
0.488** R2= 0.485
– 0.332** 0.293** 0.180*
– 0.048 (12.49) FPERF
(5.81) (5.13) 0.070* (2.74)
(0.71)
(1.94)
UTL
CUL FC 0.191**
– 0.135**
– 0.005 0.405** (3.25) R2= 0.239
(2.83)
(0.03) (8.51)
PD ING UN BV EC
CR EP
MD SHAH AZAM
** p < 0.01, * p < 0.05, CE = Cognitive evaluation, OI = Owner innovativeness, EXP = Expectation, EP = Environmental pressure,
CUL = Culture, CR = Country readiness, FC = Facilitating condition, USE = Actual usage, INT = Integration, UTL = Utilisation,
PERF = SME performance.
Fig. 9. The Comprehensive Model Estimates.

Hypotheses H2a , H2b and H2c were developed to assess the influence of
owner innovativeness on cognitive evaluation, expectation and ICT use,
respectively. The model estimation depicted a significant effect of owner
innovativeness (γ = 0:163, t = 2:21) on cognitive evaluation.
Owner innovativeness was also hypothesised to have positive correlations
with expectation
and ICT use. The effects of owner innovativeness on
expectation γ = − 0:067; t = 0:97; R2 = 0:585 and its effects on ICT use
(γ = − 0:038, t = 0:85, R2 = 0:765) were not found to be significant. Thus,
Hypothesis H2a was accepted while Hypotheses H2b and H2c were rejected.
The hypothesised relationship between environmental pressure and
expectation as well as environmental pressure and actual ICT use were pos-
tulated in Hypotheses H3a and H3b . Environmental pressure was found to
have significant effects on expectation (γ = 0:405, t = 8:51, R2 = 0:585) and
actual use (γ = 0:191, t = 3:25, R2 = 0:765). Thus, Hypotheses H3a and H3b
were accepted.
The effects of facilitating condition on expectation and actual use were
postulated in Hypotheses H4a and H4b , respectively. The SEM results pre-
sented a significant association between facilitating condition and expecta-
tion (t = 5:13, R2 = 0:585) as well as between facilitating condition and actual
use (γ = 0:070, t = 1:94, R2 = 0:765). Thus, Hypotheses H4a and H4b were
accepted.
As postulated in Hypothesis H5a , the study results supported the asso-
ciation between country readiness and expectation (γ = − 0:134, t = 2:83,
R2 = 0:585) while the results rejected the anticipated association between
country readiness and usage ðγ = − 0:005, t = 0:03, R2 = 0:765). Thus,
Hypothesis H5a was supported while Hypothesis H5b was rejected.
The relationships between culture and two endogenous variables, cogni-
tive evaluation and expectation, were postulated separately in Hypotheses
H6a and H6b . The analysis depicted a significant association between culture
and cognitive evaluation (γ = − 0:333, t = 5:81, R2 = 0:138) while it showed
the opposite result to what was postulated in Hypothesis H6b (γ = − 0:048,
t = 0:71, R2 = 0:585). Thus, Hypothesis H6a was accepted and Hypothesis
H6a was rejected.
The effects of expectation on actual use are postulated in Hypothesis H7 .
The results depicted a strong and significant effect of expectation on actual
use (γ = 0:697, t = 12:58, R2 = 0:765). Thus, Hypothesis H7 was accepted.
Hypotheses H8a , H8b and H8c , respectively, postulated the effects of ICT
usage on ICT integration, ICT utilisation and SMEs’ performance. The
model estimates showed a significant association between actual ICT usage
and ICT integration ðγ = 0:623; t = 12:58; R2 = 0:416Þ, they also showed
186 MD SHAH AZAM
significant effects on ICT utilisation (γ = 0:488, t = 2:74, R2 = 0:239). The

results indicated that actual ICT usage was not significantly associated with
SMEs’ performance (γ = 0:079, t = 1:09, R2 = 0:485). Thus, Hypotheses H8a
and H8b were accepted while Hypothesis H8c was rejected.
As postulated in Hypothesis H9 , the study results showed that ICT inte-
gration had significant effects on SMEs’ performance (γ = 0:541, t = 8:01,
R2 = 0:485). Thus, the relationship between ICT integration and SMEs’ per-
formance as postulated in Hypothesis H9 was accepted.
Hypothesis H10 anticipated significant effects of ICT utilisation on

SMEs’ performance. The study results also indicated a significant associa-
tion between ICT utilisation and SMEs’ performance (γ = 0:180, t = 2:74,
R2 = 0:485). Thus, Hypothesis H10 was accepted. The results of the hypoth-
eses tests are summarised in Table 34 and also illustrated in Fig. 9.
6.5.3. Test for Mediating Role of Integration and Utilisation

A system embraces the causes and effects among a number of variables
that produce the final outcome through a process. Like predictors or inde-
pendent variables, the process itself has some effects on the systems which,
Table 34. Results of Hypotheses Tests.

Hypotheses Comments
H1 Cognitive evaluation has direct and positive influence on expectation Accepted

H2a Owner innovativeness has direct and significant effects on cognitive Accepted
evaluation
H2b Owner innovativeness has direct and significant effects on expectation Rejected
H2c Owner innovativeness has direct and significant effects on use of ICT Rejected
H3a Environmental pressure has a significant influence on expectation Accepted
H3b Environmental pressure has a significant influence on actual use Accepted
H4a Facilitating condition has a significant influence on expectation Accepted
H4b Facilitating condition has a significant influence on actual use Accepted
H5a Country readiness has a significant influence on expectation Accepted
H5b Country readiness has a significant influence on use of ICT Rejected
H6a Culture has a significant influence on cognitive evaluation Accepted
H6b Culture has a significant influence on expectation Rejected
H7 Expectation has a significant influence on actual usage Accepted
H8a ICT usage has significant influence on integration Accepted
H8b ICT usage has a significant influence on utilisation Accepted
H8c ICT usage has a significant influence on performance Rejected
H9 ICT integration has a significant influence on performance Accepted
H10 ICT utilisation has a significant influence on performance Accepted
in turn, is reflected in the outcome. Researchers try to understand the

effects of the process on the possible outcome by mediation analysis.
Mediation exists when a predictor affects a dependent variable indirectly
through at least one intervening variable, or mediator. Fig. 10 illustrates
mediation models with single and multiple intervening variables.
The concept of mediation analysis is not new, it can be traced back to
the early genetics work of Wright (1921) followed by the stimulus response
models of Woodworth (1928) in psychology. Although the study of media-
tion has a long history (Hyman, 1955; MacCorquodale & Meehl, 1948), it
has become popular with the creation of basic measurement approaches for
assuming the effect of mediation (Baron & Kenny, 1986; Judd & Kenny,
1981; Lindley & Walker, 1993). It has, furthermore, become easier with the
application of SEM (Bollen, 1989).
In the basic measurement approach, Baron and Kenny (1986) and Judd
and Kenny (1981) discussed four characteristics of mediation that are also
used as the following steps for probing whenever there exists any mediating
relationship in a recursive model:
Step 1: The initial variable has significant correlation with the outcome.
If Y is considered as a criterion variable and X as the initial variable or pre-
dictor, the variations in levels of the initial variable, such as X, should sig-
nificantly account for the variations in the criterion variable, such as Y.
Y = b0 þ cX þ ɛ ð7Þ
where b0 = constant, X = initial variable, c = regression effect of X on Y,

ɛ = a disturbance term.
M1
M a1
b1
a b
X Y
c or c´
X Y a2
c or c´ b2
M2
Mediation model (single mediator) Mediation model (multiple mediators)
X = Independent variable, Y = Dependent variable, M = Mediating variable
Fig. 10. Mediating Model.

188 MD SHAH AZAM
Step 2: The variations in levels of the initial variable should significantly

account for variations in the presumed mediator. This step essentially
involves treating the mediating variable as if it was an outcome variable.
M = b0 þ aX þ ɛ ð8Þ
where b0 = constant, M = mediating variable, a = regression effect of X on

M, ɛ = a disturbance term.
Step 3: The variations in the mediating variable should significantly

account for variations in the criterion variable.14
Y = b0 þ bM þ ɛ ð9Þ
where b0 = constant, M = mediating variable, b = regression effect of M on

Y, ɛ=a disturbance term.
Step 4: In complete mediation, the relationship between the initial vari-
able and criterion variable is no longer significant (in control of the mediat-
ing variable).
Y = b0 þ c0 X þ bM þ ɛ ð10Þ
where b0 = constant, X = initial variable, M = mediating variable, c0 =

regression effect of X on Y (in control for M), b = regression effect of
M on Y, ɛ = a disturbance term.
The purpose of Steps 13 is to establish that zero-order relationships
among the variables exist. If all the relationships explained in Steps 13
(Eqs. 79) are seen to be significant, researchers anticipate that the third
variable M may act as a process variable which may explain the effects of
the initial variable X on the criterion variable Y. The variable M is then
called a mediating variable or process variable. Assuming there are signifi-
cant relationships from Step 1 to Step 3, one proceeds to Step 4. In Step 4,
some form of mediation is supported if the effect of M (path b) remains
significant after controlling for X. If X is no longer significant when M is
controlled, the finding supports full mediation. If X is still significant (i.e. X
and M both significantly predict Y), the finding supports partial mediation.
If the mediation process involves only one mediating variable, it is
termed simple mediation. More than one variable may perform a mediating
role in a causal model. Thus, mediation processes involving more than one
intervening variable, termed ‘multiple mediation’, may also be analysed
through the above-mentioned procedure.
6.5.3.1. Mediating Role of ICT Integration. Table 35 and Fig. 11 illustrate

that Model 1 shows a direct relationship between the initial variable ICT
use and the criterion variable SMEs’ performance (c = 0:511, t = 9:94);
Model 2 shows a relationship between the initial variable ICT use and the
mediating variable ICT integration (a1 = 0:640, t = 13:94), while Model 3

shows that the mediating variable ICT integration has significant effects on
the criterion variable SMEs’ performance ðb1 = 0:696; t = 16:21). Since all
of the above-mentioned relationships are significant, it may be assumed
that ICT integration may perform a mediating role in explaining the rela-
tionship between ICT use and SMEs’ performance. To further prove
whether any mediating relationship exists and, if it exists, what type of
mediation it is, the researcher should estimate the regression stated in
Step 4. As shown in Table 35, Model 4 illustrates a significant effect of ICT
integration on SMEs’ performance (b1 = 0:625, t = 11:24): the effects of ICT
use on SMEs’ performance is also significant (c0 = 0:111, t = 1:99). It is
important to note that, although the effects of ICT use on SMEs’ perfor-
mance are significant (in control of ICT integration), the magnitude of the
effect is reduced in comparison to the direct relationship between ICT use
and SMEs’ performance (c = 0:511, t = 9:94). Thus, in applying the basic
concept of mediation analysis (Baron & Kenny, 1986; Judd & Kenny,
1981), it is observed that ICT integration plays a partial moderating role in
explaining the relationship between ICT use and ICT integration (see
Fig. 11).
Table 35. Mediating Role of Integration in Explaining Performance.

Model 1 Model 2 Model 3 Model 4 Comments
0
c or c 0.511** 0.111* Partial mediation
(9.94) (1.99)
a1 0.640**
(13.94)
b2 0.696** 0.625**
(16.21) (11.24)
R2 0.261 0.410 484 0.491
**p < 0.01, *p < 0.05.

190 MD SHAH AZAM
INT
R2= 0.410
0.640**
USE PERF
0.511** USE (13.94)
(9.94) R2= 0.261
Model 2
Model 1
INT INT
0.625**
0.696** (11.24)
(16.21)
USE PERF
PERF 0.111*
(1.99) R2= 0.491
R2= 0.484
Model 3
Model 4
**p < 0.01, USE = ICT use, INT = Integration, PERF = Performance
Fig. 11. Mediating Effects of Integration.
The significance of indirect effects is examined by the z statistic (Sobel,

1982). The z value is formally defined as follows:
a×b
z = pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð11Þ
b × s a þ a2 × s b 2 þ s a 2 × s b 2
2 2
0:640 × 0:696
z = qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi = 10:56
ð0:696Þ2 × ð0:0459Þ2 þ ð0:640Þ2 × ð0:0429Þ2 þ ð0:0459Þ2 × ð0:0429Þ2
The results support the mediating effects of ICT integration which implies
that it has an indirect influence on SMEs’ performance. The variance
accounted for (VAF) value is used to estimate the ratio of the indirect effects.
a×b
VAF = = 87:17 ð12Þ
a×bþc
The VAF value indicates that 87.17% of the total effect of ICT use
on SMEs’ performance is explained by indirect effects through ICT
integration.
6.5.3.2. Mediating Role of ICT Utilisation. The study contained another

conceptualisation of ICT usage which was ICT utilisation, in examining
SMEs’ performance. ICT use is the measure of the assessment of using
different levels of ICT applications while ICT utilisation is the measure of
the assessment of the degree of ICT utilisation among different functional
areas of the organisation. The study assumed that using only ICT usage
may not adequately explain organisational performance in the absence
of ICT utilisation throughout the organisation. As shown in Table 36,
Model 1, Model 2, Model 3 and Model 4 illustrate the results that may be
concluded about any mediating relationship among the three latent
variables (see Fig. 12).
Model 1 shows the direct relationship between the initial variable ICT
usage and the criterion variable SMEs’ performance (c = 0:511, t = 9:94);
Model 2 shows the relationship between the initial variable and the mediat-
ing variable ICT utilisationða2 = 0:283, t = 9:22) and Model 3 shows that the
mediating variable ICT utilisation had significant effects on the criterion
variable SMEs’ performance ðb2 = 0:530, t = 10:45). Since all of the above-
mentioned relationships are significant, it may be assumed that ICT utilisa-
tion may perform a mediating role in explaining the relationship between
ICT usage and SMEs’ performance. To further prove whether any mediat-
ing relationship exists and, if it exists, what type of mediation it is, the
researcher should estimate the regression stated in Step 4 (as stated in
Section 6.5.3). As shown in Table 36, Model 4 illustrates a significant effect
of ICT utilisation on SMEs’ performance (b2 = 0:369, t = 6:78): the effects of
ICT use on SMEs’ performance is also significant (c0 = 0:333, t = 6:11). It is
important to note that although the effects of ICT use on SMEs’ perfor-
mance is significant (in control of ICT utilisation), the magnitude of the
effect is largely reduced in comparison to the direct relationship between
Table 36. Mediating Role of ICT Utilisation in Explaining Performance.

Model 1 Model 2 Model 3 Model 4 Comments
c or c0 0.511** 0.333** Partial mediation

(9.94) (6.11)
a2 0.283**
(9.221)
b2 0.530** 0.369**
(10.45) (6.78)
R2 0.261 0.233 0.230 0.365
**p < 0.01.

192 MD SHAH AZAM
USE
0.283**
(9.221)
USE PERF
0.511** UTL
(9.94) R2= 0.261 R2= 0.233
Model 1 Model 2
PERF
USE PERF
R2= 0.230 0.333**
0.530** (6.11) R2= 0.365
UTL (10.45)
0.369**
(6.78)
UTL
Model 3 Model 4
**p < 0.01, USE = ICT use, UTL = Utilisation, PERF = Performance
Fig. 12. Mediating Effects of Utilisation.
ICT usage and SMEs’ performance (c = 0:511, t = 9:94). Thus the results
indicate a partial mediation of ICT utilisation in explaining the relationship
between ICT use and SMEs’ performance (Baron & Kenny, 1986; Judd &
Kenny, 1981).
The significance of the indirect effects of ICT use on SMEs’ performance
through ICT utilisation (mediating variable) was examined by the z statistic
(Sobel, 1982). The results supported the mediating effects of ICT utilisation
(z = 5:39, VAF = :238Þ which implies that it has an indirect influence on
SMEs’ performance.
The VAF value indicated that 23.8% of the total effect of ICT use
on SMEs’ performance was explained by indirect effects through ICT
utilisation.
Through understanding the above-mentioned characteristics and apply-
ing the steps from Baron and Kenny (1986), it has been assumed that ICT
use influences the performance growth of SMEs through a process which
comprises ICT integration and ICT utilisation as intervening variables.
Baron and Kenny’s (1986) steps also helped to detect whether there was
any mediating relationship and also answered the question of whether the
target variable was feasible for mediation analysis. The process produced
reliable estimates of indirect effects and their power through multiple
regression analysis. However, researchers face difficulties when more than

one intervening variable is involved in a process in a causal model.
6.5.3.3. Mediating Roles of ICT Integration and ICT Utilisation. Different

types of mediation occur in a causal set-up such as simple mediation and
multiple mediation. Simple mediation involves the analysis of the effects of
independent variables on a criterion variable by a single intervening vari-
able. Multiple mediation involves the relatively complex analysis of the
effects of independent variables on a criterion variable by multiple media-

tors or intervening variables.
Although the analysis of mediation has received the bulk of researchers’
attention in behavioural studies, most of the previous initiatives have
focused on simple mediation analysis, only a few authors have focused on
the simultaneous testing of multiple indirect effects which is a worthwhile
method for addressing the complex mediated relationship (e.g. Bollen,
1987, 1989; Brown, 1997; MacKinnon, 2000; West & Aiken, 1997).
In multiple mediation, the specific indirect effect of X on Y via mediator
i (Brown, 1997) is defined as the product of the two unstandardised paths
linking X to Y via those mediators. For example, the specific indirect effect
of X on Y through M1 is quantified as a1 b1 . If another variable M2 is
included in the model as a simultaneous mediator, the specific indirect
effect of X on Y through M2 is quantified as a2 b2 . The complex relationship
in a simultaneous multiple mediator model is estimated by utilising the
formulas as stated in Eqs. (13) and (14).
X
j
The total indirect effect = ai bi ð13Þ
i=1
X
j
Total effect c = c0 þ ai bi ð14Þ
i=1
Total indirect effects may also be calculated by c − c0 .

The conceptual model of the study included the simultaneous multiple
mediators ICT integration and ICT utilisation to address the impact of ICT
use on SMEs’ performance. The preceding subsections have addressed
the effects of two mediators when applying Baron and Kenny’s (1986)
procedure by multiple regression separately with a single mediator.
194 MD SHAH AZAM
The emergence of SEM created remarkable progress in mediation analysis.

SEM, a second generation statistical tools, produces multiple mediators’
direct and indirect effects in a causal framework. CBSEM has been widely
used as an analytical tool for assessing mediating effects in a common and
integrated platform. As SEM produces comprehensive results, Baron and
Kenny’s (1986) four-step procedure was not entirely required for the analysis.
With the emergence of sophisticated SEM software, mediation analysis
has not only become popular, but new and easier ways have been provided
for obtaining estimates to examine mediating effects. The new procedure

considers three different models no mediation, complete mediation and
partial mediation for mediation analysis. The values of different para-
meter estimates and fit indices are the criteria for comparison to select the
preferred model. In order to prove the mediating effects, this study’s pre-
viously estimated model was restructured by developing three different
models to investigate the mediating roles of ICT integration and ICT utili-
sation on organisational performance.
In the complete mediation model, ICT use was hypothesised to have
only an indirect effect on SMEs’ performance through ICT integration and
ICT utilisation. The partial mediation model anticipated significant direct
and indirect effects while the no-mediation model involved assessment of
the predictor variables’ impact only on the outcome variable.
In CBSEM, researchers estimate the model’s extension as a no-
mediation model, full mediation model and partial mediation model. The
study thus estimated different fit indices to examine which model would
produce acceptable and high fit indices. The strength of different fit indices
was the basis for the preferred model selection which also justified the
pattern of mediation. CAIC (comprehensive Akaike information criteria),
chi-squared, comparative fit index (CFI), normed fit index (NFI), incre-
mental fit index (IFI), root mean square error of approximation (RMSEA)
and adjusted goodness-of-fit index (AGFI) statistics were estimated to
select the preferred model. CBSEM can compute the fit indices and easily
compare different models while component-based SEM only uses R2 to
anticipate the variations in the model. A different form of fit index,
Goodness of fit (GoF), a global fit measure, is conducted for PLS path
modelling. The GoF value is computed by the average communality and
average R2 for the endogenous constructs (Tenenhaus, Vinzi, Chatelin, &
Lauro, 2005).
pffiffiffiffiffiffiffiffiffiffiffi
2
GoF = AVE × R ð15Þ
The estimated GoF values may be used to assess the global validation of
PLS models. The validation of the PLS model was examined by comparing
the GoF values with different cut-off values set for the model with different
explanatory power (Wetzels et al., 2009). The GoF values followed the basic
three cut-off values as:
GoFsmall = 0:1; GoFmedium = 0:25; GoFlarge = 0:36:

The high GoF values suggested a satisfactory fit which may help to select
the preferred model for assessing the form of mediation with adequate
support to globally validate the PLS model.
As shown in Table 37 (see also Fig. 13), Model 1 (no-mediation model),
Model 2 (complete mediation model) and Model 3 (partial mediation model)
have GoF values of 0.405, 0.467 and 0.466, respectively. Thus, each of the
models exceeded the cut-off value of 0.36 for large sizes of R2 which, in turn,
provided adequate supports for global validation of the PLS models.
Model 2 showed significant direct effects of ICT use on SMEs’ perfor-
mance ðc = 0:556, t = 15:18, R2 = 0:309). The complete mediation model
showed significant effects of the initial variable on both the intervening
variables. The initial variable ICT use had a significant effect on the media-
tor, ICT integration (a1 = 0:640, t = 17:65, R2 = 0:41), and also had a
Table 37. Mediating Roles of ICT Integration and Utilisation.

Model 1 (No Model 2 (Full Model 3 (Partial Comments
Mediation) Mediation) Mediation)
c or c0 0.556** na 0.074 Complete

(15.18) (1.13) mediation
a1 na 0.640** 0.640**
(17.65) (20.80)
b1 na 0.589** 0.549**
(12.16) (8.24)
a2 na 0.494** 0.491**
(12.06) (11.86)
b2 na 0.182** 0.169**
(3.59) (2.78)
R2 0.309 0.41 (M1) 0.41 (M2)
0.244 (M2) 0.241 (M2)
0.507 (Y) 0.509 (Y)
GoF 0.405 0.467 0.466
**p < 0.01.

196 MD SHAH AZAM
USE PERF
0.556**
(15.18) R2= 0.309
Model 1 (No Mediation)
INT INT
R2= 0.410 R2= 0.410

0.640** 0.589** 0.640** 0.549**
(17.65) (12.16) (20.80) (8.24)
USE PERF USE PERF
0.074
R2= 0.507 (1.13) R2= 0.509
0.494** 0.182** 0.491** 0.169**

(12.06) UTL (3.59) (11.86) UTL (2.78)
R2= 0.244 R2= 0.241
Model 2 (Full Mediation) Model 3 (Partial Mediation)
**p < 0.01, USE = ICT use, INT = Integration, UTL = Utilisation, PERF = Performance
Fig. 13. Mediating Effects of Integration and Utilisation.
significant effect on another mediator ICT utilisation (a2 = 0:494, t = 12:06,

R2 = 2:44). The model again projected the significant effects of the media-
tors, ICT integration (b1 = 0:589, t = 12:16) and ICT utilisation (b2 = 0:182,
t = 3:59) on the criterion variable SMEs’ performance.
The partial mediation model also showed the significant effects of ICT
use on the two mediators ICT integration and ICT utilisation. The result
also indicated the significant influence of the two mediators (ICT integra-
tion and ICT utilisation) on SMEs’ performance while the effect of ICT use
on SMEs’ performance was seen to be non-significant (c0 = 0:74, t = 1:13,
R2 = 0:509). The combined model showed a reduced strength of c, which
was also insignificant. Thus, the analyses posited a complete mediating pro-
cess in assessing the effects of ICT use on SMEs’ performance through two
intervening variables, ICT integration and ICT utilisation (see Table 38).
Further indirect effects of ICT use on SMEs’ performance through ICT
integration ða1 × b1 = 0:377) were higher than the effects through ICTP utili-
j
sation (a2 × b2 = 0:09). The total indirect effect was given by i=1
ai bi = 0:467.
The hypotheses tests results for mediating effects of Integration and
Utilisation on SME performance are summarised in Table 38.
The study results showed the significant effects of ICT usage on SMEs’
performance in a complete mediational process which involved the multiple
intervening variables, ICT integration and ICT utilisation. Interestingly,
ICT usage significantly affected SMEs’ performance in a single mediational
Table 38. Summary of Hypotheses Tests for Mediating Effects of

Integration and Utilisation.
Hypotheses Comments
H11a Integration plays a significant mediation role in explaining the Partial mediation
H11b Utilisation plays a significant mediation role in explaining the Partial mediation
effects of ICT usage on SME performance
H11c Integration and utilisation jointly play a significant mediation role Complete mediation
in explaining the effects of ICT usage on SME performance
process. That is, in the no-mediation model, the effect of ICT usage on
SMEs’ performance was observed to be significant. The effects of ICT
usage were reduced in the single mediation process (either ICT integration
or ICT utilisation) which vanished in the multiple mediation process in the
presence of both ICT integration and ICT utilisation as mediators.
6.6. Summary
This section has presented the results of the quantitative analysis of a

nation-wide survey conducted in Bangladesh to study the diffusion of ICT
and its role in generating the performance of SMEs. The study employed
PLS-SEM) with a data set of 282 SMEs in Bangladesh used for data analy-
sis. The nature of the study (practical and exploratory), and the nature of
the latent variables under study (reflective and formative) justified the use
of the PLS technique as the main analytical tool. The procedures of data
collection, data screening by investigating possible biases and data analysis
were elaborated in detail.
The analysis of the data by PLS was performed in two stages (assess-
ment of the measurement model and structural model estimation) which
were reported separately. The measurement model was assessed by estimat-
ing the convergent validity and discriminant validity of the constructs.
The convergent validity of the reflective indicators was examined by item
loadings while weights of the items and absence of multicollinearity among
them were the criteria for the formative indicators. CR and AVE were
also computed to test convergent validity. The study used 0.5, 0.7 and 0.5
as threshold levels (cut-off values) for the item loadings, CR and AVE,
respectively. The study examined the square root of AVE and inter-
construct correlations for construct level discriminant validity, and the item
198 MD SHAH AZAM
cross-loading matrix for item level discriminant validity of the study con-
structs. The hypotheses of the study were tested by estimating the structural
model. The magnitude and degree of the estimates were examined by path
coefficient ðγÞ and critical ratio (t-statistic) while the nomological validity
was examined by R2 . The structural model explained 58.5% of variance in
expectation while 76.5% and 48.5% of variance were explained in ICT use
and organisational performance, respectively. The estimates nullified five
hypotheses while all other hypotheses developed in Section 5 were accepted.
The implications of these results and outcomes are discussed in Section 7.
7. DISCUSSION AND INTERPRETATION
7.1. Introduction
This section presents the discussion of the results and interpretation of the
quantitative data analysis. The premise of the data analysis was developed
by anticipating the structural relationships between endogenous and exo-
genous variables. The estimated results, through applying PLS-SEM),
showed the degree and magnitude of the relationships between the con-
structs under study. The discussion of the results and their interpretations
was performed based on the anticipated hypotheses and statistical estima-
tions. The hypotheses of the research and test results were shown in
Sections 5 and 6, respectively. This section discusses and debates the results
in light of the respective hypotheses.
7.2. Hypothesis Related to Cognitive Evaluation
7.2.1. Hypothesis H1
The TRA (Fishbein & Ajzen, 1975) suggests that overall affective evalua-
tion, that is, one’s beliefs (cognitive) towards performing a behaviour and
respective evaluations (affective) heavily influence the individual’s intention
of performing the behaviour. The evaluation is made based on one’s cogni-
tive evaluation of the consequences of that behaviour (Ajzen & Fishbein,
1980; Sheppard et al., 1988).
Based on the notion of the TRA and supported by previous studies, it
was anticipated that cognitive evaluation would have a direct and positive
influence on expectation (Mehta & Sivadas, 1995; Reynolds, 1974;
Schlosser et al., 1999). The findings of the study revealed that cognitive eva-
luation (γ = 0:155) was directly related to expectation. This finding was con-
sistent with past studies (Liao & Cheung, 2001; Mehta & Sivadas, 1995;
Reynolds, 1974; Schlosser et al., 1999) and also complied with the notion
of the TRA and the TPB.
With a strong positive and significant influence on expectation, cognitive
evaluation played an important role in ICT adoption by SMEs. The
result indicated that highly positive evaluation, that is, firms’ favourable
perception about ICT created positive expectation regarding the use or

implementation of the technology. From the model, it was observed that
the construct cognitive evaluation was a compound higher-order latent vari-
able constructed by the reflections of two perceptual beliefs, perceived ease
of use and perceived usefulness. These two factors were internal and devel-
oped through the adopter’s own knowledge and perceptions about ICT.
This result has some implications for existing and potential users of ICT
as well as for policy makers. Policy makers may be interested to know how
cognitive evaluation could be developed in a positive way to boost the
general level of ICT usage by SMEs in Bangladesh.
7.3. Hypotheses Related to Owner Innovativeness
Applying the notion of the DOI theory (Rogers, 1983, 1995, 2003), it has
been argued that the first adopters of an innovation are innovators them-
selves who usually try a new idea, concept or product first and take the
associated risks. Their good or bad experiences are transmitted to other
potential adopters through the communication channels in an innovation
diffusion process. The innovator’s first experience of an innovation works
as a reference which attracts different potential adopters and an innovation
finally gains popularity. In support of the past literature, the field study
justified three different hypotheses related to owner innovativeness.
7.3.1. Hypothesis H 2a
The estimates showed a significant positive association between owner inno-
vativeness (γ = 0:163) and cognitive evaluation. Traditionally, the literature
has suggested a positive link between adopter innovativeness and intention
and/or actual use. This hypothesis was developed through the field study
analysis. The finding of the structural model estimation was in line with the
field study and was also consistent with the basic assumption of the DOI
theory (Rogers, 2003).
200 MD SHAH AZAM
The result indicated that innovative owners or CEOs favourably evalu-

ate the usefulness and operational ease of ICT. An innovative owner or
CEO usually generates new ideas or concepts for the development or
sustainability of the organisation and has the ability to bear any risk
associated with implementing those ideas or concepts. Thus, although in
general, SME owners anticipate various constraints in accepting ICT, by
their nature, innovative owners positively evaluate the outcome of the
innovation and proceed to have actual experience. This result has some
implications for the Bangladeshi government, policy makers and concerned

organisations who are involved in promoting ICT usage among SMEs.
Innovative SME owners may play a significant role, as a reference group,
in fostering the adoption rate of ICT by SMEs in Bangladesh.
7.3.2. Hypothesis H 2b
It was anticipated that owner innovativeness would have direct and signifi-
cant effects on expectation. The estimation rejected the hypothesis by
revealing a non-significant association between owner innovativeness (γ =
− 0:067) and expectation. Surprisingly, the result was not consistent with
the field study nor with past studies (Thong, 1999; Thong & Yap, 1995).
The result indicated that owner or CEO innovativeness had no impact on
expectation. In general, the innovative owner or CEO, by their nature, initi-
ates things differently; thus, they are not hesitant to try an innovation. The
innovators bear the risks associated with the adoption of an innovation
and use the innovation at the first stage (Rogers, 2003) which implies that
innovative owners highly intend to use an innovation. This result may have
happened because the research model included expectation as an immediate
antecedent of actual usage which was explained as a state of rational inten-
tion. Unlike intention, expectation is an expressed state of willingness which
is formed by evaluating many internal, external and environmental factors.
Thus, although previous studies had found that owner innovativeness had a
strong and significant effect on intention (Thong, 1999; Thong & Yap,
1995), this study, logically, did not find a significant relationship between
them. The adoption of ICT was highly associated with the resources
and operational skills (Venkatesh et al., 2003, 2008) of the organisation as
well as the compatibility (Rogers, 2003) of the ICT with existing systems
and procedures. Thus, innovative owners, although holding new ideas
and positive evaluations about the technology, consider their ability in
terms of resources and operational skills as well as compatibility which
may produce a realistic expectation whereas innovativeness has no signifi-
cant impact.
7.3.3. Hypothesis H 2c
The estimation revealed a non-significant association between owner inno-
vativeness (γ = − 0:038) and SMEs’ actual usage of ICT. Contrary to the
field study outcome and previous studies, this finding indicated that owner
innovativeness had no impact on the usage of ICT (Thong & Yap, 1995). It
may be argued that, unlike innovations which are non-technical in nature
that may be adopted and used by the will or preference of an entrepreneur,
technological innovations particularly ICT require competent human
resources, technology resources and a compatible working environment. As

the technology involves operation of the Internet, the use, that is, the level
of ICT application and depth of use, is related to its compatibility with cus-
tomers, suppliers, regulatory authorities, partners and other stakeholders.
This study investigated SMEs in Bangladesh where small enterprises
face many problems regarding their resources and are mostly operated
in an informal manner. The Bangladeshi government, although aiming to
develop a digital-based economy by 2021, runs with a manual and tradi-
tional working environment in different departments and functional areas.
Thus, SMEs, having the required facilities and resources, are also less
inclined to use technology.
Furthermore, the SMEs that are involved in international operations are
compelled to use ICT as their foreign counterparts require their ICT com-
petency, particularly their online participation, and in this situation it is
not important whether the entrepreneurs are innovative.
7.4. Hypotheses Related to Environmental Pressure
It was anticipated that environmental pressure would have a significant
influence on expectation. The structural model estimation showed a
direct significant association between environmental pressure (γ = 0:405) and
expectation which was consistent with previous studies (Ajzen, 1985;
Kuan & Chau, 2001; Teo et al., 2003). Environmental pressure was
measured as a composite of coercive pressure (DiMaggio & Powell,
1983; Quaddus & Hofmeyer, 2007); mimetic pressure (Premkumar &
Ramamurthy, 1995; Teo et al., 2003); normative pressure (Teo et al., 2003;
Venkatesh et al., 2003) and global pressure (field study); thus, this finding
indicated the overall impact of environmental pressure on expectation. It
was argued that firms that received positive experience about the use of
ICT from their suppliers, partners or other stakeholders were highly
202 MD SHAH AZAM
inclined to adopt the same technology in their organisation to reap the

advantages of using ICT. The results also suggested that customers’ expec-
tations and competitors’ ICT usage’s trends and patterns influence SMEs’
ICT usage expectation. This study reported a unique environmental
influence which may arise from ongoing technological development and
globalisation.
Following the notion of the TOE framework (Tornatzky & Fleischer,

1990), it was anticipated that environmental pressure would have a signifi-
cant influence on SMEs’ actual use of IT. The results of the structural
model revealed a positive and significant effect of environmental pressure
(γ = 0:191) on actual usage behaviour which was consistent with the TOE
framework and other past studies (Teo et al., 2003; Tornatzky & Fleischer,
1990; Zhu & Kraemer, 2005). As with Hypothesis H3a, the overall impact
of environmental pressure was anticipated by Hypothesis H3b. The result
indicated that SMEs’ actual use of ICT was largely influenced by environ-
mental pressure arising from competitors (mimetic pressure); partners/custo-
mers (normative pressure); demands from the regulatory authority or
dominant customers/suppliers (coercive pressure); or technological develop-
ment/globalisation (global pressure). The results suggested that firms that
have had good experiences using ICT with their close stakeholders, partner
organisation and suppliers and whose customers expected to see them using
technology would use ICT. The trend, status and success of competitors’
ICT use also influenced SMEs’ level of ICT use. Most importantly, SMEs
used ICT due to pressure from their dominant customer organisations or
suppliers. This was very common in the RMG industry. Most of the
customers or suppliers of Bangladeshi RMG organisations are large
internationally reputable companies. They are completely ICT-based. They
communicate, interact and transact with all parties online. Without ICT
competency and online participation, it is not possible to interact with
them. Thus, the RMG organisations had no choice but to use ICT to
become compatible with their customers or suppliers and also to match the
competition.
Globalisation has created new opportunities for business organisations.
SMEs have become more international to gain the competitive advantage
from being active in international business. ICT, particularly the Internet,
offers easy access to the international arena and ensures effectiveness and
reliability in communication, negotiations and transactions.
7.5. Hypotheses Related to Facilitating Condition
The estimation showed a significant correlation between facilitating condi-
tion ðγ = 0:294Þ and expectation which was consistent with past research
studies (Ajzen, 1985; Taylor & Todd, 1995c; Thompson et al., 1991;
Venkatesh et al., 2008) and the field study. The operationalisation of the
construct facilitating condition emphasised external resources such as orga-
nisational, technological and human resources. Facilitating condition

referred to the degree to which an individual believed that an organisa-
tional and technical infrastructure existed to support the use of the system.
The result indicated that the SMEs with the necessary technical and organi-
sational resources (e.g. hardware and Internet connectivity, ICT-competent
employees, compatible procedures, etc.) had high expectation of using ICT.
In accordance with the UTAUT model (Venkatesh et al., 2003), it was
anticipated that facilitating condition would have a significant influence on
SMEs’ actual use of ICT. The estimation showed a significant association
between facilitating condition ðγ = 0:70Þ and actual ICT usage behaviour.
The results indicated that firms with adequate hardware and network con-
nectivity, competent employees and compatible working environments
were likely to use ICT. The results also indicated that firms lacking some
technical and organisational infrastructure were logically less inclined to
use ICT. This result supported the findings of past researchers (Gupta
et al., 2008; Venkatesh et al., 2003).
7.6. Hypotheses Related to Country Readiness
In reference to Rogers (1983), Tornatzky and Fleischer (1990), Caselli and

Coleman (2001) and Zhu and Kraemer (2005), it was anticipated that coun-
try readiness would have a significant influence on SMEs’ expectation
(Hypothesis H5a ) and actual usage (Hypothesis H5b ) of ICT.
The structural model estimation showed a significant association between
country readiness (γ = − 0:135) and SMEs’ expectation to use ICT. The
result indicated that overall country readiness was an important predictor
of expectation which was consistent with the findings of previous studies
204 MD SHAH AZAM
(Marques et al., 2011; Molla & Licker, 2005). Country readiness, an aggre-
gated variable, explained the overall effects of contextual factors as the
construct was a higher-order composite of technology infrastructure, human
infrastructure, legal infrastructure, financial infrastructure, and government
policy and supports. The overall effect of country readiness was negative
which meant that the overall infrastructure of the country was not favour-
able for the adoption of ICT particularly from the SMEs’ perspective. This
result may be debated as Bangladesh is now in its infancy in the implemen-
tation of ICT in different sectors. Although the government of the country

is dedicated to establishing a digital-based society by 2021 by applying the
ICT potential in economic development and governance, the responses
from different parties have not been very satisfactory.
For example, Internet penetration of Bangladesh on June 2012 was
calculated as 0.5% which was very low in comparison to many developed
and developing countries.15 The Bangladeshi government is trying hard to
overcome these challenges by providing subsidies and grants, initiating
favourable ICT policy and tax rebates, and also providing computer hard-
ware at a lower price. However, ICT users are facing several constraints,
such as coverage of the Internet network, Internet speed, power supply,
legislative supports in different kinds of online disputes and lack of
online banking facilities which restrict them from using ICT and establish-
ing ICT-dependent working environments.
The structural model estimation did not find any significant association
between country readiness (γ = − 0:005) and SMEs’ actual ICT usage
behaviour. The result was inconsistent with previous studies (Dewan &
Kraemer, 2000; Zhu, Dong, et al., 2006; Zhu & Kraemer, 2005).
Surprisingly, the aggregated impact of the contextual factors was not found
to be significantly related to SMEs’ ICT use while it had significant nega-
tive effects on expectation. It is argued that this result occurred because the
Bangladeshi infrastructure, namely, technology infrastructure, human infra-
structure, legal infrastructure, financial infrastructure or government policy
and supports negatively affected SMEs’ expectation of using ICT. That is,
the SMEs that were willing to adopt ICT were afraid about the outcome
of the investment when evaluating the overall infrastructural supports
required for ICT operation. Their negative perception of the infrastructure
as Internet speed was, in general, very slow, frequent power outages
occurred, banking and legal supports were incompatible, and the absence
of online consumer protection law all adversely affected their expectation.
Despite inadequate and incompatible infrastructural supports, some

organisations were compelled to use ICT due to their interactions with
international companies (many of whom were large multinational enter-
prises) as a collaborator, supplier or customer.
Organisations in foreign countries were completely computer-based
working in automated environments. They demanded that their foreign col-
laborators were compliant with their systems. Thus, SMEs in underdeve-
loped countries, although facing several constraints and having had no
expectation of adopting such technology, were using ICT to comply with

their foreign customers and collaborators and to match the competition. In
this context, although country readiness had strong negative effects on
expectation, there was no relationship with actual usage behaviour.
7.7. Hypotheses Related to Culture
Culture, as an important aspect of the external environment, affected orga-

nisational ICT usage behaviour. Based on the field study outcomes and
past studies (Bertolotti, 1984; Burn, 1995; Erez & Early, 1993; Gefen &
Straub, 1997; Hill et al., 1998; Ho et al., 1989), it was anticipated that cul-
ture would have a significant influence on cognitive evaluation (Hypothesis
H6a ) and expectation (Hypothesis H6b ).
The results of the structural equation model estimation showed a signifi-
cant negative association between culture (γ = − 0:333) and cognitive evalua-
tion. This result was consistent with the field study and prior studies
(Dolecheck & Dolecheck, 1987; Lu, Rose, & Blodgett, 1999; Ralston,
Giacalone, & Terpstra, 1994). Culture is composed of values, norms and
beliefs which individuals of a collective acquire through their long-term
interactions with their fellow community members, society and the environ-
ment. Individuals’ preferences and behaviours are largely influenced by
their own culture. The result suggested that the culture of Bangladesh was
not supportive to the operation and implementation of computer-driven
communication systems or working environments among SMEs.
Bangladesh is a developing country which is characterised by a culture
with high power distance and high in-group collectivism. Small businesses
are mostly operated in an informal manner. Individuals in this society
usually interact through face-to-face communication, maintaining close
social ties and mostly use the Bengali language as their medium of
206 MD SHAH AZAM
communication. Nepotism, political influence and interference have

become a part of the country’s culture. The overall culture of the country is
now polluted through exercising bribery and unethical ways of conducting
business in different sectors.
Utilisation of computers in the working environment requires the estab-
lishment of structured and formal working procedures, technology
resources and skilled human resources. Entrepreneurs of SMEs are usually
confused when considering the usefulness of ICT as it is not compatible
with the overall culture of the country. They may also not perceive the
technology as being easy to use as the basic interfaces of computer-driven
programs are English language-based.
This study has measured culture by applying higher-order hierarchical
modelling where culture was conceptualised as a second-order composite of
power distance, in-group collectivism, uncertainty avoidance, ethical culture
and Bengali values. The estimation revealed a gross summated effect of cul-
ture on cognitive evaluation which was negative.
The statistical estimation did not reveal any significant association between
culture ðγ = − 0:048) and expectation, which thus rejected Hypothesis H6b .
Prior studies have supported the view that culture influences organisational
ICT use (Bertolotti, 1984; Burn, 1995; Erez & Early, 1993; Gefen & Straub,
1997; Harris & Davison, 1999; Hill et al., 1998; Ho et al., 1989; Straub,
1994). The field interviews explored the view that culture was associated
with expectation and cognitive evaluation. Although interview participants
supported the link between culture and expectation, statistical analysis of
the national survey data indicated that this prediction, based as it was on a
small sample, could not be applied to a larger population.
The assumption, justified through the qualitative field study, may be
rejected by a quantitative survey. This is a common problem of mixed-
method research as some ideas primarily considered through the field
study which comprises a small sample size may be rejected through the
responses from a comparatively large sample used in the quantitative sur-
vey (Jackson, 2008; Quaddus, Islam, & Stanton, 2006). However, this
result may be argued against, according to England (1975), who stated
that culture was a key determinant in determining individual beliefs,
values and attitudes. Cognitive evaluation represented individuals’ favour-
able or unfavourable feelings towards an innovation which was comple-
tely internal and projected as attitude, thus a significant association
between culture and cognitive evaluation was justified. This result is con-
sistent with the findings of numerous past studies (Dolecheck &
Dolecheck, 1987; Lu et al., 1999; Ralston et al., 1994).
Surprisingly, culture was not found to have a significant effect on expec-
tation. One possible explanation of this result was that the comprehensive
research model included cognitive evaluation as a predictor of expectation
with culture. Cognitive evaluation was more highly influenced by culture
than expectation as culture is a determinant of individuals’ values, beliefs
and attitudes (England, 1975). Cognitive evaluation, as the relatively

expressed state of individuals’ favourableness or unfavourableness toward
an innovation, produced a strong and significant effect on expectation
while culture did not provide any significant effect on expectation but sig-
nificantly explained cognitive evaluation. As an individual’s favourableness
or unfavourableness towards an innovation formed their intention or
expectation, cognitive evaluation was found to have a strong significant
effect on expectation.
Importantly, culture, although it was not found in this study to have a
positive effect on expectation, was indirectly related to expectation through
its strong and significant influence on cognitive evaluation.
7.8. Hypothesis Related to Expectation
7.8.1. Hypothesis H 7
Applying the notion of the TRA (Fishbein & Ajzen, 1975), TPB (Ajzen,
1985) and TAM (Davis, 1989) with the outcome of a recent study
(Venkatesh et al., 2008), expectation was predicted to have a significant
positive influence on actual usage behaviour (Hypothesis H7 ). Finally, the
structural model estimation depicted expectation (β = 0:697) as a strong
and significant determinant of actual behaviour when referring to actual
use of ICT among SMEs in Bangladesh. The finding supported previous
theories and empirical studies (Ajzen & Fishbein, 1980; Azam &
Quaddus, 2009b; Chang, 1998; Fishbein & Ajzen, 1975; Mathieson, 1991;
Taylor & Todd, 1995a, 1995b; Venkatesh & Davis, 2000; Venkatesh &
Morris, 2000; Venkatesh et al., 2003, 2008). Consistent with the theoreti-
cal framework, this result indicated that a positive and strong expecta-
tion, that is, the willingness to adopt would foster the use of ICT at firm
level.
208 MD SHAH AZAM
7.9. Hypotheses Related to ICT Use
7.9.1. Hypothesis H8a

Based on the field study findings, ICT use was anticipated to have a signifi-
cant influence on ICT integration (Hypothesis H8a ). The structural model
estimation revealed a strong significant positive association between ICT
use (γ = 0:643) and integration. This result was consistent with the field
study. This study measured ICT use as a summated subjective assessment
of different levels of ICT applications, namely, use of the Internet and

homepage (capable of product cataloguing); use of interactive homepage
(capable of receiving online orders); use of interactive homepage for trans-
actions (capable of online transactions) and use of ERP. Consistent with
the field study outcome, this result indicated that the higher levels of ICT
application would speed up the process of ICT integration between differ-
ent functional areas within the organisation as well as with other organisa-
tions at industry level.
7.9.2. Hypothesis H8b

As with Hypothesis H8a , based on the field study findings, it was proposed
that ICT use would have a direct significant influence on degree of utilisa-
tion (Hypothesis H8b ). The structural model estimation revealed a strong
significant positive association between ICT use (γ = 0:488Þ and degree of
utilisation. This result was consistent with the field study. Degree of utilisa-
tion was measured through a summated subjective assessment of how
appropriately ICT was utilised in different functional areas in the firm,
namely, production, administration and accounts, and marketing and sales.
Consistent with the field study outcome, this result indicated that the higher
level of ICT applications would ensure proper utilisation of ICT in various
functional areas of the firm.
7.9.3. Hypothesis H8c

Applying the notion of the RBV of the firm (Barney, 1991) and its applica-
tion to ICT at the firm level (Bharadwaj, 2000; Powell & Dent-Micallef,
1997), it was anticipated that ICT use would have direct positive effects on
firm performance (Hypothesis H8c ). Surprisingly, the structural model esti-
mation did not find any significant association between ICT use (γ = 0:077)
and performance. This result was also not consistent with previous studies
(Bharadwaj, 2000; Powell & Dent-Micallef, 1997; Zhu, Dong, et al., 2006;
Zhu & Kraemer, 2005). This result also was contrary to the outcome of the
field study.
However, the result apparently was similar to the productivity paradox

of IT. This was argued according to the RBV (Barney, 1991) which
explains that valuable, rare, imperfectly imitable and imperfectly
substitutable resources could generate sustainable competitive advantage
for the firm with the prerequisite of heterogeneity and imperfectly mobile
resources among competing firms. The constructs, ICT integration and
degree of utilisation, combine ICT resources with different functionalities of
a firm and utilise them properly which produces unique capabilities that
are intangible and unable to be imitated. Thus, ICT use alone, without inte-
grated and properly utilised applications, may not have any significant
impact on firm performance.
7.10. Hypothesis Related to ICT Integration
ICT integration was anticipated to have a significant influence on organisa-
tional performance (Hypothesis H9 ). The structural model estimation
revealed a strong significant association between ICT integration (γ = 0:541)
and firm performance. This result was consistent with the field study out-
come and past studies (Zhu & Kraemer, 2002, 2005). Firm performance was
measured by applying higher-order hierarchical modelling where competi-
tive performance, internal operation productivity and financial performance
were the manifest variables. However, this result indicated that overall firm
performance largely depended on how ICT was integrated between different
functional areas. This result was logical as organisations would not receive
the benefit if they introduced computer-based operation only in one parti-
cular department while other departments operated through manual proce-
dures. Furthermore, organisations would not receive productive output
even in the case where computer-based operations were introduced in many
departments if they were not integrated.
7.11. Hypothesis Related to Degree of Utilisation

Based mainly on the field study findings, it was anticipated that degree of
utilisation would have a significant effect on firm performance (Hypothesis
H10 ). The structural model estimation accepted the hypothesis portraying
degree of utilisation ðγ = 0:180) as having a significant association with firm
210 MD SHAH AZAM
performance. This result was consistent with the findings of past studies
(Forster & Cornford, 1992; Ordedra et al., 1993). This result indicated that
organisations benefit from ICT use if the degree of utilisation is satisfac-
tory; that is, the technology is utilised properly. It is important to note that
ICT usage indicates the frequency or rate of usage of various ICT applica-
tions while degree of utilisation explains the extent to which the proper and
actual use of technology is ensured.
The logical interpretation of this finding is that most of the functions in
Bangladesh, from household day-to-day operations to business and social

formalities as well as government services are performed through manual
processes. Thus, numerous functions of ICT, although implemented, are
un-utilised or under-utilised which causes adverse effects on the productiv-
ity of ICT and finally on firm performance.
Furthermore, ICT, particularly the Internet, although not new in
Bangladesh, is still at an introductory level as most people are reluctant
and/or incapable of utilising it in communications, services and businesses.
As the Bangladeshi government has declared that it is seeking to achieve a
digital Bangladesh by 2021 and has set the country the target of establish-
ing e-governance by 2014, the people are being motivated and pushed
towards the adoption of technology. Unfortunately, this is confined to the
operation of entertainment-related ancillary functions, such as audio,
video, games, video streaming and chat rooms.
The application of ICT, although promoted in various sectors across the
country, unfortunately is not reasonably involved in accomplishing eco-
nomic or business functions. As a consequence, the productivity impact of
ICT at firm level has become questionable. Organisations invest and intro-
duce ICT with a view to performing some specified organisational func-
tions. Proper accomplishment of those tasks results in productivity of the
investment. Consistent with this notion, this study has revealed that organi-
sations that fail to accomplish the tasks properly would suffer from adverse
effects on firm performance.
7.12. Hypothesis Related to Mediating Effects of ICT Integration and

Degree of Utilisation
Based on the field study outcome, it was anticipated that ICT integration
and degree of utilisation would play mediation roles in explaining the effects
of ICT use on firm performance. The structural model estimation depicted
a mediational process explaining the effects of ICT use on organisational
performance which involved two intervening variables, ICT integration and

degree of utilisation. The mediational relationship was examined by apply-
ing the procedure of Baron and Kenny (1986).
It was revealed that the variation in ICT use alone would significantly
account for a variance in SMEs’ performance, the effects becoming reduced
in the presence of a single mediator (integration or utilisation) which would
finally disappear in a multiple mediation environment (with the presence of
ICT integration and degree of utilisation together). The results indicated
that the effect of ICT use on firm performance was completely mediated
through ICT integration and degree of utilisation.
It is argued that this result occurred because, due to the lack of infra-
structural supports and the lower level of ICT and Internet penetration,
organisations did not apply ICT intensively to accomplish all major organi-
sational or business functions. Moreover, although organisations used ICT,
their inability to integrate major business functionalities would impact
adversely on organisational performance. Likewise, organisations did not
attain performance growth if ICT was not utilised properly.
7.13. Summary
This section has provided a discussion of the SEM estimates presented in

Section 6. The interpretations of the findings have been carried out to sup-
port the hypotheses. This section has discussed the effects of the antecedent
factors of ICT adoption and diffusion by SMEs. In analysing the impact of
ICT on firm performance, it was found that ICT use may not generally cre-
ate enhanced firm performance. Integration and proper utilisation of ICT
were important indicators of firm performance. The discussion in this sec-
tion has indicated that firms generally experience performance growth if
ICT is integrated and properly utilised. The final section comprises the con-
clusion and presents future research directions.
8. CONCLUSIONS AND FUTURE RESEARCH

DIRECTIONS
8.1. Introduction
This section presents the conclusions of the current research. In order to

state the contribution of the research to the body of knowledge, a summary
212 MD SHAH AZAM
of the entire research process and outcomes is provided in Section 8.2.

Section 8.3 then discusses the theoretical and practical contributions of the
research. The final section states the limitations of this study and outlines
directions for future research.
8.2. Research Summary

This research project was initiated to study the impact of the diffusion of
ICT on SMEs’ performance with special attention given to the mediating
effects of ICT integration and utilisation. Previous studies have reported a
clear differentiation in the level of usage of ICT between rich and poor
countries, as well as between developed and developing countries. The
introduction of the usage of ICT, particularly Internet-dependent technol-
ogy, by individuals and organisations involves changes to the systems, pro-
cedures and processes of relevant services and also affects the way through
which customers, suppliers, regulatory bodies and other external parties
deal with business organisations. The rapid and exponential growth of ICT
usage has attracted the bulk of researchers’ attention in looking at the phe-
nomena of the adoption and diffusion of the technology. Although numer-
ous research initiatives have focused on the adoption intention (Gefen &
Straub, 2000; Kendall et al., 2001; Lal, 1999; Pavlov & Chai, 2002) and
actual usage (Anandarajan et al., 2002; Johnson & Hignite, 2000; Thatcher
et al., 2006; Venkatesh et al., 2008) of the technology, the question of
whether ICT usage impacts on firm performance in SMEs has not been
clearly answered (see Jean, 2007). Furthermore, previous studies on innova-
tion diffusion have focused on the developed country perspective and,
mostly, on the American perspective (Zhu & Kraemer, 2005). Due to the
infrastructural, environmental and cultural differences, and a wide digital
divide between developed and developing countries, the theories con-
structed from a developed country perspective may logically suffer from
their lack of generalisability and applicability in the developing country
context. It has been assumed that the theoretical framework constructed in
a developed country perspective would also fit the developing country
environment. However, it is not unlikely that the theoretical framework
would vary in different contexts. To examine the effects of the diffusion of
ICT on SMEs’ performance, this study firstly developed a comprehensive
theoretical model and then examined the research model from a developing
country perspective with survey data collected from various SMEs in
Bangladesh.
The theoretical framework of the research (described in Section 2) was

developed by combining the technology adoption-diffusion theories,
namely the TRA (Fishbein & Ajzen, 1975); Rogers’ DOI theory (Rogers,
1983); the TPB (Ajzen, 1985); the TAM (Davis, 1986); UTAUT (Venkatesh
et al., 2003) and the TOE framework (Tornatzky & Fleischer, 1990) with
institutional theory (DiMaggio & Powell, 1983) and resource-based theory/
view (Barney, 1991). The initial research model was refined and contextua-
lised by the field study and a comprehensive research model was developed.
The hypotheses for the research were formulated from the comprehensive
model.
As previously discussed in the methodological section (described in
Section 3), this study has employed a mixed-method research approach,
combining qualitative and quantitative methods of data collection and ana-
lysis to attain the research objectives. The qualitative phase of the study
extended and contextualised the initial model. It is not unlikely that due to
contextual differences the subject of this study may have a manifestation
which differs from and may be beyond the scope of the existing literature.
Furthermore, the new model developed through synpapering different
models needed to be contextualised. A field study was conducted by inter-
viewing 11 SMEs in Bangladesh (as described in Section 4). Content analy-
sis (by NVivo 10) was performed to analyse the data. The findings, in
general, supported the initial model. However, some adjustments were also
applied to the initial model to build a comprehensive and integrated
research model (Fig. 3). Importantly, the field study suggested reducing the
number of related latent variables in order to construct a higher-order
latent variable combining related first-order variables. Cognitive evaluation,
culture, environmental pressure, country readiness and performance were
developed as second-order constructs through this process. Finally, the
comprehensive model comprised owner innovativeness, cognitive evaluation,
facilitating condition, environmental pressure, culture and country readiness
as antecedents of the adoption and use of ICT by SMEs while organisa-
tional performance was predicted through the influence of ICT use, integra-
tion and degree of utilisation. Based on the comprehensive research model,
21 hypotheses were formulated under 11 groups (as described in Section 5).
The second phase of the research employed a quantitative approach to
test these hypotheses. Since this study stands under the positivist research
paradigm, the main and most voluminous work was associated with this
phase (the quantitative research) of the study. The quantitative research
study involved the development of the survey instrument, questionnaire
pre-testing, survey design, data collection, data coding, recording and
214 MD SHAH AZAM
manipulation, outlier checking and model estimation (as described in

Section 6). The questionnaire was finalised after pre-testing. Next, a pilot
study was conducted on 60 respondents. Based on the feedback, some mod-
ifications were made and the final questionnaires were distributed to SMEs
in Bangladesh seeking their responses. In total, 282 responses were gath-
ered. A PLS-SEM) technique was employed to analyse the quantitative
data with this performed through PLS-Graph Version 3.
The analysis revealed that the adoption and diffusion of ICT by SMEs
in Bangladesh were largely influenced by cognitive evaluation, facilitating

condition, environmental pressure and country readiness with owner innova-
tiveness and culture indirectly affected through cognitive evaluation.
Organisations were interested in implementing ICT for better organisa-
tional performance. Surprisingly, ICT use was not found to have any sig-
nificant effect on overall organisational performance. However, the
structural model depicted a unique inter-relationship between ICT use, inte-
gration, degree of utilisation and organisational performance. The analysis
revealed that ICT use, although not directly related to organisational per-
formance, was indirectly related through integration and the degree of utili-
sation. This result suggests that organisations may not attain better
performance by implementing ICT in their firm if it is not integrated with
different functional areas and utilised properly.
8.3. Contributions of the Research
This study successfully employed a mixed-method research approach which

has provided some unique results to address the main research questions.
As opposed to most studies in the adoption area which commonly engage
in a mono-method approach, this research applied a mixed method that
combined qualitative and quantitative approaches in the data collection
process. For the qualitative method, a field study was employed while the
quantitative method used a survey as the data collection process. The over-
all design of the study and its implementation were challenging on several
counts as the study examined the innovation diffusion phenomena from a
developing country perspective while most diffusion theories have been
constructed from a developed country perspective and, particularly, an
American perspective (Zhu & Kraemer, 2005). The mixed-method research
approach was appropriate and has made significant contributions which
included building a comprehensive research model (contextualised and fine-
tuned); exploring a few new constructs, variables and their measurement
items; exploring the causal links between the study variables; testing of
hypotheses and drawing inferences by reliable and valid research outcomes.
The main contributions of the study, both the theoretical and practical con-
tribution, are reported in the following section.
8.3.1. Theoretical Contribution

This study was initiated to examine SMEs’ ICT usage behaviour and firm
performance. This study extended its exploration from simply addressing
the innovation adoption-diffusion phenomena to analyse the effects of ICT

diffusion on organisational performance. In examining the ICT diffusion
process in SMEs, its antecedents and consequences from the viewpoint of
generating organisational performance growth, this study has reviewed
existing innovation diffusion theories and has combined those theories with
institutional theory and the RBV.
In summary, this study has reviewed the TRA (Fishbein & Ajzen, 1975);
DOI theory (Rogers, 1983); TPB (Ajzen, 1985); TAM (Davis, 1986);
UTAUT (Venkatesh et al., 2003) and the TOE framework (Tornatzky &
Fleischer, 1990) combined with institutional theory (DiMaggio & Powell,
1983) and resource-based theory/view (Barney, 1991). The results of this
synpaper have helped to construct a comprehensive model which has
focused on the process of the diffusion of ICT and its impact on organisa-
tional performance. The existing theories have mostly addressed the ICT
diffusion phenomena from a developed country perspective while their
applicability in addressing similar phenomena from a developing country
perspective is questionable. The initial theoretical model developed through
reviewing the existing literature underwent a screening process using a qua-
litative field study to fine-tune and contextualise the research model.
The field study interviews suggested some adjustments that were needed
to the model which had been primarily developed from theory. These
adjustments involved changes in the interrelationships between some con-
structs, omission of some variables and also inclusion of some new vari-
ables and measurement items. This screening process yielded an integrated
research model which enabled the study of the ICT diffusion process and
the post-adoption phenomena (i.e. the effects of ICT usage on firm perfor-
mance) in the context of SMEs in Bangladesh.
The comprehensive theoretical model includes various factors from dif-
ferent levels which involve individual, organisational, cultural, environmen-
tal and country-specific factors. The model also uses three different levels
of ICT use: ICT use, integration and degree of utilisation to address the
depth, breadth and appropriateness of ICT use, respectively. The
216 MD SHAH AZAM
theoretical model predicts that the effects of ICT use on organisational per-
formance are mediated through integration and utilisation. This prediction
was validated through analysis of the survey data. By including ICT inte-
gration and utilisation and their mediational roles, the theoretical model is
comprehensive and unique in addressing ICT diffusion and organisational
performance. This presents a major theoretical contribution.
The model measures organisational performance (SME performance) as
a higher-order construct comprising three different performance-related
manifest variables: competitiveness, internal operation productivity and

financial performance. The model also explains cognitive evaluation as a
higher-order construct which comprises two internal manifest variables:
perceived usefulness and perceived ease of use.
The model includes three different aspects of external environmental
pressure, namely, pressure arising from cultural aspects, pressure arising
from the industrial environment and pressure arising from the macro-
country environment. Each of the environment-related factors is addressed
by applying a higher-order modelling procedure. Culture is formed by the
effects of power distance, in-group collectivism, uncertainty avoidance, ethical
culture and Bengali values; environmental pressure is composed of coercive
pressure, mimetic pressure, normative pressure and global pressure; while
country readiness comprises technology infrastructure, human infrastructure,
legal infrastructure, financial infrastructure and government policy and sup-
ports. Although most of the variables being studied were borrowed from
previous literature, a number of the variables explored in the field study
were very much context-specific and were the first of their kind to examine
technology innovation diffusion phenomena from a developing country
perspective. For example, Bengali values, ethical culture and degree of utili-
sation of ICT, as well as the measurement procedure of cognitive evaluation
and country-specific variables, were new thus adding to the significance of
this study.
The theoretical framework is innovative and provides a holistic
approach for looking at the antecedents of ICT diffusion and its conse-
quences for organisational performance. More clearly, the theoretical fra-
mework offers an opportunity to examine the whole process of innovation
diffusion and its effect on organisational performance from a single plat-
form. The robustness of analysis, furthermore, suggests the suitability of
the comprehensive model for analysing the diffusion of ICT by SMEs and
its effects on organisational performance. This framework is potentially
suitable for testing similar phenomena in the large organisation environ-
ment from both a developing and developed country perspective.
8.3.2. Practical Contribution

This study has provided a clear picture of how internal and external factors
affect ICT usage by SMEs (in a developing country) and the consequential
effects of this usage on organisational performance mediated through inte-
gration and utilisation. By categorising SMEs by their different levels of
ICT use, it was revealed that SMEs that have positive cognitive evaluation
towards ICT and adequate technological resources, and that receive posi-
tive pressure from their institutional environment are more likely to use
ICT while the overall culture and country infrastructure were not suppor-
tive of the use of ICT.
Previous studies investigating adoption or usage of ICT have assumed
that the use of ICT would generate performance growth. Interestingly, this
study has found that ICT use did not immediately increase performance
growth in SMEs. Organisations needed to acquire integration and proper
utilisation of the technology in order to achieve performance growth. This
was consistent with the notion of the RBV of ICT which emphasises that
combining ICT into different organisational functionalities (i.e. integration
with front-end functionalities or back-end integration) can create IT cap-
abilities that are rare, unable to be imitated, valuable and sustainable,
thereby contributing to value generation. Managers, interested stakeholders
and policy makers may follow these findings and focus on ICT integration
and the degree of utilisation with the view of attaining superior firm
performance.
This study may also suggest some changes and modifications in
Bangladesh’s existing policies and strategies for promoting ICT adoption
and ICT use by SMEs in order to achieve economic development.
Bangladesh is a developing country in which SMEs play a significant
role in its development processes. The Bangladeshi government is dedicated
to the utilisation of the potential of ICT in its economic development and
to the establishment of a digitally based society by 2021. In light of the
country’s Vision 2021, the government has already revised the ICT policy
and has made provision for various supports to promote ICT use in differ-
ent sectors of the country. For example, the government has withdrawn all
taxes from computer hardware, provides interest-free loans for ICT (soft-
ware) businesses, and launches pro-ICT policies, grants, subsidies and
motivational programs.
Government offices are gradually developing and becoming equipped
with ICT devices to introduce e-governance systems. The last few years
have witnessed significant physical and infrastructural development in
terms of the installation of a large number of computer devices in different
218 MD SHAH AZAM
government offices, business enterprises and households. Frustratingly,

most of these technological facilities are either un-utilised or under-utilised.
For example, the majority of ICT users only access audio-visual entertain-
ment programs, such as using their MP3 players, listening to audio or video
music and playing online games. In recent years, computer users have
increasingly participated in online social networks, audio and video chat
rooms and other forms of communication. The patterns of ICT usage by
individuals, households and organisations are similar. Thus, although a
good number of SMEs use ICT, the productivity of the technology is

questionable.
Entrepreneurs are apprehensive about a reduction in productivity as a
result of the installation of computers in the workplace. They fearfully
anticipate the risk that employees will participate in entertainment pro-
grams, games or personal communication during business hours using the
technologies which have been installed to increase firm performance.
Furthermore, organisations, in general, use ICT to conduct some of
their tasks while a number of functions are still carried out by the tradi-
tional means which, in many cases, are not compatible with the compu-
terised online environment. The lack of implementation of completely
computerised operations (i.e. the installation of computers to accomplish
the tasks of all functional areas) prevents the attainment of the full benefits
of ICT usage. This study’s findings may provide a lesson for entrepreneurs
who are confused about the possible outcome of computer device installa-
tion at their workplace as the findings suggest that partial implementation
of ICT in the workplace may not result in organisational productivity. The
costs of computerisation, that is, operational, maintenance and installation
costs, are substantial. Only very high performance growth can make it
worthwhile to invest in computerisation. In reality, partial use of the com-
puterised operation, although producing some enhanced performance, is
not enough to cover the significant costs associated with ICT-based opera-
tions. However, organisation-wide usage of the technology, although seem-
ing to involve very high costs, may regain the reduction in productivity lost
through these costs.
Organisation-wide complete ICT-based operations are not possible until
and unless other closely associated organisations and important stake-
holders use the same system. For example, organisations cannot gain
benefits from ICT-based operations if their suppliers or customers are not
ICT-based. In these circumstances, organisations, although they have
highly sophisticated ICT devices, will need to use manual operations to
interact or communicate with their stakeholders as ICT will not provide
any help. The introduction of ICT-dependent communication and exchanges

between organisations and external parties, for example, customers,
suppliers, government regulatory authorities, partners and trade bodies,
may help to develop this difficult situation.
Although some organisations perform some financial functions through
ICT, most are confined to managing employees’ salaries and account-
keeping tasks. Organisations generally employ one or a few computer
operators to perform computer-related tasks which means that a large
number of functions remain external to their computers. It is not unlikely

that some offices have computer hardware which is not used for operations
but simply for social prestige as a showpiece. Moreover, the reality is that
organisations, although using ICT, generally fail to ensure its proper utili-
sation and integrated use. Thus, the productivity of ICT usage has become
questionable. In this situation, the Bangladeshi government may take the
initiative to develop an integrated plan to promote the proper and inte-
grated use of ICT which may include motivational programs, training and
supports. This new initiative may help to develop the productivity of ICT
usage and economic growth of the country.
The study thus calls for the initiation of a new ICT research direction
exploring the status of ICT integration and ICT utilisation by SMEs and
finding a way for ICT to contribute positively to organisational
performance.
This study has also explored the patterns of ICT use by SMEs and has
forecast the degree and magnitude of the effects of various individual, orga-
nisational and environmental factors on the adoption and usage of differ-
ent levels of ICT applications. Cognitive evaluation emerged as a strong
predictor of the expectation of SMEs’ ICT use. The study revealed that
organisations with adequate knowledge and a favourable perception of the
usefulness and operational ease of the technology were more likely to use a
higher level of ICT. Organisations that held a higher level of what was
termed facilitating condition were also more likely to implement technolo-
gical innovations than those who were lacking in this attribute. The facili-
tating condition comprised adequate resources, such as technological
resources, human resources or systems compatible with computerised
operations which facilitated the use of ICT in an organisation.
Entrepreneurs or the government may utilise this notion to develop poli-
cies or strategies to foster the growth of ICT use by SMEs. Providing useful
information about the applicability of ICT to performing various organisa-
tional functions, to operational know-how and to technological facilities
will potentially attract SMEs to using ICT for organisational
220 MD SHAH AZAM
competitiveness. The government and policy makers should undertake

mass promotional programs to disseminate information about the various
uses of computers and their contribution in developing better organisa-
tional performance. Organising specialised training programs for SMEs’
capacity building may succeed in attaining SMEs’ motivation to use ICT.
Ensuring easy access to the technology by providing loans, grants and sub-
sidies is also important.
The comprehensive model also forecast significant environmental pres-
sure on SMEs’ use of ICT. It suggested that organisations were highly

influenced by various institutional stakeholders such as peer groups, com-
petitors, dominant suppliers, customers or regulatory authorities, and by
the changing environment arising from the globalisation process.
Surprisingly, SMEs used ICT mainly to comply with the demands of their
significant suppliers or dominant customers’ requirements. Most of the
SMEs, particularly in the RMG sector, have implemented ICT-based
operations as their suppliers and dominant customers allow only online
communications and exchanges. Thus, they have installed computers and
ICT devices in their organisation in accordance with their customers’
requirements in order to maintain their business deals while all other inter-
nal functions and domestic communications are performed through tradi-
tional means of operations. In reality, the technology is only used to
accomplish a part of their organisational functionalities. Although the tech-
nology ensures the success of the most important task of the business being
the carrying out of its orders and deals, this is still not considered produc-
tive in terms of the costs and investment associated with the technology
installation and operation.
This study urges business enterprises’ management and the government
to look at the environmental and cultural factors with the view of achieving
ICT usage success which may foster the country’s economic development.
Initiating strategic adjustments to improve organisational practices by
eliminating the traditional power distance inside the organisation as well as
introducing formal working procedures by implementing necessary rules
and regulations may result in productive use of technology. Furthermore,
the application of a Bengali user interface may help to ensure the produc-
tivity of ICT use by SMEs in Bangladesh.
Finally, the government, various stakeholders and regulatory authorities
should rationally adapt their policy and strategies focusing on the inte-
grated and appropriate use of technology rather than only promoting the
use of technology. However, the question may be asked regarding how a
rational strategy to move towards appropriate and integrated utilisation of
ICT could be implemented before its promotion for mass usage. My argu-
ment is that the government and many financial and non-government agen-
cies have already made significant efforts and provided enormous support
promoting the usage of ICT at the government and non-government orga-
nisational level including in SMEs. Unfortunately, most of these organisa-
tions are using the computer as a typewriter or, in some cases, as an
alternative to postal mail or telephone. The situation of most small busi-
nesses is similar. The quantitative analysis indicated that organisations with
high ratings with regard to the ease or usefulness of the technology were
more likely to use ICT. However, the field study and subsequent survey
have suggested that numerous SMEs, although currently ICT users, were
not able to accomplish the expected tasks by using the technology. The
practical implication of this study is that the integration and proper utilisa-
tion of ICT should be implemented for superior organisational
performance.
Interestingly, the study has explained SMEs’ ICT usage phenomena by
employing both organisational and consumer behaviour theories. However,
the growth and economic potential of SMEs demand the development of a
theoretical branch which will address the adoption and usage behaviour of
SMEs.
8.4. Limitations of the Study
Despite this study’s substantial contribution by studying SMEs’ ICT usage

(from a developing country perspective) and its analysis of the effects of a
range of variables on the adoption and diffusion of ICT and its conse-
quences for organisational performance, it has some limitations. By its nat-
ure, this study has analysed the aggregated effects of macro-environmental
variables on ICT use and has explored how this affects aggregated firm per-
formance. The results have implications for industry as well as for
Bangladesh. However, as the study does not clearly answer how the first-
order or manifest variables (the variables which construct a higher-order
hybrid variable) affect ICT use and organisational performance, this may
be considered a limitation of the study.
Secondly, this study focuses on the manufacturing industry sector in
Bangladesh. As the RMG industry is the major contributor to the coun-
try’s economy (contributing 75% of export earnings), the ICT diffusion
phenomena have been analysed mainly from this industry’s perspective.
The study has also included responses from the leather industry which is an
222 MD SHAH AZAM
emerging industry in the manufacturing industry sector in terms of its

export contribution. The number of responses from the leather industry
was proportionate to the RMG industry responses, determined through the
respective industry’s contributions in export earnings. The proportion of
export earnings by the RMG industry and leather industry in Bangladesh
is roughly 80:20. The application of a proportionate stratified random sam-
pling (in fact, given the disproportion between the respective industries, a
disproportionate stratified random sampling) technique for data collection
was therefore methodologically sound. Moreover, in its design, the services

industry was beyond the study’s scope which may be considered to be a
limitation in drawing general implications for SMEs. For example, the IT
industry, software and telecommunications industries, media and education
industries, financial institutions, and tourism industry have high potential
and make a significant contribution to the national economy.
Finally, this study utilised cross-sectional data which was collected at a
single point in time. Thus, assessing the impact of the antecedent factors on
the diffusion of ICT and its role in explaining SMEs’ performance at differ-
ent points in time was not possible.
This study has estimated the ICT adoption and diffusion phenomena by
introducing a new conceptualisation of ICT use which was constructed by
a composite measure of various levels of ICT applications. Although deal-
ing with the aggregated use of ICT by SMEs is an innovative measure and
has great importance in developing policy strategies at entrepreneur, indus-
try and government level, this may be considered as a limitation as there is
significant variation in the diffusion of different levels of ICT applications,
such as ERP, e-commerce and e-business.
8.5. Future Research Directions
The limitations of this study may provide direction for new research inves-
tigating the adoption-diffusion of ICT by SMEs in Bangladesh. Analysis of
the direct effects of the various dimensions of a higher-order composite,
with these used as manifest variables or first-order variables, may provide
clearer understanding of the phenomena. The effects of various dimensions
of culture and the institutional environment as well as country-specific fac-
tors may help in developing policy and strategies to foster SMEs’ ICT
usage. Future research could compare the aggregated results with the speci-
fic outcomes of different dimensions of performance such as competitive-
ness, internal operation productivity and financial performance.
The changes in patterns of ICT usage and the effects of various antece-
dent factors could be examined by the analysis of longitudinal data collected
from the same or similar panels of organisations at different points in time.
The inclusion of diversified industries in the sample would enhance the
validity of the predictions. Future studies could include the services industry
as well as the manufacturing industry to produce representative results.
Finally, a comparison between the diffusion phenomena of specific ICT
systems such as customer relationship management (CRM), e-commerce or
ERP could provide a detailed picture of the adoption-diffusion patterns of

specific systems. A future research initiative could validate this theoretical
foundation by comparing the aggregated results with the adoption beha-
viour of various specific ICT systems. This could help in developing appro-
priate policies and strategies to foster the usage growth of various levels of
ICT applications by SMEs in Bangladesh with the view of ensuring ICT’s
positive impact on organisational performance.
The final research model comprises numerous latent variables for study-
ing SMEs’ ICT usage behaviour and its impact on organisational perfor-
mance. Due to the scope of the chapter, the influence of control variables,
such as size and experience (age of the organisation) have not been exam-
ined. Future research could add various control variables to the existing
model to examine variations in the effects of the study constructs in the
presence of different control variables.
This study has explored some new constructs that arose from the field
study, namely Bengali values, ethical culture, global pressure and degree of
ICT utilisation. By using higher-order conceptualisation, the study has also
introduced different measurement procedures for some latent variables such
as cognitive evaluation, culture, environmental pressure, country readiness
and organisational performance. Various indices explored through the field
study have been employed to estimate the composite of the higher-order
latent variables, with the exception of cognitive evaluation for which a well-
recognised measurement scale is applicable for measuring perceived useful-
ness and perceived ease of use. Development of a set of reliable measurement
scales for these newly developed constructs could significantly contribute to
future research initiatives thus providing a logical grounding for future
research.
8.6. Conclusion
The modelling approach employed in this study was based on the premise
that ICT usage by SMEs ensures improved organisational performance.
224 MD SHAH AZAM
However, this study, in particular, has addressed these questions: (1) does
ICT usage affect enhanced organisational performance; and (2) how does
ICT usage by SMEs contribute to enhanced organisational performance?
The short answers to these questions are that ICT usage by SMEs in
Bangladesh does not immediately ensure firm performance. Integration and
proper utilisation of the technology play the key role in ensuring enhanced
performance. The study also forecast the strong and significant effects of
cognitive evaluation, facilitating condition, country readiness and environmen-
tal pressure on SMEs’ expectation of using ICT which is an antecedent of

ICT use. In addition, the study forecast a strong and positive impact of
expectation on actual usage behaviour. The impacts of culture and owner
innovativeness were found to be very strong on cognitive evaluation.
This study has several implications for SME owners, the coordinating
authorities or trade bodies, and the Bangladeshi government who are inter-
ested in enhancing the pace of computerisation and the implementation of
various ICT applications in SMEs with the view that new technology
implementation will enhance organisational performance. One methodo-
logically oriented finding was that the structural model estimations clearly
indicated that integration and utilisation of ICT are more important than
simply deciding on the use of ICT. Integration and utilisation, although
sounding very normal in general terms, are not easy to ensure in the
ICT-based working environment within SMEs. Integration and proper
utilisation of ICT are heavily dependent on the firm’s resources,
technical capability and operational skill, and compatible intra- and inter-
organisational systems, availability of hardware and connectivity. The gov-
ernment, regulatory authorities and concerned organisations need to show
their support by providing competent infrastructure, resources and policy
guidelines to ensure proper utilisation and integrated use of ICT in SMEs.
Expectation has been revealed as an important construct which has a
very strong and significant impact on ICT usage. The findings also indi-
cated that those SME owners who had favourably evaluated the technol-
ogy and who had the resources and skills were more receptive, that is, they
had a higher expectation to use ICT. Culture provided a negative impact
on expectation. Interestingly, country readiness provided a positive impact
on expectation while providing a negative impact on ICT usage. The find-
ing implied that when SMEs are in the process of deciding to use ICT, they
assume that various supports and concerns related to the expected use and
outcome exist; therefore, they favourably evaluate the country’s overall
development, infrastructure and supports. However, during the implemen-
tation phase, that is, actual use, they experience different services and
supports. It is the reality that Bangladesh, as is the case in many developing

countries, has a lack of technological, legal and financial infrastructure.
Thus, the SMEs that use ICT-based applications suffer from various con-
straints which include lack of Internet speed; connectivity; availability of
experts and resources; a compatible legal and financial environment, insti-
tutional supports, etc. Interestingly, although use of ICT depends on the
choice made by the SME owner or manager, in reality, ICT use is seriously
affected by the government and regulatory authorities. The government
could build capable technology, legal and financial infrastructure and for-
mulate policies and supports to motivate various concerned authorities and
SME owners to ensure an integrated and properly utilised ICT operational
environment.
NOTES
1. Particular ICT advantages include: service accuracy (quality service); immedi-

acy (real-time capability); service ubiquity (i.e. services are conducted ‘anywhere’);
efficiency (low-cost service capability); privacy (services conducted are personal and
private); and service customisation (person-specific services). Important emerging
ICT applications include: communications (email, web browsing, social networks);
management (accounts management, human resource management, production
management); business (homepage, product cataloguing, order processing); transac-
tions (interactive homepage, online payment processing); finance (e-banking); ser-
vices (entertainment, information, tourism and facilitation of services in relation to
gas, water, electricity supply, etc.) and governance (governmentcitizen,
governmentbusiness and governmentgovernment communication, interaction
and transactions).
2. According to the International Telecommunication Union (ITU, 2013), in
2013, over 2.7 billion people are using the Internet which corresponds to 39% of the
world’s population. In the developing world, 31% of the population is online, com-
pared with 77% in the developed world. The Internet penetration in developed and
developing countries is significantly different with developed countries in an advan-
tageous position in various ICT applications. However, at 30 June 2012, the
Internet penetration of North America, Europe, Australia, Africa and Asia stood at
78.6%, 63.2%, 67.6%, 15.6% and 27.5% while the growth of Internet usage
between 2000 and 2012 was estimated as 153.3%, 393.4%, 218.7%, 3,606.7% and
841.9%, respectively (Internet World Stats, 2012). The high Internet usage growth
in African and Asian countries spells out the potential of ICT usage in developing
and LDCs.
3. The parsimonious research model attracts researchers seeking accuracy in
outcomes and ease of estimations. For instance, the TAM has been found to be the
most frequently used model across multidisciplinary fields of study around the
world due to the parsimony of the conceptual framework.
226 MD SHAH AZAM
4. Inclusion of additional relevant variables can improve the explanatory power

of the model and also cause changing patterns of the explanatory variables. A parsi-
monious model (such as TAM), although it ensures robustness in the analysis, may
not be able to adequately explain the phenomena due to not including all relevant
variables in the model. Thus the recent adoption-diffusion studies have employed a
relatively large number of diversified variables for comprehensive and reliable
results (viz., Tornatzky & Fleischer, 1990; Venkatesh et al., 2003, 2008; Zhu &
Kraemer, 2005; Zhu et al., 2004, 2006).
5. Information source: Internet World Stats (http://www.internetworldstats.
com/stats.htm, accessed on 18 July 2013).
6. Dhaka is the capital city of the People’s Republic of Bangladesh.

7. Recently, a significant development has occurred through the participation of
the mobile service provider in providing Internet connectivity to remote areas.
However, urban inhabitants are still gaining more benefits on several counts than
rural inhabitants.
8. SEA-ME-WE means South East Asia, Middle East and West Europe.
9. The DOI theory, widely used in the past few decades to study the adoption
of an innovation, is popularly known as Rogers’ (1983) model. The theory was
mostly cited from the book ‘Diffusion of Innovations’ by E. M. Rogers published
by the Free Press, New York in 1983. The publisher released two more editions of
the same book in 1995 and 2003 which analyse the theory in relation to contempor-
ary issues. However, the basic model (Rogers, 1983) appears unchanged in the three
editions of the book. Although much of the contemporary literature refers to the
model as Rogers (1995) or Rogers (2003), this study refers to this theoretical frame-
work as Rogers’ (1983) model to indicate its origin and applicability at different
points of time.
10. A detailed review of some important literature has been provided in Table 3.
The reviews take a tabular form to clearly address various aspects of research in
detail. The literature are organised according to their contribution and relevance
with the current study. Sections 2.5 and 2.6 also deal with literature review for the
construction of a primary research model.
11. Orlikowski and Baroudi’s (1991) classification of research paradigm is not
too dissimilar from Guba and Lincoln’s (1994) suggestion: positivist, post-positivist,
constructivist and critical paradigm. Critical analysis is an emerging stream of
research in IS which usually focuses upon the opposites, conflicts and contradictions
(Myers, 1997). However, positivist research paradigm is still a dominant paradigm
which guides 75% researches in IS while 17% and 5% researches apply interpretive
and critical research, respectively (Mingers, 2003).
12. See a Global Corruption Perceptions Index at http://www.transparency.org/
research/cpi/
13. Hair et al. (1998) suggest 200 sample units to estimate any multivariate analy-
sis which was widely executed when using CBSEM. However, researchers have fol-
lowed different suggestions for sample size determination for the PLS-SEM
estimation. According to Gefen et al. (2000), the sample size should be at least 10
times the number of items within the most complex, formative construct of the
model. Barclay et al. (1995) and Chin (1998a) stated that the sample size must be 10
times the indicators within the most complex formative construct.
14. Although ‘zero’ level regression is logically required to test the influence of
the mediation variable on the criterion variable as in Eq. (9), Baron and Kenny
(1986) suggested that the same equation be used to ascertain the relationship stated
in Step 3 and Step 4 by multiple regression as modelled in Eq. (6.8). Thus, the influ-
ence of mediating variables on the outcome variable (in control of the initial vari-
able) is also assessed through Eq. (10).
15. Internet penetration is higher in Australia (88.8%), USA (78.1%), the UK
(83.6%), Egypt (35.6%), South Africa (17.4), Malaysia (60.7%), China (40.1%) and
India (11.4%) in comparison to Bangladesh (Internet World Stats, 2013).
ACKNOWLEDGEMENT
Every reasonable effort has been made to acknowledge the owners of the
copyright materials. I would be pleased to hear from any copyright owner
who has been omitted or incorrectly acknowledged.
REFERENCES
Aaker, D. A., Kumar, V., & Day, G. S. (2004). Marketing research (8th ed.). New York, NY:
Wiley.
Agarwal, R., & Prashad, J. (1997). The role of innovation characteristics and perceived volun-
tariness in the acceptance of information technologies. Decision Science, 28(3),
557582.
Agarwal, R., & Prashad, J. (1998). The antecedents and consequences of user perceptions in
information technology adoption. Decision Support Systems, 22(1), 557582.
Agarwal, R., & Prashad, J. (1999). Are individual differences Germane to the acceptance of
new information technologies? Decision Sciences, 30(2), 361391.
Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl &
J. Beckman (Eds.), Action-control: From cognition to behavior (pp. 1139). Heidelberg:
Springer.
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision
Processes, 50, 179211.
Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior.
Englewood Cliffs, NJ: Prentice Hall.
Ajzen, I., & Madden, T. J. (1986). Prediction of goal-directed behavior: Attitudes, intentions,
and perceived behavioral control. Journal of Experimental Social Psychology, 22,
453474.
Akter, S., D’Ambra, J., & Ray, P. (2010). Service quality of mHealth platforms: Development
and validation of a hierarchical model using PLS. Electron Markets, 20, 209227.
Akter, S., D’Ambra, J., & Ray, P. (2011). Trustworthiness in mHealth information services:
An assessment of a hierarchical model with mediating and moderating effects using
228 MD SHAH AZAM
Partial Least Squares (PLS). Journal of the American Society for Information Science
and Technology, 62(1), 100116.
Alreck, P. L., & Settle, R. B. (1995). The survey research handbook (2nd ed.). Burr Ridge, IL:
Irwin.
Anandarajan, M., Igbaria, M., & Anakwe, U. P. (2002). IT acceptance in a less-developed
country: A motivational factor perspective. International Journal of Information
Management, 22, 4765.
Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: A review
and recommended two-step approach. Psychological Bulletin, 103(3), 411423.
Anderson, R. E. (1973). Consumer dissatisfaction: The effect of disconfirmed expectancy on
perceived product performance. Journal of Marketing Research, 10(1), 3844.

Andreev, P., Maoz, T., Heart, T., & Pliskin, N. (2009). Validating formative Partial Least
Square (PLS) models: Methodological review and empirical illustration. Paper pre-
sented at the International Conference on Information Systems.
Attewell, P. (1992). Technology diffusion and organizational learning The case of business
computing. Organization Science, 3(1), 119.
Avgerou, C. (2008). State of the art information systems in developing countries: A critical
research review. Journal of Information Technology, 2(3), 133146.
Azam, M. S. (2004, July 2125). Factors influencing the adoption of Internet: An inquiry into
the perceptions of university academics in Bangladesh. Paper presented at the 10th
International Conference on Information Systems Analysis and Synpaper: ISAS 2004
and the International Conference on Cybernetics and Information Technologies,
Systems and Applications, Orlando, FL, USA.
Azam, M. S. (2005, July 2830). Adoption of personal computers in Bangladesh: The effects
of perceived innovation characteristics. Paper presented at the 2nd International
Conference of the Asian Academy of Applied Business (AAAB), Padang, Indonesia.
Azam, M. S. (2006a). E-commerce in Bangladesh: Adoption intention and strategic options.
M. Phil Paper. Bangladesh: Institute of Bangladesh Studies (IBS), University of Rajshahi.
Azam, M. S. (2006b). E-commerce in Bangladesh: Understanding SMEs’ intention and explor-
ing barriers. Journal of Business Studies, 2, 3758.
Azam, M. S. (2006c). Implementation of B2C e-commerce in Bangladesh: The effects of buy-
ing culture and e-infrastructure. Advances in Global Business Research, 3, 5566.
Azam, M. S. (2007). Adoption and usage of internet in Bangladesh. Japanese Journal of
Administrative Science, 20(1), 4354.
Azam, M. S. (2013). Towards digital communication and transaction: An inquiry into the indi-
viduals’ Internet acceptance and usage behaviour in Bangladesh. Journal of
International Technology and Information Management, 22(1), 123140.
Azam, M. S., & Lubna, N. (2008a). Concerns and constraints of e-commerce: An investigation
into the service industry and manufacturing industry in Bangladesh. In D. S.
Chundawat, K. Saxena, & S. S. Bhadu (Eds.), Managing global competition: A holistic
approach (pp. 101121). Delhi: Macmillan India Ltd.
Azam, M. S., & Lubna, N. (2008b). Implementation of e-commerce in Bangladesh: Does it
benefit SMEs? In M. Z. Mamun & S. M. Jahan (Eds.), Small medium enterprise in
Bangladesh: Issues involving enterprise competitiveness (pp. 79100). Dhaka:
Association of Management Development Institutions of Bangladesh (AMDIB).
Azam, M. S., & Quaddus, M. (2009a, April 1112). Adoption of B2B e-commerce by the
SMEs in Bangladesh: An empirical analysis. Paper presented at the Asian Business
Research Conference, Dhaka, Bangladesh.
Azam, M. S., & Quaddus, M. (2009b, November 29December 2). Adoption of B2B
e-commerce by the SMEs in Bangladesh: The effects of innovation characteristics and
perceived risk. Paper presented at the Proceedings of Australia New Zealand
Marketing Academy Conference (ANZMAC), Melbourne, Australia.
Azam, M. S., & Quaddus, M. (2009c, December 26). How organizational characteristics
explain the adoption of e-commerce by the SMEs in Bangladesh. Paper presented at the
Australian Association for Information Systems (ACIS) Conference, Melbourne,
Australia.
Azam, M. S., & Quaddus, M. (2009d, December 911). Internet usage behavior in
Bangladesh: A test of the theory of planned behaviour through structural equation
modeling. Paper presented at the Curtin International Business Conference (CIBS),
Mirri, Malaysia.
Babbie, E. R. (2001). The practice of social research (9th ed.). Belmont, CA: Wadsworth
Publishing Co.
Bagozzi, R. P. (1981). Attitudes, intentions and behaviour: A test of some key hypotheses.
Journal of Personality and Social Psychology, 41, 607627.
Bagozzi, R. P. (1982). A field investigation of causal relations among cognitions, affect, inten-
tion, and behaviour. Journal of Marketing Research, 19(4), 562583.
Bagozzi, R. P., & Yi, Y. (1998). On the evaluation of structural equation models. Journal of
the Academy of Marketing Science, 16(1), 7494.
Baldwin, J. R., & Sabourin, D. (2001). Impact of the adoption of advanced information and com-
munication technologies on firm performance in the Canadian manufacturing sector.
Micro-Economic Analysis Division. Retrieved from http://www.statcan.gc.ca/pub/
11f0019m/11f0019m2001174-eng.pdf. Accessed on September 10, 2009.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.
Englewood Cliffs, NJ: Prentice Hall.
Bang, H., Ellinger, A. E., Hadjimarcou, J., & Traichal, P. A. (2000). Consumer concern,
knowledge, belief, and attitude toward renewable energy: An application of the rea-
soned action theory. Psychology & Marketing, 17(6), 449468.
Barclay, D., Higgins, C., & Thompson, R. (1995). The Partial Least Squares (PLS) approach
to causal modeling: Personal computer adoption and uses as an illustration. Technology
Studies, 2(2), 285309.
Barney, J. B. (1986). Strategic factor markets: Expectations, luck and business strategy.
Management Science, 32(10), 12311241.
Barney, J. B. (1991). Firm resources and sustained competitive advantage. Journal of
Management, 17(1), 99120.
Barney, J. B. (2001). Is the resource-based “view” a useful perspective for strategic manage-
ment research? Yes. Strategic Management Journal, 26(1), 4156.
Barney, J. B., Wright, M., & Ketchen, D. (2001). The resource-based view of the firm: Ten
years after 1991. Journal of Management, 27(6), 625641.
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social
psychological research: Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51(6), 11731182.
230 MD SHAH AZAM
Bassellier, G., Benbasat, I., & Reich, B. H. (2003). The influence of business managers’ IT
competence on championing IT. Information Systems Research, 14(4), 317336.
Bayo-Morionesa, A., & Lera-López, B. F. (2007). A firm-level analysis of antecedents of ICT
adoption in Spain. Technovation, 27, 352366.
Bazeley, P. (2002). The evolution of a project involving an integrated analysis of structured
qualitative and quantitative data: From N3 to NVivo. International Journal of Social
Research Methodology, 5(3), 229243.
Bazeley, P. (2007). Qualitative data analysis with NVivo. Thousand Oaks, CA: Sage.
Beard, T. R., Madden, G., & Azam, M. S. (2014). Blended traditional and virtual seller market
entry and performance. In J. Alleman, Á. Nı́Shúilleabháin, & P. Rappoport (Eds.),
Demand for communication services Insights and perspectives: Essays in honor of

Lester D. Taylor. New York, NY: Springer-Verlag New York, Inc.
Beckett, H. (2003). Half of SMEs have no IT strategy. Computer Weekly, 1, 34.
Berg, B. L. (2001). Qualitative research methods for the social sciences (4th ed.). Boston, MA:
Allyn and Bacon.
Berg, B. L. (2004). Qualitative research method for the social sciences (5th ed.). Boston, MA:
Pearson Education.
Bertolotti, D. S. (1984). Culture and technology. Bowling Green, OH: Bowling Green State
University Press.
Bharadwaj, A. S. (2000). A resource-based perspective on information technology capability
and firm performance: An empirical investigation. MIS Quarterly, 24(1), 169196.
Blalock, H. M. (1960). Social statistics. New York, NY: McGraw-Hill Book Company Inc.
Bock, G., Zmund, R. W., Kim, Y., & Lee, J. (2005). Behavioral intention formation in knowl-
edge sharing: Examining the roles of extrinsic motivators, social-psychological forces,
and organisational climate. MIS Quarterly, 29(1), 87111.
Boes, D. C., Graybill, F. A., & Mood, A. M. (1974). Introduction to the Theory of Statistics.
Series in probabili.
Bollen, K. A. (1987). Total, direct, and indirect effects in structural equation models. In C. C.
Clogg (Ed.), Sociological methodology 1987 (pp. 3769). Washington, DC: American
Sociological Association.
Bollen, K. A. (1989). Structural equations with latent variables. New York, NY: Wiley.
Bollen, K. A., & Lennox, R. (1991). Conventional wisdom on measurement: A structural
equation perspective. Psychological Bulletin, 110, 305314.
Brancheau, J. C., & Wetherbe, J. C. (1990). The adoption of spreadsheet software: Testing
innovation diffusion theory in the context of end-user computing. Information Systems
Research, 1(2), 115143.
Broderick, R., & Boudreau, J. (1992). Human resource management, information technology,
and the competitive edge. The Executive, 7(2), 717.
Brown, R. L. (1997). Assessing specific mediational effects in complex theoretical models.
Structural Equation Modeling, 4(2), 142156.
Brynjolfsson, E. (1993). The productivity paradox of information technology. Communication
of the ACM, 35, 6667.
Brynjolfsson, E., & Hitt, L. (1996). Paradox lost? Firm-level evidence on the returns to infor-
mation systems spending. Management Science, 42(4), 541558.
Burn, J. M. (1995). The new cultural revolution: The impact of EDI on Asia. Journal of Global
Information Management, 3(4), 623.
Burns, A. C., & Bush, R. F. (2000). Marketing research (3rd ed.). Sydney, Australia: Prentice
Hall.
Burt, R. S. (1982). Toward a structural theory of action: Network models of social structure, per-
ception, and action. New York, NY: Academic Press.
Byrne, B. M. (2001). Structural equation modeling with AMOS: Basic concepts, applications,
and programming. London: Lawrence Erlbaum Associates.
Carmines, E. G., & Zeller, R. A. (1979). Reliability and validity assessment. Sage University
Paper Series on Quantitative Applications in the Social Sciences. No. 07-017. Beverly
Hills, CA: Sage.
Carr, N. G. (2003). IT doesn’t matter. Harvard Business Review, 81(5), 4149.
Caselli, F., & Coleman, W. J. I. (2001). Cross-country technology diffusion: The case of com-
puters. The American Economic Review, 91(2), 328335.
Celuch, K., Taylor, S. A., & Goodwin, S. (2004). Understanding insurance salesperson internet
information management intentions: A test of competing models. Journal of Insurance
Issues, 27(1), 2240.
Chang, M. K. (1998). Predicting unethical behaviour: A comparison of the theory of reasoned
action and the theory of planned behaviour. Journal of Business Ethics, 17, 18251834.
Chatfield, A. T., & Bjørn-Anderson, N. (1997). The impact of IOS-enabled business process
change on business outcomes: Transformation of the value chain of Japanese airlines.
Journal of Management Information System, 14(1), 1340.
Chau, P. Y. K. (1997). Reexamining a model for evaluating information center success using a
structural equation modelling approach. Decision Sciences, 28(2), 309334.
Chau, P. Y. K., & Hu, P. J. H. (2002). Investigating healthcare professionals’ decisions to
accept telemedicine technology: An empirical test of competing theories. Information
and Management, 39(4), 197311.
Cheung, M. W. L. (2007). Comparison of approaches to constructing confidence intervals for
mediating effects using structural equation models. Structural Equation Modeling: A
Multidisciplinary Journal, 14(2), 227246.
Cheung, W., Chang, M. K., & Lai, V. S. (2000). Prediction of internet and world wide web
usage at work: A test of an extended Triandis model. Decision Support Systems, 30(1),
83101.
Chin, W. W. (1995). Partial least squares is to LISREL as principal components analysis is to
common factor analysis. Technology Studies, 2, 315319.
Chin, W. W. (1998a). Issues and opinion on structural equation modeling. MIS Quarterly,
22(1), viixvi.
Chin, W. W. (1998b). The partial least squares approach to structural equation modeling. In
G. A. Marcoulides (Ed.), Modern methods for business research (pp. 295336).
Mahwah, NJ: Lawrence Erlbaum Associates, Publisher.
Chin, W. W. (2002). PLS-Graph user’s guide. Retrieved from http://www.pubinfo.vcu.
edu/carma/Documents/OCT1405/PLSGRAPH3.0Manual.hubona.pdf. Accessed on
February 12, 2010.
Chin, W. W. (2010). How to write up and report PLS analyses. In V. E. Vinzi, W. W. Chin, J.
Henseler, & H. Wang (Eds.), Handbook of partial least squares: Concepts, methods and
application (pp. 645689). Heidelberg: Springer.
Chin, W. W., & Gopal, A. (1995). Adoption intention in GSS: Importance of beliefs. Data
Base Advances, 26, 4264.
232 MD SHAH AZAM
Chin, W. W., Marcolin, B. L., & Newsted, P. R. (2003). A partial least squares latent variable
modeling approach for measuring interaction effects: Results from a Monte Carlo
simulation study and an electronic-mail emotion/adoption study. Information Systems
Research, 14(2), 2141.
Chin, W. W., & Newsted, P. R. (1999). Structural equation modeling analysis with small sam-
ples using partial least squares. In R. Hoyle (Ed.), Statistical strategies for small sample
research (pp. 307341). Thousand Oaks, CA: Sage.
Cho, J. (2004). Likelihood to abort an online transaction: Influences from cognitive evalua-
tions, attitudes, and behavioral variables. Information & Management, 41, 827838.
Churchill, J. G. A. (1991). Marketing research: Methodological foundation. Chicago, IL:
Dryden Press.
Churchill, J. G. A., & Iacobucci, D. (2002). Marketing research: Methodological foundations.
Mason, OH: SouthWestern Cengage Learning.
Clemons, E., & Weber, B. (1990). London’s big bang: A case study of information technology,
competitive impact, and organizational change. Journal of Management Information
Systems, 6(4), 4160.
Clemons, E. K., & Row, M. C. (1991). Sustaining IT advantage: The role of structural differ-
ences. MIS Quarterly, 15(3), 275292.
Cockburn, C., & Wilson, T. D. (1996). Business use of the world wide web. International
Journal of Information Management, 16, 83102.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ:
Lawrence Erlbaum Associates.
Coltman, T., Devinney, T. M., Midgley, D. F., & Venaik, S. (2008). Formative versus reflec-
tive measurement models: Two applications of formative measurement. Journal of
Business Research, 61(12), 12501262.
Cool, A. L. (2000, January 28). A review of methods for dealing with missing data. Paper pre-
sented at the Annual Meeting of the Southwest Educational Research Association,
Dallas, TX.
Cooper, R. B., & Zmund, R. W. (1990). Information technology implementation research: A
technological diffusion approach. Management Science, 36(2), 123139.
Cox, E. P. (1980). The optimal number of response alternatives for a scale: A review. Journal
of Marketing Research, 17(November), 407422.
Cragg, P., King, M., & Hussin, H. (2002). IT alignment and firm performance in small manu-
facturing firms. Journal of Strategic Information Systems, 11, 109132.
Crawley, F. E. (1988). Antecedents of physical science teachers’ intention to use investigative
teaching methods: A test of the theory of reasoned action. Paper presented at the 61st
Annual Meeting of the American Educational Research Association, New Orleans, LA
(ERIC Document Reproduction Service No. Ed 243 672).
Crawley, F. E., & Coe, A. S. (1990). Antecedents of middle school students’ intention to enroll
in a high school science course: An application of the theory of reasoned action.
Journal of Research in Science Teaching, 27(5), 461476.
Creswell, J. W. (2003). Research design: Qualitative, quantitative, and mixed methods approach.
Thousand Oaks, CA: Sage.
Dada, D. (2006). E-readiness for developing countries: Moving the focus from the environ-
ment to the users. The Electronic Journal of Information Systems in Developing
Countries, 27(6), 114.
Daily Star. (2006). The Daily Star (Dhaka), A national English daily, May 21.
Davenport, T. (1994). Saving IT’s soul: Human-centered information management. Harvard

Business Review, 72(2), 119131.
Davila, A., Gupta, M., & Palmer, R. J. (2002). Moving procurement systems to the Internet:
The adoption and use of e-procurement technologies models. Stanford GSB Research
Paper No. 1742.
Davis, D. (2000). Business research for decision making (5th ed.). Belmont: Duxbury Press.
Davis, F. D. (1986). A technology acceptance model for empirically testing new end-user infor-
mation systems: Theory and results. PhD dissertation, Massachusetts Institute of
Technology, USA.
Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of infor-
mation technology. MIS Quarterly, 1(3), 319340.

Davis, F. D. (1993). User acceptance of information technology: Systems characteristics, user
perception and behavioral impacts. International Journal of Man-Machine Studies, 38,
475487.
Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1989). User acceptance of computer technol-
ogy: A comparison of two theoretical models. Management Science, 35(8), 9821003.
Day, G. S. (1994). The capabilities of market-driven organizations. Journal of Marketing,
58(October), 3752.
Deal, T. E., & Kennedy, A. A. (1982). Corporate culture: The rights and rituals of corporate
life. Reading, MA: Addison-Wesley Publishing Company.
Denzin, N. K., & Lincoln, Y. S. (1994). Introduction: Entering the field of qualitative research.
In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (pp. 117).
Denzin, N. K., & Lincoln, Y. S. (2005). The sage handbook of qualitative research. Thousand
Oaks, CA: Sage.
Deshpande, R. (1983). Paradigms lost: On theory and method in research marketing. Journal
of Marketing, 47(4), 101110.
Dess, G. G., & Robinson, R. B., Jr. (1984). Measuring organizational performance in the
absence of objective measures: The case of the privately held firm and conglomerate
business unit. Strategic Management Journal, 5, 265273.
Devaraj, S., & Kohli, R. (2003). Performance impact of information technology: Is actual
usage the missing link? Management Science, 49(3), 273289.
Dewan, S., & Kraemer, K. L. (2000). Information technology and productivity: Evidence from
country-level data. Management Science, 46(4), 548562.
Diamantopoulos, A., & Schlegelmilch, B. B. (1994). Linking manpower to export perfor-
mance: A canonical regression analysis of European and US data. Advances in
International Marketing, 6, 161181.
Diamantopoulos, A., & Siguaw, J. A. (2006). Formative versus reflective indicators in organi-
zational measure development: A comparison and empirical illustration. British Journal
of Management, 17, 263282.
Diamantopoulos, A., & Winklhofer, H. M. (2001). Index construct with formative indicators:
An alternative to scale development. Journal of Marketing Research, 38(2), 269277.
DiMaggio, P. J., & Powell, W. W. (1983). The iron cage revisited: Institutional isomorphism
and collective rationality in organizational fields. American Sociological Review, 48(2),
147160.
Dolecheck, M. M., & Dolecheck, C. C. (1987). Business ethics: A comparison of attitudes of
managers in Hong Kong and the United States. The Hong Kong Managers, 1, 2843.
234 MD SHAH AZAM
Doolin, B., McLeod, L., McQueen, B., & Watton, M. (2003). Internet strategies for establish-
ing retailers: Four New Zealand case studies. Journal of Information Technology Cases
and Applications, 5(4), 319.
Doukidis, G. I., Lybereas, P., & Galliers, R. D. (1996). Information systems planning in small busi-
ness: A stages of growth analysis. The Journal of Systems and Software, 33(2), 189201.
Dvir, D., Segev, E., & Shenhar, A. (1993). Technology’s varying impact on the success of stra-
tegic business units within the miles and snow typology. Strategic Management Journal,
14, 155162.
Edwards, J. R. (2001). Multidimensional constructs in organizational behaviour research: An
integrative analytical framework. Organizational Research Methods, 4(2), 144192.
Edwards, J. R., & Bagozzi, R. P. (2000). On the nature and direction of relationships between
constructs. Psychological Methods, 5(2), 155174.
Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Orlando, FL: Harcourt Brace
Jovanovich College Publishers.
Eisenhardt, K. M. (1989). Building theories from case study research. The Academy of
Management Review, 14(4), 532550.
England, G. (1975). The manager and his values—An international perspective. Cambridge,
UK: Ballinger Publishing Co.
Erez, M., & Early, P. C. (1993). Culture, self-identity and work. New York, NY: Oxford
University Press.
Erumban, A. A. A., & Jong, S. B. D. (2006). Cross country differences in ICT adoption: A
consequence of culture? Journal of World Business, 41, 302314.
Falk, R. F., & Miller, N. B. (1992). A primer for soft modelling. Akron, OH: University of
Akron Press.
Fazio, R. H., & Zanna, M. P. (1978). Attitudinal qualities relating to the strength of the atti-
tude-behavior relationship. Journal of Experimental Social Psychology, 14(4), 398–408.
Feeny, D., & Ives, B. (1990). In search of sustainability: Reaping long-term advantage from
investments in information technology. Journal of Management Information Systems,
7(1), 2746.
Fichman, R. G. (2000). The diffusion and assimilation of information technology innovations.
In R. W. Zumond (Ed.), Framing the domains of IT management: Projecting the future
through the past (pp. 105127). Cleveland, OH: Pinnaflex Publishing.
Filstead, W. J. (1979). Qualitative methods: A needed perspective in evaluation research. In
T. D. Cook & C. S. Reichardt (Eds.), Qualitative and quantitative methods in evaluation
research (pp. 3348). Beverly Hills, CA: Sage.
Fishbein, M., & Ajzen, I. (1975). Beliefs, attitude, intention, and behavior: An introduction to
theory and research. Reading, MA: Addison-Wesley.
Fitzgerald, B., & Howcroft, D. (1998). Towards dissolution of the IS research debate: From
polarisation to polarity. Journal of Information Technology, 13(4), 313326.
Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable
variables and measurement error. Journal of Marketing Research, 18(1), 3950.
Fornell, C., Lorange, P., & Roos, J. (1990). The cooperative venture formation process: A
latent variable structural modeling approach. Management Science, 36, 12461255.
Forster, D., & Cornford, T. (1992). Evaluation of information systems: Issues, models and
case studies. In S. Bhatnagare & M. Odedra (Eds.), Social implications of computers in
developing countries (pp. 304314). New Delhi: Tata McGraw Hill.
Galliano, D., & Roux, P. (2007). Organisational motives and spatial effects in internet
adoption and intensity of use: Evidence from French industrial firms. The Annals of
Regional Science, 42(2), 425448.
Gefen, D., & Straub, D. (2000). The relative importance of perceived ease of use in IS adop-
tion: A study of E-commerce adoption. Journal of the Association for Information
Systems, 1(8), 128.
Gefen, D., Straub, D., & Boudreau, M. (2000). Structural equation modeling techniques and
regression: Guidelines for research practice. Communications of the Association for
Information Systems, 7(7), 178.
Gefen, D., & Straub, D. W. (1997). Gender differences in the perceptions and use of e-mail:
An extension of the technology acceptance model. MIS Quarterly, 21(4), 389400.

Gehrt, K., & Carter, K. (1992). An exploratory assessment of catalog shopping orientation:
The existence of convenience and recreational segments. Journal of Retailing, 6(1),
2939.
George, J. F. (2002). Influence on the intent to make internet purchases. Internet Research:
Electronic Networking Applications and Policy, 12(2), 165180.
George, J. F. (2004). The theory of planned behavior and Internet purchasing. Internet
Research, 14(3), 198212.
Gibbs, J. L., & Kraemer, K. L. (2004). A cross-country investigation of the Antecedents of
scope of e-commerce use: An institutional approach. Electronic Markets, 14(2),
124137.
Givon, M. M., & Shapira, Z. (1984). Response to rating scales: A theoretical model and its
application to the number of categories problem. Journal of Marketing Research, 21(4),
410419.
Goldsmith, R. E. (1990). The validity of scale to measure global innovativeness. Journal of
Applied Business Research, 7(2), 8997.
Government of Bangladesh (GOB). (1999). The industrial policy 1999. Dhaka: Ministry of
Commerce and Industry of the People’s Republic of Bangladesh.
Greene, J. C., Caracelli, V. J., & Graham, W. F. (1989). Toward a conceptual framework for
mixed-method evaluation designs. Educational Evaluation and Policy Analysis, 11(3),
255274.
Gretton, P., Gali, J., & Parham, D. (2004). The effects of ICTs and complementary innova-
tions on Australian productivity growth. In OECD (Ed.), The economic impact of ICT:
Measurement, evidence and implications (pp. 105130). Paris: OECD.
Guba, E. G., & Lincoln, Y. S. (1994). Competing paradigms in qualitative research. In N. K.
Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (pp. 105117).
Gupta, B., Dasgupta, S., & Gupta, A. (2008). Adoption of ICT in a government organization
in a developing country: An empirical study. Journal of Strategic Information Systems,
17, 140154.
Hair, J. F., Anderson, R. I., Tatham, R. L., & Black, W. C. (1998). Multivariate data analysis
(5th ed.). Englewood Cliffs, NJ: Prentice Hall.
Hair, J. F., Ringle, C. M., & Sarstedt, M. (2011). PLS-SEM: Indeed a silver bullet. Journal of
Marketing Theory and Practice, 19(2), 139151.
Hanlon, D. (2001). Vision and support in new venture start-ups. Paper presented at the
Frontiers of Entrepreneurship Research Conference, Jonkoping, Sweden.
236 MD SHAH AZAM
Harris, R., & Davison, R. (1999). Anxiety and involvement: Cultural dimensions of attitudes
toward computers in developing societies. Journal of Global Information Management,
7(1), 2638.
Hill, C., Loch, K., Straub, D. W., & El-Sheshai, K. (1998). A qualitative assessment of Arab
culture and information technology transfer. Journal of Global Information
Management, 6(3), 2938.
Hitt, L., & Brynjolfsson, E. (1996). Productivity, business profitability, and consumer surplus:
Three different measures of information technology value. MIS Quarterly, 20(2),
121142.
Ho, T. H., Raman, K. S., & Watson, R. T. (1989). Group decision support systems: The cul-
tural factors. Paper presented at the Proceedings of the 10th Annual International
Conference on Information Systems, Boston, MA.
Hoelter, J. W. (1983). The analysis of covariance structures: Goodness-of-fit indices.
Sociological Methods and Research, 11, 325344.
Hofstede, G. (1984). Culture’s consequences: International differences in work related values.
London: Sage.
Hofstede, G. (2001). Culture’s consequences comparing values, behaviours, institutions, and orga-
nizations across the nations (2nd ed.). London: Sage.
Hohental, J. (2006). Integrating qualitative and quantitative methods in research on interna-
tional entrepreneurship. Journal of International Entrepreneurship, 4(4), 175190.
Holak, S. L., & Lehman, D. R. (1990). Purchase intentions and the dimensions of innovation:
An exploratory model. Journal of Product Innovation Management, 7, 5973.
Hong, W., Thong, J. Y. L., Wong, W. M., & Tam, K. Y. (1999). Antecedents of user accep-
tance of digital libraries: An empirical examination of individual differences and system
characteristics. Journal of Management Information System, 16(2), 91112.
Hong, W., & Zhu, K. (2006). Migrating to internet-based e-commerce: Factors affecting
e-commerce adoption and migration at the firm level. Information & Management, 43,
204221.
Hoppe, R., Newman, P., & Mugera, P. (2001, October 17). Factors affecting the adoption of
internet banking in South Africa: A comparative study. Paper presented to the
Department of Information Systems, University of Cape Town.
Horton, N. J., & Kleinman, K. (2007). Much ado about nothing: A comparison of missing
data methods and software to fit incomplete data regression models. The American
Statistician: Statistical Software Review, 61(1), 7990.
Hossain, N. (2000). E-commerce in Bangladesh: Status, potential and constraints. Retrieved
from http://www.iris.umd.edu/publications/detail.asp?ID=bd&number=6. Accessed on
September 23, 2003.
House, R. J., Hanges, P. J., Javidan, M., Dorfman, P. W., & Gupta, V. (2004). Culture, leader-
ship, and organizations: The GLOBE study of 62 societies. London: Sage.
Hudson, L. A., & Ozanne, J. L. (1988). Alternative ways of seeking knowledge in consumer
research. Journal of Consumer Research, 14(4), 508521.
Huisman, M., Krol, B., & Sonderen, E. V. (1998). Handling missing data by re-approaching
non-respondents. Quality and Quantity, 32, 7791.
Hulland, J. (1999). Use of Partial Least Squares (PLS) in strategic management research: A
review of four recent studies. Strategic Management Journal, 20(2), 195204.
Huy, L. V., & Filiatrault, P. (2006). The adoption of e-commerce in SMEs in Vietnam: A
study of users and prospectors. Proceedings of the 10th Asia-Pacific Conference on
Information Systems (pp. 13351344).
Hyman, H. H. (1955). Survey design and analysis. Glencoe, IL: Free Press.
Iacovou, C., Benbasat, I., & Dexter, A. (1995). Electronic data interchange and small organi-
zations: Adoption and impact of technology. MIS Quarterly, 19(4), 465485.
Igbaria, M., Guimaraes, T., & Davis, G. B. (1995). An assessment of structure and causation
of IS usage. The Database for Advances in Information Systems, 27(2), 6167.
Igbaria, M., Zinatelli, N., Cragg, P., & Cavaye, A. L. M. (1997). Personal computing accep-
tance factors in small firms: A structural equation model. MIS Quarterly, 21(3),
279302.
Internet World Stat. (2012). Internet world stats: Usage and population statistics. Retrieved
from http://www.internetworldstats.com/stats.htm. Accessed on December 29, 2012.
ITRC. (2000). E-commerce in Bangladesh: A readiness assessment. ITRC Research Brief.
Retrieved from http://www.itrc.techbangla.org. Accessed on February 21, 2005.
ITU. (2010). World telecommunication/ICT development report 2010: Monitoring the WSIS tar-
gets: A mid-term review. Geneva: International Telecommunication Union.
ITU. (2013). The world in 2013 ICT facts and figures. Geneva: International Telecommunica-
tion Union.
Jackson, E. L. (2008). Behavioral antecedents of the adoption of forward contracts by Western
Australia wool producers. Perth: Curtin University of Technology.
Jarvis, C. B., MacKenzie, S. B., & Podsakoff, P. M. (2003). A critical review of construct indi-
cators and measurement model misspecification in marketing, and consumer research.
Journal of Consumer Research, 30, 199218.
Jean, R. B. (2007). The ambiguous relationship of ICT and organizational performance: A lit-
erature review. Critical Perspectives on International Business, 3(4), 306321.
Jehangir, M., & Downe, A. G. (2011). Technology resources and e-commerce impact on
business performance. In V. Snasel, J. Platos, & E. El-Qawasmeh (Eds.), ICDIPC 2011
(pp. 440447). Berlin Heidelberg: Springer. Part II, CCIS 189, 440447.
Jeon, B. N., Han, K. S., & Lee, M. J. (2006). Determining factors for the adoption of
e-business: The case of SMEs in Korea. Applied Economics, 13(16), 19051916.
Jiang, J. J., Hus, M. K., Klien, G., & Lin, B. (2000). E-commerce user behavior model: An
empirical study. Human System Management, 19(4), 265276.
Johnson, R. A., & Hignite, M. A. (2000). Applying the technology acceptance model to the
WWW. Academy of Information and Management Sciences Journal, 3(2), 130142.
Jones, E., Sundaram, S., & Chin, W. (2002). Factors leading to sales force automation use: A
longitudinal analysis. Journal of Personal Selling & Sales Management, 22(3), 145157.
Judd, C. M., & Kenny, D. A. (1981). Process analysis: Estimating direction in evaluation
research. Evaluation Research, 9, 602618.
Kaplan, B., & Maxwell, J. A. (1994). Qualitative research methods for evaluating computer
information systems. In J. G. Anderson, C. E. Aydin, & S. J. Jay (Eds.), Evaluating
health care information systems: Methods and applications (pp. 4568). Thousand Oaks,
CA: Sage.
Kappos, A., & Rivard, S. (2008). A three-perspective model of culture, information systems,
and their development and use. MIS Quarterly, 32(3), 601634.
238 MD SHAH AZAM
Karahanna, E., Straub, D. W., & Chervany, N. L. (1999). Information technology adoption
across time: A cross-sectional comparison of pre-adoption and post-adoption beliefs.
MIS Quarterly, 23(2), 183213.
Kendall, J. D., Tung, L. L., Chua, K. H., Ng, C. H. D., & Tan, S. M. (2001). Receptivity of
Singapore’s SMEs to electronic commerce adoption. Journal of Strategic Information
Systems, 10, 223242.
Kettinger, W., Grover, V., Guha, S., & Segars, A. (1994). Strategic information systems revis-
ited: A study in sustainability and performance. MIS Quarterly, 18(1), 3158.
Kim, D., Cavusgil, S. T., & Calantone, R. J. (2006). Information system innovations and sup-
ply chain management: Channel relationships and firm performance. Journal of the
Academy of Marketing Science, 34(1), 4054.

King, G., Honaker, J., Joseph, A., & Scheve, K. (2001). Analyzing incomplete political science
data: An alternative algorithm for multiple imputation. American Political Science
Review, 95(1), 4969.
Kleinbaum, D. G., Kupper, L., Muller, K. E., & Nizam, A. (1998). Applied regression analysis
and other multivariable methods (3rd ed.). Pacific Grove, CA: Duxbury Press.
Kline, R. B. (2010). Principles and practices of structural equation modelling (3rd ed.). New
York, NY: Guilford Press.
Konings, J., & Roodhooft, F. (2002). The effect of e-business on corporate performance: Firm
level evidence from Belgium. De Economist, 150, 569581.
Kraemer, K. L., & Dedrick, J. (1994). Payoffs from investment in information technology:
Lessons from the Asia-Pacific region. World Development, 22(12), 19211931.
Kroeber, A. L., & Kluckhohn, C. (1952). Culture: A critical review of concepts and definitions.
Papers of the Peabody Museum of Archaeology & Ethnology, Harvard University.
Kuan, K., & Chau, P. (2001). A perception-based model of EDI adoption in small businesses
using a technology-organizational-environmental framework. Information and
Management, 38(8), 507521.
Kuhn, T. (1962). The structure of scientific revolutions. Chicago, IL: University of Chicago
Press.
Kurnia, S., & Peng, F. (2010). Electronic commerce readiness in developing countries: The
case of the Chinese grocery industry, E-commerce. In K. Kang (Ed.), E-commerce (pp.
203227). Shanghai, China: Intech.
Lal, K. (1999). Antecedents of the adoption of information technology: A case study of electri-
cal and electronic goods manufacturing firms in India. Research Policy, 28, 667680.
Law, K. S., & Wong, C. (1999). Multidimensional constructs in structural equation analysis:
An illustration using the job perception and job satisfaction constructs. Journal of
Management, 25(2), 143160.
Law, K. S., Wong, C., & Mobley, W. H. (1998). Towards taxonomy of multidimensional con-
structs. Academy of Management Review, 23(4), 741755.
Lederer, A., Mirchadani, D. A., & Sims, K. (1997). The link between information strategy and
EC. Journal of Organizational Computing and Electronic Commerce, 7(1), 1734.
Lee, C., & Shim, J. (2007). An exploratory study of Radio Frequency Identification (RFID)
adoption in the healthcare industry. European Journal of Information Systems, 16,
712724.
Leinder, D. E., & Kayworth, T. (2006). Review: A review of culture in information systems
research: Toward a theory of information technology culture conflict. MIS Quarterly,
30(2), 357399.
Li, M., & Ye, L. R. (1999). Information technology and firm performance: Linking with
environmental, strategic and managerial contexts. Information & Management, 35(1),
351.
Liao, Z., & Cheung, M. T. (2001). Internet-based e-shopping and consumer attitudes: An
empirical study. Information and Management, 38, 299306.
Lim, P., & Speece, M. W. (2002). The effects of perceived characteristics of innovation on
e-commerce adoption by SMEs in Thailand. Retrieved from http://www.blake.montclair.
edu/~cibconf/conference/DATA/Theme7/Thailand.pdf. Accessed on October 11, 2004.
Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry. Beverly Hills, CA: Sage.
Lindley, P., & Walker, S. N. (1993). Theoretical and methodological differentiation of modera-
tion and mediation. Nursing Research, 42(5), 276279.

Lu, J., Yu, C. S., Liu, C., & Yao, J. E. (2003). Technology acceptance model for wireless
Internet. Internet Research, 13(3), 206222.
Lu, L. C., Rose, G. M., & Blodgett, J. G. (1999). The effects of cultural dimensions on ethical
decision making in marketing: An exploratory study. Journal of Business Ethics, 18(1),
91106.
MacCallum, R. C., & Browne, M. W. (1993). The use of causal indicators in covariance struc-
ture models: Some practical issues. Psychological Bulletin, 114(3), 533541.
MacCorquodale, K., & Meehl, P. E. (1948). On a distinction between hypothetical constructs
and intervening variables. Psychological Review, 55, 95107.
MacKenzie, S. B., Podsakoff, P. M., & Jarvis, C. B. (2005). The problem of measurement
model misspecification in behavioral and organizational research and some recom-
mended solutions. The Journal of Applied Psychology, 90(4), 710730.
MacKinnon, D. P. (2000). Contrasts in multiple mediator models. In J. S. Rose, L. Chassin,
C. C. Presson, & S. J. Sherman (Eds.), Multivariate applications in substance use
research: New methods for new questions (pp. 141160). Mahwah, NJ: Erlbaum.
Madden, G., Azam, M. S., & Beard, T. R. (2013). Small firm performance in online markets.
Economics of Innovation and New Technology, 22(1), 99111.
Malhotra, N. K. (2002). Basic marketing research—Applications to contemporary issues. Upper
Saddle River, NJ: Prentice Hall.
Malhotra, N. K. (2004). Marketing research An applied orientation (4th ed.). Upper Saddle
River, NJ: Prentice Hall.
Malhotra, N. K., Agarwal, J., & Peterson, M. (1996). Methodological issues in cross-cultural
marketing research: A state-of-the-art review. International Marketing Review, 13,
743.
Mandal, T. (2007). Small and medium enterprises in BIMSTEC synergies and emerging issues
for cooperation. CSIRD Discussion Paper #22. Centre for Studies in International
Relations and Development, Kolkata.
Marin, G., & Marin, B. V. O. (1991). Research with hispanic populations: Applied social
research series. Beverly Hills, CA: Sage.
Marques, A., Oliveira, T., Dias, S. S., & Martins, M. F. O. (2011). Medical records system
adoption in European hospitals. The Electronic Journal of Information Systems
Evaluation, 14(1), 8999.
Marsden, D., & Littler, D. (1996). Evaluating alternative research paradigms: A market-
oriented framework. Journal of Marketing Management, 12(7), 645655.
Mata, F. J., Fuerst, W. L., & Barney, J. B. (1995). Information technology and sustained com-
petitive advantage: A resource-based analysis. MIS Quarterly, 19(4), 487505.
240 MD SHAH AZAM
Mathieson, K. (1991). Predicting user intention: Comparing the technology acceptance model
with the theory of planned behavior. Information Systems Research, 2(3), 173191.
Mathieson, K., Peacock, E., & Chin, W. W. (2001). Extending the technology acceptance
model: The influence of perceived user resources. The Data Base for Advances in
Matlay, H., & Addis, M. (2003). Adoption of ICT and e-commerce in small businesses: An
HEI-based consultancy perspective. Journal of Small Business and Enterprise
Development, 10(3), 321335.
Maykut, P., & Morehouse, R. (1994). Beginning qualitative research: A philosophical and a
practical guide. Washington, DC: Falmer Press.
McKnight, P. E., McKnight, K. M., Sidani, S., & Figueredo, A. J. (2007). Missing data: A gen-
tle introduction. New York, NY: The Guilford Press.
Mehta, R., & Sivadas, E. (1995). Direct marketing on the internet: An empirical assessment of
consumer attitudes. Journal of Direct Marketing, 8(3), 2132.
Merriam, S. B. (1998). Qualitative research and case study applications in education. San
Francisco, CA: Jossey-Bass.
Mingers, J. (2003). The paucity of multi-method research: A review of the information systems
literature. Information Systems Journal, 13(3), 233249.
Miniard, P., & Cohen, J. B. (1979). Isolating attitudinal and normative influences in beha-
vioural intentions models. Journal of Marketing Research, 16, 102110.
Miniard, P., & Cohen, J. B. (1981). An examination of Fishbein-Ajzen behavioral-intentions
model’s concepts and measures. Journal of Experimental Social Psychology, 17,
309339.
Miniard, P., & Cohen, J. B. (1983). Modelling personal and normative influences on beha-
viour. Journal of Consumer Research, 10, 169179.
Molla, A., & Licker, P. S. (2005). eCommerce adoption in developing countries: A model and
instrument. Information & Management, 42, 877899.
Moore, G. C., & Benbasat, I. (1991). Development of an instrument to measure perceptions of
adopting an information technology innovation. Information Systems Research, 2(3),
192222.
Mukhopadhyay, T. S., Kekre, S., & Kalathur, S. (1995). Business value of information tech-
nology: A study of electronic data interchange. MIS Quarterly, 19(2), 137156.
Mustamil, N. M. (2010). The influence of culture and ethical ideology on the ethical decision
making process of Malaysian managers. Perth, Australia: Doctor of Business
Administration (DBA), Curtin University.
Myers, M. (1997). Qualitative research in information systems. MIS Quarterly, 21(2),
241242.
Myers, M. D., & Klein, H. K. (2011). A set of principles for conducting critical research in
information systems. MIS Quarterly, 35(1), 1736.
Nejadirani, F., Behravesh, M., & Rasouli, R. (2011). Developing countries and electronic com-
merce: The case of SMEs. World Applied Sciences Journal, 15(5), 756764.
Neuman, W. L. (2006). Social research methods Qualitative and quantitative approaches.
Boston, MA: Allyn and Bacon.
Nieswiadomy, R. M. (2011). Foundations of nursing research (6th ed.). New York, NY:
Pearson Education.
Nunnally, J. C. (1994). Psychometric theory (3rd ed.). New York, NY: McGraw-Hill.
O’Donnell, A., Gilmore, A., Carson, D., & Cummins, D. (2002). Competitive advantage in
small and medium-sized enterprises. Journal of Strategic Marketing, 10, 205223.
Oliveira, T., & Martins, M. F. O. (2010). Firms’ patterns of e-business adoption: Evidence for
the European Union-27. The Electronic Journal of Information Systems Evaluation,
13(1), 4756.
Ordedra, M., Lawrie, M., Bennett, M., & Goodman, S. (1993). Sub-Saharan Africa: A techno-
logical desert. Communications of the ACM, 35(2), 2529.
Orlikowski, W. J., & Baroudi, J. J. (1991). Studying information technology in organizations:
Research approaches and assumptions. Information Systems Research, 2(1), 128.
Oxley, J. E., & Yeung, B. (2001). E-commerce readiness: Institutional environment and inter-
national competitiveness. Journal of International Business Studies, 32(4), 705723.

Parker, P. M., & Sarvary, M. (1996). A cross cultural study of consumer innovativeness.
A working paper in the INSEAD working paper series. INSEAD, Fontainebleau,
France.
Patton, M. Q. (1999). Enhancing the quality and credibility of qualitative analysis. Health
Services Research, 34(5), 11891208.
Pavlov, P. A., & Chai, L. (2002). What drives electronic commerce across cultures? A cross-
cultural empirical investigation of the theory of planned behavior. Journal of Electronic
Commerce Research, 3(4), 240253.
Perry, C. (1998). Processes of a case study methodology for postgraduate research in market-
ing. European Journal of Marketing, 32(910), 785802.
Peteraf, M. A. (1993). The cornerstones of competitive advantage: A resource-based view.
Strategic Management Journal, 14(3), 179191.
Peteraf, M. A., & Barney, J. B. (2003). Unravelling the resource-based tangle. Managerial and
Decision Economics, 24(4), 309323.
Petter, S., Straub, D., & Rai, A. (2007). Specifying formative constructs in information systems
research. MIS Quarterly, 31(4), 623656.
Pham, L. T., & Jordan, E. (2007). Information technology capability, the effects on organiza-
tional performance. Proceedings of the 13th Asia-Pacific Management Conference,
Melbourne, Australia (pp. 261269).
Pilat, D., & Wyckoff, A. W. (2005). The impact of ICT on economic performance: An interna-
tional comparison of three-level analysis. In W. H. Dutton, B. Kahin, R. O’Callaghan, &
A. W. Wyckoff (Eds.), Transforming enterprise, the economic and social implications
of information technology (pp. 77108). Cambridge, MA: Massachusetts Institute of
Technology.
Porter, M. E. (1985). The competitive advantage. New York, NY: The Free Press.
Porter, M. E. (1998). The competitive advantage of the nation. New York, NY: Palgrave.
Powell, T. C., & Dent-Micallef, A. (1997). Information technology as competitive advantage:
The role of human, business, and technology resources. Strategic Management Journal,
18(5), 375405.
Premkumar, G., & Potter, M. (1995). Adoption of Computer Aided Software Engineering
(CASE) technology: An innovation adoption perspective. ACM SIGMIS Database,
26(23), 105124.
Premkumar, G., & Ramamurthy, K. (1995). The role of interorganizational and organizational
factors on the decision mode for adoption of interorganizational systems. Decision
Science, 26(3), 303336.
242 MD SHAH AZAM
Premkumar, G., Ramamurthy, K., & Nilakanta, S. (1994). Implementation of electronic data
interchange: An innovation diffusion perspective. Journal of Management Information
Systems, 11(2), 157179.
Prothom Alo. (2006). The Prothom Alo (Dhaka), A Bengali national daily, June 16.
Quaddus, M., & Hofmeyer, G. (2007). An investigation into the factors influencing the adop-
tion of B2B trading exchanges in small business. European Journal of Information
Systems, 16(3), 202215.
Quaddus, M., Islam, N., & Stanton, J. (2006, August 1218). An investigation of significant
factors influencing Western Australian wool producers to produce wool: A Structural
Equation Modelling approach. In Proceedings of The International Association of
Agricultural Papers presented at the Economists Conference, Gold Coast, Australia.

Quaddus, M., & Tung, L. L. (2002). Explaining cultural differences in decision conferencing.
Communications of the ACM, 45(8), 9398.
Raghunathan, T. E. (2004). What do we do with missing data? Some options for analysis of
incomplete data. Annual Review of Public Health, 25, 99117.
Rahim, M. A., Antonioni, D., & Psenicka, C. (2001). A structural equation model of leader
power, subordinate styles of handling conflict, and job performance. International
Journal of Conflict Management, 12(3), 191211.
Rahman, M. L. (2002). E-commerce and concerns for e-commerce in Bangladesh. Journal of
the Institute of Bankers Bangladesh, 48, 88113.
Rahman, M. L. (2003). Global context of ICT development and Bangladesh. In Proceedings of
the National Conference of Inter-University IT Professionals in Bangladesh, 30
September1 October 2003, Dhaka, Bangladesh (pp. 122).
Rai, A., Patnayakuni, R., & Seth, N. (2006). Firm performance impacts of digitally enabled
supply chain integration capabilities. MIS Quarterly, 30(2), 225246.
Ralston, D. A., Giacalone, R. A., & Terpstra, R. H. (1994). Ethical perceptions of organiza-
tional politics: A comparative evaluation of American and Hong Kong managers.
Journal of Business Ethics, 13(12), 918999.
Ramayah, T., Ignatius, J., & Aafaqi, B. (2005). PC usage among students in a private institu-
tion of higher learning: The moderating role of prior experience. Educators and
Education Journal, 20, 131152.
Ramayah, T., Jamaludin, N., & Azam, M. S. (2007, May 2125). Applying theory of planned
behavior (TPB) in predicting intention to use Internet tax filing. In Proceedings of the
7th International Conference of Asian Academy of Management, Penang, Malaysia (pp.
102109).
Ramayah, T., Jantan, M., Noor, N., Razak, R. C., & Ling, K. P. (2003). Receptiveness of
Internet banking by Malaysian consumers. Asian Academy of Management Journal,
8(2), 129.
Ramayah, T., Noor, M. N. M., Nasurdin, A. M., & Sin, Q. B. (2004). The relationships
between belief, attitude, subjective norm, intention, and behavior towards infant food
formula selection: The views of the Malaysian mothers. Gadjah Mada International
Journal of Business, 6(3), 405418.
Ravichandran, T., & Lertwongsatien, C. (2005). Effect of information systems resources and
capabilities on firm performance: A resource-based perspective. Journal of Management
Raymond, L., & Magnenat-Thalmann, N. (1982). Information systems in small business: Are
they used in managerial decisions? American Journal of Small Business, 6(4), 2023.
Regan, D. T., & Fazio, R. (1977). On the consistency between attitudes and behavior: Look to
the method of attitude formation. Journal of Experimental Social Psychology, 13(1),
2845.
Reynolds, F. D. (1974). An analysis of catalog buying behavior. Journal of Marketing, 38,
4751.
Reynolds, F. D., & Neter, J. (1982). How many categories for respondent classification?
Journal of Market Research Society, 24(4), 345346.
Reynolds, N. L., Simintiras, A. C., & Diamantopoulos, A. (2003). Theoretical justification of
sampling choices in international marketing research: Key issues and guidelines for
researchers. Journal of International Business Studies, 34(1), 8089.
Rhodes, R. E., & Courneya, K. S. (2003). Investigating multiple components of attitude, sub-
jective norm, and perceived control: An examination of the theory of planned beha-
viour in the exercise domain. The British Journal of Social Psychology, 42, 129146.
Rigdon, E. E., Ringle, C. M., & Sarstedt, M. (2010). Structural modeling of heterogeneous
data with partial least squares. In N. K. Malhotra (Ed.), Review of marketing research
(Vol. 7, pp. 255296). Armonk, NY: M. E. Sharpe.
Rockart, J. F., Earl, M. J., & Ross, J. W. (1996). Eight imperatives for the new IT organiza-
tion. Sloan Management Review, 38(1), 4355.
Roessner, J. D. (1988). Innovation policy in the United States: An overview of the issues. In J.
D. Roessner (Ed.), Government innovation policy: Design, implementation, evaluation,
New York, NY: St. Martin’s Press, Inc.
Rogers, E. M. (1983). Diffusion of innovation (3rd ed.). New York, NY: Free Press.
Rogers, E. M. (1995). Diffusion of innovation (4th ed.). New York, NY: Free Press.
Rogers, E. M. (2003). Diffusion of innovation (5th ed.). New York, NY: Free Press.
Rossiter, J. R. (2002). The C-OAR-SE procedure for scale development in marketing.
International Journal of Research in Marketing, 19, 305335.
Sager, M. (1988). Competitive information systems in Australian retail banking. Information
and Management, 15, 5967.
Saleh, M. A. (2006). Antecedents of commitment to an import supplier. PhD dissertation,
Queensland University of Technology, Brisbane, Queensland, Australia.
Sandelowski, M. (1995). Sample size in qualitative research. Research in Nursing & Health,
18(2), 179183.
Sanders, N. R., & Premus, R. (2005). Modeling the relationship between firm IT capability,
collaboration, and performance. Journal of Business Logistics, 26(1), 123.
Santosa, P. I., Wei, K. K., & Chan, H. C. (2005). User involvement and user satisfaction with
information-seeking activity. European Journal of Information Systems, 14, 361370.
Sathye, M., & Beal, D. (2001). Adoption of electronic commerce by SMEs: Australian evi-
dence. Journal of E-Business, 1(1), 111.
Schlosser, A. E., Shavitt, S., & Kanfer, A. (1999). Survey of internet users’ attitudes toward
Internet advertising. Journal of Interactive Marketing, 13(3), 3454.
Schwarz, G. (1978). Estimating the dimensions of a model. The Annals of Statistics, 6(2),
461464.
Schwarzer, B. (1995). Organizing global IS management to meet competitive challenges:
Experiences from the pharmaceutical industry. Journal of Global Information
Management, 3, 516.
Scupola, A. (2003a). Adoption of e-commerce in SMEs: Lessons from stage models. In IFIP
international federation for information processing (pp. 291308). Springer.
244 MD SHAH AZAM
Scupola, A. (2003b). The adoption of internet commerce by SMEs in the south of Italy: An
environmental, technological, and organizational perspective. Journal of Global
Information Technology Management, 6(1), 5271.
Segars, A. H., & Grover, V. (1998). Strategic information systems planning success: An investi-
gation of the construct and its measurement. MIS Quarterly, 22(2), 139163.
Sethi, V., & King, W. R. (1994). Development of measures to assess the extent to which an
information technology application provides competitive advantage. Management
Science, 40(12), 16011627.
Shamsuddoha, A. K. (2004). Antecedents of firm export performance: The role of export promo-
tion programs. PhD dissertation, Queensland University of Technology, Brisbane,
Queensland, Australia.
Sheppard, B. H., Hartwick, J., & Warshaw, P. R. (1988). The theory of reasoned action: A
meta-analysis of past research with recommendations for modifications and future
research. Journal of Consumer Research, 15(3), 325343.
Shih, C., Dedrick, J., & Kraemer, K. L. (2005). Rule of law and the international diffusion of
e-commerce. Communications of the ACM, 48(11), 5762.
Shih, H. (2004). Extended technology acceptance model of internet utilisation behaviour.
Information & Management, 41, 719729.
Shih, Y., & Fang, K. (2004). The use of a decomposed theory of planned behavior to study
internet banking in Taiwan. Internet Research, 1(3), 213223.
Shih, Y., & Fang, K. (2006). Effects of network quality attributes on customer adoption inten-
tions of Internet banking. Total Quality Management & Business Excellence, 17(1),
6177.
Shimp, T., & Kavas, A. (1984). The theory of reasoned action applied to coupon usage.
Journal of Consumer Research, 11, 795809.
Siltaoja, M. E. (2006). Value priorities as combining core factors between CSR and reputation
A qualitative study. Journal of Business Ethics, 68(1), 91111.
Sobel, M. (1982). Asymptotic confidence intervals for indirect effects on structural equation
modeling. In S. Leinhardt (Ed.), Sociological methodology (pp. 290312). New York,
NY: Jossey-Bass.
Solomon, S. (1986). Small business USA: The role of small enterprises in sparking America’s
economic transformation. New York, NY: Crown.
Song, G., & Mueller-Falcke, D. (2006). The economic effects of ICT at firm-levels. In
M. Torero & J. Braun (Eds.), Information and communication technologies for develop-
ment and poverty reduction: The potential of telecommunications (pp. 166184).
Baltimore, MD: John Hopkins University Press.
Steinhoff, D., & Burgess, J. F. (1986). Small business management fundamentals (4th ed.).
Singapore: McGraw-Hill.
Stewart, K. A., & Segars, A. H. (2002). An empirical examination of the concern for informa-
tion privacy instrument. Information Systems Research, 13(1), 3649.
Straub, D. (1989). Validating instruments in MIS research. MIS Quarterly, 13(2), 147169.
Straub, D. W. (1994). The effects of culture on IT diffusion: E-mail and fax in Japan and the
US. Information Systems Research, 5(1), 2347.
Strauss, A., & Corbin, J. (1998). Basics of qualitative research techniques and procedures for
developing grounded theory. London: Sage.
Subramanian, G. H. (1994). A replication of perceived usefulness and perceived ease of use
measurement. Decision Science, 25(56), 863874.
Suppe, F. (Ed.). (1977). The structure of scientific theories (2nd ed.). Urbana, IL: University of
Illinois Press.
Szajna, B. (1996). Empirical evaluation of the revised technology acceptance model.
Management Science, 42(1), 8592.
Szymanski, D. M., Bharadwaj, S. G., & Varadarajan, P. R. (1993). An analysis of the market
share profitability relationship. Journal of Marketing, 57(3), 118.
Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics. New York, NY: Harper
Collins College Publishers.
Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics (4th ed.). Boston, MA:
Allyn and Bacon.
Tan, M., & Teo, T. S. H. (2000). Factors influencing the adoption of internet banking. Journal
of the Association for Information Systems, 1, Article 5, 142.
Tashakkori, A., & Teddlie, C. (1998). Mixed methodology: Combining qualitative and quantita-
tive approaches. Thousand Oaks, CA: Sage.
Taylor, S., & Todd, P. A. (1995a). Assessing IT usage: The role of prior experience. MIS
Quarterly, 19(2), 561570.
Taylor, S., & Todd, P. A. (1995b). Understanding information technology usage: A test of
competing models. Information Systems Research, 6(4), 144176.
Taylor, S., & Todd, P. A. (1995c). Decomposition and crossover effects in the theory of
planned behavior: A study of consumer adoption intentions. International Journal of
Research in Marketing, 12, 137155.
Tenenhaus, M., Vinzi, V. E., Chatelin, Y.-M., & Lauro, C. (2005). PLS path modeling.
Computational Statistics and Data Analysis, 48(1), 159205.
Teo, H. H., Wei, K. K., & Benbasat, I. (2003). Predicting intention to adopt interorganiza-
tional linkages: An institutional perspective. MIS Quarterly, 27(1), 1949.
Teo, T. S. H., & Ranganathan, C. (2004). Adopters and non-adopters of business-to-business
electronic commerce in Singapore. Information and Management, 42, 89102.
Teo, T. S. H., Tan, M., & Buk, W. K. (1998). A contingency model of internet adoption in
Singapore. International Journal of Electronic Commerce, 2(2), 89102.
Thatcher, S. M. B., Foster, W., & Zhu, L. (2006). B2B e-commerce adoption decisions in
Taiwan: The interaction of cultural and other institutional factors. Electronic
Commerce Research and Applications, 5, 92104.
The New Nation. (2008). The New Nation, A national English daily, January 2. Retrieved
from http://povertynewsblog.blogspot.com/2008/01/bangladesh-bank-report-says-devel-
opment.html. Accessed on June 3, 2008.
Thompson, R. L., Higgins, C. A., & Howell, J. M. (1991). Personal computing: Toward a con-
ceptual model of utilization. MIS Quarterly, 15(1), 124143.
Thong, J. Y. L. (1999). An integrated model of information systems adoption in small busi-
nesses. Journal of Management Information Systems, 15(4), 187214.
Thong, J. Y. L., & Yap, C. S. (1995). CEO characteristics, organizational characteristics and
information technology adoption in small businesses. Omega: The International Journal
of Management Science, 23(4), 429442.
Tippins, M. J., & Sohi, R. S. (2003). IT competency and firm performance: Is organizational
learning a missing link? Strategic Management Journal, 24(8), 745761.
Tolbert, P. S. (1985). Institutional environments and resource dependence: Sources of adminis-
trative structure in institutions of higher education. Administrative Science Quarterly,
30(1), 113.
246 MD SHAH AZAM
Tornatzky, L. G., & Fleischer, M. (1990). The process of technological innovation. Lexington,
MA: Lexington Books.
Triandis, H. C. (1979). Values, attitudes, and interpersonal behavior. In Nebraska Symposium
on Motivation, Beliefs, attitudes, and values (pp. 195259). Lincoln, NE: University of
Nebraska Press.
Twati, J. M., & Gammack, J. G. (2006). The impact of organisational culture innovation on
the adoption of IS/IT: The case of Libya. Journal of Enterprise Information
Management, 19(2), 175–191.
United Nations Conference on Trade and Development (UNCTAD). (2006). Using ICTs to
achieve growth and development. Background paper by the UNCTAD Secretariat. New
York, NY: United Nations.

Venkatesh, V., Brown, S. A., Maruping, L. M., & Bala, H. (2008). Predicting different concep-
tualizations of system use: The competing roles of behavioral intention, facilitating con-
ditions, and behavioral expectation. MIS Quarterly, 32(3), 483502.
Venkatesh, V., & Davis, F. D. (2000). A theoretical extension of the technology
acceptance model: Four longitudinal field studies. Management Science, 46(2),
186204.
Venkatesh, V., & Morris, M. D. (2000). Why don’t men ever stop to ask for directions?
Gender, social influences and their role in technology acceptance and usage behaviour.
MIS Quarterly, 24(1), 115139.
Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of infor-
mation technology: Toward a unified view. MIS Quarterly, 27(3), 425478.
Venkatesh, V., & Zhang, X. (2010). Unified theory of acceptance and use of technology: U.S.
vs. China. Journal of Global Information Technology Management, 13(1), 527.
Venkatraman, N. (1990). Performance implications of strategic co-alignment: A methodologi-
cal perspective. Journal of Management Studies, 27(2), 1941.
Venkatraman, N., & Zaheer, A. (1990). Electronic integration and strategic advantage: A
quasi-experimental study in the insurance industry. Information Systems Research, 1(4),
377393.
Verbeek, M. (2000). A guide to modern econometrics. New York, NY: Wiley.
Vijayasarathy, L. R. (2004). Predicting consumer intentions to use on-line shopping: The case
for an augmented technology acceptance model. Information and Management, 41(6),
747762.
Voss, G. B., & Voss, Z. G. (2000). Strategic orientation and firm performance in an artistic
environment. Journal of Marketing, 64(January), 6783.
Wahab, E. (2010). Perceived organizational support and organizational commitment in medium
enterprises in Malaysia. PhD dissertation, Curtin University, Perth, WA, Australia.
Wang, Y., Wang, Y., Lin, H., & Tang, T. (2003). Antecedents of user acceptance of Internet
banking: An empirical study. International Journal of Service Industry Management,
14(5), 501519.
Warner, T. N. (1987). Information technology as a competitive burden. Sloan Management
Review, 29(1), 5561.
Warshaw, P. R., & Davis, F. D. (1984). Self understanding and the accuracy of behavioural
expectations. Personality and Social Psychology Bulletin, 10(1), 111118.
Warshaw, P. R., & Davis, F. D. (1985). Disentangling behavioral intention and behavioral
expectation. Journal of Experimental Social Psychology, 21, 213228.
Watson, R. T., Kelly, G., Galliers, R. D., & Brancheau, J. C. (1997). Key issues in information
systems management: An international perspective. Journal of Information Management
Systems, 13(4), 91115.
Weber, D. M., & Kauffman, R. J. (2011). What drives global ICT adoption? Analysis and
research directions. Electronic Commerce Research and Applications, 10, 683701.
West, S. G., & Aiken, L. S. (1997). Toward understanding individual effects in multicompo-
nent prevention programs: Design and analysis strategies. In K. J. Bryant, M. Windle, &
S. G. West (Eds.), The science of prevention: Methodological advances from alcohol and
substance abuse research (pp. 167209). Washington, DC: American Psychological
Association.
Werner, O., & Campbell, D. T. (1970). Translating, working through interpreters, and the pro-
blem of decentering. In R. Naroll & R. Cohen (Eds.), A handbook of method in cultural
anthropology (pp. 398–420). New York, NY: The Natural History Press.
Wetzels, M., Odekerken-Schröder, G., & van Oppen, C. (2009). Using PLS path modeling for
assessing hierarchical construct models: Guidelines and empirical illustration. MIS
Quarterly, 33(1), 177195.
Woodworth, R. S. (1928). Dynamic psychology. In C. Murchison (Ed.), Psychologies of 1925.
Worcester, MA: Clark University Press.
World Bank. (2010). Country level Internet usage statistics. Retrieved from http://data.world-
bank.org/indicator/IT.NET.USER.P2?cid=GPD_44. Accessed on June 28, 2008.
Wright, S. (1921). Correlation and causation. Journal of Agricultural Research, 20, 557585.
Wu, F., Yeniyurt, S., Kim, D., & Cavusgil, S. T. (2006). The impact of information technology
on supply chain capabilities and firm performance: A resource-based view. Industrial
Marketing Management, 35(4), 493504.
Wymer, S., & Regan, E. (2005). Factors influencing e-commerce adoption and use by small
and medium businesses. Electronic Markets, 15(4), 438453.
Xu, J., & Quaddus, M. (2005). Adoption and diffusion of knowledge management systems:
Field studies of factors and variables. Knowledge-Based Systems, 18, 107115.
Xu, S., Zhu, K., & Gibbs, J. (2004). Global technology, local adoption: A cross-country inves-
tigation of internet adoption by companies in the United States and China.
Globalization and Electronic Commerce, 14(1), 1324.
Yap, C. S., Soh, C. P. P., & Raman, K. S. (1992). Information systems success factors in small
business. OMEGA International Journal of Management Science, 20(56), 597609.
Yoon, C. (2009). The effects on national cultural values on consumer acceptance of
e-commerce: Online shoppers in China. Information & Management, 46, 294301.
Zhang, C., Cui, L., Huang, L., & Zhang, C. (2007). Exploring the role of government in infor-
mation technology diffusion: An empirical study of IT usage in Shanghai firms. In
T. McMaster, D. Wastell, E. Femeley, & J. DeGross (Eds.), Organizational dynamics of
technology-based innovation: Diversifying the research agenda (Vol. 235, pp. 393407).
Boston, MA: Springer.
Zhu, K. (2004). The complementarity of information technology infrastructure and
e-commerce capability: A resource-based assessment of their value. Journal of
Management Information Systems, 21(1), 167202.
Zhu, K., Dong, S., Xu, S. X., & Kraemer, K. L. (2006). Innovation diffusion in global con-
texts: Antecedents of post-adoption digital transformation of European companies.
European Journal of Information Systems, 15(6), 601616.
248 MD SHAH AZAM
Zhu, K., & Kraemer, K. L. (2002). E-commerce matrics for net-enhanced organisations:
Assessing the value of e-commerce to firm performance in the manufacturing sector.
Information Systems Research, 13(3), 275295.
Zhu, K., & Kraemer, K. L. (2005). Post-adoption variations in usage and value of e-business
by organizations: Cross-country evidence from the retail industry. Information Systems
Research, 16(1), 6184.
Zhu, K., Kraemer, K. L., & Xu, S. (2003). Electronic business adoption by European firms: A
cross-country assessment of the facilitators and inhibitors. European Journal of
Information Systems, 12, 251268.
Zhu, K., Kraemer, K. L., & Xu, S. (2006). The process of innovation assimilation by firms in
different countries: A technology diffusion perspective on e-business. Management

Science, 52(10), 11571576.
Zhu, K., Kraemer, K. L., Xu, S., & Dedrick, J. (2004). Information technology payoff in
e-business environments: An international perspective on value creation of e-business in
the financial services industry. Journal of Management Information Systems, 21(1),
1754.
Zikmund, W. G. (2000). Business research methods. Orlando, FL: The Dryden Press.
Zikmund, W. G. (2003). Business research methods. Cincinnati, OH: Thomson/South-Western.
Zmud, R. W. (1984). An examination of ‘push-pull’ theory applied to process innovation in
knowledge work. Management Science, 30(6), 727738.
APPENDIX A
Sample Design
(Stratified random sampling)
Bangladesh
(Industry sector)
Informal Formal
ICT = 0 ICT = Small amount
Small and Medium

Large Enterprise
Enterprise (SME)
Urban Rural
SME SME
Dhaka
(80% Internet Chittagong Khulna Rajshahi
Penetration)
Manufacturing SMEs Services SMEs

[50 < M < 100 [10 < M < 25
20 < S < 50] 5 < S < 10]
Leading Emerging
RMG Leather
Random Random
240 60
250 MD SHAH AZAM
APPENDIX B
Graduate School of Business

Interview Guide
Participant Information Sheet .
My name is Md. Shah Azam. I am currently conducting a research entitled

‘Diffusion of ICT and SMEs Performance: The Mediating Effects of
Integration and Utilisation’ under Doctor of Philosophy Programme,
Graduate School of Business (GSB), Curtin University of Technology,
Australia.
Purpose of Research
I am investigating how diffusion of ICT links with organisational perfor-

mance. The study focuses Small and Medium-Sized Enterprises (SMEs) in
Bangladesh. Exploring the effects of different factors the study looks for
the strategies that may help develop the SMEs in Bangladesh.
Your Role
I am interested in finding out the factors that influence your company’s

ICT use, related decision and performance.
I would like to talk about your status of ICT use, your internal desire, atti-
tude towards ICT improvements, exchanges with the internal peoples and
external parties. The issues related to ICT resources, capabilities, national
infrastructure, as well as national culture will further be discussed as to
explore how they may link with organisational ICT decision. I would also
like to discuss about any indicators you use to determine ICT performance
as well as overall organisational performance. I will use a set of semi-
structured questions to organise a formal and fluent discussion with you
which brings in required information for the study.
The interview process will take approximately 60 minutes.
Consent to Participate
Your involvement in the research is entirely voluntary. You have the right
to withdraw at any stage without it affecting your rights or my responsibil-
ities. When you have signed the consent form I will assume that you are
agreed to participate and allow me to use your data in this research.
Confidentiality
The information you provide will be kept separate from your personal
details, and only I will have access to this. The interview transcript will not
have your name or any other identifying information on it and in adher-
ence to university policy, the interview tapes and transcribed information
will be kept in a locked cabinet for five years, before it is destroyed.
Further Information
This study has been approved by the Curtin University Human Research
Ethics Committee. If needed, verification of approval can be obtained by
either writing to the Curtin University Human Research Ethics Committee,
c/- Office of Research & Development, Curtin University of Technology,
GPO Box U1987, Perth 6845, or telephone +618-92662784. If you would
like further information about the study, please feel free to contact me on
+610432224840 or by email: [email protected]. Alternatively, you
can contact my supervisor Professor Mohammed Quaddus on +618-
92662862 or by email: [email protected]
Thank you very much for your involvement in this research, your participa-
tion is highly appreciated.
252 MD SHAH AZAM
Interview Guide
Diffusion of ICT and SMEs Performance:
The Mediating Effects of Integration and Utilisation .
PhD Programme,
Curtin University of Technology, Australia
This interview guide will help explore theoretical constructs and their inter-
relationship; and fine-tune proposed model for the study entitled ‘Diffusion
of ICT and SMEs Performance: The Mediating Effects of Integration and
Utilisation’ under PhD programme, Graduate School of Business, Curtin
University of Technology, Australia.
GENERAL
Q.1.1. Do you use any type of ICT applications in your organisation at
present?
Q.1.2. What ICT system do you use in your organisation? Please describe
the level of use of the system (High, Medium or low, etc.).
OWNER AND CEO CHARACTERISTICS

Q.2.1. How would you describe the characteristics of the owner/CEO of
the enterprise that may influence organisational ICT use?
Q.2.2. Please describe the owners’/CEOs roles in ICT adoption and other
ICT-related decision.
BELIEFS AND EVALUATIONS

Q.3.1. How would you describe your beliefs and evaluations about ICT
which may influence you to use or not use ICT in your organisation?
Q.3.2. Describe how your belief or evaluation about ICT would relate with
the firms’ overall ICT use or ICT-related decision? Please give example.
ENVIRONMENTAL PRESSURE
Q.4.1. Please describe the ICT status of your different stake holders (such
as supplier, distributor, competitor, customer and govt. agencies). Please
state if you have received any encouragement or constraints from them to
use or not use or enhance the current mode of ICT use in your organisa-
tion? Please give examples.
Q.4.2. Please describe how above-mentioned issues would link with the
firms’ ICT-related decisions?
BEHAVIOURAL INTENTION
Q.5.1. If you have all the resources (and no constraint at all) to implement
ICT, how would you express your intention in next few months?
Q.5.2. How does your above-mentioned intention (internal desire) relate

with ICT use at firm level? Please give example.
BEHAVIOURAL EXPECTATION
Q.6.1. Considering external environment, infrastructure, law and culture
can your intention be put into practice?
Q.6.2. Please describe how expectation would relate with firms’ ICT use or
associated decisions.
MACRO AND MICRO FACILITATING CONDITION

Q.7.1. What are the resources and skills essential to implement ICT in an
organisation like you? Please state the strengths and weaknesses of your
organisation to implement ICT. Please detail as much as possible.
Q.7.2. How would you describe the resources and facilities at national level
that may encourage you to implement or not implement ICT in your orga-
nisation. Please detail it as much as possible.
Q.7.3. How would you describe the way the above-mentioned facilities or
condition link with your firms’ ICT use or related decisions.
NATIONAL CULTURE
Q.9.1. How would you describe the cultural issues and its impact on orga-
nisational technology usage behaviour?
Q.9.2. How would you describe other national cultural issues (as you see it)
in addition to the above dimensions.
Q.9.3. Please describe how the national cultural issues are linked with ICT
use or ICT-related decisions.
254 MD SHAH AZAM
ICT USE, INTEGRATION AND DEGREE OF UTILISATION

Q.10.1. What types of organisational ICT supports do you have in place
currently (ICT resources, its application, etc.)?
Q.10.2. How long have you been using ICT system in your organisation.
Q.10.3. How do you describe the way ICT use and its level of use relate
with firms’ overall performance?
ORGANISATIONAL PERFORMANCE
Q.11. Would you kindly let me know how could you evaluate the overall
performance of your organisation? Give examples.
DEMOGRAPHIC PROFILE OF THE INTERVIEWEE

Q.12.1. Please mention your demographic information:
a. Name: …………………………………………………………………
b. Position: …………………………………………………………………
c. Age:
(i) below 25 Years (ii) 2534 Years (iii) 3544 Years

(iv) 4560 Years (v) 60 < Years
d. Gender:
(i) Male (ii) Female
e. Education:
(i) PhD (ii) Post graduate (iii) Graduation (iv) HSC

(v) SSC (vi) Primary school (vii) Literate (viii) Others
f. Monthly income:
(i) <10,000 Tk. (ii) 10,00020,000 Tk. (iii) 20,00030,000 Tk.

(iv) 30,00050,000 Tk. (v) 50,00080,000 Tk. (vi) above 80,000 Tk.
g. ICT experience: …………………………………………………………….

h. How long you been in this organisation: ………………………………….
i. Your field of Specialisation: ……………………………………………….
j. How long you are working in this position: ……………………………...
Q.12.2. Please brief about your organisation (Business type, nature etc.).
a. Type of industry: ……………………………………………………………
b. Nature of business (Manufacturing industry, Whole sale, retail etc.)
…………………………………………………………………………….
………………………………………………………………………………
Q.12.3. Please let me know the numbers of worker employed in the

organisation.
Full time: ……………………………………………………………………

Part-time/casual: …………………………………………………………….
Q.12.4. Please state about your market position in terms of the completion:
very bad 1 2 3 4 5 6 7 very good
Q.12.5. Please state about your revenue in last year:
substantially decreased 1 2 3 4 5 6 7 substantially increased

256 MD SHAH AZAM
Q.12.6. Please state about your sales in last year:
Q.12.7. Please state about your profit in last year:

Q.12.8. Please state about your expenditure in last year:
Q.13.1. Have you received any kind of privilege or benefits for your organi-
sation as SMEs?
(i) Yes (ii) No
Q.13.2. Have you received any grants or subsidies for ICT development in
your organisation?
(i) Yes (ii) No
Q.14. Does your organisation have 9001 or 14001 certification? Please

describe about acquiring the certification (9001:14001 Certificate) and its
usefulness (Please mention whether you have been refused or have any
desire to get 9001:14001 certification in the case of no certification).
APPENDIX C
Questionnaire code Surveyor code

Survey Questionnaire
Participation information sheet
Dear survey participant,
This survey is part of an academic research project. In response to the

Government’s recent move towards establishing digital Bangladesh, this
study is being undertaken to discover the level of small and medium-sized
enterprises’ preparations and willingness to adopt ICT in their organisation
and the resultant outcomes in terms of organisational performance. The
outcome of the survey will provide the basis for a chapter towards my PhD
degree in the Graduate School of Business, Curtin University, Western
Australia.
Your role in this survey is to answer the questions to the best of your
knowledge. There is no right or wrong answer. We are only interested in
your opinions on the issues.
Your responses will be kept strictly confidential and in no cases will your
personal or organisational identity be disclosed. The outcome of the study
will be used solely for academic purposes. However, your participation in
this study is completely voluntary. You can participate or you can with-
draw yourself anytime from the research.
It is important to be noted here that the study has been approved by the
Curtin University Human Research Ethics Committee. If needed, verifica-
tion of approval can be obtained by either writing to the Curtin University
Human Research Ethics Committee, Office of Research and Development;
Curtin University, GPO Box U1987, Perth, WA 6000, Australia, or con-
tacting to +61-8-92662784.
258 MD SHAH AZAM
My contact details are provided below if you would like further informa-
tion about the study. Alternatively, you can contact my supervisor,
Professor Mohammed Quaddus at +61-8-92662862, or +61-8-92667147, or
[email protected].
It is my pleasure to receive the completed questionnaire directly or via
post, email or fax at the address below.
Thank you for your co-operation and valuable response. Your participa-
tion is highly appreciated.

Kind regards,
Md Shah Azam
PhD Candidate
Graduate School of Business,
Curtin University
78 Murray Street, Perth 6000, Western Australia
Phone: +61-8-92881171 or +61-0432224840
Fax: +61-8-92663368
Email: [email protected]
Questionnaire
Name of the respondent: …………………………………………………….
Name of the organisation:…………………………………………………...
Nature of Business (wholesaler, retailer etc.)…………………………….

Industry category:……………………………………………………………….
Address: ……………………………………………………………………….
……………………………………………………………….
Phone: ……………………………………….. email: ……………………
[Kindly put tick (✓) marks in the appropriate check boxes]
Please specify the category of your business:
Small business Medium Business
Other (please specify)………………………………………………........
Are you currently using any kind of ICT?
No Yes
Please indicate the level of ICT you are currently using in your organisation:
Basic Computer operation
Internet with own static home page
Interactive homepage which supports product cataloguing and order processing

(Online order receiving and processing)
260 MD SHAH AZAM
Interactive home page which supports online transaction and account management
(E-business)
Complete digital communication and exchanges within and outside the organisation
(ERP or digitisation)
Please indicate which types of software you are currently using:
Basic computing software packages Human resource management

Finance/accounting software Supply chain management
Inventory management E-commerce
Customer relations management Resource planning
[Please circle the number that best matches your views on the statements]
Section A: Please indicate your perceptions about the use of ICT in your
organisation (where 1 = Strongly disagree, 2 = Disagree, 3 = Neither agree
nor disagree, 4 = Agree and 5 = Strongly agree)
Perceived usefulness
a. I find ICT useful in our organisation 1 2 3 4 5
b. Using the technology enables us to accomplish tasks more quickly 1 2 3 4 5
c. Using the technology increases our productivity 1 2 3 4 5
d. I feel that using the technology will increase our chances of growing or 1 2 3 4 5
developing the organisation
Perceived ease of use
a. I think our interaction with the technology is clear and understandable 1 2 3 4 5
b. It is easy for us to become skilful at using the technology 1 2 3 4 5
c. We find the technology easy to use 1 2 3 4 5
d. Learning to operate the technology is easy for us 1 2 3 4 5

Section B: Please state your opinion about the characteristics of the owner or
CEO of this organisation (where 1 = Strongly disagree, 2 = Disagree,
3 = Neither agree nor disagree, 4 = Agree and 5 = Strongly agree)
Owner characteristics
a. He often bears risk in doing things differently 1 2 3 4 5
b. He has original ideas 1 2 3 4 5

Section C: Please indicate your estimates of various pressures you face from
the external environment that influence your interest or rate of, ICT adoption
in your organisation (where 1 = strongly disagree, 2 = Disagree, 3 = Neither
agree nor disagree, 4 = Agree and 5 = Strongly agree)
Coercive pressure
a. Our parent company directs us to use ICT 1 2 3 4 5
b. Our major customers demand we use ICT 1 2 3 4 5
c. Our major suppliers demand we use ICT 1 2 3 4 5
Competitive pressure
a. All of our competitors have already adopted ICT 1 2 3 4 5
b. Our main competitors that have adopted 1 2 3 4 5

ICT have benefitted a great deal
Normative pressure (where 1 = None has

adopted, and 5 = All have
adopted)
a. What is the current extent of ICT adoption 1 2 3 4 5

by your firm’s customers?
b. What is the current extent of ICT adoption 1 2 3 4 5

by your suppliers?
c. What is the current extent of ICT adoption 1 2 3 4 5

by your important business partners?
d. What is the extent of ICT adoption 1 2 3 4 5

by your regulatory bodies?
262 MD SHAH AZAM
Global pressure
a. We are encouraged to adopt ICT in accessing 1 2 3 4 5

the opportunities arising from globalisation
c. It has become necessary to embrace 1 2 3 4 5

ICT for sustainability
Section D: Please indicate your estimates of the facilitating condition of your

organisation (where 1 = Strongly disagree, 2 = Disagree, 3 = Neither agree

nor disagree, 4 = Agree 5 = Strongly agree)
a. ICT is compatible with other existing systems we use 1 2 3 4 5
b. Our employees have the skill and competency 1 2 3 4 5

necessary to handle the technology
c. We have the resources necessary to use the technology 1 2 3 4 5
Section E: Please indicate your estimation on the status of the following

infrastructural supports and their ability to motivate you or your organisation
in adopting ICT (where 1 = Highly dissatisfactory, 2 = Dissatisfactory,
3 = Neither dissatisfactory nor satisfactory, 4 = Satisfactory and 5 = Highly
satisfactory)
Technology Infrastructure
a. Availability and coverage of Internet service in the country 1 2 3 4 5
b. Speed and sophistication of the Internet 1 2 3 4 5
c. Availability of necessary hardware and accessories 1 2 3 4 5
d. Availability and continuity of power supply 1 2 3 4 5
Human Infrastructure
a. Peoples knowledge and skill on IT and Internet-based business 1 2 3 4 5
b. IT and Internet-based business operational knowledge and skill at 1 2 3 4 5

institutional level
c. Institutional support for computer education 1 2 3 4 5

Legal Infrastructure
a. Legal support for digital communication and E-business 1 2 3 4 5
b. Law relating to online consumer protection 1 2 3 4 5
Financial Infrastructure
c. Online banking facility 1 2 3 4 5
d. Credit and debit card usage trend in the country 1 2 3 4 5

Government Policy and Supports
a. Government tax and customs policy 1 2 3 4 5
b. Government’s motivational programme 1 2 3 4 5
c. Government grants 1 2 3 4 5
d. Governments subsidies 1 2 3 4 5
Section F: Please indicate your estimate on the following statements about

the cultural issues (where 1 = Strongly disagree, 2 = Disagree, 3 = Neither
agree nor disagree, 4 = Agree, 5 = Strongly agree)
Power distance
a. In this society employees are expected to obey and follow their leader 1 2 3 4 5
without question
b. Power is concentrated at the top management 1 2 3 4 5
c. Employees should obey their leaders without questions 1 2 3 4 5
d. Employees don’t question any of the decisions taken by their leaders 1 2 3 4 5
a. In this society organisations have orderliness and consistency to face any 1 2 3 4 5

future uncertain events
b. Rules and regulations are important because they inform what the 1 2 3 4 5
organisation expects
In-group collectivism
a. In this society, group members take pride in the individual 1 2 3 4 5

accomplishments of their group managers
264 MD SHAH AZAM
b. Group managers take pride in the individual accomplishments of their 1 2 3 4 5

group members
c. Managers encourage group loyalty even if individual goals suffer 1 2 3 4 5
Bengali value
a. In this society, face-to-face communication is vital among the employees 1 2 3 4 5

in day-to-day operation
b. Employees are comfortable and like to interact through Bengali 1 2 3 4 5

language
c. Employees have an intimate and tight social bond with other fellow 1 2 3 4 5
members.
Ethical culture
a. Bribes and corruption are seen as common phenomena in the main 1 2 3 4 5

institutions of the society
b. Politicisation and Nepotism are seen as common phenomena in the main 1 2 3 4 5

institutions of the society
Section G: Please indicate your expectation or plans for ICT use in your
organisation (where 1 = Current user, 2 = Going to use by 1 year, 3 =
Within 1-2 years, 4 = Within 2-3 years, 5 = We are not going to use ICT in
near future)
Behavioural Expectation
a. Basic computer operation 1 2 3 4 5
b. Internet with own static home page 1 2 3 4 5
c. Interactive homepage which supports product cataloguing and order 1 2 3 4 5

processing (Online order receiving and processing)
d. Interactive home page which supports online transaction and account 1 2 3 4 5

management (E-business)
e. Complete digital communication and exchanges within and outside the 1 2 3 4 5

organisation (ERP or digitisation)
Section H: Please indicate how often you use the following ICT technology
(where 1 = Not use at all, and 5 = Use quite often)
ICT Use
a. Basic computer operation 1 2 3 4 5
b. Internet with own static home page 1 2 3 4 5
c. Interactive homepage which supports product cataloguing and order 1 2 3 4 5

processing (Online order receiving and processing)
d. Interactive home page which supports online transaction and account 1 2 3 4 5

management (E-business)
e. Complete digital communication and exchanges within and outside the 1 2 3 4 5

organisation (ERP or digitisation)
Section I: Please indicate your evaluation about ICT utilisation in your orga-
nisation (where 1 = ICT not at all utilised; 2 = utilised to conduct 20% of
functions; 3 = utilised to conduct 2050% of functions; 4 = utilised to con-
duct 5080% of functions and 5 = utilised to conduct more than 80% of
functions)
a. ICT is properly utilised in production and service-related functions 1 2 3 4 5
b. ICT is properly utilised in admin and accounts-related functions 1 2 3 4 5
c. ICT is properly utilised sales and marketing-related functions 1 2 3 4 5
Section J: Please indicate your estimates of the level of ICT integration in

your organisation (where 1 = Strongly disagree, 2 = Disagree, 3 = Neither
agree nor disagree, 4 = Agree and 5 = Strongly agree)
Level of Integration
a. ICT is used in the entire working areas in the organisation 1 2 3 4 5
b. All departments and functional areas of this organisation are integrated 1 2 3 4 5

through a single ICT system
266 MD SHAH AZAM
c. Our website is well developed for front-end functionality which supports 1 2 3 4 5

information, product catalogue, customer customisation and account
management
d. Our web applications are electronically integrated with back-office 1 2 3 4 5

systems and data bases
e. Company’s data bases are electronically integrated with suppliers and 1 2 3 4 5

partners
Section K: Please indicate your estimate about the performance of your orga-
nisation on the following counts since starting the technology (ICT) use
(where 1 = Strongly disagree, 2 = Disagree, 3 = Neither agree nor disagree,
4 = Agree and 5 = Strongly agree)
Competitiveness
a. Our sales area has been widened 1 2 3 4 5
b Interaction with customers has been increased 1 2 3 4 5
c. Competitive position has been improved 1 2 3 4 5
Internal operation productivity
a. Internal operation of the organisation became transparent 1 2 3 4 5
b. Internal operation of the organisation became structured 1 2 3 4 5
c. Productivity of the employees has been improved 1 2 3 4 5
a. Overall performance is increased 1 2 3 4 5
b. Profitability of the company has increased 1 2 3 4 5
c. Sales of the company have increased 1 2 3 4 5

Demographic questions
1. Please indicate your role in the organisation:
Owner Owner-manager Manager Employee
Other (please specify) …………………………..

2. Your age group?
Under 25 Years, 2534 Years,
3544 Years, 4560 Years,
Over 60 Years
3. Your gender? Male Female
4. Please indicate the level of your educational attainment:
Postgraduate, Graduate, HSC, SSC,
Primary school, Literate,
Others (please specify)………………………………
5. Your monthly income:
Less than10,000 Tk. 10,00020.000 Tk.
20,00030,000 Tk. 30,00050,000 Tk.
50,00080,000 Tk. above 80,000 Tk.

268 MD SHAH AZAM
6. Do you have any prior experience in using ICT?
No Yes
7. Number of employees engaged in your firm?
Full time: ————————————————————————.

Part-time/casual——————————————————————
8. Number of computer literate employees engaged in your firm?
Computer specialist: ——————————————————

(System analyst/programmer, etc.)
Computer operator: ————————————————————
General employees having computing skills:————————————
9. Percentage of employees who are involved with ICT use in your orga-
nisation? ————————————————————————————
10. Please indicate the types of internet connectivity in your

organisation:
Dial-up/DSL Broadband cable
Mobile broadband Other (please specify).................
11. How long have you been connected to the internet?
——————————————————————————— Years.
12. Kindly state the monthly average expenditure on internet and online
communications in your organisation:
Expenditure for Internet:————————————————————
Expenditure for homepage:———————————————————
Others (please specify):...........................................................................

13. Do you employ any specialised employees for ICT operation and
homepage management in your organisation?
No Yes [if no go to question 15]
14. Please state the average monthly expenditure that you are incurring
for salaries and other benefits to the specialised employees (Please give
your best estimate)?
Salary———————————————————————————
Other benefit———————————————————————
15. Please provide the total ICT investment/expenditure in your organi-

sation (Please give your best estimates)?
ICT investment/expenditure in last year:—————————— Tk
ICT investment/expenditure in last three years:——————— Tk.
16. How many offices/sales centres/stores does your organisation have?
—————————————————————
17. Is your entire business setup located in the same region?
No Yes
18. Are your customers geographically dispersed?
No Yes
19. Is your company involved in international business operations?
No Yes
20. When did you start ICT operations in your organisation?
Year: ………………
270 MD SHAH AZAM
21. When was your company established?
Year: ……………….
22. Have you received any kind of privilege or benefits for your organi-
sation as an SME?
No Yes
23. Have you received any grants or subsidies for ICT development in
your organisation from any source?
No Yes
24. Does your organisation have 9001 or 14001 certification?
No Yes
25. Would you kindly state your annual turnover in the Year
20102011?
Up to 200,000 Tk. 200,001500,000 Tk
500,0011,000,000 Tk.
1,000,0012,000,000 Tk.
2,000,0015,000,000 Tk 5,000,00110,000,000 Tk.
10,000,000 Tk.
26. Are you operating online banking?
No Yes
27. Are you communicating with your regulatory authorities online?
No Yes
28. Are you communicating with your major customers online?
No Yes
29. Are you communicating with your major suppliers online?
No Yes
272
MD SHAH AZAM

273
274
MD SHAH AZAM

275
276
MD SHAH AZAM

277
278
MD SHAH AZAM

279
280
MD SHAH AZAM

281
282
MD SHAH AZAM

283
284
MD SHAH AZAM

285
286
MD SHAH AZAM

287
288 MD SHAH AZAM
APPENDIX E
Table E1. Decision Rules for Formative or Reflective Measurements.

Formative Model Reflective Model
1. Direction of causality from Direction of causality is Direction of causality is

construct to measure implied by from items to construct from construct to items
the conceptual definition
Are the indicators (items) (a) Indicators are defining Indicators are
defining characteristics or (b) characteristics of the manifestations of the
manifestations of the construct? construct construct
Would changes in the indicators/ Changes in the indicators Changes in the indicator
items cause changes in the should cause changes in the should not cause changes in
construct or not? construct the construct
Would changes in the construct Changes in the construct Changes in the construct do
cause changes in the indicators? do not cause changes in the cause changes in the
indicators indicators
2. Interchangeability of the Indicators need not be Indicators should be
indicators/items interchangeable interchangeable
Should the indicators have the Indicators need not have Indicators should have the
same or similar content? the same or similar content same or similar content
theme
Do the indicators share a Indicators need not share a Indicators should share a
common theme? common theme common theme
Would dropping one of the Dropping an indicator may Dropping an indicator
construct indicators alter the alter the conceptual should not alter the
conceptual domain of the domain of the construct conceptual domain of
construct? theconstruct
3. Covariation among the Not necessary for Indicators are expected to
indicators indicators to covary with covary with each other
each other
Should a change in one of the Not necessarily Yes
indicators be associated with
changes in the other indicators?
4. Nomological net of the Nomological net for the Nomological net for the
construct indicators indicators may differ indicators should not differ
Are the indicators/items expected Indicators are not required Indicators are required to
to have the same antecedents and to have the same have the same antecedents
consequences? antecedents and and consequences
consequences
Source: Jarvis et al. (2003).

ACRONYMS
AGFI Adjusted goodness-of-fit index

AVE Average variance extracted
B2B Business to business
BBS Bangladesh Bureau of Statistics
BDT Bangladeshi Taka
BGMEA Bangladesh Garment Manufacturers and Exporters Association
BKMEA Bangladesh Knitwear Manufacturers and Exporters Association

CAIC Comprehensive Akaike information criteria
CATI Computer-aided telephone interview
CBIS Computer-based information systems
CBSEM Covariance-based structural equation modelling
CEO Chief Executive Officer
CFI Comparative fit index
CIS Commonwealth of Independent States
C-TAM-TPB Combined TAM and TPB
DOI Diffusion of innovation
DSS Decision support system
DTPB Decomposed theory of planned behaviour
EB Electronic business
ECaTT Electronic Commerce and Telework Trends
EDI Electronic data interchange
ERP Enterprise resource planning
GDP Gross domestic product
GLOBE Global Leadership and Organisational Behaviour Effectiveness
GoF Goodness of fit
GSS Group support system
HOTE Human, organisation, technology and environment
HOT-fit Human, organisation and technology fit
ICT Information and communications technology
IFI Incremental fit index
IS Information systems
IT Information technology
ITRC Information Technology Research Cell
ITU International Telecommunication Union
LDCs Least developed countries
LLMEB Leather and Footwear Manufacturers and Exporters Association of
Bangladesh
MDG Millennium Development Goals
ML Maximum likelihood
MM Motivational model
MPCU Model of PC utilisation
NFI Normed fit index
PLS Partial least squares
PLS-SEM Partial least squares-based structural equation modelling
RBV Resource-based view
RMG Ready-made garments
290 MD SHAH AZAM
RMSEA Root mean square error of approximation

SCCS Supply chain communication systems
SCT Social cognitive theory
SEA-ME- South East Asia, Middle East and West Europe
WE
SEM Structural equation modelling
SME Small and medium-sized enterprise
SPSS Statistical package for the social sciences
TAM Technology acceptance model
TOE Technology, organisation and environment
TPB Theory of planned behaviour

TRA Theory of reasoned action
UK United Kingdom
UNCTAD United Nations Conference on Trade and Development
USA United States of America
UTAUT Unified theory of acceptance and use of technology
VAF Variance accounted for
VIF Variance inflation factor
WB World Bank
WSIS World Summit on the Information Society

(2016) Diffusion of Ict and Sme Performance

Uploaded by

Copyright:

Available Formats

(2016) Diffusion of Ict and Sme Performance

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(2016) Diffusion of Ict and Sme Performance

Uploaded by

Copyright:

Available Formats

E-Services Adoption: Processes by Firms in Developing

Permanent link to this document:

Access to this document was granted through an Emerald subscription provided by

Information and communications technology (ICT) offers enormous

E-Services Adoption: Processes by Firms in Developing Nations

on the research objectives. Following a positivist research paradigm, this

which involve both reflective and formative measures. Partial least

improving organisational practices and procedures by eliminating the tra-

existing policies and strategies. The government and policy makers

Information and communications technology (ICT) has become an integral

business performance has not been clearly identified. Similarly, another

diverse use and utilisation of ICT in developing countries. The changing

1.2. Research Questions

The success of the use of ICT as a tool for increasing productivity in

has offered robust estimates and an increased ability to handle a large

RQ-2. How does behavioural expectation correlate with ICT use by

The main objective of this study is to formulate a theoretical framework

RO-2. Delineate the effects of culture and environmental pressure on the

RO-3. Measure the effects of facilitating condition and country readiness

RO-4. Examine the role of behavioural expectation in explaining ICT

RO-5. Examine the contribution of ICT usage to SME performance.

RO-6. Examine the mediating effects of ICT integration and utilisation

1.4. Research Background

The Internet penetration of those countries was more recently calculated as

• Urban and rural populations

Despite suffering from multidimensional problems in terms of Internet

1.5. Definition of Terms

Behavioural expectation: Refers to an individual’s self-reported subjective

Cognitive evaluation: Refers to an individual’s overall evaluation towards

ICT usage: Refers to an individual’s self-reported subjective assessment

1.6. Research Significance

the limitations of a mono-method application, that is, either a qualitative

1.6.1. Contribution to Theory

The theoretical framework now offers an opportunity to examine the

1.6.2. Contribution to Practice

1.7. Organisation of the Chapter

The chapter is structured as eight sections followed by references and

discriminant validity of the constructs and measures are discussed in detail.

future research directions. This section provides an overview of the study

This section provides background information and an overview of the

This section focuses on reviewing the literature, both the theories

a comprehensive model which finally employs individual, organisational,

cusses the Antecedents of firm performance. Finally, a testable research

2.2. Diffusion Process

The adoption of a new product is a decision process that moves through

are projected into an integrated environment. However, according to the

confirmation) is called diffusion.

An innovation comes to the market to satisfy specific needs of indivi-

diffusion or use generates improved organisational performance.

2.3. Theoretical Framework

This research has investigated SMEs’ ICT diffusion behaviour by extending

2.3.1. DOI Theory (Rogers, 1983)9

starts. An innovation may be confirmed and incorporated into the culture

Table 1. Empirical Evidence in Favour of Rogers’ (1983) Model.

2.3.2. TRA (Fishbein & Ajzen, 1975)

explain usage behaviour by adding more variables as antecedents of BI.

2.3.3. TPB (Ajzen, 1985)

Ramayah, Ignatius, & Aafaqi, 2005; Ramayah et al., 2007; Ramayah,

it has an indirect effect on actual behaviour through significantly contribut-

2.3.4. TAM (Davis, 1986)