OPERATIONS AND SUPPLY CHAIN MANAGEMENT

Vol. 4, No. 1, January 2011, pp. 31-41

ISSN 1979-3561|EISSN 1979-3871
31

Conducting a Case Study in Supply Management

Anni-Kaisa Kähkönen
Lappeenranta University of Technology, School of Business
P.O. Box 20, FI 53851 Lappeenranta, Finland
Email: [email protected]

Abstract
The aim of this article is to enhance the understanding about case studies as one of the research methods used
in qualitative research especially in the research field of supply management. By using an example case study
this article illustrates how a case study can be conducted successfully and how the validity and reliability of
the research should be evaluated. The study shows that a case study as a research method is suitable for the
research of supply chain management and purchasing and supply management. Because case study enables
the collection of rich and profound data, an in-depth analysis of the phenomenon in question can be reached
which can be highly critical for example in the development of supply relationships. The article has practical
relevance especially for academic researchers in the fields of supply chain management, logistics, supply
management and operations management by offering initial insights into the importance of the research
process in case studies. Case studies have been criticized for the lack of rigor and thus it is important to
highlight the significance of the accuracy and systematic way of action in the case research.
Keywords: qualitative research, case study, interview, data analysis

1. Introduction possibility of locating meanings and the natural

Case study is one of the research strategies that can setting are some of the advantages of qualitative
be chosen when conducting a qualitative research. research. McCutcheon and Meredith (1993) have
Case studies have been criticized for the lack of rigor argued that case studies can be seen as a particularly
(Ellram, 1996) and for having a soft nature especially useful when assessing “real world” examples and
when compared with quantitative research with Ellram (1996) continues that case studies are excellent
statistical methods. However, many researchers in providing detailed explanations of “best
argue that both qualitative and quantitative methods practices”.
are needed in order to ensure the diversity and Seuring (2008) states that the use of case studies
profundity of the academic research. Qualitative and would be particularly suitable for approaching
quantitative research may aim at different objectives several stages of a supply chain because they allow
and for example case research does not purport to direct observation of the field. In addition to Seuring
produce findings that can be generalized in a (2008), several authors have found case studies
statistical sense, but as Yin (2003) suggests, the useful and valuable in the research made in the fields
findings will be valuable in an analytical sense if they of supply chain management, logistics, operations
clarify or extend understanding of the existing management, and purchasing and supply
theory. According to Miles and Huberman (1994), management. According to Halinen and Törnroos
flexibility, richness, holism, causality assessment, the (2005) and Järvensivu and Törnroos (2010), a case
 Anni-Kaisa Kähkönen : Conducting a Case Study in Supply Management
32 Operations & Supply Chain Management 4 (1) pp 31-41 © 2011

strategy is the most suitable for studying business As Figure 1 shows, the first phase of the research
networks. Voss, Tsikriktsis and Frohlich (2002) argue process is the literature review which is the
that case research has consistently been one of the beginning of the research. As Stuart et al. (2002, p.
most powerful research methods in the research 423) state, “all research starts from an examination
made in the field of operations management. Ellram of existing theory”, and thus the process should
(1996) states that excellent opportunities exist for always start with an extensive literature review
using case study research methodology in many which gives a preunderstanding of the research
areas of logistics and purchasing. topic. A literature review creates the basis for the
The aim of this article is to describe how a case research and it enables researcher to identify possible
study can be conducted successfully and how the research gaps which moreover, defines the research
research process should proceed. The description of questions and research strategy used.
the research process is based on a literature review
on case method and on an extensive case study 2.1 Choosing the research strategy
conducted in the research field of supply The choice of methodology and data collection
management by the author. The topic of this example method should be based on the research questions
case study was a supply network containing various
(Stuart et al., 2002). Yin (2003) has argued that
buyer–supplier relationships. defining the research questions is probably the most
important step to be taken in a research. The aims
and objectives of a research are defined in the form
2. How to Conduct a Case Study? of research questions. The researcher needs to be
This section gives an overview of how a case study aware of the substance (what is the study about) and
can be conducted and how the research process form (what kind of question is asked, i.e. “who”,
develops during the research. In this article, the “what”, “where”, “why” or “how”) of the questions
research process for case studies is presented (Yin, 2003). Koulikoff-Souviron and Harrison (2005)
through eight different phases. Figure 1 shows the have stated that for example in the research of supply
eight phases for case research based on the literature chains, the questions can be asked at different levels
review of previous studies on case method and the of analysis depending on the scope of the study.
research process conducted in the example case Harland (1996) has further determined that in the
study by the author. research of purchasing and supply management
these levels are the internal chain, the dyadic
relationship, the external chain, and the network.
Figure 1. The research process for case studies
According to Voss et al. (2002), research questions
may evolve over time and the constructs may be
Literature review
modified, developed or abandoned. This is a
potential strength because it allows the acquisition
Research question definition
of more knowledge than if the questions were fixed.
Selecting the research methodology
Yin (2003) has defined that because “how” and
“why” questions are explanatory in nature, they are
Selecting the cases
likely to lead to the use of case studies.
Yin (2003) defines a case study as an empirical
Choosing data collection method inquiry that investigates a contemporary
phenomenon within its real-life context, when the
Data collecting boundaries between the phenomenon and the
context are not evident and multiple sources of
Data analysis evidence are used. The different aspects of the case
are understood in relation to one another as a
Conclusion drawing coherent whole or configuration, rather than as
modular assemblages of variables (Dubois & Araujo,
2007). According to Eisenhardt (1989, p. 534), “a case
Anni-Kaisa Kähkönen : Conducting a Case Study in Supply Management
Operations & Supply Chain Management 4 (1) pp 31-41 © 2011 33

study is a research strategy which focuses on management, four case companies were selected to
understanding the dynamics present within single be the case companies of the study. They were selected
settings.” Case study allows investigation of a because i) they represented different nodes of the
contemporary phenomenon that is difficult to network in question and hence had different roles in
separate from its context, and necessary to study it, ii) there were different kinds of relationships
within it in order to understand the dynamics in the between them, iii) they were all significant actors in
setting (Halinen & Törnroos, 2005). the industry in question, and iv) analyzing them
Moreover, the fact that a case study combined would produce a comprehensive picture of the
with interviewing allows more in-depth discussion network. Furthermore, the fact that they had different
in which why and how questions can be asked and roles, power positions and relationships in the
the fundamental reasons behind and the background network was relevant to the research aim of the study,
nature of the issues can be explained further justifies and supported the selection of case companies. There
the choice of method in several studies. Case design were four different buyer–supplier relationships
makes it possible to gather rich empirical data and between these case companies and these relationships
thereby to gain a deep understanding of the were used as units of analysis in the example case.
phenomenon in question. Yin (2003) suggests four types of design for case
As Table 1 shows, case studies may be studies: single or multiple cases examined from the
exploratory, explanatory or descriptive (see Yin, holistic perspective with single units of analysis on
2003). The aim in an exploratory case study is to the one hand, and from an embedded perspective
define questions and hypotheses for a subsequent with multiple units of analysis on the other (see
study, or to determine the feasibility of the desired Figure 2). In the example case study, the case under
research procedure. A descriptive case study gives a the investigation was a supply network containing
complete description of the phenomenon within its four different buyer–supplier relationships. Thus, it
context, whereas an explanatory case study was defined that the case study was an embedded
comprises data based on cause-effect relationships, single case study. Because the main issue of the
explaining how events happened (see also Seuring, research was a network and because there was only
2008). one network that was studied, the case was clearly a
single case. However, the network contained four
2.2 Case selection different buyer–supplier relationships which were
After defining the research questions and choosing the units of analysis in the study. Thus, there were
the research strategy, a suitable sample for the study multiple units of analysis and the perspective could
can be selected. In the case study, the selection of the be defined to be an embedded one. For example,
case and case companies as well as the number of Dubois and Araujo (2007) suggest that the use of
cases, units of analysis and the time perspective for multiple units of analysis strengthens the study. This
the research are in highly critical role. In the example case situation is illustrated in Figure 2 as an
case study from the research field of supply embedded single-case design.

Table 1. Three types of case studies

Exploratory Explanatory Descriptive

Aim To define questions and hypotheses To explain how events happened by To present a complete description of
for a subsequent study, or to determine comprising data based on cause- the phenomenon within its context
the feasibility of the desired research effect relationships
procedure

Questions How, Why How, Why Who, What, Where

Example case studies Miemczyk (2008), Peck (2005), Elango (2008), Soroor et al. (2009) Kähkönen and Lintukangas (2010),
Seuring (2003) Murray et al. (2008)
 Anni-Kaisa Kähkönen : Conducting a Case Study in Supply Management
34 Operations & Supply Chain Management 4 (1) pp 31-41 © 2011

Figure 2. Four types of case design

Sing le- ca se de sign Multiple -case design

Con text Context Context
C as e: a n et work Ca se : a n etw ork
Un it of an al ys is: U n it o f a na lys is :
a n etw ork a ne tw or k

Holistic Ca se : a n etw ork

Context Context
Un it of a na lys is : a n etw ork
C as e: a n et work Ca se : a n etw ork
Un it of an al ys is: U n it o f a na lys is :
a n etw ork a ne tw or k

Con text Context Context

Case: a network Case: a network
Ca se : a n etw ork Unit of analysis: Unit of analysis:
U n it of a n aly si s: a r ela tion sh ip a relationship a relationship
Relati o n- Relati o n- Re latio n- Relat io n -
sh ip 1 sh ip 2 sh ip 1 sh ip 2

E mbedde d Re lati on - Re lati on-

s h ip 1 s h ip 2 Context Context
Case: a network Case: a network
Unit of analysis: Unit of analysis:
Re lati on - Re lati on- a relationship a relationship
s h ip 3 s h ip 4 Relati o n- Relat io n - Re latio n- Relat io n -
sh ip 1 sh i p 2 sh ip 1 sh ip 2

However, if the researcher would have selected should always be able to justify and explain why
that the research phenomenon under the certain decisions and selections have been made.
investigation should be only a supply network and Equally important is the rationale behind the
used the whole network, not the relationship, as a selection of the data collection method and key
unit of analysis, then the situation of the example informants.
case would have been single case with holistic
perspective (see Figure 2). In the situation with 2.3 Data collection
several networks, the case design would have been
Yin (2003) suggests that multiple sources of evidence
a multiple case with a holistic or embedded
should be used to establish construct validity and
perspective depending on the units of analysis
reliability of the study. According to Dubois and
(network or relationship as a unit of analysis).
Araujo (2007), the use of multiple respondents makes
In addition to the number of cases and the it possible to capture a variety of perceptions and
number of units of analysis, the time perspective for meanings, which is vital to the understanding of
the case study needs to be determined. For example,
complex business relationships. Furthermore,
Halinen and Törnroos (2005) have defined that the having numerous and highly knowledgeable
problem of time is central in network research. Voss informants who view the focal phenomena from
et al. (2002) have made a distinction between
diverse perspectives should limit interview bias
retrospective and current cases, and further (Eisenhardt & Graebner, 2007). Therefore, several
separated the historical or long-time period studies research informants and data from different sources
to retrospective and longitudinal cases. Table 2 shows
(data triangulation) should be used in the case
the characteristics of single and multiple cases, and studies in order to limit biases.
retrospective and longitudinal cases based on the
It is common that the data for a case study is
literature review and table presented earlier in Voss
collected through interviews (see Yin, 2003).
et al. (2002). Because different choices lead to
According to Eisenhardt and Graebner (2007), the
different advantages and limits, the researcher
Anni-Kaisa Kähkönen : Conducting a Case Study in Supply Management
Operations & Supply Chain Management 4 (1) pp 31-41 © 2011 35

Table 2. Choosing the number and type of cases (modified from Voss et al., 2002)

Single cases Multiple cases Retrospective cases Longitudinal cases

Typical One case Several cases Study of the past Two or more points in time

When to use When the case: When the aim is to develop a When the aim is to look When the aim is to define how
- is a critical case to test a rich, theoretical framework backwards certain conditions change over
well-formulated theory by replication time
- is an extreme or unique - predict similar or to show
- reveals a previously contracting results among
inaccessible phenomenon replications

Advantages Depth of the study - Augment external validity - Allow collection of data on - Overcome the problems of
- Can richly describe the - Smaller observer bias historical events retrospective cases
existence of a phenomenon - Allow case comparisons - Overcome the problems of
- Provide a stronger base for retrospective cases
theory building

Limits - Generalizability - Less depth per case - Difficult to determine - May be difficult to do
- The risk of misjudging the - May require extensive cause and effect because of long time
representativeness resources and time - Participants may not recall period
- The risk of exaggerating important events
easily available data

interview is a highly efficient way of gathering rich, questions remained unchanged for all respondents.
empirical data. Interview as a data collection method The interviewees were nevertheless allowed to talk
was used also in the example case study. Because of freely, and the questions were adjusted accordingly
the research aims and questions of the example case with due consideration to the informant’s role and
study, it was significant to collect rich and profound position in the company. Questions concerning
data, and thus the interview as a data collection purchasing and supply were used mainly when the
method was seen to be the most suitable one. informant represented the supply side, and those
Given the approach of the example case study, covering information sharing and market knowledge
the questions concerned issues on the organizational were asked of informants representing marketing
(i.e. how the organization works) rather than the and sales.
individual (i.e. individual behaviour) level (see Yin, The participation of six interviewers in the data
2003). Yin (2003) and Hirschman (1986) have collection process facilitated investigator
recommended that theory should be used in case triangulation. According to Eisenhardt (1989), the
studies and for example interview questions should use of multiple investigators has two key advantages:
be formed on the basis of the existing theory. The it enhances the creative potential of the study
questions in the example study were based on the because researchers often have complementary
literature review and existing theory and they insights that add to the richness of the data, and it
covered both the supply management perspective increases confidence in the findings of the
and the perspective of marketing because both the convergence of observations. The use of multiple
buyer and supplier from each relationship need to researchers was beneficial not only in the data
be interviewed. collection process but also in the data analysis: the
The interview questions were validated in a researchers discussed the interpretations and
review process involving several researchers findings, for example, and hence brought to light
(researchers of a certain research project) who different perspectives and observations.
checked and modified the questions. They were then In the example case 29 interviews were
tested on a few key informants in the case companies, conducted with the case companies’ personnel, of
and the ones in which problems arose were modified. which 27 were face-to-face, semi-structured
If possible, the wording and the order of the interviews. The other two were email-based
 Anni-Kaisa Kähkönen : Conducting a Case Study in Supply Management
36 Operations & Supply Chain Management 4 (1) pp 31-41 © 2011

Table 3. Interviews in the example case companies selling (i.e. customer directors, sales managers)
represented companies in the supplier role. In order
Case Number of Number of Number of Company’s to obtain additional information and complementary
company interviews interviews interviewees role in the
network perspectives interviews were also conducted with
informants from other company levels and in other
Company A 16 7 3 Buyer, Supplier
positions (i.e. CEOs, commercial directors, quality
Company B 6 4 2 Buyer
managers). Given that Company A was the focal
Company C 5 1 2 Supplier
company in the case network and was in the role of
Company D 2 1 1 Supplier both buyer and supplier, and in order to cover both
Total
otal 29 13 perspectives, interviews were conducted with the
purchasing director and supply manager, as well as
with the customer director, for example.
interviews, the structured interview form being Given the focus of the example case study on
emailed to the interviewees and the questions buyer–supplier relationships, the interviews covered
focusing on the volumes of sales and purchases both the buyer and the supplier side. Take the case
between the case companies. This information was of the relationship between Company D and
collected in written form in order to ensure accuracy Company A, for example. One interviewee
of the numerical data and to minimize the possibility represented sales and marketing management in
of researcher misperception. Table 3 shows the Company D, which was in the role of supplier in
numbers of interviews and interviewees in each case this relationship, and another represented supply
company of the example case, the number of management in Company A because of its buyer role.
relationships (from the case network) in which the Company A also represented the viewpoint of sales
company was involved, and its role in the network and marketing because it was in the supplier role in
and in the relationships. its relationship with Company B, whereas the
As the table shows, there were 16 interviews in interviewees from Company B represented the
Company A, six in Company B, five in Company C, supply side. It was considered necessary to interview
and two in Company D. Company A was the focal both sides from each relationship. This yielded
company of this network and was involved in three interesting results on account of the different
relationships. It also represented both the buyer and reactions in the companies to the same issue
the supplier side, which is why more interviews were depending on the role they had.
carried out there. Given that Company B and The study also incorporated data triangulation
Company C were both involved in two relationships in the use not only of several informants but also of
and Company D only in one, the point at which new various types of data. The sources of the empirical
interviews did not provide new evidence (saturation data included: i) interview data, ii) data on sales and
point) was reached earlier in Company D than in purchases, iii) company information taken from a
the other case companies. In fact, the saturation point database, iv) the case companies’ annual reports, v)
was used in determining the sufficiency of market information, and vi) reports and studies
interviews. This is characteristic of case studies, and concerning the business of the case industry
as Eisenhardt (1989) states, data saturation occurs worldwide.
when corroborating the existing interview data with According to Eisenhardt (1989), case study data
more interviews would not bring any new may be qualitative, quantitative or both. As Tesch
information relevant to the research questions. (1992, p. 55) posits, “qualitative data is any
The research informants were selected and information the researcher gathers that is not
interviewed on the basis of their active involvement expressed in numbers”, whereas in the view of Miles
in buyer–supplier relationships, supply management and Huberman (1994, p. 1) qualitative data is “a
and marketing. The interviewees from the buyer source of well-grounded, rich descriptions and
companies mostly represented purchasing and explanations of processes in identifiable local
supply management (i.e. purchasing directors, contexts.” Both qualitative and quantitative data
supply managers), whereas those in marketing and were utilized in the example case study, and as
Anni-Kaisa Kähkönen : Conducting a Case Study in Supply Management
Operations & Supply Chain Management 4 (1) pp 31-41 © 2011 37

suggested by earlier studies, quantitative data was was colour-coded based on the interview themes. As
used for validation purposes. The numerical Griggs (1987) suggests, abbreviations as codes were
quantitative data was gathered from a database that also used and marked in the margins of the
provides company information, including basic transcripts.
information about the case companies (turnover and The pre-coding provided the basis for the final
numbers of employees, etc.), and other financial coding. This followed the interview themes because
information. Numerical data on sales and purchases it seemed to be a logical, systematic and reliable way
was collected directly from the case companies by of doing, it and secured the chain of evidence. The
means of email-based interviews. This and the coding was done manually, and it was checked several
interview data comprised the primary data of the times afterwards. Codes were used to bring order,
study. The secondary data comprised the database structure and meaning to the raw data (see Shaw, 1999;
information as well as all other material mentioned Strauss & Corbin, 1998). Colour-coding, involving the
above. allocation of a colour to each topic or theme, was used,
and a total of 11 themes were identified.
2.4 Data analysis The coding provided a basis for the classification
Miles and Huberman (1994) and Griggs (1987) suggest of the data, which was categorized by themes.
that data reduction, data display, and conclusion According to Dey (1993), categorization causes the
drawing and verification can be used as data-analysis data to lose its original shape following its division
techniques in the case research. In fact, one of the most into relevant blocks for analysis. Nevertheless, it
critical phases in the process of case research, is the facilitates more effective comparison. As Strauss and
analysis of the data. After the data is collected, the Corbin (1998) state, comparison has an important role
researcher should be aware of data analysis in data analysis because it allows identification of
techniques. However, in many academic articles this variations in the patterns found. This process also
critical phase is often discussed in too general terms. represents a method of data reduction and provides
Thus, next this article illustrates how the process of an initial level of analysis (Miles & Huberman, 1994;
data analysis was conducted in the example case Tesch, 1992). According to Miles and Huberman (1994,
study and how it can be reported in academic p. 10), data reduction refers to “the process of
writings. selecting, focusing, simplifying, abstracting, and
In the example case, all the interviews, except the transforming the data” and, as Griggs (1987, p. 20)
adds, “should not be confused with throwing away
two that were email-based, were audio-recorded, and
notes were taken during them. The recordings were or ignoring data.” Data display, on the other hand,
transcribed literally, and non-verbal occurrences such refers to the use of quotations, tables and matrices,
for instance (Griggs, 1987).
as laughter and long breaks were also indicated. The
transcriptions produced 375 pages of data, 14 pages In the example case, the coded and categorized
per interview. The two email-based interviews data was subjected to qualitative content analysis.
produced seven pages of data. After the interviews According to Tesch (1992), the units in content analysis
were transcribed literally, the resulting qualitative data may be words, sentences, paragraphs or themes (see
was read several times and notes were taken and pre- also Neuendorf, 2002). In this case they were
coding was done. The pre-coding of the data was sentences: a word was considered too narrow because
based on the interview themes. This was done by the aim was to analyze the meanings qualitatively
using colour-coding and other notes in the rather than to separate words from larger entities.
transcriptions. Paragraphs also formed units of analysis because
Reading and re-reading the data helped to sometimes the meanings and the interviewee’s key
familiarize the researcher with it and started the thoughts required more than one sentence to be fully
understood.
process of structuring and organizing the data into
meaningful units. This process heightens the According to Yin (2003), the reliability of a study
researcher’s awareness of the patterns, themes and can be increased through the use of a case study
categories of meaning embedded in the data (Shaw, protocol, developing a database and ensuring the
1999). After reading the data thoughtfully, the data chain of evidence and this advice was followed in the
 Anni-Kaisa Kähkönen : Conducting a Case Study in Supply Management
38 Operations & Supply Chain Management 4 (1) pp 31-41 © 2011

example case. The creation of a case study database reliability of the study were considered from the
facilitated the organization of the qualitative data and perspective of qualitative research. Therefore, the
the maintenance of a chain of evidence. The coded definitions of Yin (2003) for example were approved
and categorized data were put in separate documents and followed.
in order to make the analysis easier. However, the Voss et al. (2002) refer to construct validity as the
original file containing all the interview data was extent to which the correct operational measures of
preserved. The case study database comprised all this the studied concepts are established. According to Yin
data, i.e. the original transcriptions, the coded and (2003) and Stuart et al. (2002), construct validity can
categorized data, the data from the email-based be ensured through the use of multiple sources of
interviews as well as the summaries, memos, matrices evidence, establishing a chain of evidence and asking
and mind maps. key informants to review case study reports (see also
The data was scrutinized separately in accordance Beverland & Lindgreen, 2010). Multiple sources of
with the main theoretical constructs. However, the evidence (data triangulation) were used in order to
same data was used in several documents in order to strengthen the construct validity of the example study.
confirm the relationships between the constructs. First, a variety of informants were interviewed.
Mind maps also helped in identifying and analyzing Secondly, data was obtained from several sources.
the relations between the different constructs and Investigator triangulation was evident in that six
themes, and in highlighting the relations between the researchers were involved in the interview and data-
case companies. This phase helped in creating a analysis processes. Furthermore, the chain of evidence
deeper understanding of the big picture. The data- strengthened the construct validity of the study: the
analysis phase produced several summaries, mind steps of the research were well documented, and all
maps, matrices and memos. The interview data was the original evidence is available and carefully
compared and complemented with the other documented. As Seuring (2008) notes, even if the case
empirical (secondary) data. Finally, all the data was study is often criticized for its lack of rigor (see also
considered in the light of the theoretical insights and Ellram, 1996), case study research will continue to
interfaces were sought. This involved going back to allow in-depth analysis of contemporary phenomena
the literature review and collecting complementary if the research process is carried out in a structured
material in order to find support for the empirical way and is well documented.
results or to explain them. Tables and summaries were As suggested for example by Jauch, Osborn and
used in comparing the empirical data with the Martin (1980), the interviewees of the example case
theoretical framework. reviewed the case reports and the results were
presented to the informants in an attempt to prevent
interview bias. It was highly critical that the
3. Evaluating the Validity and interviewees were able to comment on possible
Reliability of the Case Study researcher misperceptions but they could not change
the results in their favour. Miles and Huberman (1994)
3.1 Validity of the case study recommended reviewing the results and
One of the most important issues in research in interpretations with informants, but nevertheless
general, is the evaluation of the validity and retaining them in their original form. Furthermore,
reliability of the study. According to Yin (2003), the according to Yin (2003) this process enhances
quality of case research depends on the construct construct validity, and according to Hirschman (1986)
validity, internal validity, external validity and this can be used to determine the credibility of the
reliability of the research design. Stuart et al. (2002) study.
suggest that the most important ones for the case Yin (2003) defines external validity as the domain
study are construct validity and internal validity. It in which a study’s findings can be generalized, in other
has to be noted that validity and reliability are words how accurately the results represent the
considered from different standpoints in qualitative studied phenomenon (see Ellram, 1996). Neuendorf
and quantitative research. Given the qualitative (2002) determines generalizability as the extent to
nature of the example case study, the validity and which the findings may be applied to other cases.
Anni-Kaisa Kähkönen : Conducting a Case Study in Supply Management
Operations & Supply Chain Management 4 (1) pp 31-41 © 2011 39

According to Stuart et al. (2002), one of the most Smircich, 1980). This often refers to the role of the
serious charges against case research is that the researcher. Hirschman (1986) highlights the meaning
samples are too small to allow generalization of the of confirmability, suggesting that outside auditors
results, and that this criticism stems from confusion should be used to judge whether or not the
between two types of generalization - statistical interpretations are drawn in a logical and
generalization on which survey research relies and unprejudiced manner. Yin (2003) also mentions the
analytical generalization on which case studies rely. utilization of several researchers in the context of
According to Stuart et al. (2002, p. 430), “with case study bias: in order to test for bias and the tolerance
research, generalization is from each case to broader of contrary findings the preliminary findings should
theory not from samples to populations.” The aim of be reported to two or three critical colleagues, who in
the example case study was at analytical turn should offer alternative explanations. In the
generalization, and in terms of external validity the example case study several researchers read and
recommendations of Yin (2003) were followed. Yin commented on the case reports. The research team
(2003) also recommends increasing external validity also discussed the interpretations and the findings,
through the use of theory in single-case studies, and and the results were also discussed with the case
thus the coding and analysis of the empirical data companies. As Yin (2003, p. 37) states, “The goal of
should be theory-based (see also Hirschman, 1986). reliability is to minimize the errors and biases in a
According to Yin (2003, p. 34), “internal validity study.”
refers to establishing a causal relationship, whereby Again according to Yin (2003), using a case study
certain conditions are shown to lead to other protocol and developing a case study database could
conditions, as distinguished from spurious increase the reliability of the study (see also Beverland
relationships”, and further that it is only a concern & Lindgreen, 2010): these recommendations were
in causal or explanatory case studies and is irrelevant followed in the example case study. According to
in descriptive and exploratory studies. Hence, given Ellram (1996, p. 104), “a case study protocol includes
that the example case study was mainly descriptive the interview guide, as well as the procedures to be
in nature, internal validity was not considered to be followed in using the test instrument.” As mentioned
as important as other forms of validity. However, above, a case study database was created: the evidence
according to Stuart et al. (2002), with regard to a obtained, the stages of the research process and the
single case, pattern matching ensures the matching interview questions and procedures were all carefully
of the data patterns with the proposed patterns, documented, thereby increasing the reliability.
thereby providing good evidence for a given
proposition and increasing internal validity. Account
was therefore taken of internal validity, and 4. Conclusions
discussions with key informants helped in ensuring Despite the criticism towards case research, several
that a particular result was attributable to a particular researchers have found it interesting and useful to
phenomenon. In other words the reasons behind the conduct a case study in order to analyze a certain
issues were discussed, which was considered research topic. This article as well as many others
necessary given that the main research question was argue that if a case study is conducted in a proper
a “how” question. way and the instructions given in previous literature
are followed, important and interesting research
3.2 Reliability of the case study results can be found and there is no need to aim at
statistical generalization but the analytical
In addition to the validity of the study, reliability is
generalization is as valid as the statistical one.
another evaluation criterion for judging research
quality. According to Yin (2003), the reliability of a By using an example case study from the
study reflects the extent to which its operations can research field of supply management, this article has
be repeated and yield the same results. Näslund (2002) demonstrated how a case study can be conducted
refers to the common definition of qualitative research and how the validity and reliability of the case study
as more subjective than quantitative research, which can be evaluated. Based on the literature review
is consequently more objective (see also Morgan & made and the research process of the example case
 Anni-Kaisa Kähkönen : Conducting a Case Study in Supply Management
40 Operations & Supply Chain Management 4 (1) pp 31-41 © 2011

study a framework of the case study process was Halinen, A., and Törnroos, J.-Å. (2005). Using case methods
created. This framework was presented in Figure 1 in the study of contemporary business networks. Journal
of Business Research 58 (9), pp. 1285 – 1297.
and it described the process of case studies in detail.
The process starts from the literature review and Harland, C.M. (1996). Supply chain management:
Relationships, chains and networks. British Journal of
through eight different phases the final goal is the
Management 7 (1), pp. 63 – 80.
conclusion drawing (see Figure 1).
Hirschman, E. (1986). Humanistic inquiry in marketing
This article enhances the understanding research: Philosophy, method, and criteria. Journal of
concerning the process of case studies and highlights Marketing Research 23 (3), pp. 237 – 249.
the significance of qualitative case studies in Jauch, L.R., Osborn, R.N., and Martin, T.N. (1980).
especially in the research fields of supply chain Structured content analysis of cases: A complementary
management, logistics, purchasing and supply method for organizational research. Academy of
management and operations management. In many Management. The Academy of Management Review 5 (4),
cases the research topic in these fields requires the pp. 517 – 525.
presence of “how” questions. Because rich and Järvensivu, T., and Törnroos, J.-Å. (2010). Case study
profound data can be collected through a case study research with moderate constructionism:
Conceptualization and practical illustration. Industrial
and interviews, the case study as a research method Marketing Management 39 (1), pp. 100 – 108.
facilitates the in-depth analysis of the phenomenon
Koulikoff-Souviron, M., and Harrison, A. (2005). Using case
under investigation. Moreover, this can be highly study methods in researching supply chains. In Research
critical in these research areas and for example in methodologies in supply chain management, H. Kotzab, S.
the development of supply relationships, chains and Seuring, M. Müller, and G. Reiner (eds.), Physica-Verlag,
networks. Heidelberg, pp. 267 – 282.
Kähkönen, A-K., and Lintukangas, K. (2010). Dyadic
relationships and power within a supply network
References context. Operations and Supply Chain Management: An
International Journal 3 (2), pp. 59 – 69.
Beverland, M., and Lindgreen, A. (2010). What makes a McCutcheon, D.M., and Meredith, J.R. (1993). Conducting
good case study? A positivist review of qualitative case case study research in operations management. Journal
research published in Industrial Marketing of Operations Management 11 (3), pp. 239 – 256.
Management, 1971–2006. Industrial Marketing
Management 39 (1), pp. 56 – 63. Miemczyk, J. (2008). An exploration of institutional
constraints on developing end-of-life product recovery
Dey, I. (1993), Qualitative data analysis – A user-friendly guide capabilities. International Journal of Production Economics
for social scientists, Routledge, London. 115 (2), pp. 272 – 282.
Dubois, A., and Araujo, L. (2007). Case research in Miles, M.B., and Huberman, M.A. (1994), Qualitative data
purchasing and supply management: Opportunities and analysis - An expanded sourcebook, Sage Publications,
challenges. Journal of Purchasing & Supply Management Thousand Oaks, California.
13 (3), pp. 170 – 181.
Morgan, G., and Smircich, L. (1980). The case for qualitative
Eisenhardt, K.M. (1989). Building theories from case study research. Academy of Management Review 5 (4), pp. 491 –
research. Academy of Management. The Academy of 500.
Management Review 14 (4), pp. 532 – 550.
Murray, G.J., Rentell, P.G., and Geere, D. (2008).
Eisenhardt, K.M., and Graebner, M.E. (2007). Theory Procurement as a shared service in English local
building from cases: Opportunities and challenges. government. International Journal of Public Sector
Academy of Management Journal 50 (1), pp. 25 – 32. Management 21 (5), pp. 540 – 555.
Elango, B. (2008). Using outsourcing for strategic Neuendorf, K.A. (2002), The content analysis guidebook, Sage
competitiveness in small and medium-sized firms. Publications, Thousand Oaks, California.
Competitiveness Review: An International Business Journal
18 (4), pp. 322 – 332 Näslund, D. (2002). Logistics needs qualitative research –
Especially action research. International Journal of Physical
Ellram, L.M. (1996). The use of the case study method in Distribution & Logistics Management 35 (5), pp. 321 – 338.
logistics research. Journal of Business Logistics 17 (2), pp.
93 – 138. Peck, H. (2005). Drivers of supply chain vulnerability: an
integrated framework. International Journal of Physical
Griggs, S. (1987). Analysing qualitative data. Journal of the Distribution & Logistics Management 35 (4), pp. 210 – 232
Market Research Society 29 (1), pp. 15 – 34.
Anni-Kaisa Kähkönen : Conducting a Case Study in Supply Management
Operations & Supply Chain Management 4 (1) pp 31-41 © 2011 41

Seuring, S.A. (2008). Assessing the rigor of case study Strauss, A., and Corbin, J. (1998), Basics of qualitative research:
research in supply chain management. Supply Chain Techniques and procedures for developing grounded theory,
Management: An International Journal 13 (2), pp. 128 – Sage publications, Thousand Oaks, California.
137. Stuart, I., McCutcheon, D., Handfield, R., McLachlin, R.,
Seuring, S.A. (2003). Outsourcing into service factories - An and Samson, D. (2002). Effective case research in
exploratory analysis of facility operators in the German operations management: A process perspective. Journal
chemical industry. International Journal of Operations & of Operations Management 20 (5), pp. 419 – 433.
Production Management 23 (10), pp. 1207 – 1223 Tesch, R. (1992), Qualitative research: Analysis types and
Shaw, E. (1999). A guide to the qualitative research process: software tools, The Falmer Press, New York.
Evidence from a small firm study. Qualitative Market Voss, C., Tsikriktsis, N., and Frohlich, M. (2002). Case
Research 2 (2), pp. 59 – 70. research in operations management. International Journal
Soroor, J., Tarokh, M.J., and Shemshadi, A. (2009). of Operations & Production Management 22 (2), pp. 195 –
Theoretical and practical study of supply chain 219.
coordination. Journal of Business & Industrial Marketing Yin, R.K. (2003), Case study research: Design and methods, Sage
24 (2), pp. 131 – 142. Publications, Thousand Oaks, California.

Anni-Kaisa Kähkönen, D.Sc (Econ. & Bus.Adm.), is a researcher in Lappeenranta University of Technology,
School of Business. Her current areas of interest include buyer-supplier collaboration, power relations, value
nets and supply strategies, especially in the food industry. She is the corresponding author and can be contacted
at: [email protected]