Theeffect of Recruitment and Selection Process On The Performance of Organization (Cooperative Bank of Oromia)

American Journal Of Multidisciplinary Research & Review (AJMRR) 2023
American Journal of Multidisciplinary Research & Review (AJMRR)

Volume-02, Issue-05,
Pp-01-27
www.ajmrr.com
Research Paper Open Access
THEEFFECT OF RECRUITMENT AND SELECTION PROCESS

ON THE PERFORMANCE OF ORGANIZATION
(COOPERATIVE BANK OF OROMIA)
DESALEGN FEKADU ETEFA (PhD)
ADDIS ABABA, ETHIOPIA
Ambo University: College of Business and Economics, Department of Cooperative Studies,
ABSTRACT: In this study, the main objective of the study was to assess the effect of recruitment & selection processes
on the performance of Cooperative Bank of Oromia. To do so, the methodology employed was descriptive &
explanatory research design. Both qualitative & quantitative research approaches were followed. To conduct this
study, primary and secondary data were collected. Majority of the data were collected from 300 employees working in
82 branches and the size was determined by formula of Yamane 1967. To proportionally select employees, Cochran
1963 formula was used and then respondents were randomly selected. The tools used were questionnaires, interview,
focus group discussion, observation & document review. To collect the data, the tools were pre-tested and then pilot
was taken to check for errors and any ambiguity. Questionnaires were prepared in 5-point likert scale. Analyses were
done using both descriptive and inferential statistical tools. The descriptive statistics include, mean & standard
deviation, while the inferential statistics were chi-square, ordinal logistic regression &Spearman’s Rank Correlation.
The results of the research were markedly reported & summarized. Accordingly, the total descriptive statistics of the
respondents have been described as feeling alright with the mean values >3 > 60%> 180 for both predicted &
predictor variables. Besides, both the primary and secondary data ratified that the bank’s performance is promising.
Moreover, the inferential statistics computed indicated the degree of relationship between the variables. The chi-
square values of all explanatory variables were significant at P-values<α=.05 & Chi-Square at α=.05, df 4 were > the
Table value (9.49). This shows there is association between the variables & performance of the bank. Together with
this, ordinal logistic regression was applied. Consequently, computed values verified that there was strong links
between the practices & performance of the bank. The correlation between performance of the bank & the variable
was above 0.5, which is labeled strong. Besides the odds ratio of the explanatory variable was>1, which means a unit
increase in the variables cause > 1 entity increase on the performance of the bank. The same result was also indicated
by the inferential statistics. Accordingly, recruitment & selection has strong effect on the performance of Cooperative
Bank of Oromia. Thus, it was concluded that proper handling & execution of practices of human resource management
(recruitment and selection) has eloquent value on the performance of the bank. Lastly, the proposed intervention
strategies identified were advertising properly recruitment criteria to invite competent employees and announcing the
result of selection to candidates on time.
Keywords: Cooperative Bank;Effect;Performance; Recruitment and Selection; Statistics
I. INTRODUCTION
1.1. Background of the Study
Recruitment is the method of inviting proficient persons to submit their document to an organization so as
to get job. Likewise, selection is the method by which directors and others utilize particular tools to select from a
pool of candidates an individual or people more likely to succeed within the job(s) given administration objectives
and legitimate necessities (Bratton & Gold, 2017). The topic investigated in this study was the performance of the
Cooperative Bank of Oromia (PCBO) in line to the effect of recruitment and selection process of human resource.
Scholars in the field of HRM stated that the success and survival of the organization is contingent on the
AJMRR Journal Page |1

several factors namely capital, equipment, and human power and other resources (Egbu, 2004; Hsieh & Chen, 2011;
Asiaei & Jusoh, 2015). Of all these resources, the human factor is the most significant one, since it is the people that
use all other resources (Rudrabasavaraj, 2005 cited in Seidu, 2012). So that it is confessed that the workers of the
organization greatly contribute for the attainment of organizational missions (Mehta, 2016). This means without the
productive efforts of workers, the material resources of an institution would be of no use. Thus, the intention of the
researcher was to show Cooperative Bank of Oromia to appropriately apply human resource management practices
to enhance its performance by filling the prevailing gaps.
1.2. Objectives of the Study
1. To explore the effect of recruitment & selection processes on the performance of organization.
2. To assess the performance of Cooperative Bank of Oromia.
3. To identify the associations between the variables.
1.3. Research Hypotheses
Research hypothesis is an expected relationship between variables, which is clear, specific, testable and falsifiable
(Toledo et al., 2011; Biesecker, 2013). Thus, to address the above identified research objectives, the research
hypotheses were formulated, which were also tested. Consequently, it is supposed that there are positive and
significant harmonious associations among the dependent variable, in this case performance of Cooperative Bank of
Oromia and the predictor variable recruitment and selectionby the alternative hypotheses (HA) while the opposite
nature of relationships between the variables was predicted by the null hypotheses (H0).
II. LITERATURE REVIEW

Authors such as Bell et al (2018); Shahsavar & Kourepaz (2020) assert that the reason for writing a literature review
is to give avocation to an exploration question and to help the researcher's decision of ensuing research design. Thus,
this section attempted to present related literature pertaining to the research questions introduced in chapter one.
2.1. The Link between Recruitment & Selection processes and Organizational Performance
Recruitment and selection play decisive role in determining organizations’ effectiveness and performance.
Organizations acquire workers who already possess applicable knowledge, skills and aptitudes so as
to make correct prediction concerning their future talents. Evidence from the findings of Lynch & Smith (2010)
reveals that recruitment and selection are the initial procedure to measure employees. Empirical findings show that
recruitment and selection processes are positively associated with the employee’s performance (Ferris et al., 2002).
In connection to this, in their studies, Chand & Katou (2007); Marques (2007) confirmed that recruitment and
selection processes are related with the profitability and then they suggested that organization’s
management should emphasis on these human resource management practices leading to an improved profit. Due to
this, to make the practice industrious, the involvement of HR Department in employment procedure ropes to build
up the knowledge of the workforce that enables to possess the desirable skills. Thus, positive enabling conditions
have to be created that has an ultimate effect on the virtue of the employees.
III. RESEARCH METHODOLOGY

In line with this, under this topic, the study area’s description, research design, sources of data and data
types, population and sampling design method, sample size determination, proportional sampling techniques, data
collection methods, data analysis approach, validity and reliability were presented.
In this research, both descriptive and explanatory research design was used. It used descriptive since it
included surveying and fact finding(Creswell, 2014). The surveying used was cross-sectional survey. A cross-
sectional study is a sort of observational investigation that analyses the gathered information from the source
(population, or from a subgroup), at a fixed point in time. Explanatory research, on the other hand, helps in the
attempt to connect ideas to understand cause and effect.An explanatory design is used since it examined the
magnitude and relationship among variables(Schmidt & Kohlmann, 2008).With regard to the research approach, the
research followed mixed research methods.In connection to the sources of data, they were gathered from both
primary and secondary sources that were used as an input for the research findings. The sample size of the studywas
1,206 employees from 82 branches functioning as a city branches. In statistics, a sampling frame is the source
material from which units are drawn for the sample. Consequently, these branches were clustered into 10 areas to
make ease the management of data collection and its supervision. Accordingly, city branches were selected
purposively from the other branches existing in the country. After identifying the sampling frame with convenience,
branches of the bank in Addis Ababa were given a chance to be incorporated in the sample according to their cluster
proportionally which provided equivalent chance of inclusion in the sample. In this process, random sampling
procedure was employed where at the first place from 82 branches; employees of the bank from the branches were
selected (Table 3.1). In this sampling, branch managers were excluded to reach them through other means
particularly interview while other employees were covered through questionnaire.
In this dissertation, sample size was decided by the help of Yemane 1967: n = N / [1 + N (e) 2], where n represented
sample size, N represented population, and e was the desired level of precision (0.05) i.e., the likelihood that the
answer fallen outside that range (the probability not to represent).Therefore, 1,206/ [1+1,206 (0.05)2], which is equal
to 1206/4.015300.This method was utilized to calculate the published tables for sample sizes determination. At 95%
confidence level and P = 0.5 were assumed and the following formula was used. This formula is supposed to be easy
method to determine sample size for social science research since 1% - 10%level of precision errors are acceptable
for the study purpose (Kothari, 2004 & 2007).
So as to have fair and unbiased knowledge of the existing situation, randomly selected respondents from
each branch was encompassed in the sample. In this regard, the findings of researchers support that taking (10 - 30
%)from the population is the optimum sample (Levy & Lemeshow, 2013). Again, to ensure equal inclusion of the
employee respondents in the sample, respondents were taken from the branches proportionate to their size using the
formula proposed by Cochran (1963): Where, n = Sample size of population, NC = Total population i.e. number of
employees and N = Total population.
Table 3.1: Sample Frame of the Study
No. Names of Clusters No. of branches No. of employees from No. of sample Remark
within the which sample was selected respondent
cluster employees
M F T M F T
1 Cluster A 9 119 38 157 30 9 39 157/1206*300
2 Cluster B 8 94 34 128 23 9 32 128/1206*300
3 Cluster C 8 92 26 118 23 6 29 118/1206*300
4 Cluster D 9 92 27 119 23 7 30 119/1206*300
5 Cluster E 9 85 24 109 21 6 27 109/1206*300
6 Cluster F 8 88 33 121 22 8 30 121/1206*300
7 Cluster G 9 85 28 113 21 7 28 113/1206*300
8 Cluster H 8 82 32 114 20 8 28 114/1206*300
9 Cluster I 7 83 28 111 21 7 28 111/1206*300
10 Cluster J 7 84 32 116 21 8 29 116/1206*300
T 82 904 302 1206 225 75 300 N / [1 + N (e) 2]
Source: Compiled from Cooperative Bank of Oromia, 2021

Tools of data collection were questionnaire, interview, focus group discussion, observation and document analysis.
However, of all tools, questionnaire was the major data collection instrument in this study.The questionnaires for the
collection of the data were developed based upon the type of the models utilized. For instance, a five point Likert
scale questionnaire ranging from strongly disagree, disagree, neither agree nor disagree, agree and strongly agree for
close ended questionnaires (Likert, 1932;Chyung et al., 2017). Consequently, “1 to 5” values were given to the
scales, where “1” represented Strongly Disagree, “2” Disagree, “3” Neutral, “4” Agree and “5” stood for Strongly
Agree respectively. The questionnaire was self-distributed at the Head Office and the selected branches.List of
names of employees were provided to the researcher by Human Resource Manager at the head office. Lottery
method was applied in selecting respondents.The questionnaire had, as its cover page, an explanation about the
research to clarify points on the objective, the topic under study, and future use of the study.
The validity and reliability of the instrument was tested. Ultimately, after assuring the collection of the necessary
data, further process was launched, including data processing and analysis. Following the completion of data
collection, data processing was conducted through filtering inaccuracy, inconsistency; incompleteness and
illegibility of the raw data to make analysis very easy. To solve such problems, editing, coding, data entry and
consistency checking were done. Responses to the questions in the questionnaires were checked for errors and then
codes were given. This involved allocation of numbers to the responses given by respondents for the ease of data
entry and analysis (Williams, 2003; Utts & Heckard, 2021). The data were entered into the computer using the
Statistical Package for Social Sciences (SPSS) software Version 21.0. To facilitate further computations, particularly

that of the total average, Microsoft Excel was used. Once the process of data entry accomplished, cleaning of the
data was started. Data cleaning and editing focused on checking whether the assigned value for each case was
legitimate, on the logical consistency and structure of cases.
To perform required calculations, therefore, questionnaires were coded and entered into the Statistical Package for
Social Sciences and Microsoft Excel. The Statistical Package for Social Sciences can accept data from more or less
any form of file and apply them to produce organized reports, charts and conspiracies of spreading and tendencies,
descriptive statistics, and inferential statistical analyses.
The data collected from questionnaire were analyzed through quantitative descriptive statistical tools such as
percentages and frequencies, mean and standard deviations using SPSS version 21.0 computer software. Inferential
statistics were also used to experiment the hypotheses. Since the data set are not normally distributed, chi-square test
of association and ordinal logistic regression model were employed among the inferential statistics. To check for the
muli-collinearity problems and to further test the strength of association between the variables,Spearman’s Rank
Correlation Coefficient was also used.
Then, by the help of Neglerke Coefficient of determination, the goodness of the model was tested. This is in line
with the suggestions of Hussain & Nassir (2015). While qualitative data obtained through interviews and focus
group discussions were analyzed qualitatively in sentence form. Finally, the results were discussed and interpreted to
draw important implications, conclusions, and recommendations.
IV. RESULTS AND DISCUSSIONS

In this chapter, discussions and analysis of the feed-backs gathered from the respondents through
questionnaire, interview and focus group discussion are presented. Data analysis for this study was done in two
steps; the preliminary analysis and the main analysis. The preliminary analysis involved mainly descriptive statistics
summarizing the demographic characteristics of the respondents. By the aid of descriptive statistics, either the
characteristics of a sample or the rapport among the variables in a sample are described. Descriptive statistics is the
discipline that describes the notable features of the data collected quantitatively (Cramer & Howitt, 2004; Kemp et
al., 2018).In this study, this analysis was used to ascertain frequencies and percentages, mean and standard deviation
of responses that were given to all the questions in the questionnaire and was also used to express the features of
selected variables.
The other analysis tool was inferential statistics. This tool moves further than the description of explicit
observations to create conclusions about the entire population from which the sample observations were drawn. It is
used to test hypotheses statistically, in doing so; it has shown the relationship between variables.
For that matter, Statistical Package for Social Sciences (SPSS) software program version 21.0 was used for
processing the data and ultimately computing both descriptive and inferential statistics (Utts & Heckard, 2021).
Besides, Microsoft Excel was used to compute the output obtained from SPSS. In this section, the principal
outcomes of the research were analyzed and discussed corresponding to the literature reviews, stated specific
objectives and logically constructed questioners that have led to draw conclusions and recommendations.
Moreover, the results are presented in tables, pi-chart, histogram and graph to show frequency counts, percentage
and statistical descriptive such as mean and standard deviation for each question and followed by narrative
explanations by supporting the explanation with the interview results as well as the necessary secondary data
triangulation has been made to support the explanation.
4.1. Validity and Reliability Test Result
4.1.1 Validity Test Result
Validity explains how well the collected data covers the actual area of investigation (Ghauri & Gronhaug, 2005;
Petter et al., 2007). Furthermore, Field (2005); Oluwatayo (2012); Engellant et al (2016) briefed that validity
basically measures what is intended to be measured. In this dissertation, of the types of validity an emphasis is given
to the content and constructs validity.
Content validity is defined as the degree to which items in an instrument reflect the content universe to which the
instrument will be generalized (Straub et al., 2004; Oluwatayo, 2012). It involves evaluation of a new survey
instrument in order to ensure that it includes all the items that are essential and eliminates undesirable items to a
particular construct domain (Boudreau et al., 2001; Choudrie & Dwivedi, 2005; Engellant et al., 2016). The second
validity test was checking for the construct validity, which needs to do with the convergent and discriminant validity
by the help of SPSS. Construct validity refers to how well you translated or transformed a concept, idea, or behavior
that is a construct into a functioning and operating reality, the operationalization. Construct validity has two
components: convergent and discriminant validity (Drost, 2011; Taherdoost, 2016; Clark & Watson,
2019).Convergent validity, a parameter often used in sociology, psychology, and other behavioral sciences, refers to
the degree to which two measures of constructs that theoretically should be linked are in fact linked (Strauss &
Smith, 2009; Edwards, 2011; Taherdoost, 2016). In brief, Convergent validity tests that constructs that are expected
to be associated are, in fact, associated.With the determination of corroborating the construct validity (discriminant
and convergent validity), a factor analysis can be conducted employing principal component analysis (PCA) with
varimax rotation method (Koh & Nam, 2005; Wee & Quazi, 2005).
In connection to exploitation of this method, there are assumptions that have to be fulfilled such as (Pohar et al,
2004; Park et al., 2010; Alfayad & Arif, 2017) as per the view of some scholars. These are: Composite reliability
(CR) is better if it is more than 0.7, Average Variance Extracted (AVE) to be equal or more than 0.5 and also
Composite reliability (CR) has to be above Average Variance Extracted (AVE). In addition, correlation between the
items that is the Average factor loading should be greater than 7 (Cassidy et al., 2005; Veale, 2014; Samuels, 2017.
Discriminant validity test that should have no relationship do, in fact, no have any relationship (No correlation
between two constructs), Average Variance Extracted (AVE) should be more than Square of correlation (Farrell,
2010; Zaiţ & Bertea, 2011; Henseler et al., 2015;Franke & Sarstedt, 2019).
Hence, to check construct validity (convergent and discriminant validity) of the data collection tool used in this
dissertation, initial assessment of convergent and discriminate validity was conducted using factor analysis
(Varimax rotation). Based on this, as it is demonstrated in Table 4.1, CR > 0.7, AVE > 0.5 and CR > AVE, which is
in line with the assumptions.
Table 4.1: Factor Analysis Results (Varimax Rotation)
Variables Average Loading Factors Average Variance Extracted
Recruitment & Selection 0.984592 0.758655
Performance of CBO 0.793682 0.511644
Source: Researcher’s Computation from SPSS, 2021
The computed result indicated that all the items within the same constructs have almost high correlation with each
other while they have different correlation with constructs from different items. In general, the average of each
variable regarding the average variance extracted was 0.98 (recruitment and selection), and 0.86 (demographic
variables). AVE of performance of Cooperative Bank of Oromia was 0.79.
According to Crawford & Henry (2003); Zaiţ & Bertea (2011); Henseler et al (2015); Voorhees et al (2016), Franke
& Sarstedt (2019) discriminant validity has to be tested by examining the correlation coefficients of each item within
and among constructs. As to them, each item should correlate highly with its intended construct, but not with other
constructs. Acceptable discriminant validity is evidenced when the correlations within each construct exceed the
correlations with all other constructs (Carlson, 2012; Henseler et al., 2015; Voorhees et al., 2016; Ab Hamid et al.,
2017). As shown in Table 4.2, (with the bolded variable scores) this condition holds for all items, suggesting that the
scales or constructs themselves have a high degree of discriminant validity.With this regard, researchers Kasper-
Brauer & Leischnig (2016); Lacap (2019); Lacap & Tungcab (2020), Muhammadet al (2020); Ekanayake et al
(2021); Wang et al (2021) suggest that the diagonal values (square of AVE) should be greater than the vertical
values (correlational values). Ultimately, construct discriminant validity was assessed by comparing correlations
between all pairs of constructs with the square root of AVE of each construct. Correlations that are greater than the
square root of AVE are indicative of poor discriminant validity between the constructs involved (Farrell & Rudd,
2009; Ghadi et al., 2012; Henseler et al., 2015). The results (Table 4.1) indicated that the square root of AVE is
larger than the correlation between any construct pair as shown by the bolded square root of AVE scores along the
diagonal. From this, AVE > Correlation square, hence discriminant validity established. In short, based on the
results of the analyses of reflective items and related constructs, the survey items indicate satisfactory convergent
and discriminant validity. This result is consistent with the opinion of Hung et al (2010); Lou et al (2013); Palladan
et al (2016); Franke& Sarstedt (2019). Therefore, the results indicate clear convergence and divergence of the
reflective indicators along construct lines.
Table 4.2: Discriminant Validity for Reflective Constructs
R&S PCBO
Recruitment & Selection 0.98
Performance of CBO 0.42** 0.79
Based on the result computed in Table 4.2, the diagonal values, which are the squares of AVE, were greater than the
vertical values that are the correlational values. This means, all the bolded values were greater than the other values
that are the correlational values. Thus, this signifies the existence of discriminant validity.

Moreover, the result of KMO and Bartlett's Test is computed. KMO is a test conducted to examine the strength of
the partial correlation (how the factors explain each other) between the variables. In connection to this, it shows that
the data collection tool is more valid if it is more than 0.5. In this regard, the finding of Yong & Pearce (2013);
Yavuz & Dogan (2014); Ramdani (2018), KMO values closer to 1.0 are considered ideal while values less than 0.5
are unacceptable. Likewise, the finding of researchers (Hadi et al., 2016; Arifin et al., 2018; Prasetyo et al., 2019;
Vembriliya et al., 2019) indicate that if KMO and Bartlett's Test is above 0.5, it is good indicator of the rationality
of the instrument. So, the result of Kaiser-Meyer-Olkin Measure of Sampling Adequacy 0.854 is > 0.5 and also sig.
is 0.000 < alpha (0.05) is better reflector of validity.
Table 4.3: KMO and Bartlett's Test
Kaiser-Meyer-Olkin Measure of Sampling Adequacy .854
Bartlett's Test of Sphericity Approx. Chi-Square 10449.861
df 406
Sig. .000
Moreover, validity can be tested by other method other than by the stated methods above. This method in composes
having the total data of respondents, browsing to Correlate under Analyze then opening Bivariate, then selecting
Pearson Product Momentum, we can compute the result of correlation coefficient against the critical (table) value.
By this premise, if the calculated value is greater than the Table value, it is significant evidence that the questions
are valid (Fisman & Svensson, 2007; Harrison & Rutström, 2008; Taber, 2018). Based on this, the calculated values
of all the variables were above the critical value at level of significance 0.05, df at 2 and, two tailed (N-2). Since the
sample size is 300, it should be checked at 300-2 = 298. They all were significant (sig < alpha = 0.05). So, since the
calculated values are more than the Table value, it is significant proving valid questions.
Table 4.4: Validity check by Correlation
V ARI AB LE S CO R R E LAT IO N T OT AL
Recruitment & Selection Pearson Correlation 0.643**
Performance of CBO 0. 539**
Source: Researcher’s Computation by SPSS, 2021
4.1.2. Reliability Test Results
Reliability concerns the extent to which a measurement of a phenomenon provides stable and consist result (Ghauri
& Grönhaug, 2005; Ghsuri et al., 2020). Reliability is also concerned with repeatability. For example, a scale or test
is said to be reliable if repeat measurement made by it under constant conditions will give the same result (Mahon et
al., 2005). So, testing for reliability is important as it refers to the consistency across the parts of a measuring
instrument (Huck, 2007). A scale is said to have high internal consistency reliability if the items of a scale “hang
together” and measure the same construct (Huck, 2007; Robinson, 2009). The most commonly used internal
consistency measure is the Cronbach Alpha coefficient (Ali & Raza, 2017). It is viewed as the most appropriate
measure of reliability when making use of Likert scales (Whitley, 2002; Robinsson, 2009).
The usual array of Cronbach’s coefficient alpha value ranges between 0-1 where the larger value reveals the larger
degree of internal consistency. Different writers reached on consensus in taking different values of this test so as to
attain internal reliability. In fact, no absolute rules exist for internal consistencies, however most agree on a
minimum internal consistency coefficient of 0.70 (Whitley, 2002; Robinson, 2009). Thus, the most largely
recognized value to be considered sufficient for exploratory studies is equal to or upper than 0.70 as it would be to
reach internal reliability (Cavana et al., 2001; Hair et al., 2003; Nunnally, 2010). On this regard, the detail
explanation provided by George & Mallery (2003) shown alpha value (> 0.9 = Excellent, > 0.8 = Good, > 0.7 =
Acceptable, > 0.6 = Questionable, > 0 .5 = Poor, and < 0.5 = Unacceptable). Furthermore, Hinton et al (2004) have
suggested four cut-off points for reliability, which includes excellent reliability (0.90 and above), high reliability
(0.70-0.90), moderate reliability (0.50-0.70) and low reliability (0.50 and below).
To verify the reliability of the data collection instrument is in line, data on the various multi-items constructs
representing the different components of dependent and that of independent variable were first tested for reliability
by computing Cronbach’s alpha values. Therefore, in this research, the inter-item consistency measure of
Cronbach’s alpha value was used to measure the reliability of the instruments.

Table 4.5: Cronbach's Alpha Result of Dependent & Independent Variables

Variables Number of Items Cronbach'sAlpha Test
Recruitment & selection 9 0.74

Performance of Coop Bank of Oromia 11 0.79
Total 20 0.77
The Cronbach’s alpha was computed for each field of the questionnaire. Table 4.5 depicted that the values of
Cronbach’s Alpha for each field of the questionnaire and the entire questionnaires. As it can be seen from the table
above, for each field value of Cronbach's Alpha was in the range of 0.74 to 0.79. This range is considered as high;
the result ensures the reliability of each field of the questionnaire. The computed Cronbach's Alpha result was 0.77.
This result indicated very good reliability.
4.2. Demographic Characteristics of the Respondents
This part commences with the analysis of the demographic data gathered from the respondents using
frequencies and percentages. The reason behind considering these variables is that the performance of organization,
which is the result of employees’ performance, is affected by demographic characteristics (Adio & Popoola, 2010;
Wakene et al., 2020). Accordingly, of the various demographic characteristics, gender, age, marital status,
educational level, field of study, work experience, salary level and position of respondents are presented below
figures.
Based on the exploration of the data, 75 (25%) and 225 (75%) of the employee respondents were females and males
respectively indicating the higher opportunity the males have got to be employed by the bank. Regarding gender,
researchers revealed that males are preferred even if they are in the minority or majority, whereas females are
impoverished, particularly, while they are in the majority, and rarely even when they are in the minority (Ebenuwa-
Okoh, 2010; Robb, 2012). In this study, the prevalence of males’ dominance is reflected in the stated bank. Added to
this, the interview conducted with the Human Resource Manager indicated the same result. Consequently, of the
total number of employees of the bank, males constitute (3,601 = 70%) and females are (1,573 = 30%). This again
reveals males’ higher access to employment opportunity, which in turn reflects the higher opportunity the males got
in income generation in the bank. The result is consistent with the finding of the same researcher (Desalegn, 2021).
As to his finding, females constitute about 50 percent of the world population and contribute about 75 percent to
subsistence production. On the other hand, they earn one tenth of the world’s income and they own one hundredth of
the world’s property including land. Therefore, the researcher has come up with the premises that females’
participation in economy is very low, particularly in cooperatives.
With reference to age classification, the researcher categorized the employees into three (Young, Middle and Old
age). With this regard, people from the age of 15 to 44 are classified as young and those in the range of 45 to 59 are
middle-aged. While with age equal to and above 60 are grouped as old age (Laditka et al., 2004; Ebner et al., 2010).
Accordingly, in this research 206 (68.7%) were in the category below 45 years that are categorized as young aged
employees, while 76 (25.3%) of them are in the age group from 45-59 ages that are termed as middle aged in this
research. The rest 18 (6%) are categorized as old aged, that are in the age group above 59 years. Of the stated 206
respondents who were grouped as young age, 154 of them were males while 52 respondents were females. In the
category of middle-aged groups, 56 respondents were males while 20 respondents were females. Likewise, of those
classified as old-aged, 15 and 3 respondents were males and females respectively.
Moreover, the age category and the tenure of employees indicated that the bulk of the staffs were in the generation
of young age which revealed an energetic workforce of the industry although there is controversy among
researchers. Generally, great numbers of them were young aged by taking ILO report (2013) in Kapsos &
Bourmpoula (2013) as a reference implying that they could have good productive prospects. As per Bedard &
Dhuey (2006); Pellizzari & Billari (2012), age category has significant effect on the performance of individuals.
Related to this, more than 85% of the respondents’ term in office did not exceed 15 years. This seems to be due to
the horizontal and vertical expansion of the banking business after the liberalization of the economy and its effect on
creating employment opportunity. In short, the cumulative effect shows employees of the bank are dominated by
young age.
As far as the matrimonial prestige of respondents is concerned, 81 (27%) of them are single and only 6% and 0.3%
of them are divorced and widowed respectively. While majority 200 (66.7%) of them are in the family engagement.
In the investigation of Saif et al (2012), it is suggested that individual workers who are married are much devoted to
their existing organization, they are employed in than their unmarried counterparts.

Educational background of the employees is an imperative element in undertaking their respective responsibilities
and to make critical decisions in their activities. In this area, existing literature confirms that the qualification is
identified to be positively linked with performance demonstrating that the better educated workers are, the better
likely to execute well (Borman & Mehay, 2009).The educational level of employee respondents indicated that 214
(71.3%) were identified to be first degree holders. In the contrary, 34 (11.3%) were master’s holders. Whilst the rest
employees, 52 (17.3%) were at the level of diploma and below. The less amount of workforces with advanced
educational level (MA and above) is to reduce cost and as a consequence the bank is focused on employment of first
degree holders who have the minimum know how in the provision of its service operations.
Concerning the respondent’s field of study, majority of them (98%) studied business related fields of study. Of
these, Accounting constitutes 32%, Economics 26.3% and followed by Management, which its total covers 41.8%
(Business Management - 22.7%, Business Administration = 7%, Marketing = 6.3%, Public Management = 3.8%).
Contrariwise, other fields of study as mentioned, General Management and Agricultural Economics constitute 1%
each. According to Griffin et al (2011), appropriate specialization areas contribute highly to the performance of
organization than the counterparts.Thus,this is reliable with the prevailing literature, which signified the field of
study of employees is highly relevant to the type of works expected of them to perform.
Meanwhile, pertaining to the service period of the employees, the result shown that large number of respondents 196
(65.3) of them were within a period of 1-5 service years in the banking industry. Likewise, about one-third of the
respondents, 101(33.7%) had relatively longer service experience, which is in the period that ranges from 5-10
years. The rest, only 3 (1%) of respondents had service years above 10 years. From this, it can be deduced that the
respondents’ large number were new entrants with an experience below 5 years.
Long service year/experience in a particular sector can lead to high level of commitment (Affum et al., 2015). This
infers that the banking operation has fallen on the shoulders of the new entrants because of either high turnover or
low interest of the bank to invite experienced personnel. In other words, this denotes that the sector is not in the right
track in capturing and developing well experienced staffs.
Finally, the salary level of respondents indicated that 53.3% of them earn between 10,001 to 15,000 ETB whereas
7.7% and 1.7% of them earn more than 20,000 and less than 5,000 ETB respectively. This indicates the bank’s
employees are getting better salary when compared to other organizations’ workers in Ethiopia. Existing literature
addresses that the more salary employees get, the more achievement they attain (Ertanto & Suharnomo, 2011). So,
the salary level of the employees is in a good condition. Anyways, it cannot be generalized that the total
performance of organization is not merely the share of those who get high salary; rather it is the cumulative effect of
all employees working at different positions.
4.3. Descriptive Statistics of Respondents
In this part, to illustrate the feedback of the respondents, descriptive statistics in the mode of frequency and percent,
mean and standard deviation (SD) have been presented.Mean is the average of a group of numbers whereas,
standard deviation (SD) is the most useful measure of variability. It tells us what is happening between the minimum
and maximum scores. In other language, it states how much the scores in the data set vary around the mean. The
more spread apart the data is, the higher the deviation. Moreover, standard deviation was used to indicate variation
from the mean. A small standard deviation show that points are nearby the mean, while large standard deviation
directs that the data is spread in excess of a big variety of values. In other words, extreme values occur more
frequently.
Meanwhile, some indicators of the performance of the opinion of the respondents for the variables indicated below
were measured on five point Likert scale with measurement value 1= Strongly disagree (SD); i.e. very much
dissatisfied with the case described; 2 = Disagree (D), i.e. unhappy with the case described; 3 = Neutral, i.e.,
uncertain with the case described; 4 = Agree (A), i.e., feeling all right with the case described and considered as
satisfy; and 5 = Strongly Agree (SA), i.e. very much supporting the case described and considered as highly satisfied
with regard to the variables.
Cooperative Bank of Oromia in terms of return on asset, return on deposits, return of equity, in terms of maximizing
dividend of shareholders, branch expansion, amount of dividend shared to the shareholders, bank’s credit
performance in lending money to the Coop societies, the status of the return of the money lent to the customers, the
status of number of customers, the status of saving deposit accounts of the customers, and bank’s performance in
competitive market) were also measured on five point Likert scale with measurement value 1= Very Low (VL) with
the case described; 2 = Low (L), 3 = Medium (M), 4 = High (H), and 5 = Very High (VH). To make easy
interpretation, the following ranges of values were reassigned to each scale: equal to 1= strongly disagree or very
low; 1.01 - 2 = Disagree or Low; 2.01 - 3 = Neutral or Medium; 3.01 - 4 = Agree or High; and 4.01 - 5 = Strongly

Agree or Very High, which was later re-categorized into 3. Consequently, strongly disagree & disagree = Disagree =
not satisfied and agree & strongly agree = agree and Neutral/ medium so as to compute X 2.
In short, a mean of equal/above 3 is regarded to measure satisfaction on the test variables. To analyze the collected
data consistent with the overall objective of the research undertaking, statistical procedures were carried out using
SPSS version 21.0 software. While the outcomes of the interview questions and focus group discussions were
integrated with the responses obtained through questionnaire.
4.3.1. Employees’ Perception Regarding Recruitment and Selection Process
Human Resource, especially the skilled one is an important asset for the development of any organization. The
political, economic and social-cultural development is depending on the quantity of qualified and skilled human
resource beyond all. In this regard, Foot & Hook (2008) stated that more adequate supply of physical and financial
resources does not alone ensure the economic advancement of any country; unless efficient and effective human
power is available in the country. In connection to this, a good human resource management is typified by its open
and transparent recruitment and selection system.
Regarding the practices of recruitment and selection in Cooperative Bank of Oromia, the survey study on operation
level employees found out that respondents came to an agreement with the mean scored value of 3.72 with (Standard
deviation (SD) 0.62 that the bank has well skilled and competent manpower. They also agreed that the recruitment
& selection outcome affects the performance of the bank with mean value 3.62 and SD 0.82. Their feed-back is also
alright that the bank has competitive selection system that attracts competent people with mean 3.42 and SD 0.82.
With regard to whether the bank has employee recruitment policy linked with its business strategy and the bank
disseminates information about recruitment & selection processes with mean 3.38 (SD 0.82) and 3.29 (SD 0.91)
respectively.
Likewise, respondents agreed that the processes of recruitment and selection are conducted by trained people of the
bank with the mean scored value of 3.24 (SD 0.87). Even if respondents agreed that follows up on the adaptation of
employees to their functions is there, the mean score is low as compared with others’ mean values and high standard
deviations, which is an indication for the requirement of strong follow up.
The overall respondents’ perception regarding the processes of recruitment and selection were rated with a mean
scored value of 3.06 = 61.2% = 184 respondents and SD 0.84 level of agreement that the practice was aligned with
corporate/ business strategies and have been carried out in accordance with the organizational policies even if some
discrepancies have been identified regarding line managers’ participation in selection process. In general, perception
of majority of respondents indicated that the process of recruitment and selection in the bank is somewhat good;
however, the respondents were not satisfied on items such as line managers’ participation in the selection process
enabled banks to obtain the required quality of staff with mean 2.87 (SD 1.06). In addition, whether the recruitment
and selection process of the bank is free of bias or sabotage, respondents’ feedback is low (mean = 2.65). The
descriptive result of the item of the process, which is announcing the result of the selection process on time to the
candidate was very low of all with (mean = 1.36= 27.2% =82 and SD = 0.68). This means, 72.8% = 218 respondents
indicated that the bank is not in a position to apply the stated issue on time.
Table 4.6: Respondents’ Perception on Recruitment and Selection
Recruitment and Selection N Mean SD
The bank has well skilled and competent manpower 300 3.72 0.68
The recruitment & selection outcome affects the performance 300 3.62 0.84
The bank has competitive selection system that attract competent people 300 3.42 0.87
The bank has employee recruitment policy linked with its business strategy 300 3.38 0.82
The bank disseminates information about recruitment & selection processes 300 3.29 0.91
Recruitment & selection processes are conducted by trained personnel 300 3.24 0.87
In the bank, line managers participate in selection process 300 2.87 1.06
Recruitment and selection process of the bank is free of bias 300 2.67 0.83
The bank notifies the result of the process within short period of time 300 1.36 0.68
Total 300 3.06 0.84
Source: Survey, 2021
4.3.7. Performance of Cooperative Bank of Oromia
Organizations today are continuously facing external and internal forces that drive them to change due to the world
are more competitive in times. Performance in banking can be mainly of financial. Commercial bank’s financial
performance can be measured by profitability; however there are other indicators such as non-financial measures. In
this study, performance of Cooperative Bank of Oromia was assessed by both financial and non-financial indicators,

which include: in terms of return on asset (ROA), return on deposits (ROD), return of equity (ROE), maximizing
dividend (wealth) of shareholders, the bank’s branch expansion, and amount of dividend shared to the shareholders
were analyzed by the use of the perception of the respondents as a measure for analyzing the performance of the
bank.
In addition, perception of the employees towards the bank’s credit performance in lending money to the Coop
societies, the status of the return of the money lent to the customers, the status of number of customers, the status of
saving deposit accounts of the customers and the overall bank’s performance in competitive market were the other
attention areas investigated in Table 4.7.
Table 4.7: Descriptive Statistics of Employees’ Perception towards PCBO

Descriptive Statistics on Performance of the Bank N Mean SD
Performance of the bank in competitive market 300 3.94 .63
Market performance of the bank in terms of branch expansion 300 3.61 .84
The status of saving deposit accounts of the customers 300 3.60 .87
The status of number of customers 300 3.58 .86
Performance of the bank in terms of maximizing dividend of shareholders 300 3.51 .79
The status of the return of the money lent to the customers 300 3.50 .88
Credit Performance of the bank in lending money to the Coop societies 300 3.47 .83
Amount of dividend shared to the shareholders 300 3.47 .85
Performance of the bank in terms of Return on Deposits 300 3.43 .87
Performance of CBO in terms of Return on Asset 300 3.43 .89
Performance of the bank in terms of Return of Equity 300 3.33 .76
Total 300 3.53 .82
Source: Field Survey, 2021
As depicted in Table 4.7 above, the bank is competitive in the market that was confirmed by the computed mean
result equal to 3.94 and SD 0.63. Majority of respondents also agreed by the mean value 3.61 that market
performance of the bank in terms of branch expansion is better. Next, they again admitted that the status of saving
deposit accounts of the customers with the mean value 3.6. The status of number of customers and performance of
the bank in terms of maximizing dividend of shareholders were also alright (mean = 3.58 & 3.51).
In the same manner, based on the perception of the respondents, the bank is at better condition on the following
performance indicators. The status of the return of the money lent to the customers, credit performance of the bank
in lending money to the Coop societies and the amount of dividend shared to the shareholders with mean values 3.5,
3.47 and 3.47 respectively. Performance of the bank in terms of Return on Deposits and its performance with regard
to Return on Asset were rated as 3.43 mean values. Likely, performance of the bank in terms of Return of Equity is
also alright (mean = 3.33). In general, the mean of all the statements demonstrating the performance of Cooperative
Bank of Oromia are rated above 3, which indicated respondents are satisfied with the statements. The SD of all the
statements are almost near to each other that can be termed as high values showing difference perception of
respondents. Depending on these values, it can be judged that the bank’s performance is relatively better.
Furthermore, the above employees’ feedbacks were ratified by the secondary data that was taken from the annual
reports of the bank. In this regard, the annual reports of the bank for the five consecutive periods from 2016/2017 to
2020/2021 were computed. Accordingly, indicators of both the financial and non-financial performance of the bank
during the stated periods for the items identified is presented. Some of the performance indicators which were taken
into consideration were: total deposits, revenues/earnings, profit after tax, assets and capital. In addition, number of
customers, dividend shared, loans, balance sheet, customers’ deposits and status of the bank’s branches expansion
were incorporated in the analysis.
i. Deposits of CBO from 2016 to 2020
Figure 4.1: Total Deposit status of CBO, 2021
AJMRR Journal P a g e | 10
Source: Computed from CBO Annual Reports, 2021

The above graph (Fig. 4.1) indicated that the bank’s total deposit has grown from 14.28 billion in the fiscal year
2016 to 71.12 billion ETB as of June 2021 showing 398 growth rates.
ii. Revenues of CBO from 2016 to 2020
Figure 4.2: Revenues/ Earnings of the Bank, 2021

Source: Computed from CBO Annual Report, 2021
In connection to the total revenues generated from 2016/2017 to 2020/2021, it was increased from 1.53 billion to
8.03 billion ETB signifying 424% rate of growth in the periods.
iii. Assets of CBO from 2016 to 2020
Figure 4.3: Total Assets of CBO, 2021

The above figure verified that the bank’s assets was reached 8.32 billion in 2020/2021 from the base year
2016/2017, which was 17.72 billion. The growth rate of the period was 359%.
iv. Capital of CBO from 2016 to 2020
Figure 4.4: Capital of CBO, 2021

As it is displayed in Figure 4.4, the capital of the bank has grown from 1.52 billion to 8 billion ETB showing 426%
rate of growth from 2016/2017 to 2020/2021.
v. Loans of CBO from 2016 to 2020
Figure 4.5: Total Loans of the Bank, 2021

The above pie chart indicated that the loans of the CBO increased from 9.92 billion to 54.5 billion ETB from the
year 2016/2017 to 2020/2021. Based on this figure, there is enough evidence that the bank’s capital was increasing
from one period to another.
vi. Balance Sheet and Customer’s Deposit of CBO form 2016 to 2020
Figure 4.6: The Bank’s Balance Sheet and Customers’ Deposit, 2021
As far as the bank’s balance sheet was concerned, it has increased from 23 billion in 2016 to 81 billion ETB in 2020.
Whereas, deposit of the customers of the bank grown from 14 billion in 2016 to 43 billion ETB in 2020. Thus, the
balance sheet and customer’s deposit of the bank were grown by the rate of 252% and 207% respectively from
2016/2017 to 2020/2021.
vii. Net Profit of CBO and Dividend Shared to Shareholders
Figure 4.7: Profit of the Bank after Tax and Dividend Shared, 2021
As shown in Figure 4.7, the net profit and amounts of dividend shared to the shareholders were increasing overtime.
The net profit of the bank was 338.2 million in 2016, which was increased to 1.327 billion ETB in 2020. Likewise,
the amount of dividend shared to shareholders has grown from 700,000 to almost 1 billion ETB in the year 2020.
This has shown an increase of rate of 292% and 500% growth respectively.
Last, but not least performance indicator of the bank was to increase its market share. In connection to this, the
CBO’s numbers of branches were 256 during 2016/2017 and reached 469 in the fiscal year 2020/2021 showing an
aggregate growth rate of 83%. This confirms the bank is increasing its share in the market by expanding its branch
to access its clients so as to make expansive its market base.
In short, based on the reports of the bank, the performance of the bank from 2019/2020 to 2020/2021 was very high
as compared with other years. In all the cases, the computed results ratified that the bank is performing well even if
there are issues to be considered to further assure its sustainability in the tougher market competition.
4.4. Inferential Statistics of Respondents
As it is stated in the model specification, inferential statistical tools were used. Inferential statistics infer about the
population depending on the sample data, based on estimation and test of hypothesis. Inferential method falls into
two major categories: parametric and non-parametric approaches. Though parametric tests are highly influential than
non-parametric statistics, in the inquiry of this data, the second method was used. In case where the hypotheses of
normalcy and constant variance of residuals are not satisfied, Onyango & Odebero (2009), proposed the use of non-
parametric methods.
Non-parametric is not dependent on the shape of the dispersal of the population and hence are known as distribution
free tests.In many ways, the non-parametric tests are relatively advantageous to parametric tests. First, they could be
used in circumstances where strict assumptions are requested by parametric tests do not have to be happened.
Second, they are not more difficult to use as much as conceptual methods are concerned. Third, they are best
suitable when dealing with qualitative values with data that could be ordered in ranks. Finally, if the size of sample
is small, the parametric tests are not appropriate except the distribution of population is accurately known which is
usually not likely.
Therefore, prior to decide using this model, assumptions of other continuous models were checkedin condition
where the statistical significance was set at α = 0.05. In SPSS, p-values are labeled as “sig”. In line to this, the nature
of the data was not fulfilling the presume of continuous data. Hence, in this study, non-parametric model was used
where chi-square (Χ2) test of association and ordinal logistic regression were applied that fulfill the assumption of
the likert scale data under consideration. In addition, Spearman’s Rank Correlation Coefficient was applied to find
the strength of association between two variables and also to check for the multicollinearity problem. Before
computing the result by any type of inferential statistical model, data testing was the first determinant process.
Accordingly, in the following sections this issue is going to be presented.
4.4.1. Data Testing
The most critical assumptions related to chi-square test of association and ordinal logistic regression are tested. In
this regard, normality, multicollinearity, and model specification tests have been made to make the data available to
give reliable result and make the model fit the data. These assumptions were tested so as to reach at a conclusion
about the population. When the assumptions are met, the likelihood that the model obtained from a sample of being
the same as the population of concern (the coefficients and parameters of the regression equation are thought to be
unbiased) is increased.
Consequently, all the assumptions for using the models specified were critically checked. In this regard, the data
nature of set was not normally distributed, which recommended to use non-parametric method. Of the non-
parametric methods, chi-square test of association, ordinal logit regression and Spearman’s Rank Correlation were
specified to be applicable. For instance, the data nature used in this research fulfilled the criteria to use chi-square
test of association since the two variables are measured at an ordinal level. Added to this, the major assumptions for
applying ordinal logit regression model too were thoroughly tested.
Testing normality of data can be done either through graphs (Q-Q plot and frequency distribution/ normality tests),
D’Agostino-Pearson omnibus, Shapiro-Wilk or Kolmogorov-Smirnov (Alfayad, 2017). In general, this was done by
computing the statistic and standard error of skewness and kurtosis result, Kolmogorov-Smirnov, histogram, Q-Q
plot with tests and test of assumption of proportional odds.
According to Ikhlas et al (2021); Obumneke Ezie (2021); Onyeiwu et al (2021), normality distribution of data by the
help of the results of skewness and kurtosis, Kolmogorov-Smirnov and other criteria have to be thoroughly tested. In
this context, the results of all the issues have shown that the data is not normally distributed. For instance, the result
derived from the descriptive statistics where much emphasis lays on skewness and kurtosis. Therefore, to determine
the nature of the data by this normality test measure requires calculating the value of each. Consequently, Skewness
or Kurtosis z-value = Skewness or Kurtosis measure ∕ standard error of Skewness or Kurtosis. In connection to this,
by the help of this formula, the results of each Skewness and Kurtosis z- value is computed and displayed in Table
4.17 as the following:
Table 4.8. Check of Normality Distribution of Variables by Skewness & Kurtosis
Variable Statistic SE Stat. of ske or kur/ SE
Recruitment & selection Skewness -1.096 0.141 -7.80
Kurtosis -1.428 0.281 -5.10
Performance of CBO Skewness 0.339 0.141 2.40
Kurtosis -0.574 0.281 -2.04
Source: Computed by Researcher, 2021
From the above table, the statistic and standard error of skewness and kurtosis identified that if the data were
normally distributed. According to the verification given byDoane & Seward (2011); Ikhlas et al (2021); Obumneke
Ezie (2021); Onyeiwu et al (2021), when the statistic of kurtosis z-value of each variable is divided to its standard
error and its result falls somewhere in the span of -1.96 and +1.96 which is the z-value, then we can conclude that
the data is normally distributed which works for skewness too. Otherwise if the output is somewhere outside of the
noted span, it is not normally distributed which encourages the use of non-parametric statistics.
Accordingly, the calculated results of skewness z-value when divided by the skewness measure to its standard error
for each variable (recruitment & selection, training & development, employees’ satisfaction, employees’ relation,
performance appraisal, demographic variables and performance of CBO) was: -7.8, -2 and 2.4 respectively that are
outside of the z-value i.e., -1.96 and +1.96 (Table 4.8).
In the same manner, the calculated results of kurtosis z-value when divided by the kurtosis measure to its standard
error for each variable (recruitment and selection and performance of CBO) was -5.1 and -2.3 respectively which
again signifying the results falling outside of the stated z-values. In short, the computed values of both the Skewness
and Kurtosis for all the variables (all the seven z-values were >/< +/-1.96). Thus, the result obtained in the above
Table (4.8) confirmed the use of non-parametric statistics. This can further be tested from the variables’ log of their
skewness and kurtosis.
Table 4.9. Check of Normality Distribution of Log of Variables by Skewness & Kurtosis
Variable Statistic SE Stat. of ske or kur/ SE
Log of Recruitment & selection Skewness -0.464 0.141 -3.29
Kurtosis -0.130 0.281 -0.46
Log of Performance Skewness -0.254 0.141 -1.80
of CBO Kurtosis -0.639 0.281 -2.27
Based on the computed data (Table 4.9), by taking the result of the performance of Cooperative bank of Oromia as
an example; using parametric method is not advisable. The data are not normally distributed with the skewness of -
0.254 (SE = 0.141) and a kurtosis of -0.639 (SE = 0.281). The skewness and kurtosis z-value when divided by the
respective measures to their standard errors were -1.8 and -2.27 respectively. From this, no z-values were in the
range of +/- 1.96, which was also the same for the rest computed variables.In general, the computed values of the
log of all the variables their skewness and kurtosis result were not falling within the z-value. Thus, the result
obtained in the above Tables (Table 4.13 and Table 4.14) suggested the use of non-parametric statistics since their z-
values were not falling within +/-1.96.
In addition, in order to be certain the test of normality distribution of data, additional confirmation can be carried out
by the help of Kolmogorov-Smirnovand Shapiro-Wilk. According to Lofgren (2013); Joyner (2019), if the data set
is less than 100, Shapiro-Wilk is used otherwise Kolmogorov-Smirnov is applied if the data is above 100. In this
case, instead of Shapiro-Wilk, Kolmogorov-Smirnov is taken into consideration since the data is more than 100. In
Kolmogorov-Smirnov computation, if p-values are below 0.05, then the data are not normally distributed (Razali &
Wah, 2011; Hartgerink et al., 2020). To do this, it is a need to get the mean of each computed variable that has to be
done by transforming each of the variables. Performing this necessitates going to “transform” in the SPSS dialog
box and compute variable and then following all the steps. Based on the results computed in Table 4.15, the
transformed means of the p-values of all the variables were below 0.05, which recommends applying non-parametric
regression particularly order logit regression model.
Table 4.10: Test of Normality by Kolmogorov-Smirnov Statistic
Variables Kolmogorov-Smirnov
Statistic df Sig.
Performance of CBO 0.075 300 0.000
Recruitment & Selection 0.070 300 0.001
Source:Researcher’s Computation from SPSS 21.0

Furthermore, before performing analysis detail computation can be run by transforming the log of each variable.
Accordingly, the computed results of allthe p-values of the variables are by far less than 0.05. Thus, based on the
transformed log and statistic of Kolmogorov-Smirnov, p-valuesare less than 0.05. Thissuggests using order logit
regression other than any other model such as linear and multiple regressions in particular. This suggestion is in
consistent to the finding of Okeniyi & Okeniyi (2012); Ogundapo et al (2018).Null hypothesis states that P-values
are > 0.05, which means the values are sampled from a population that follows a normal distribution. Whereas: -
Alternative hypothesis states that P-values are < 0.05, which implies the values are not sampled from a population
that follows a normal distribution. Thus, depending on the p-values computed above, the researcher rejected the null
hypothesis and accepted the alternative hypothesis, since the calculated p-values for all variables were below 0.05.
Table 4.11: Test of Normality by Log of Kolmogorov-Smirnov Statistic
Variables Log of Kolmogorov-Smirnov
Statistic df Sig.
Log Performance of CBO 0.071 300 0.000
Log Recruitment & Selection 0.094 300 0.000
Source: Researcher’s Computation from SPSS 21.0

Thus, based on the results computed, since the p-values of both the Kolmogorov-Smirnov Statistic and its logs were
less than the alpha level (0.05), it is suggested to use non-parametric model; ordinal regression in particular.
As mentioned earlier, normality can also be tested by the help of either Q-Q plot, Histogram or Box plots whose
results are presented in Figure 4.8 to 4.9. The quantile-quantile plot (Q-Q plot), where the X-axis represent the
actual data while Y-axis on the other portray the predicted data. When the slot line on the diagram is equal, it
verifies normal distribution and otherwise, when it deviates from the two axes, it shows not normally distributed. As
per Joyner (2019), in case of histogram, when visually observed, if it is not long tailed to one side (i.e., right-skewed
or left-skewed) which implies normally distributed around the mean (i.e., symmetric  Mean = Median = Mode), it
can be judged that the data is normally distributed. In consequence, the results of all described that the data are not
normally distributed.
The following figures either the Q-Q plot or Histogram (frequency distribution) (recruitment and selection variable
in the following Figures) further verified the nature of the data that it is not normally distributed.
a. Histogram showing Recruitment and Selection Processes
Figure 4.8:Histogram showing the Distribution of Recruitment and Selection Data

Source: Computed from SPSS, 2021
As it is shown on the above figures, the residuals show signs of abnormal distribution. Furthermore, as it is observed
from the histograms it can be thought the residuals are not symmetrically normally distributed and the data are
somewhat skewed. Therefore, based on this evidence it can be concluded that the data is not normally distributed,
which recommended using non-parametric tools.
b. Q-Q plot showing the Distribution of Recruitment and Selection Processes
Figure 4.9: Q-Q plot showing the Distribution of Recruitment and Selection Data
Source: Computed from SPSS, 2021
Likewise, as shown from the above Q-Q plots, the distribution deviates from the normal (dispersed across
axes). Existing literature supposes that when the slot line on the diagram is equal, it verifies normal distribution and
otherwise, when it deviates from the two axes, it shows not normally distributed. So, based on the tests conducted by
skewness and kurtosis result (>/<1.96), both Kolmogorov-Smirnov and log of Kolmogorov-Smirnov normality test
(p < 0.05) and visual inspection of histograms and Q-Q plots showed that the data were not normally distributed.
Therefore depending on all these verifications, it was right to use non-parametric models.
4.4.1.1. Chi-Square Test of Association

Chi-Square (Χ2) is among the non-parametric statistical approaches that illustrates the association between the
variables. Chi-square has different curves based on the degrees of freedom. It is tilted to the right for small degrees
of freedom and becomes highly symmetric as the degrees of freedom increases. As the test statistic comprises
squaring the differences, the test statistics remain totally positive. A chi-squared test for independence is often right
tailed (Gagunashvili, 2010; Wickens, 2014). The objective here is to decide if there is a relationship among variables
or not. Therefore, by using chi-square test of association, it is easy to determine the relationships among the
variables and test hypotheses where statistical significance was set at α = 0.05. In this analysis, since the two
variables are measured at categorical (ordinal level), chi-square is applied which is in line to Kothari (2004; 2007).
To compute the data in this model, the 5-likert scale was merged into three where strongly disagree and disagree
were merged to disagree and agree and strongly agree to agree. While neutral remained the same. In doing so, value
0 was given for disagree, 1 for neutral and 2 for agree. Similar re-categorization was also done for the measure of
variables of PCBO.
Table 4.12: The Result of Test of Association by Chi-Square Test Statistics
Variables Pearson’s Chi-Square df Sig.
Recruitment and Selection 74.367 4 0.000*
*sig at 5% level of significance
Source: Researcher’s Computation, 2021
As can be observed from the Table above, it is clearly observed that recruitment and selection variableis
significantly associated with performance of Cooperative Bank of Oromia at 5% level of significance since the
values of all the predictor variables were less than the alpha level (P-values < α = 0.050). Thus, the swift of the
calculated values for recruitment and selection, training and development, employees’ satisfaction, employees’
relations and performance appraisal for chi-square = 74.367 and 9.52 at df 4, and significance level of 0.000 and
0.04 respectively were greater than the Table value which is 9.49 at α = 0.05 as shown in preceding Table 4.12.
Hence, depending on this there is adequate prove to fail to accept the null hypotheses and in favor of the alternative
hypotheses. This implies there is strong relationship among the practices of human resource management and the
Coop Bank of Oromia’s performance.
The chi-square test shows only, whether on the grounds of sample data, alternative hypothesis could be concluded or
not. In order to decide the scope to which dependent and independent variables are related to each other, ordinal
logistic regression was applied to the data.
4.7.1.2. Correlation Analysis
Correlation analysis was used to examine the extent the independent variables associates with the dependent variable
and also to check multicollinearity problem by checking the relationship between each variable. It can also be used
to test the hypotheses formulated.Unlike the Pearson’s Product Moment Correlation Coefficient (PMCC),
Spearman’s Rank Correlation Coefficient was applied since the nature of the data was relevant to it. Thus,to find the
strength of association between two variables, it is likely to look at the results of correlation. Correlation coefficients
show the trend and intensity of the relationship between variables unlike the chi-square. Bivariate Correlations were
used to know the nature, direction and significance of the bivariate relationship of the variables of this study. A
correlation matrix was constructed using the variables in the questionnaire to show the strength of relationships
among the variables considered in the questionnaire.
According to Higham (2002); Myers & Sirois (2004), correlation matrix is stated as a set of correlation coefficients
between numbers of variables.The result of the coefficient ranges from +1 to -1, where +1 indicates perfect positive
correlation, -1 indicates perfect negative correlation and 0 indicates no correlation.Positive linkages take place once
a change in one variable causes a movement in the other related variable in similar direction (i.e. an increase in one
variable causes an increase in the other and vice-versa).
A negative association takes place while changes in related variable are happening in the reverse direction (i.e. an
increase in one variable causes a decrease in the other and vice- versa). Based on the suggestion of Myers & Sirois
(2004); Pallant (2010);Farres et al (2015), the interpretation of r value indicating the intensity of the link between
two variables is as follows: r = 0.10 to 0.29 or r = -0.10 to -0.29 is Weak, r = 0.30 to 0.49 or r = -0.30 to -0.49
Medium and r = 0.50 to 1.0 or r = -0.50 to -1.0 is High.In this analysis, this is done by the help of Spearman’s rank
correlation analysis after the data have been proved to be not normally distributed as presented in Table 4.13. In this
regard, the relationship between the dependent and independent variables is identified at first step as the following:
Table 4.13: Correlation Coefficient of the Dependent and Independent Variables

V ARI AB LE S CO R R E LAT IO N P CB O
Recruitment & Selection Spearman Correlation 0.631**
Source: Researcher’s Computation by SPSS Output from Survey Data, 2021

Based on the values calculated by Spearman’s, the direction of the linkage among predictors and predicted variables
are displayed. Spearman’s rank correlation finds the correlation between the two ordinal scaled random variables.
Accordingly, the bank’s performance and demographic variables are correlated at 0.32, n = 300 and p-value = 0.000,
which was statistically significant. This signified negative and medium correlation between the two variables.
Conversely, the correlation of recruitment and selection was high/strong and positive at r = 0.63, and n = 300, alpha
< 0.05 level of significance.
Furthermore, detail investigation was carried out to look at the relationship among all the variables by the help of the
same model at the second place. In this step, multicollinearity problem could be identified. For instance, as
obviously specified above, the link between the dependent and independent variables was above 0.5, which is
considered as strong relationship.
Table 4.14: Spearman’s Correlation Coefficient between each Variable
PCBO Recruitment & Selection
PCBO 1.000 .631
Recruitment & Selection .631 1.000
Source: Computed by Researcher, 2021
Depending on the above table (Table 4.14), the correlation between the variables is denoted as strong since the
correlation coefficients are above 0.5.
4.4.1.3. Ordinal Logistic Regression Analysis
Regression helps to predict the value of a variable based on the value of another variable. It is a powerful statistical
method that allows examining the relationship between two or more variables of interest. Regression analysis is a
statistical method to deal with the formulation of mathematical model depicting relationship amongst variables
which can be used for the purpose of prediction of the values of dependent variable, given the values of the
independent variable (Kothari, 2004 & 2007). Logistic regression model can be classified as multinomial, ordinal
and binary (Meron, 2011).In this investigation ordinal logistic regression model was used. The ordinal logistic
regression procedure empowers one to select the predictive model for ordered dependent variables. It describes the
relationship an ordered response variable and a set of explanatory variables.
Since the data sets have not been normally distributed as confirmed earlier, the researcher used ordinal logistic
regression to perform the analysis instead of linear regression. Ordinal logistic regression is applied to estimate an
ordinal dependent variable provided one or extra independent variables. The procedure used concurs with the work
of Desalegn et al (2021). To analyze data by the help of this tool involves looking for: Model Fitting Information,
Goodness-of-Fit, Pseudo R-Square, Omnibus test, Parameter Estimates and Test of parallel lines as suggested by
(Adejumo & Adetunji, 2013).
In this model, five (5) kinds of link functions are there i.e., logit, probit, Complementary log-log, Negative log-log
and Cauchy (inverse Cauchy). From these, logit link function, which is a transformation of the cumulative
probabilities that allows estimation of the model, is preferred in the analysis as per Hahn & Soyer (2005). In the
analysis, logit link function is applied because it is evenly distributed categories and is reasonable choices when the
changes in the cumulative probabilities are gradual and dichotomy the response scale (Das & Rahman, 2011).
Appropriateness of ordinal logistic regression is tested by the help of checking model (Model Diagnostic) where the
results are presented in the following tables:
Table 4.15: Model Fitting Information
Model -2 Log Likelihood Chi-Square df Sig.
Intercept Only 1784.576
Final 1562.083 222.493 6 0.000

The model fitting information table expresses how well the model fits the data as per the opinion of Cai & Monroe
(2014); Onyeiwu et al (2021). As one can observe from the computation, the p-value is significantly less than 5%,
which signifies the model used fits the data very well and it is also evident that the model gives better predictions
than if we just guessed based on the marginal probabilities for the outcome categories.
Table 4.16: Goodness-of-Fit
Chi-Square df Sig.
Pearson 7490.049 7170 .858
Deviance 1562.083 7170 1.000
As clearly indicated by Person and Deviance Chi-square test results, the p values are greater than 5% level of
significance for both Pearson and Deviance Chi-square. Depending on the end result of goodness of the fit indicated
above, the results for the analysis suggested the model does fit very well (p > 0.05) (i.e. fail to castoff the null
hypothesis depending on the observed data. The result is in line to the finding of Archer & Lemeshow (2006). So, it
was promising to infer that the model fitted the data well since the p-value was statistically not significant.
Table 4.17: Results of Model Summary: Pseudo R-Square
Cox and Snell .524
Nagelkerke .641
McFadden .125
Pseudo R squares were additional measures of goodness of fit for ordinal logistic regression. They are used in
ordinal regression to estimate the variance explained by the independent variable (Adejumo & Adetunji, 2013;Smith
& McKenna, 2013). The results here were nearly comparable to R2 for linear regression. However, for ordinal
logistic regression models, it is not possible to compute the same R2 statistic as in linear regression; instead, three
approximations are computed. These are: Cox and Snell, Nagelkerke and McFadden whose values are computed as
0.524, 0.641 and 0.125 respectively for final model. The results supported the conclusion that the final model fit the
data well. In this case, more focus is given to Nagelkerke = 0.641), which verified that 64.1% changes in the
performance of Cooperative Bank of Oromia is the outcome of the stated explanatory variables. While the rest
percentages are explained by the other factors other than the five variables identified.
Moreover, in order to ensure the assumptions are not violated about the statistical test of parallel lines, the notion
that all logit surfaces are parallel must be tested. Test of parallel lines helps to determine whether it is reasonable to
assume that the results of the location parameters are constant across categories of the response as supposed by
Adejumo & Adetunji (2013). In connection to this, the test of parallelism contains: –2 Log-Likelihood for the
constrained model; the model that assumes the planes or surfaces are parallel and–2 log-likelihood for the General
model, the model that assumes planes or surfaces are separated.
One of the suppositions of ordinal regression is that regression coefficients are alike for all categories. If the
postulate of parallelism is detested, it is preferable to imagine employing multinomial regression, which estimates
distinct coefficients for every category O’Connell (2006). In case the null hypothesis is excluded at a specific level
of significance, it infers the relationship function designated is wrong for the data, which implies the associations
between the independent variables and logits are asymmetrical for all logits.
The chi-square statistic is the log-likelihood variance between the two models. If the lines or planes are parallel, the
observed significance level for the change would be large, as the common model does not improve the fit very
ample and the parallel model is sufficient. If there is indication to fail to accept null hypothesis, it is likely that the
linkage function selected is wrong or that the associations between the independent variables and logits are not equal
for all logits. Decision regarding the test of parallel lines is elaborated in successive table (Table 4.18).
Table 4.18: Test of Parallel lines
Model -2 Log Likelihood Chi-Square df Sig.
Null Hypothesis 1562.083
General .000 1562.083 6 .346
The table above (Table 4.18) shows parallel line test for final model with chi-squarevalue 1562.083 and p-value =
0.346 which is more than the 5% level of significance. In the recommendation of Adejumo &Adetunji (2013); Erkan
& Yildiz (2014), if the assumption of parallelism is rejected, it is better to consider using multinomial regression that
estimates separate estimates for each category. If the null hypothesis is rejected at specific level of significance, this
implies that it is possible that the link function selected is incorrect for the data or the relationships between the
explanatory variables and logits are not for all logits. Therefore, there is definitely not ample evidence not to accept
the null hypothesis for final model. Thus, the proportional odds notion appears to have held for final model, which
supposes to apply ordinal logistic regression since the test of proportional odd is not violated. In addition, the result
of Omnibus test, which compares the fitted model against the intercept only model, suggests the use of ordinal
regression since its p-value was < α = 0.05.
In summary, reliant on the computed results practically tested by Model Fitting Information
Goodness-of-Fit, Pseudo R-Square and Test of Parallel lines which were described in the above tables and Omnibus
test), the use of ordinal logistic regression model was found as the most applicable non-parametric method.
Moreover, the transformed log of all the variables with regard to Model Fitting Information
Goodness-of-Fit, Pseudo R-Square, Test of Parallel lines and Omnibus test (D’Agostino-Pearson omnibus)
displayed in the appendices part, indicated in line with the existing assumptions. Therefore, based on these
evidences there was an ample evidence to apply ordinal logistic regression model.
Table 4.19: Parameter Estimates of Ordinal Logistic Regression based on Route-One Result
Variables β SE Wald df Sig. 95% CI
Lower Upper
[Performance of CBO = .00] -3.014 1.379 4.780 1 .000* -5.716 -.312
Threshold
[Performance of CBO = 1.00] 2.035 1.034 3.873 1 .000* .008 4.061
Location Recruitment & Selection 2.284 .3262 43.397 1 .000* 1.509 2.788
Depending on the values in Table 4.19 above, that is the parameter estimates of ordinal logistic regression; the
coefficients, their standard errors, the Wald test and associated p-values (Sig.), and the 95% confidence interval of
the coefficients were computed. These values revealed that there is a strong association among practices of human
resource management and demographic variables and performance of Cooperative Bank of Oromia.From the
table,the Wald statistic is the square of the ratio of the coefficient to its standard error. The significance of the Wald
statistic in the column with heading sig < 0.05 stipulates the implication of the predictor variables in the model (fail
to accept H0 = 0) and high a value of the Wald statistic shows that the comparable predictor variable is significant as
stated by Brant 1990, cited in Adejumo & Adetunji (2013).
In the same table (Table 4.19), threshold and location terms are found at the left corner of the table, where the two
terms represent predicted and predictor variables respectively. In line to this, the threshold estimate for [performance
= 0] is the cutoff value between low and middle performance of Cooperative Bank of Oromia while the threshold
estimate for [performance = 1] is the cutoff value between middle and high performance of Cooperative Bank of
Oromia. The threshold coefficients are representing the intercepts, specifically the point (in terms of a logit) where
the status of the bank was predicted.
Besides, β is computed, which expresses ordinal logistic regression coefficients (estimates).It is also called beta
coefficient, which measures the degree of change in the outcome variable for every one-unit of change in the
predictor variable.It measures the goodness of fit of the line. Logistic regression coefficients are indicating the trend
and intensity of the linkage among independent variables and the log odds of dependent variable. In the exploration
of ranked data, the constants of the predictor variables in the model are interpreted as logarithm of the ratio of the
odds of response variable while the proportionate odds model is used. This means that estimates of this odds ratio,
and corresponding confidence intervals, can be easily found from the fitted model.
They are simply inferred as the projected shift in the log odds of being in the higher (as contrasting to lower)
category on the predicted variable (controlling for the residual independent variables) per unit rise on the
independent variable. In this regard, positive and negative constants have different inferences. Positive coefficients
are interpreted as for each one-unit increase on an independent variable; there is a predicted increase of a certain
value in the log odds of falling at an upper level on the predicted variable.
Contrary to this, the negative estimates are interpreted as for each one-unit increase on independent variables, there
is a predicted decrease of some quantity in the log odds of being in a higher level on the dependent variables, which
simply mean that as the results of independent variables increase, there is a decreased probability of falling at an
upper level on the predicted variable (Ikhlas et al., 2021; Obumneke Ezie, 2021; Onyeiwu et al., 2021). The table of
Parameter estimates tells us specifically about therelationship between the explanatory variables and the outcome
variables (Adejumo &Adetunji, 2013).
Based on this premises, the results obtained from the computation of parameter estimation by using ordinal logistic
regression coefficients are interpreted as a following: Consequently, from Table 4.29, values of -3.014 or less on the
primary covert variable that gave rise to dependent variable are classified as low given they were all reference
variables in the explanatory categories (holding all other category of visible explanatory variables are evaluated at
zero).
On the other hand, [Performance = 1] is the estimated cut point on the latent variable used to differentiate low and
medium from high performance when values of the forecaster variables are evaluated at zero. The response variable
that had a value of 2.04 or greater on the primary covert variable that gave rise to performance variable are classified
as high achievement given they were all reference variables in the predictor categories (holding all other category of
visible explanatory variables are evaluated at zero). Depending on the explained values that is, the values that range
from -3.014 to 2.04 on the primary covert variable would be classified as medium performance.
The computed values for the forecaster variables are also presented in the same table. When looking at demographic
variables, the result of the coefficient also shown positive result and also statistically significant (p-value = 0.020,
which is < 0.05). Therefore, concentrating on the route-one result, demographic variables were positive significant
predictor of performance of Cooperative Bank of Oromia. In this case, 1.09 value shows that for each one entity rise
in the demographic variables, there is a predicted increase of 1.09 in the log odds of being on an upper level on the
PCBO.
In a different way, the positive estimates of the prognosticator variable in case of recruitment and selection
depending on the route-one result, the interpretations are as follows: In connection to recruitment and selection, it
was positive significant predictor where the end product of the coefficient shows 2.28, which implies for each one
item increase in recruitment and selection, there is a predicted increase of 2.28 in the log odds of being on higher
level on the performance of Cooperative Bank of Oromia. This in turn shows the contribution of individuals who
pass through recruitment and selection process, their performance to the bank is 2.28 times more than individuals
who did not pass through the process when the other variables in the model are detained constant. Likewise, the
transformed log of all the variables indicated the same, which is consistent with the normally transformed variables’
result.
In the same manner, the computations made so far can further be interpreted by the help of route-two result. The key
difference between Route-Two Result and Route-One Result is the Exp (β) column and confidence interval (Ita,
2020; Ikhlas et al., 2021; Obumneke Ezie, 2021; Onyeiwu et al., 2021). The Exp (β) called exponential parameter
estimates column contains odds ratio reflecting the multiplicative change in the odds of being in a higher
classification on the predicted variable for each one item increase on the independent variable, keeping the residual
independent variables constant. The odds ratio is a value which measures the strength of effect of each independent
variable in the model on the log odds of the predicted variable. Thus, an odds ratio more than 1 proposes an
increasing chance of being in a higher level on the predicted variable as values on an independent variable increase.
On the contrary, an odds ratio below 1 suggests a decreasing probability with increasing values on independent
variables which also mean, the occurrence is less likely than non-occurrence. Whereas an odds ratio equivalent to 1
recommends no predicted change in the likelihood of being in a higher category as values on independent variables
increases.
The odds ratio is a value which measures the strength of effect of each independent variable in the model on the log
odds of the predicted variable. Based on the above principle, the odds ratios for the explanatory variables were
interpreted according to its result displayed in Table 4:30, which was computed in the SPSSAnalyzeGeneralized
Liner Models Generalized Liner Models again Ordinal Logistic by transforming the dependent and independent
variables in their proper place and then completing all the steps, which is presented in the following table as:
Table 4.20: Parameter Estimates of Ordinal Logistic Regression based on Route-Two (Exp.β)
Variables Exp.(β) 95% Wald Confidence Interval for Exp.(B)
Lower Upper
[Performance of CBO = .00] .49 -5.716 .312
Threshold
[Performance of CBO = 1.00] 7.649 .008 4.061
Location Recruitment & Selection 8.575 4.524 16.250
In view of that the results of Route-Two of the explanatory variables are indicated in the above table. To this end,
the odds of being in a higher level on the bank’s performance increases by a factor of 8.58 for every one unit
increases on recruitment & selection. As indicated from the values of odds ratios, the values for demographic
variables was positive in contrast to the beta value (B) computed above and the values of the odds ratios for
recruitment &selection was also> 1, implying increasing probability of being in a higher level on the bank’s
performance as values increase for the stated predictor variables.
Here again taking into consideration the computed values of Parameter Estimates of Ordinal Logistic Regression [β
& Exp. (β)], conclusion can be made about the practice.
4.4.1.4. Hypotheses Testing
Test of the research hypotheses were made based on the relationship between the dependent and independent
variables. Accordingly, the formulated research hypotheses were checked depending on the result of P-values
(statistical significance was set at α = 0.05)and the models identified based on the nature of the data.
H01: Recruitment and selection processes do not positively and significantly affect the P CBO.
HA1: Recruitment and selection processes positively and significantly affect the bank’s performance.
H02: Cooperative Bank of Oromia is not performing well.
HA2: Cooperative Bank of Oromia is performing well.
H03: There is no positive and significant association between the variables.
HA3: There is positive and significant association between the variables and the PCBO.
In connection to the above stated hypotheses, chi-square test of association and ordinal logistic regression results are
presented. Spearman’s Rank Correlation Coefficient was also applied to determine the relationships between the
variables as the following.
Based on the chi-square test of association, all the variables were significantly associated with PCBO at 5% level of
significance since the values of all the predictor variables were less than the alpha level (P-values < α = 0.05). In the
same way, the values of Pearson’s Chi-Square at df 4 and level of significance at 5%, all the computed values were
above the table value (9.49). Thus, depending on this computation, null hypotheses were rejected for all the
variables.Likewise, by the help of Spearman’s rank correlation all the variables were discovered to be positive,
medium and strong relationship and also statistically significant. Thus, depending on this model all the null
hypotheses were rejected.
On the other hand, by the use of ordinal logistic regression, the hypotheses set were tested. Accordingly, the
computed variables found indicated the following. With this regard, other than the direction and strength of
associations, all the computed results signified there was a relationship between the variables. They all were also
statistically significant at P-values < α = 0.05 and the direction of association too was positive. Therefore, depending
on this model, all the alternative hypotheses were accepted. This also implied that the all predictor variables have an
effect on the performance of the bank.
In summary, based on the computations made by the above three models (chi-square, correlation and ordinal logistic
regression), all the independent variables have an effect on the dependent variable. Hence, there was enough
evidence to reject the null hypotheses and in favor of the alternative hypotheses.
V. CONCLUSION AND RECOMMENDATIONS

5.1. Conclusion
The primary objective of this study was to assess the effect of recruitment and selection practice on the
PCBO. The effect of demographic variables and the performance of CBO were also analyzed. These days, it is
obvious that organization cannot exist unless it manages its human resource well, which is capable of properly
exploiting the material resources to generate surplus products for survival, growth and diversification into new
fields. This is true of banking and marketing organizations, where physical capital is only a fraction of the total
resources mobilized by the organization. These are service organizations where human capital makes the difference
between success and failure.
Currently, the Ethiopian economic sector is dominated by government-owned commercial bank (CBE) and
the industry is characterized by recently privatized new entrant banks which are in stiff competition with each other
to lead the market. Having the core competency in regards to attracting, developing and capitalizing on the most
valuable human resource of the organization is unquestionable. Cooperative Bank of Oromia is one among the 16
Private Banks currently in operations in Ethiopia. It is established in 2004 by primary Cooperatives, Cooperative
Unions operating in Oromia regional states, different organizations and individuals. The bank has a great vision in
order to benefit the grass-root members of the cooperatives and enhancing their social and cultural values, which is
realized in practice.
To obtain related data on the issues, it largely focused on the perception of 300 employees working in
Addis Ababa who provided related data through questionnaire. In addition, branch Managers of the selected banks,
Human Resource Manager of the bank and Board Members delivered complementary data through interview and
focus group discussions regarding the specified objectives or the effect human resource management practices have
on the performance of Cooperative Bank of Oromia. Employees working in Addis Ababa (city branches) were
selected randomly as respondents. The results were analyzed through descriptive and inferential statistics.
Consequently, the results verified that the bank needs to pay much more attention to the proper management of its
human resource in order to realize its full potential and create competitive advantage. For instance, with regard to
recruitment and selection processes, the descriptive statistics computed on the issues regarding notifying the result
of the selection on time, partiality and the participation of line managers in the process is low (mean value = 1.36,
2.65 and 2.87) respectively.
Similarly, the inferential statistics was computed to test the hypotheses and to show the relationship
between the variables. Accordingly, of the inferential statistics, chi-square, ordinal logistic regression and
Spearman’s Rank Correlation Coefficient were employed. As a consequence, the results of descriptive statistics and
the inferential statistics revealed that recruitment and selectionpractice appeared to have greater power to shape the
bank’s performance. Ultimately, the formulated hypotheses were tested by chi-square test of association, ordinal
logistic regression & Spearman’s rank correlation. Accordingly, by the help of these models all the alternative
hypotheses were accepted and all the null hypotheses were rejected. Overall, by the help of chi-square, ordinal
logistic regression and Spearman’s rank correlation coefficient, all the predictor variables were proved to be
positively affecting the performance of the bank where the degree of the relationships were labeled as medium and
strong. Eventually, the research concluded that proper handling and execution of the practices of human resource
management has enormous value on the performance of Cooperative Bank of Oromia.
5.2. Recommendations
Depending on the results and conclusions of the study already made, the researcher provided way forward
that aims at bringing human resource management of the Cooperative Bank of Oromia to establish sustainable
advantage and improve its performance. For that matter, management of this bank should start to incorporate human
resource management thinking into business strategies and make the human resource management function
strategically proactive. It is imperative for the bank to ensure that investment in its human resources and human
resource management practices will attract and retain talented people. In general, the findings addressed the Bank
should work on the major human resource management practices and other related issues to achieve more than what
it is performing:
Whatever the mode of recruitment (external/ internal) are favored by the bank, it should be visibly
communicated in what the company wants in an understandable way by prospective employees at large so as to
avoid prejudice/ Nepotism. Thus, the bank has to check its requirements before posting recruitment; disclose its
evaluation criteria procedures; and update results with justification using online Medias. Then after the completion
of recruitment criteria in a clear manner, the next step is doing the selection process in the same way. Those
committee or sub-committee who take part in selection of appropriate applicants, should run it in a free and fair way
and announce the result as immediate as possible.
Mechanisms to sustain long experienced and higher professionals have to be devised in the bank because
employees whose long service year and higher qualifications were few as compared with others from the
demographic characteristics. A worker who spends a longer length at any specific organization is accustomed to the
rules, strategies and dogmas of the organization and consequently can adapt in a better way. In the same way, highly
qualified employees that can fill the gap of professionalism and make the bank competitive in the market have to be
appealed.
Moreover, the participation of females as compared to male is by far low in the bank. So, females have to
be encouraged to take part in the employees and at all decision making positions in the bank. Government has to
continue the cooperative sector development and the concession to cooperative business like free legal and audit
services, provision of land and tax exemptions. Currently CBO is paying tax to the Government as equal as private
banks.The National Bank of the country or the regulatory and supervisory authorities should develop appropriate
legal framework for controlling, guiding and supervising the Cooperative banking system.
Eventually, corruption is a bottle neck in different organizations hindering themfrom attaining their missions and
visions. Thus, to enhance the PCBO such mal-practices should be removed.
[1]. REFERENCES
[2]. Ab Hamid, M. R., Sami, W., & Sidek, M. M. (2017, September). Discriminant validity assessment: Use of
Fornell
[3]. & Larcker criterion versus HTMT criterion. In Journal of Physics: Conference Series (Vol. 890, No. 1, p.
[4]. 012163). IOP Publishing.
[5]. Adejumo, A. O., & Adetunji, A. A. (2013). Application of ordinal logistic regression in the study of
students’ performance. Mathematical Theory and Modeling, 3(11), 10-19.
[6]. Adio, G., & Popoola, S. O. (2010). Demographic variables and self-efficacy as factors influencing career
commitment of librarians in federal university libraries in Nigeria. Library philosophy and practice, 1.
[7]. Alfayad, Z., & Arif, L. S. M. (2017). Employee voice and job satisfaction: An application of Herzberg two-
factor theory. International Review of Management and Marketing, 7(1), 150-156.
[8]. Ali, M., & Raza, S. A. (2017). Service quality perception and customer satisfaction in Islamic banks of
Pakistan: the modified SERVQUAL model. Total Quality Management & Business Excellence, 28(5-6),
559-577.
[9]. Archer, K. J., & Lemeshow, S. (2006). Goodness of-fit test for a logistic regression model fitted using
survey sample data. The Stata Journal, 6(1), 97-105.
[10]. Arifin, Z., Kartono, K., & Supriyadi, S. (2018). Development of authentic assessment instrument for
performance in learning mathematics in linear program. Journal of Research and Educational Research
Evaluation, 7(2), 154-162.
[11]. Asiaei, K., & Jusoh, R. (2015). A multidimensional view of intellectual capital: the impact on
organizational performance. Management decision.
[12]. Bedard, K. & Dhuey, E. (2006). The persistence of early childhood maturity: International evidence of
long-run age effects. Quarterly Journal of Economics, 121(4):1437-1472.
[13]. Bell, E., Bryman, A., & Harley, B. (2018). Business research methods. Oxford University Press.
[14]. Biesecker, L. G. (2013). Hypothesis-generating research and predictive medicine. Genome research, 23(70,
1051-1053.
[15]. Borman, W. R., & Mehay, S. L. (2009). Graduate education and employee performance: evidence.
Economics of Education Review3, 453-463.
[16]. Bratton, J., & Gold, J. (2017). Human resource management: theory and practice. Palgrave.
[17]. Cai, L., & Monroe, S. (2014). A New Statistic for Evaluating Item Response Theory Models for Ordinal
Data. CRESST Report 839. National Center for Research on Evaluation Standards, and Student Testing
(CRESST).
[18]. Cassidy, D. J., Hestenes, L. L., Hegde, A., Hestenes, S., & Mims, S. (2005). Measurement of quality in
preschool child care classrooms: An exploratory and confirmatory factor analysis of the early childhood
environment rating scale-revised. Early Childhood Research Quarterly, 20(3), 345-360.
[19]. Cavana, R., Delahaye, B., & Sekeran, U. (2001). Applied business research: Qualitative and
a. quantitative methods. John Wiley & Sons.
[20]. Chand, M., & Katou, A. A. (2007). The impact of HRM practices on organizational performance in the
Indian hotel industry. Employee relations.
[21]. Choudrie, J. & Dwivedi, Y. K. (2005). Investigating Broadband Diffusion in the Household: Towards
Content
[22]. Validity and Pre-Test of the Survey Instrument. Proceedings of the 13 th European Conference on
Information
[23]. Systems (ECIS), Regensburg, Germany.
[24]. Chyung, S. Y., Roberts, K., Swanson, I., & Hankinson, A. (2017). Evidence‐based survey design: The use
of a midpoint on the Likert scale. Performance Improvement, 56(10), 15-23.
[25]. Clark, L. A., & Watson, D. (2019). Constructing validity: New developments in creating objective
measuring instruments. Psychological Assessment, 31(12), 1412.
[26]. Cochran, W. G. (1963). Sampling Technique, Indian Edition, Bombay. Asian Publishing House, 413.
[27]. Cramer, D., & Howitt, D. (2004). The SAGE dictionary of Statistics. London: SAGE.
[28]. Creswell, J. (2014). Qualitative inquiry and research design: Choosing among five traditions. London:
Sage.
[29]. Das, S., & Rahman, R. M. (2011). Application of ordinal logistic regression analysis in determining risk
factors of child malnutrition in Bangladesh. Nutrition Journal, 10(1), 1-11.
[30]. Desalegn, F. E. (2021). Exploration of contributions of women in rural development and
[31]. determinant factors influencing their participation, the case of agricultural cooperatives in Ethiopia. Int. J.
Polit. Sci. Develop. 9(1) 4-16.
[32]. Desalegn, Z., Deyessa, N., Teka, B., Hailemariam, D., … & Abebe, T. (2021). Evaluation of COVID-19
related knowledge and preparedness in health professionals at selected health facilities in a resource-limited
setting in Addis Ababa. PloS one, 16(2), e0244050.
[33]. Doane, D. P., & Seward, L. E. (2011). Measuring Skewness. Journal of Statistics Education, 19(2), 1-18.
[34]. Drost, E. A. (2011). Validity and reliability in social science research. Education Research and
Perspectives, 38(1), 105-123.
[35]. Ebenuwa-Okoh, E. E. (2010). Influence of age, financial status, and gender on academic performance
among undergraduates, Journal of Psychology, 1(2), 99-103.
[36]. Edwards, J. R. (2011). The fallacy of formative measurement. Organizational research methods, 14(2),
370-388.
[37]. Egbu, C. O. (2004). Managing knowledge and intellectual capital for improved organizational innovations
in the construction industry: an examination of critical success factors. Engineering, Construction and
Architectural Management.
[38]. Ekanayake, E. M. A. C., Shen, G., Kumaraswamy, M., Owusu, E. K., & Xue, J. (2021). Capabilities to
withstand vulnerabilities and boost resilience in industrialized construction supply chains: a Hong Kong
study. Engineering, Construction and Architectural Management.
[39]. Engellant, K., Holland, D. & Piper, R. (2016). Assessing Convergent and Discriminant Validity of the
Motivation Construct for the Technology Integration Education (TIE) Model. Journal of Higher Education
Theory and Practice 16, 37-50.
[40]. Erkan, A., & Yildiz, Z. (2014). Parallel lines assumption in ordinal logistic regression and analysis
approaches. International Interdisciplinary Journal of Scientific Research, 1(3), 8-23.
[41]. Ertanto, Y.D., & Suharnomo, S. (2011). Effect of salary on employee performance with self-esteem as
intervening variable (Studies in PDAM Grobogan). (Doctoral dissertation, Universitas Diponegoro).
[42]. Farrell, A. M. (2010). Insufficient discriminant validity: A comment on Bove, Pervan, Beatty, and Shiu
(2009). Journal of Business Research, 63(3), 324-327.
[43]. Farrell, A. M., & Rudd, J. M. (2009). Factor analysis and discriminant validity: A brief review of some
practical issues. Anzmac.
[44]. Ferris, G. R., Berkson, H. M., & Harris, M. M. (2002).The recruitment interview process: persuasion and
organization reputation promotion in competitive labor markets, Human Resource Management Review,
12(3), 359-375.
[45]. Field, A. P. (2005). Discovering Statistics Using SPSS, Sage Publications Inc.
[46]. Franke, G., & Sarstedt, M. (2019). Heuristics versus statistics in discriminant validity testing: a comparison
of four procedures. Internet Research.
[47]. Gagunashvili, N. D. (2010). Chi-square tests for comparing weighted histograms. Nuclear Instruments and
Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated
Equipment, 614(2), 287-296.
[48]. George, D., & Mallery, P. (2003). SPSS for Windows step by step: answers to selected exercises. A simple
guide and reference, 63, 1461-1470.
[49]. Ghauri, P. & Grönhaug, K. (2005). Research methods in business studies: A Practical Guide. 3 rd Ed.
London: Prentice Hall.
[50]. Ghauri, P., Grønhaug, K., & Strange, R. (2020). Research methods in business studies. Cambridge
University Press.
[51]. Ghadi, I., Alwi, N. H., Bakar, K. A., & Talib, O. (2012). Construct Validity Examination of Critical
Thinking Dispositions for Undergraduate Students in University Putra Malaysia. Higher Education
Studies, 2(2), 138-145.
[52]. Hadi, N. U., Abdullah, N., & Sentosa, I. (2016). An easy approach to exploratory factor analysis:
Marketing perspective. Journal of Educational and Social Research, 6(1), 215.
[53]. Hair Jr Joseph, F., Bush Robert, P., & Ortinau David, J. (2003). Marketing research: within a changing
[54]. environment.
[55]. Hartgerink, C. H., van Aert, R. C., Nuijten, M. B., Wicherts, J. M., van Assen, M. A., & Gruber, C. W.
(2020). P-Values Less Than 0.05 in Psychology: What is going on? In The Theory of Statistics in
Psychology (pp. 35-50). Springer, Cham.
[56]. Henseler, J., Ringle, C. M., & Sarstedt, M. (2015). A new criterion for assessing discriminant validity in
variance-based structural equation modeling. Journal of the Academy of Marketing Science, 43(1), 115-
135.
[57]. Higham, N. J. (2002). Computing the nearest correlation Matrix-a problem from finance.
[58]. Journal of Numerical Analysis, 22(3), 329-343.
[59]. Hsieh, Y. H., & Chen, H. M. (2011). Strategic fit among business competitive strategy, human resource
strategy, and reward system. Academy of Strategic Management Journal, 10(2), 11.
[60]. Huck, S. W. (2007). Reading Statistics and Research, United States of America, Allyn & Bacon.
[61]. Hung, M. L., Chou, C., Chen, C. H., & Own, Z. Y. (2010). Learner readiness for online learning: Scale
development and student perceptions. Computers & Education, 55(3), 1080-1090.
[62]. Hussain, J. N., & Nassir, A. J. (2015). Cluster Analysis as a Strategy Of Grouping to Construct Goodness–
Of–Fit Tests when the Continuous Covariates Present in the Logistic Regression Models.
[63]. Ikhlas, M., Kuswanto, K., Sakunti, S. R., Debi, M. R., & Collantes, L. M. (2021). A Sequential Explanatory
Investigation in using ICTs on Arabic Language and Islamic Values Education: Teacher-Students
Perspective. Indonesian Journal on Learning and Advanced Education (IJOLAE), 3(3), 153-167.
[64]. Ita, E. O. (2020). Business Ethics in a Constrained Environment: Ethical Dilemmas of Entrepreneurship in
Nigeria (Doctoral dissertation, Dublin, National College of Ireland).
[65]. Joyner, H. S. (2019). Semisolid Food Tribology. In Rheology of Semisolid Foods (pp. 133-165). Springer,
Cham.
[66]. Kapsos, S., & Bourmpoula, E. (2013). Employment and economic class in the developing world. Geneva:
ILO.
[67]. Kasper-Brauer, K., & Leischnig, A. (2016). Yes, we can! A fuzzy-set analysis of challenges, skills, and
enjoyment of work. Journal of Business Research, 69(11), 5286-5291.
[68]. Kemp, S. E., Ng, M., Hollowood, T., & Hort, J. (2018). Introduction to descriptive analysis. Descriptive
analysis in sensory evaluation, 1.
[69]. Koh, C. E. & Nam, K. T. (2005). Business use of the internet: a longitudinal study from a value chain
perspective. Industrial Management & Data Systems, 105 85-95.
[70]. Kothari, C. R. (2004). Research methodology: Methods and techniques. New Age International.
[71]. Kothari, C. R. (2007). Quantitative techniques. New Delhi, UBS Publishers LTD.
[72]. Lacap, J. P. (2019). The effects of food-related motivation, local food involvement, and food satisfaction on
destination loyalty: the case of Angeles city, Philippines. Advances in Hospitality and Tourism Research
(AHTR), 7(2), 238-257.
[73]. Lacap, J. P. G., & Tungcab, A. P. (2020). The Influence of Brand Experience on Brand Loyalty among
Mobile Phone Users in Pampanga, Philippines: A Mediation Analysis. Asia-Pacific Social Science
Review, 20(3).
[74]. Laditka, S. B., Fisher, M., Laditka, J. N., & Segal, D. R. (2004). Attitudes about aging and gender among
young, middle age, and older college-based students. Educational Gerontology, 30(5), 403-421.
[75]. Levy, P. S., & Lemeshow, S. (2013). Sampling of Populations: Methods and Applications. John Wiley &
Sons.
[76]. Likert, R. (1932). A technique for the measurement of attitudes. Archives of Psychology, 140, 1-55.
[77]. Lou, J., Fang, Y., Lim, K. H., & Peng, J. Z. (2013). Contributing high quantity and quality knowledge to
online Q & A communities. Journal of the American Society for Information Science and
Technology, 64(2), 356-371.
[78]. Lynch, S., & Smith, K. (2010). The dilemma of judging unpaid workers. Personnel review.
[79]. Mahon, N. E., Yarcheski, A., & Yarcheski, T. J. (2005). Happiness as related to gender and
[80]. health in early adolescents. Clinical nursing research, 14(2), 175-190.
[81]. Marques, J. (2007). HR in all its glory. Human resource management international Digest.
[82]. Mehta, E. (2016). Literature review on HR practice in banking sector. International Research Journal of
Engineering, IT & Scientific Research, 2(7), 90-97.
[83]. Meron Desalegn (2011).Ordinal Logistic Regression Analysis of Correlates of Crime Severity: The case of
Tigray Region, Ethiopia, (MA Thesis).
[84]. Muhammad, S., Anjum, T., Afridi, F. K., & Alasan, I. I. (2020). Impact of transformational leadership on
project success criteria: The moderating role of work engagement. PalArch's Journal of Archaeology of
Egypt/Egyptology, 17(6), 7460-7476.
[85]. Myers, L., & Sirois, M. J. (2004). Spearman correlation coefficients, differences between. Encyclopedia of
statistical sciences, 12.
[86]. Nunnally, J. C. (2010). Psychometric Theory, 2 ndEd. New York: McGraw-Hill.
[87]. O'Connell, A. A. (2006). Logistic regression models for ordinal response variables (Vol. 146). sage.
[88]. Ogundapo, O., Nche, C., Duffy, A., & Zhang, G. (2018). ’Application of the Feature Selective Validation
Method
[89]. and Kolmogorov-Smirnov Test to Evaluate Handling Effects on Crosstalk of Ethernet Cables’. American
[90]. Journal of Electrical and Electronic Engineering, 6(1), 32-37.
[91]. Okeniyi, J. O., & Okeniyi, E. T. (2012). Implementation of Kolmogorov-smirnov p-value computation in
Visual
[92]. Basic®: Implication for Microsoft Excel® library function. Journal of Statistical Computation and
[93]. Simulation, 82(12), 1727-1741.
[94]. Oluwatayo, J. (2012). Validity and reliability issues in educational research. Journal of Educational and
Social Research 2, 391-400.
[95]. Onyeiwu Ph D, C., Muoneke, O. B., & Abayomi, A. M. (2021). Effects of Micro-finance Bank Credit on
Small and Medium Scale Businesses: Evidence from Alimosho LGA, Lagos State. The Journal of
Entrepreneurial Finance, 22(2), 4.
[96]. Palladan, A. A., Abdul Kadir, K., & Yen, W. C. (2016). Effects of strategic leadership, organizational
innovativeness and information technology capability on effective strategy implementation. International
Journal of Organizational and Business Excellence, 1(2), 1-18.
[97]. Pallant, J. (2010). SPSS survival manual: a step by step guide to data analysis using SPSS.
[98]. Park, C. W., MacInnis, D. J., Priester, J., Eisingerich, A. B., & Iacobucci, D. (2010). Brand attachment and
brand attitude strength: Conceptual and empirical differentiation of two critical brand equity
drivers. Journal of Marketing, 74(6), 1-17.
[99]. Pellizzari, M. & Billari, F. C. (2012). The younger, the better? Age related differences in academic
performance at University. Journal of Population Economics, 25(2), 697- 739.
[100]. Petter, S., Straub, D., & Rai, A. (2007). Specifying formative constructs in information systems research,
MIS Quarterly31(4), 623-656.
[101]. Pohar, M., Blas, M., & Turk, S. (2004). Comparison of logistic regression and linear discriminant analysis:
a simulation study. Metodoloski zvezki, 1(1), 143.
[102]. Prasetyo, K., Masrukan, M., & Sunawan, S. (2019). The Development of Mathematical Literation
Instruments Based on Class IV Geometry Material Conservation. Journal of Research and Educational
Research Evaluation, 8(1), 1-13.
[103]. Ramdani, Z. (2018). Construction of academic integrity scale. International Journal of Research Studies in
Psychology, 7(1), 87-97.
[104]. Robb, A. M., & Watson, J. (2012). Gender differences in firm performance: Evidence from new ventures in
the United States. Journal of Business Venturing, 27(5), 544-558.
[105]. Robinson, G. (2009). In order to grow, must the founder go: a comparison of performance between founder
and non-founder managed high-growth manufacturing firms? Journal of Business Venturing, 7(3), 181-
194.
[106]. Saif, S. K., Nawaz, A., & Jan, F. A. (2012). Predicting job-satisfaction among the academicians of
universities in KPK, Pakistan. Industrial Engineering Letters, 2(2), 34-45.
[107]. Samuels, P. (2017). Advice on exploratory factor analysis.
[108]. Schmidt, C. O., & Kohlmann, T. (2008). When to use the odds ratio or the relative risk? International
Journal of Public Health, 53(3), 165.
[109]. Shahsavar, Z., & Kourepaz, H. (2020). Postgraduate students’ difficulties in writing their theses literature
review. Cogent Education, 7(1), 1784620.
[110]. Straub, D., Boudreau, M.-C. & Gefen, D. (2004). Validation guidelines for IS positivist research.
Communications of the Association for Information Systems, 13, 380-427.
[111]. Strauss, M. E., & Smith, G. T. (2009). Construct validity: Advances in theory and methodology. Annual
Review of Clinical Psychology, 5, 1.
[112]. Taherdoost, H. (2016). Validity and reliability of the research instrument; how to test the validation of a
questionnaire/survey in a research. How to test the validation of a questionnaire/survey in a research
(August 10, 2016).
[113]. Toledo, A. H., Flikkema, R., & Toledo-Pereyra, L. H. (2011). Developing the research hypothesis. Journal
of Investigative Surgery, 24(5), 191-194.
[114]. Utts, J. M., & Heckard, R. F. (2021). Mind on statistics. Cengage Learning.
[115]. Veale, J. F. (2014). Edinburgh handedness inventory–short form: a revised version based on confirmatory
factor
[116]. analysis. Laterality: Asymmetries of Body, Brain & Cognition,19(2), 164-177.
[117]. Vembriliya, Z. A., Khumaedi, M., & Masrukan, M. (2019). Development of Assessment Performance
Materials Cylinder Students Grade IX SMP Adiwiyata-Based. Journal of Research and Educational
Research Evaluation, 8(1), 30-38.
[118]. Voorhees, C. M., Brady, M. K., Calantone, R., & Ramirez, E. (2016). Discriminant validity testing in
marketing: an analysis, causes for concern, and proposed remedies. Journal of the academy of marketing
science, 44(1), 119-134.
[119]. Wakene, D. K., Munapo, E., & Olufemi, A. P. (2020). The effect of demographic factors on intention to
use banking services: evidence from Ethiopia. GSJ, 8(6).
[120]. Wang, L., Ali, M., Kim, H. J., Lee, S., & Hernandez Perlines, F. (2021). Individual entrepreneurial
orientation, value congruence, and individual outcomes: Does the institutional entrepreneurial environment
matter? Business Strategy & the Environment, 30(5), 2293-2312.
[121]. Wee, Y. S. & Quazi, H. A. (2005). Development and validation of critical factors of environmental
management. Industrial Management & Data Systems, 105, 96-114.
[122]. Whitley, B. E. (2002). Principals of Research and Behavioral Science, Boston, McGraw-Hill.
[123]. Wickens, T. D. (2014). Multi-way contingency tables analysis for the social sciences. Psychology Press.
[124]. Williams, A. (2003). How to write and analyze a questionnaire? Journal of Orthodontics, 30,245-252.
[125]. Yamane, T. (1967). Statistics: An introductory analysis (No. HA29 Y2 1967).
[126]. Yavuz, G., & Dogan, N. (2014). Maslach burnout inventory-student survey (MBI-SS): a validity
study. Procedia-Social and Behavioral Sciences, 116, 2453-2457.
[127]. Yong, A. G., & Pearce, S. (2013). A beginner’s guide to factor analysis: Focusing on exploratory factor
analysis. Tutorials in Quantitative Methods for Psychology, 9(2), 79-94.
[128]. Zaiţ, A., & Bertea, P. S. P. E. (2011). Methods for testing discriminant validity. Management & Marketing
Journal, 9(2), 217-224.
ACKNOWLEDGEMENTS
The execution of such most important research is never the effort of anyone alone. The contributions of many
different people and organizations, in their different ways, have made this possible. Therefore, after all these years
of hard work, it is necessary to express my gratitude to those who in one way or another contributed and extended
their support and valuable backing in the preparation and completion of this academic work. First and foremost, I
would like to thank Almighty God, for his blessing, mercy bestowed on me abundantly and for this wonderful
opportunity he has given me to write this dissertation. Next, I would like to take immense pleasure to extend my
heartfelt thankfulness toProfessor S. Nakkiran, my PhD supervisor. Besides, I thank all the participants in my
survey, who have shared their precious time all through the progression of data collection. Finally, I express my
appreciation to all writers of those materials I referred. May God bless you all!

Theeffect of Recruitment and Selection Process On The Performance of Organization (Cooperative Bank of Oromia)

Uploaded by

Copyright:

Available Formats

Theeffect of Recruitment and Selection Process On The Performance of Organization (Cooperative Bank of Oromia)

Uploaded by

Document Information

Original Title

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

Theeffect of Recruitment and Selection Process On The Performance of Organization (Cooperative Bank of Oromia)

Uploaded by

Copyright:

Available Formats

American Journal Of Multidisciplinary Research & Review (AJMRR) 2023

American Journal of Multidisciplinary Research & Review (AJMRR)

Research Paper Open Access

THEEFFECT OF RECRUITMENT AND SELECTION PROCESS

Keywords: Cooperative Bank;Effect;Performance; Recruitment and Selection; Statistics

AJMRR Journal Page |1

II. LITERATURE REVIEW

III. RESEARCH METHODOLOGY

Source: Compiled from Cooperative Bank of Oromia, 2021

AJMRR Journal Page |3

IV. RESULTS AND DISCUSSIONS

AJMRR Journal Page |5

AJMRR Journal Page |6

Table 4.5: Cronbach's Alpha Result of Dependent & Independent Variables

Recruitment & selection 9 0.74

AJMRR Journal Page |7

AJMRR Journal Page |8

AJMRR Journal Page |9

Table 4.7: Descriptive Statistics of Employees’ Perception towards PCBO

Figure 4.1: Total Deposit status of CBO, 2021

Source: Computed from CBO Annual Reports, 2021

Figure 4.2: Revenues/ Earnings of the Bank, 2021

Figure 4.3: Total Assets of CBO, 2021

iv. Capital of CBO from 2016 to 2020

Figure 4.4: Capital of CBO, 2021

Figure 4.5: Total Loans of the Bank, 2021

Source:Researcher’s Computation from SPSS 21.0

Source: Researcher’s Computation from SPSS 21.0

Figure 4.8:Histogram showing the Distribution of Recruitment and Selection Data

4.4.1.1. Chi-Square Test of Association

Table 4.13: Correlation Coefficient of the Dependent and Independent Variables

Source: Researcher’s Computation by SPSS Output from Survey Data, 2021

Source: Researcher’s Computation, 2021

V. CONCLUSION AND RECOMMENDATIONS

You might also like