Anda di halaman 1dari 7

© 2019 IJRAR March 2019, Volume 6, Issue 1 www.ijrar.

org (E-ISSN 23481269, P- ISSN 2349-5138)

TESTING THE RELIABILITY AND VALIDITY OF


HRM MEASUREMENT SCALE USING
CONFIRMATORY FACTOR ANALYSIS: AN
EMPIRICAL EVIDENCE FROM ETHIOPIAN
PUBLIC SERVICE ORGANIZATIONS
Assefa Tsegay Tensay1, Professor Manjit Singh2
1
PhD Candidate, 2 Professor
1
School of Applied Management,
1
Panjabi University Patiala, India
Abstract: Although HRM is strategically vital in organizational functions, there is lack of appropriate instrument to measure employees’
perception about HR practices. The purpose of this study is to validate the appropriateness of the HRM measurement scale adopted from
Vermeeren (2014) in public service organizations of Ethiopian situation. Based on a sample of 340 federal employees of nine government
organizations, data analysis was done using Structural Equation Modelling. The finding of the study revealed that the six factor constructs
have a high level of reliability coefficient. In addition, the Confirmatory Factor Analysis result demonstrated that both convergent and
divergent validity were confirmed. Thus, the measurement scale is both reliable and valid which is appropriate in Ethiopian context.
Key Works: HRM, CFA, Ethiopian Public Service Organizations
1. Introduction
The most challenging task of organizations in today’s complex business environment is improving organizational performance. It was argued
that the productivity and efficiency of organizations is typically the result of HR (Kim, 2005; Armstrong, 2009). This implies that the quality
and quantity of Human Resource in the workplace is critically important for any type of organizations (private, public or Non-profit
organizations). Although HRM is strategically vital in organizational functions, there is lack of appropriate instrument to measure employees’
perception in the workplace about the HR practices (Demo et al, 2012). Recently, several scholars tried to understand the mechanism how
HRM related to Organizational Performance. Nevertheless, most of the studies utilized different HR practices and measurement instruments
(Boselie et al., 2005; Combs et al., 2006; Paauwe, 2009).
It is widely recognized that the Public Sector is one of the most important sectors that contribute towards the economy growth of nations
(Vermeeren, 2014), less attention is given by HRM scholars particularly in developing context. Regarding measurement of HRM, prominent
scholars (Huselid, 1995; Demo et al, 2012) argued that there is very limited work on appropriate instrument to measure perception of HRM
practices. Based on the work of the well-known HRM scholars (Ahmad & Schroeder, 2003; Appelbaum et al., 2000; Gould-Williams, 2003;
Huselid, 1995; Wright et al., 2005), Vermeeren (2014) developed HRM measurement scale in Netherlands, which is known as high-
commitment HR practices consists of six HR practices such as Recruitment and Selection; Training and Development; Performance
Appraisal; Compensation and Reward; Autonomy and Employee Participation. Commitment based HR practices are critically important in
encouraging employees to take more responsibility and ultimately benefits their organization (Combs et al, 2006). The purpose of this study
is to validate the appropriateness of the HRM measurement scale in Ethiopian context adopted from Vermeeren (2014) through Confirmatory
Factor Analysis. This study contributes to the knowledge of Human Resource Management by extending the literature on the measurement
instruments of HR practices in the public sector. In terms of organization of the article, it has the following sections. First, the article presents
introduction and literature review on the study issue. Next, methodology of the study was discussed. Third, results of the finding are presented.
Based on the results of the study, this article put the discussion by providing empirical evidences in fourth section. Finally, conclusion was
made.
2. Theoretical Background
Although HRM is strategically relevant in organizations, there is no consistency in the definition and measurement (Paauwe, 2009; Coms et
al, 2006; Boselie et al, 2005). In the HRM research, different researchers applied different sets of practices in the HRM-Performance study.
According to Savaneviciene and Stankeviciute (2012), there is no commonly acknowledged theoretical rationale for selecting HR practices.
Recent meta-analysis reviews uncovered that the utilization of HR practices used in HRM systems differ dramatically from one study to
another (Boselie et al, 2005; Subramony, 2009; Jiang and Messersmith, 2017). Nonetheless, there is ample commonality as investigates
typically shield a significant variety of HR practices (Wall and Wood, 2005). In their meta-analysis, Boselie et al (2005) mentioned the top
four HR practices used by many researchers as Recruitment and Selection, Training and Development, Performance Appraisal, Compensation
and Reward. In his literature review, Khan (2010) also mentioned the five key HRM practices as recruitment and selection; training and
development; compensation and reward; performance appraisal, and employee participation. As Kehoe and Wright (2013) claimed that
autonomy is critical part of the HRM practices in service organizations. In strategic HRM research, Batt (2002) argued that the top four HRM

IJRAR19J3057 International Journal of Research and Analytical Reviews (IJRAR) www.ijrar.org 1106
© 2019 IJRAR March 2019, Volume 6, Issue 1 www.ijrar.org (E-ISSN 23481269, P- ISSN 2349-5138)
practices (Recruitment and Selection, Training and Development, Performance Appraisal, Compensation and Reward) can reflect the major
objectives of strategic HRM.
Remarkably, researchers were advised to follow AMO model to select HR practices to be included in the HRM System instead of randomly
selecting them (Boxall and Purcell, 2008; Lepak et al., 2006; Jiang et al, 2012). Previous studies used AMO model in selecting and classifying
the different HR practices. For example, in their conceptual review of HRM System, Lepak and colleagues (2006) categorized a number of
HR practices in three Dimensions: knowledge, ability and skill HR domain (recruitment, selection and training); motivation and effort HRM
domain (job security, performance appraisal, compensation and reward) whereas opportunity to contribute HRM domain (employee
participation, empowerment, job rotation, job enrichment, autonomy, grievance, information sharing). Similarly, prominent researchers in
HRM literature (Subramony, 2009; Jiang et al, 2012) classified the different HR practices into Skill/Ability-enhancing, Motivation-enhancing
and Opportunity/Empowerment-enhancing HR practices following the AMO Model.
In our review of literature, opportunity-enhancing practices were missing in the HRM-Performance research (Wood and de Menezes, 2011).
It is widely recognized that if employees have good participation in the process of decision making as well as autonomous in their work,
employees become more engaged and become productive (Solomon and Sridevi, 2010). Boxall and Macky (2009) argued that employee
participation is critically relevant practice to improve HRM’s contribution to organizational effectiveness. Based on the above theoretical
arguments, the current study selected the Vermeeren’s HR measurement scale as it includes the most commonly used HR practices in the
public sector and fulfills the AMO theory.
In the HRM literature, the outcomes of HRM are categorized into HR outcomes (commitment, motivation, OCB, job satisfaction,
engagement, trust); operational outcomes (productivity, quality, customer satisfaction, turnover); financial outcomes and marketing outcomes
(Dyre and Reeves, 1995). It is widely recognized that HRM has a key role in enhancing performance. They argued that HRM practices first
affects intermediary outcomes (HR and operational outcomes), which are known as proximal outcomes and in return affects the distal
outcomes (financial and marketing outcomes). Taken together, HRM and the appropriate implementation of the practices is strategically
crucial in enhancing performance of organizations in all forms. But, studies revealed that there is inconsistency in the relationship between
HRM and Organizational Performance due to lack of agreement on the practices and measurement scale of the HRM construct. From the
literature, we can infer that proper measurement of HRM is relevant as its contribution to enhance performance is magnificent.
3. Materials and Methods
Sample
The article was based on the perception of 340 employees from different departments and sections of the nine randomly selected public
service organizations (20% of the population study organizations) at the federal level. Employees were selected randomly from different
departments. The demographic profile of the respondents is presented as follows. The data result indicated that the majority of the respondents
are male, married, holds bachelor’s degree, matured, experienced and fall under the category of professional science job grade. Thus, the
study can be generalized because respondents are representative of the population in terms of gender, age, education and service year which
is in line with the government report (MoPH&HRD, 2016).
Measurement Scale and Pilot Study
The questionnaire survey has two parts. Part one is about the demographic characteristics and the second part is the HRM practices. This
study used standard questionnaire to measure the HRM practices developed by Vermeeren (2014) with 31 items categorized into six HR
practices. Before going to final data collection, we made discussion with experts from academic institution and industry. Following the
translation from English to Amharic, a pilot study was conducted in three public service organizations to check the understandability of the
questionnaire (see Table 1).

Table 1: Operationalization of Study Variables


S.No Variable Attributes # of Items Source
1 HRM Strongly Disagree (1), 31 items classified into six Vermeeren
Disagree (2), Neutral (3), sub-dimensions: (2014)
Agree (4) and Strongly
Agree (5)

The reliability of the measurement scale revealed that the Cronbach’s alpha values of each of the variables is greater than the minimum
threshold (Hair et al, 2010), indicating that the instrument is reliable and can be used for final data collection.

IJRAR19J3057 International Journal of Research and Analytical Reviews (IJRAR) www.ijrar.org 1107
© 2019 IJRAR March 2019, Volume 6, Issue 1 www.ijrar.org (E-ISSN 23481269, P- ISSN 2349-5138)

Table 2: Pilot study reliability coefficient of the overall HRM and its sub-dimension scales
Variables #Respondents #Items Cronbach’s Alpha
Recruitment and Selection 45 4 0.776
Training and Development 45 8 0.727
Performance Appraisal 45 6 0.885
Compensation and Reward 45 3 0.732
Autonomy 45 6 0.822
Employee Participation 45 4 0.771
Overall HRM 45 31 0.916
Source: compiled from own survey
Data Collection and Analysis
In order to validate the scale, a questionnaire survey with 31 items categorized into six HR practices were distributed and self-administered
to 400 respondents of the sample organizations. In this respect, 340 properly filled questionnaires were returned. Regarding SEM, sample
size determination is critical because it is a large sample size statistical technique (Kline, 2016). According to the author, large sample size
are necessary to improve statistical power and trustworthiness of the results. Sekaran and Bougie (2013) identified research objective, level
of confidence, amount of variability in the population, cost and time constrained and size of the population itself as major factors to determine
sample size. And also Hair et al (2010) asserted that the minimum sample size for a particular SEM model depends on the complexity of the
model, value of communality and other factors. They suggested that researchers can have five to ten cases for each parameter estimates. In
multivariate research, Sekaran and Bougie (2013) argued that sample size to be ten times as large as the number of variables in the study. In
line with scholars (for example, Hair et al, 2010; Sekaran and Bougie, 2013), the present study took sample of 400 which is above the
minimum sample requirement. In this study, the multivariate assumptions such as missing data, normality, outlier and multicollinearity were
tested and it was found that they are not serious problems. Data was analyzed using Exploratory Factor Analysis through SPSS version 24 to
test the factorability of the measurement items and Confirmatory Factor Analysis through AMOS version 23 was conducted to test the
construct validity of the constructs. To ensure the model fitness with the sample data, several model fit parameters were used. For this study,
both absolute and relative indices (Chi Square ratio (χ2/df), Standardized Root Mean Square Residual, Goodness fit index, comparative fit
index, Tucker-Lewis index, Root Mean Square Error of Approximation) which are widely reported in reputable journals (Hair et al, 2010)
are used. The cut-off points for Chi Square ratio (χ2/df) is less than five; for RMR less than 0.05; for GFI, CFI, and TLI greater than 0.90
and RMSEA less than 0.08 (Hair et al, 2010; Byrne, 2016).
4. Results
Correlation and Descriptive Result Analysis
As it is indicated in table 4, the mean score of the six HR practices is above the midpoint (3) which is encouraging in the public service
organization. As it is expected, the result of Pearson correlation coefficients shows that there was a positive and a strong relationship between
the dimensions.

Table 3: Correlations and Descriptive Result Analysis


Mean RS TD PA CR AU EP
Recruitment and Selection 3.72
Training and Development 3.38 .510**
Performance Appraisal 3.57 .387** .471**
Compensation and Reward 3.04 .331** .362** .576**
Autonomy 3.19 .297** .347** .406** .282**
Employee Participation 3.28 .294** .383** .450** .370** .477**
**. Correlation is significant at the 0.01 level (2-tailed).

Reliability Analysis
To assess the reliability of the data, we used the internal consistency approach (Cronbach’s alpha). Kline (2016) suggested that the value of
alpha with 0.7 or higher shows better reliability of the data.

Table 4: The reliability of sub-dimensions and overall measurement scale

Measurement Scale Number of Number of Items Cronbach’s Alpha


Respondents
Recruitment and Selection 340 4 0.88
Training and Development 340 8 0.91
Performance Appraisal 340 6 0.81
Compensation and Reward 340 3 0.77
Autonomy 340 6 0.86

IJRAR19J3057 International Journal of Research and Analytical Reviews (IJRAR) www.ijrar.org 1108
© 2019 IJRAR March 2019, Volume 6, Issue 1 www.ijrar.org (E-ISSN 23481269, P- ISSN 2349-5138)

Employee Participation 340 4 0.82


Overall reliability 340 31 0.92

As it is displayed in table 4 above, the finding of the study revealed that the instrument is reliable both dimension-wise and overall coefficient
of the measurement scale. The finding of this study revealed that the six HR practices are reliable as the Cronbach’s alpha for each practices
is above the minimum threshold. In addition to alpha coefficient, this study also considered composite reliability as a measurement of
reliability (see table 5)
Validity Analysis
In order to ensure the construct validity of the dimensions, CFA was conducted to validate the measurement model. Before going into the
overall measurement model, the individual measurement model was run to test the convergent validity. In this respect, the CFA results of
each dimensions was discussed below.

Table 5: CFA Result for the six first order dimensions of HR practices
First order factors Standardized Level of AVE CR
factor loading significance
RS1 <--Recruitment and Selection 0.755 *** 0.6581 0.8849
RS2 <--Recruitment and Selection 0.835 ***
RS3 <--Recruitment and Selection 0.807 ***
RS4 <--Recruitment and Selection 0.845 ***
(χ2/df = 3.16; RMR= 0.012; GFI = 0.991; CFI = 0.994; TLI = 0.982; RMSEA = 0.08
TD1 <--Training and Development 0.530 *** 0.5537 0.9065
TD2 <--Training and Development 0.857 ***
TD3 <--Training and Development 0.889 ***
TD4 <--Training and Development 0.822 ***
TD5 <--Training and Development 0.712 ***
TD6 <--Training and Development 0.749 ***
TD7 <--Training and Development 0.688 ***
TD8 <--Training and Development 0.638 ***
(χ2/df = 3.17; RMR= 0.031; GFI = 0.960; CFI = 0.976; TLI = 0.963; RMSEA = 0.08
PA1 <--Performance Appraisal 0.631 *** 0.5093 0.8600
PA2 <--Performance Appraisal 0.642 ***
PA3 <--Performance Appraisal 0.702 ***
PA4 <--Performance Appraisal 0.852 ***
PA5 <--Performance Appraisal 0.642 ***
PA6 <--Performance Appraisal 0.784 ***
(χ2/df = 2.76; RMR= 0.024; GFI = 0.975; CFI = 0.972; TLI = 0.953; RMSEA = 0.072
CR1 <--Compensation and Reward 0.764 *** 0.5670 0.7924
CR2 <--Compensation and Reward 0.578 ***
CR3 <--Compensation and Reward 0.885 ***
(χ2/df = just identified; RMR= 0.000; GFI = 1.00; CFI = 1.000. Perfectly fit with the data.
AU1 <--Autonomy 0.751 *** 0.5111 0.8620
AU2 <--Autonomy 0.695 ***
AU3 <--Autonomy 0.723 ***
AU4 <--Autonomy 0.688 ***
AU5 <--Autonomy 0.787 ***
AU6 <--Autonomy 0.636 ***
(χ2/df = 3.315; RMR= 0.028; GFI = 0.974; CFI = 0.978; TLI = 0.958; RMSEA = 0.083
EP1 <--Employee Participation 0.721 *** 0.5331 0.8173
EP2 <--Employee Participation 0.634 ***
EP3 <--Employee Participation 0.895 ***
EP4 <--Employee Participation 0.640 ***
(χ2/df = 2.05; RMR= 0.009; GFI = 0.997; CFI = 0.998; TLI = 0.988; RMSEA = 0.054
*** P< 0.0010.01 level.

Table 5 shows that factor loading, average variance explained and composite reliability of the six dimensions of the HRM measurement
scale. The finding established that factor loading are all greater than 0.5 (Hair et al, 2010). For example, factor loading for Recruitment and
Selection ranges from 0.755 to 0.845, implying that the items are strongly related with the latent construct at significant level with p<0.001.

IJRAR19J3057 International Journal of Research and Analytical Reviews (IJRAR) www.ijrar.org 1109
© 2019 IJRAR March 2019, Volume 6, Issue 1 www.ijrar.org (E-ISSN 23481269, P- ISSN 2349-5138)
Similarly, the factor loading and significance level are specified in table 5 above. Accordingly, the Average variance explained (AVE) and
Composite reliability (CR) for each of the construct meet the minimum threshold, indicating that AVE is greater than 0.5 and that of CR is
greater than 0.7 (Hair et al, 2010). In this article, the AVE for each of the constructs is greater than 0.5 and that of CR greater than 0.7,
convergent validity was properly established for each of the dimensions.

To validate the distinctiveness (divergent validity) of the six dimensions, CFA for the overall measurement model, by combing all the six
dimensions together, was performed. The overall CFA result of the measurement model was evaluated and presented in figure 1. Results of
the overall CFA shown that the sample data properly fit with the hypothesized model (χ2/df = 1.918; RMR= 0.041; CFI = 0.928; TLI = 0.920;
and RMSEA = 0.052).

Note: RSD= Recruitment and Selection Dimension; TDD= Training and Development Dimension; CRD= Compensation and Reward
Dimension; EPD= Employee Participation Dimension; PAD= Performance Appraisal Dimension and AUD= Autonomy Dimension.

Figure 1: CFA result of the overall measurement model

IJRAR19J3057 International Journal of Research and Analytical Reviews (IJRAR) www.ijrar.org 1110
© 2019 IJRAR March 2019, Volume 6, Issue 1 www.ijrar.org (E-ISSN 23481269, P- ISSN 2349-5138)

5. Discussion
In measurement scales, both reliability and validity are critically relevant issues. The objective of this article is to test the reliability and
validity of the HRM measurement scale developed by Vermeeren (2014) in Ethiopian context. The finding of this study confirmed that the
measurement scale is both reliable and valid, which can be applicable in Ethiopian context.
In terms of reliability analysis, the six dimensions of the measurement scale have Cronbach’s alpha greater than the threshold, ranging from
0.77 to 0.92, indicating that the measurement instrument has high level of internal reliability. It is widely known that Cronbach’s alpha is not
enough to measure the reliability of constructs (Hair et al, 2010). In this respect, the composite reliability is considered as part of the reliability
measurement. In view of that, the composite reliability for each constructs are above 0.7 and this implies that the instrument is reliable enough
to be applicable for in other contests. Regarding the validity analysis, the finding of this study established that construct validity was
confirmed. As it is displayed in table 6 above, the average variance extracted and composite reliability of each of the six dimensions of the
measurement scale are greater than 0.5 and 0.7 respectively, which indicates that convergent validity is confirmed. This implies that the
indicators are significantly and positively related with the constructs. That means the six dimensions of the measurement scale are uniquely
related to each other. So, this study contributes to the HRM literature as most of the study were done in western countries where there is
limited validity in developing countries particularly in the public sector.
In the HRM-performance debate, prominent scholars (Boselie et al, 2005; Combs et al, 2006) argued that there is no uniformity of research
results in the HRM –performance equation because different researchers utilized different types of measurement scale. This implies that it is
not possible to compare the results of the various studies done so far. Different scholars utilized different HR practices and instruments,
which is believed as the major challenge to build an appropriate HRM model especially in the public sector. Following this, some HRM
scholars (Boxall and Purcell, 2008; Lepak et al., 2006; Jiang et al, 2012) introduced AMO as a theoretical framework in selecting the HR
practices instead of randomly selecting the practices. The idea behind the Ability, Motivation and Opportunity (AMO) model is that to
examine the contribution of HRM, at least the practices to be employed in research studies has to be enable employees’ ability of employees,
motivation and opportunity to contribute at workplace. In this respect, the adopted HRM measurement scale is in line with AMO theory and
contains the basic HR practices in the public sector.
6. Conclusion
The study established that the HRM measurement scale is both reliable and valid. In terms of reliability and validity analysis, the 31 item
HRM measurement scale consisting of six factors (Recruitment and Selection; Training and Development; Performance Appraisal;
Compensation and Reward; Autonomy and Employee Participation) originally developed by Vermeeren (2014) in public sector of
Netherlands is applicable in Ethiopian context. Thus, the finding of this study enable future researchers to utilize the instrument in
investigating the role of Human Resource Management practices in improving performance of organizations in the public sector.
7. Reference

[1] Ahmad, S. and Schroeder, R. G. (2003). The impact of human resource practices on operational performance: Recognizing country
and industry differences. Journal of Operations Management, 21, 19-43.
[2] Appelbaum, E., Bailey, T., Berg, P., and Kalleberg, A. (2000). Manufacturing advantage:Why high performance work systems pay
off. Ithaca, NY: ILR Press.
[3] Armstrong, M. and Taylor (2014).Armstrong’s handbook of human resource management practice (13th ed.). London: Kogan Page.
[4] Armstrong,M. (2009).Armstrong’s handbook of human resource management practice(11 th ed.). London: Kogan Page.
[5] Arthur, J.B. (1994), “Effects of human resource systems on manufacturing performance and turnover”, Academy of Management
Journal, 37(3), 670-87.
[6] Batt, R. (2002). Managing customer services: Human resource practices, quit rates, and sales growth. Academy of Management
Journal, 45(3): 587-597.
[7] Boselie, P., Dietz, G. and Boon, C. (2005). ‘Commonalities and contradictions in HRM and performance research’. Human Resource
Management Journal, 15, 67–94.
[8] Boxall, P., and Macky (2009). Research and Theory on High-Performance Work Systems: Progressing the High-Involvement
Stream, HRM Journal 19(1):3 - 23 ·
[9] Boxall, P., and Purcell, J. (2008). Strategy and human resource management (2nd ed.). Basingstoke: Palgrave Macmillan.
[10] Byrne, B. M. (2016). Structural Equation Modeling with AMOS: Basic Concepts, Applications, and Programming (3rd Edition).
New York: Routledge.
[11] Combs, J., Liu Y., Hall, A., & Ketchen, D. (2006). How much do high-performance work practices matter? A meta-analysis of their
effects on organizational performance. Personnel Psychology, 59(3), 501-528.
[12] Delaney, J.T and Huselid, M. A. (1996). The impact of human resource management practices on perceptions of organizational
performance. Academy of Management Journal, 39: 919–969.
[13] Demo, Gisela, Neiva, Elaine Rabelo, Nunes, Iara, & Rozzett, Kesia. (2012). Human resources management policies and practices
scale (HRMPPS): Exploratory and Confirmatory Factor Analysis. BAR - Brazilian Administration Review, 9(4), 395-420.
[14] Dyer, L. and Reeves, T. (1995). ‘Human resource strategies and firm performance: what do we know and where do we need to go?’
International Journal of Human Resource Management, 6(3), 656–670
[15] Gould-Williams, J. S. (2003). The importance of HR practices and workplace trust in achieving superior performance: A study of
public-sector organizations. International Journal of Human Resource Management, 14(1), 28-54.
[16] Hair, J., Black, W., Babin, B., and Anderson, R. (2010). Multivariate data analysis (7th ed.): Prentice-Hall, Inc. Upper Saddle River,
NJ, USA.
IJRAR19J3057 International Journal of Research and Analytical Reviews (IJRAR) www.ijrar.org 1111
© 2019 IJRAR March 2019, Volume 6, Issue 1 www.ijrar.org (E-ISSN 23481269, P- ISSN 2349-5138)
[17] Huselid, M. A. (1995). The impact of human resource management practices on turnover, productivity, and corporate financial
performance. Academy of Management Journal, 38(3): 635–672.
[18] Jiang, K., Lepak, D. P., Hu, J., & Bear, J. C. (2012). How does human resource management influence organizational outcomes? A
meta-analytic investigation of mediating mechanism. Academy of Management Journal, 55(6), 1264–1294
[19] Kaifeng Jiang & Jake Messersmith (2017): On the shoulders of giants: a meta-review of strategic human resource management, The
International Journal of Human Resource Management, DOI: 10.1080/09585192.2017.1384930
[20] Kehoe, R. and Wright, P.M. (2013). The impact of high-performance human resource practices on employee attitudes and behavior.
Journal of Management, 39(2), 366-391
[21] Khan, A. M. (2010). Effects of human resource management practices on organizational performance: An empirical study of oil and
gas industry in Pakistan. European Journal of Economics, Finance and Administrative Sciences, 24, 157-174.
[22] Kim, S. (2005). Individual-Level Factors and Organizational Performance in Government Organizations. Journal of Public
Administration Research and Theory, 15(2), 245-261.
[23] Kline, R. B. (2016). Principles and practice of structural equation modeling (4rd ed.). New York: The Guilford Press.
[24] Lepak, D., Liao, H., Chung, Y. and Harden, E. (2006). ‘A conceptual review of human resource management systems in strategic
human resource management research’. Research in Personnel and Human Resources Management, 25, 217–271.
[25] Markos, S., & Sridevi, M. S. (2010). Employee engagement: The key to improving performance. International Journal of Business
and Management, 5(12), 89-96.
[26] Ministry of Public Service and Human Resource Development Statistical Report (2016).
[27] Paauwe, J. (2009). HRM and performance: Achievements, methodological issues and prospects. Journal of Management Studies,
46(1), 129-142.
[28] Vermeeren B. (2014). Variability in HRM Implementation among Line Managers and its Effect on Performance: A 2-1-2
Meditational Multilevel Approach, International Journal of Human Resource Management, DOI:10.1080/09585192.2014.934891
[29] Wall, T. D., & Wood, S. J. (2005). The romance of human resource management and business performance, and the case of big
science. Human Relations, 58(4), 429-462
[30] Wright, P., Gardner, T., Moynihan, L. & Allen, M. (2005). The relationship between HR practices and firm performance: examining
causal order. Personnel Psychology, 58(2), 409-446.
[31] Zainudin Awang (2012). Research Methodology and Data Analysis (2nd) Edition.

IJRAR19J3057 International Journal of Research and Analytical Reviews (IJRAR) www.ijrar.org 1112

Anda mungkin juga menyukai