Index

Abstract

The performance of public healthcare facilities is critical, due to the impact on human lives. However, in Punjab, the infant and maternal mortality rates are 88 per 1000 and 227 per 100,000, respectively, while Pakistan ranks 149th in healthcare worldwide. Against these figures, the targets set in the 2018 Public Health Sector Plan, Building a Healthier Punjab seem overoptimistic. Hence, this study aimed to determine the mediating effect of integrated systems on the relationship between supply chain management practices and healthcare performance. Adopting a quantitative methodology, a survey questionnaire administered to 200 respondents, selected through cluster sampling from a target population 2899 in Punjab. SPSS and AMOS were used for the exploratory and confirmatory factor analyses and structural equation modeling. Following validation of the measurement model, structural equation modeling found integrated systems exerted a significant and full mediating on the supply chain management practices–healthcare performance relationship. The indications are, therefore, that integrated systems and efficient supply chain management practices enhance patients care while minimizing healthcare costs. These findings should be useful to both public and private healthcare facilities, as well as other public organizations and supply chain professionals, by providing a fuller understanding of the issues. Future research studies could further broaden this knowledge through an investigation into the impact of backlog inventories on financial performance.

Keywords: Hospitals, Information technology (IT), Inventory control, Organizational performance, Healthcare management, Service quality, Stockout, Structural equation modeling (SEM), AMOS.

Received: 4 June 2020 / Revised: 7 July 2020 / Accepted: 10 August 2020/ Published: 2 September 2020

Contribution/ Originality

This study is one of very few investigating the mediating effect of integrated systems on the relationship between supply chain management practices and healthcare performance. The study also contributes a second-order construct model, identifying the dimensions within those constructs, to the existing literature.

1. INTRODUCTION

The performance of public (not-for-profit) healthcare facilities is fundamentally different from the private healthcare sector (Cheon, 2016): there are 13 million deaths per year worldwide due to infectious diseases, and higher numbers are common in developing countries (Rashid, Amirah, & Yusof, 2019). Therefore, public healthcare facilities purchase and stockpile the most expensive medications for future consumption, despite the risks of using out-of-date or obsolescent medicines and being unable to replenish out-of-stock items (McKelvey, 2013). For instance, with a population of 101.6 million, the State Government of Punjab allocated 168 billion Pakistani rupees to primary and secondary healthcare facilities, yet the mortality rate continues to rise due to expired and out-of-stock critical medicines: infant mortality increased from 78 per 1000 in 2008 to 88 per 1000 in 2016; maternal mortality rate was 227 per 100,000 in 2006–07, while Pakistan’s ranked 149th in the World Health Organization’s (WHO) 2016 Maternal Mortality Ratio Index. In fact, it was discovered that manual errors and poor inventory control in most healthcare facilities led to patients being prescribed expired medications (Rashid & Amirah, 2017). The goal of the 2018 Public Health Sector Plan, Building a Healthier Punjab, to bring the maternal mortality rate below 140 per 100,000 and infant mortality rate below 40 per 1000 thus seems overoptimistic (Rashid et al., 2019).

1.1. Research Question and Objectives

This study aimed to discover the strategic importance of supply chain management practices (SCMP), integrated systems (IS) healthcare performance (HP) in healthcare facilities of Punjab, Pakistan. To answer this question, empirical evidence was sought for each of the following objectives:

  1. To determine the significant positive effect between SCMP and IS.
  2. To determine the significant positive effect between SCMP HP.
  3. To determine the significant positive effect between IS and HP.
  4. To determine the mediating effect of IS between SCMP and HP.

2. THEORETICAL FRAMEWORK

A quantitative research method was used to examine the three variables: SCMP, IS mediator, and HP (Creswell, 2014).  The study then adopted the resource-based view (RBV) theory to test the hypothesized relationships between these variables.

According to Peteraf and Barney (2003), RBV theory evaluates resources and performance divergence at an organizational level, with an emphasis on those resources and capabilities that improve efficiency. According to Porter (1981), those that contribute either indirectly or directly to enhancing organizational performance are viable resources (Shibin, Gunasekaran, & Dubey, 2017), which involve technology, effective processes, employee knowledge, machinery, and capital (Wernerfelt, 1984). In fact, Wade and Hulland (2004) proved that as technology improves processes, it indirectly adds value to an organization. Consequently, RBV theory provides an appropriate theoretical framework (Hsu, Tan, Kannan, & Keong Leong, 2009), as already confirmed by resource synergy in association with SCMP (Lee & Hughey, 2001).

3. LITERATURE REVIEW

Over time, SCM has become pivotal in every organization, but, as it involves immense capital and affects performance, requires managerial oversight (Samuel & Ondiek, 2014). Performance can be enhanced through increasing value, reducing costs, and monitoring operational expenses (Pritchard, Gracy, & Godwin, 2010); moreover, economies of scale, cost-effectiveness, and efficiency are essential in a competitive environment. In this case, SCMP provides a sustained and efficient supply (Baily, David, David, & David, 2005), although the failure to link the financial and functional units of SCMP will ruin performance (Shah & Shin, 2007). However, Dhodi (2018) empirically proved that IS greatly improve planning, scheduling, and production, and thus enhance stock availability and eliminate shortages. Many researchers believe integration cuts costs and develops customer care (Samuel & Ondiek, 2014), and more attention is being paid to how integration can result in benefits for enterprises (Attom, 2013).

Consequently, a model was developed based on relevant literature and RBV theory (Hsu et al., 2009; Samuel and Ondiek, 2014)), as depicted in Figure 1. Here, there is one independent variable (SCMP), one dependent variable (HP), and one mediator (IS) with which to examine the mediation effect on the relationship between the independent and dependent variables.

Figure-1. Proposed research model.

Source:  Hsu et al. (2009); Samuel and Ondiek (2014).

3.1. Research Hypotheses

Based on the proposed research model, the following hypotheses were developed:
H1: SCMP exerts a significant positive effect on IS.
H2: SCMP exerts a significant positive effect on HP.
H3: IS exerts a significant positive effect on HP.
H4: IS exerts a mediation effect on the relationship between SCMP and HP.

4. DATA ANALYSIS AND RESULTS

A positivist paradigm is used in the current study, as it is appropriate for quantitative research (Creswell, 2014), which is suitable for empirical investigations (Given & Saumure, 2008). A questionnaire was administered to collect data on the three variables through a 5-point Likert scale (Amirah, Asma, Muda, Amin, & Him, 2019). Exploratory factor analysis (EFA) followed by confirmatory factor analysis (CFA) were then performed, using IBM® SPSS® V22.0 and IBM® SPSS® AMOS™ 22, respectively, to validate a second-order construct model before testing the structural equation modeling (SEM). According to Worthington and Whittaker (2006),  this two-stage process involving different sample sizes acts as a scale validation procedure. Therefore, the EFA was conducted on 100 and the CFA on 200 respondents selected via multistage cluster sampling, dividing the population into mutually homogeneous but internally heterogeneous groups (Sekaran & Bougie, 2010). Hence, one district from each of the nine Punjabi divisions provided a sampling frame of 343 healthcare facilities for this study, from which 200 were randomly selected (Lomax & Schumacker, 2004).

4.1. Exploratory Factor Analysis (EFA)

For the EFA, orthogonal rotation (i.e., Varimax) with principal axis factoring on 28 items, was used (Tabachnick & Fidell, 2014). Table 1 presents the summarized results of the rotated factor matrix for each item (from highest to lowest), of which three with factor loadings < 0.60 (SCMP4, HP6, and HP5) were excluded, leaving 25 items. Eventually, two factors for each variable with cross-loadings < 75% of any other item were reached (Field, 2013; Hair et al., 2010). Furthermore, factors were extracted on the basis of the eigenvalues-greater-than-1.0 rule (Hair,  Black, Babin, & Anderson, 2010; Rashid et al., 2019).

Certain statistical assumptions were considered in the EFA. Kaiser–Meyer–Olkin (KMO) values > 0.60 were considered excellent: SCMP = 0848, IS = 0.852, and HP = 0.800. Bartlett’s test of sphericity should be significant and was shown to be p < 0.001 (Awang, 2014; Beavers, Lounsbury, Richards, & Huck, 2013; Field, 2013). Communalities value should be > 0.2 for each item and did indicate that additional factors were not signified for each item (Child, 2006). The total variance should be > 50%: first two factors each of SCMP = 61%, IS = 60%, and HP = 61%; moreover, the variance for the first factor of each was < 50% (Yusof, Awang, Jusoff, & Ibrahim, 2017). As a result, the final factors extracted via EFA were appropriately renamed.

Table-1. Summarized results of rotated factor matrix.

Items
SCMP Factors
IS Factors
HP Factors
1
2
1
2
1
2
SCMP1
0.806
SCMP2
0.798
SCMP5
0.797
SCMP3
0.720
SCMP4
0.477
SCMP7
0.824
SCMP8
0.821
SCMP6
0.781
SCMP9
0.762
IS2
0.825
IS1
0.808
IS4
0.791
IS3
0.655
IS7
0.778
IS6
0.727
IS5
0.713
IS8
0.678
HP9
0.846
HP8
0.834
HP10
0.750
HP11
0.746
HP7
0.698
HP6
0.491
HP3
0.949
HP4
0.887
HP2
0.816
HP1
0.759
HP5
0.470

Notes: Extraction method: principal axis factoring; rotation method: varimax (orthogonal) rotation; factor loadings < 0.60 were excluded from further analysis; rotation converged in three iterations.

4.2. Demographic Attributes of Respondents

Table-2. Demographic attributes of respondents.


 Demographic Profile
Category
Frequency (n = 200)
%
Marital Status
Single
154
77
Married
46
23
Age (years)
20–30
141
70.5
31–40
51
25.5
41–50
8
4
Gender
Male
183
91.5
Female
17
8.5
Years of Experience
15+
15
7.5
11–15
33
16.5
6–10
53
26.5
2–5
87
43.5
< 2
12
6
Educational Level
High School
23
11.5
Diploma
52
26
Bachelors
94
47
Masters
17
8.5
Other
14
7

Notes: n: number of respondents.

Table 2 shows the demographic attributes of the respondents: a far greater proportion was single (154, 77%) than married (46, 23%); the majority were aged 20–30 years (141, 70.5%), then 31–40 (51, 25.5%) and 41–50 (8, 4%); almost all were male (183, 91.5%), with very few females (17, 8.5%); their experience, in descending order of numbers, ranged from 87 (43.5%) with 2–5 years, 53 (26.5%) with 6–10 years, 33 (16.5%) with 11–15 years, 15 with over 15 years, and 12 (6%) less than 2 years experience; and finally, 23 (11.5%) completed high school, while 52 (26%), 94 (47%) and 17 (8.5%) earned their diploma, bachelor’s, and master’s degrees, respectively, with 14 (7%) receiving other qualifications.

4.3. Descriptive Analysis

The results from the descriptive analysis of each item and variable are shown in Table 3. The means (M) and standard deviations (SD) for SCMP (M = 4.02, SD = 0.760), IS (M = 3.91, SD = 0.834), and HP (M = 4.01, SD = 0.849) revealed that the majority of respondents “agreed” on the essential role played by SCMP and IS in HP.

Table-3. Descriptive analysis results.


 Demographic Profile
Category
Frequency (n = 200)
%
Marital Status
Single
154
77
Married
46
23
Age (years)
20–30
141
70.5
31–40
51
25.5
41–50
8
4
Gender
Male
183
91.5
Female
17
8.5
Years of Experience
15+
15
7.5
11–15
33
16.5
6–10
53
26.5
2–5
87
43.5
< 2
12
6
Educational Level
High School
23
11.5
Diploma
52
26
Bachelors
94
47
Masters
17
8.5
Other
14
7

Notes: M: mean; SD: standard deviation; EDI: electronic data interchange; EPOS: electronic point of sale; ERP: enterprise resource planning.; InvS: inventory stocks; InvA: inventory accuracy; IMS: information management system; ITC: collaboration; HPS: hospital performance service quality; HPC: hospital performance cost.

4.4. Measurement Model Assessment

The assessment model involved a second-order construct model comprising three constructs and six sub-constructs, with which a pooled CFA was undertaken, being able to address any identification issues—even if a construct consists of less than four indicators (Awang, 2015). Figure 2 depicts the measurement model for assessing unidimensionality, validity, and reliability prior to SEM (Henseler, Ringle, & Sarstedt, 2014).

Figure-2. Measurement model for assessment.

Note: Chisq/df: chi square/degrees of freedom.

4.4.1. Unidimensionality

According to Hair et al. (2010), unidimensionality is a set of dummy variables that explain a construct. First, maximum likelihood estimation (MLE) of factor loadings was conducted (Awang, 2015), and then the two items (InvS4 and HPS5) calculated as < 0.50 were removed (see Figure 3) (Hair et al., 2010). The factor loadings of all the other items were high (> 0.70), confirming there were similarities between indicators within a single construct (Hair, Hult, Ringle, & Sarstedt, 2013).

Figure-3. Measurement model after achieving unidimensionality.

Validity and reliability were then assessed through average variance extracted (AVE) and composite reliability (CR). As the results sufficed the recommended levels, those items with factor loadings between 0.40 and 0.70 did not need to be removed (Hair, Ringle, & Sarstedt, 2011).

4.4.2. Validity

Three types of validity tests were conducted on the measurement model: construct, convergent, and discriminant (Henseler et al., 2014).

4.4.2.1. Construct Validity and Reliability

Construct validity measures accuracy through three categories of goodness-of-fit indices: parsimonious, incremental, and absolute fit (Hair et al., 2010). The most cited and recommended indices are listed in Table 4; Figure 3 also suggests indicators specific to the measurement model for each of these indices. All the indicators in each category confirmed goodness of fit for the model: AGFI = 0.85 ≥ 0.80 and RMSEA = 0.05 ≤ 0.08 verified absolute fit; TLI, IFI, and CFI = 0.97 ≥ 0.90 and NFI = 0.92 ≥ 0.90 validated incremental fit; while parsimonious fit was proved by chisq(x2)/df = 1.52 (1.00–5.00) and PNFI = 0.80 > 0.05. Nevertheless, various other tests were necessary, which are detailed later in this paper.

Table-4. Goodness of fit indices for measurement model.

Category Index Level Cited
Absolute fit AGFI ≥ 0.80 Chau and Hu (2001); Jöreskog and Sörbom (1993)
RMSEA ≤ 0.08 Hair et al. (2010)
Incremental fit TLI ≥ 0.90 Bagozzi and Yi (1988); Tucker and Lewis (1973)
IFI ≥ 0.90 Bollen (1990)
NFI acceptable ≥ 0.80 Bentler and Bonett (1980)
CFI ≥ 0.90 Bagozzi and Yi (1988); Hair et al. (2010); Byrne (2010)
Parsimonious fit x2/df 1.00–5.00 Kline (2010)
PNFI > 0.05 Bentler and Bonett (1980)

Notes: AGFI: adjusted goodness of fit index; RMSEA: root mean square error of approximation; TLI: Tucker–Lewis index; IFI: incremental fit index; NFI: normed fit index; CFI: comparative fit index; x2/df: chi square/degrees of freedom; PNFI: parsimony normed fit index. At least one fitness index should be selected from each category (Hair et al., 2010).

Figure-4. Structural equation modeling.

Table-5. Validity and reliability test results for full model.

Construct
Sub-Constructs and Indicators
α
Std β
CR
AVE
SCMP
InvA:
0.932
0.894
0.947
0.721
InvA5
0.825
InvA6
0.862
InvA7
0.850
InvA8
0.848
InvS:
0.959
InvS1
0.857
InvS2
0.891
InvS3
0.807
IS
IMS:
0.923
0.853
0.954
0.722
IMS1
0.807
IMS2
0.846
IMS3
0.760
IMS4
0.879
ITC:
0.829
ITC5
0.852
ITC6
0.863
ITC7
0.852
ITC8
0.821
HP
HPC:
0.941
0.877
0.960
0.751
HPC1
0.858
HPC2
0.850
HPC3
0.796
HPC4
0.919
HPS:
0.905
HPS6
0.784
HPS7
0.863
HPS8
0.895
HPS9
0.952

Notes: α: Cronbach’s alpha; Std β: standardized estimate; CR: composite reliability; AVE: average variance extracted. AVE for the second-order model = averaging the squared multiple correlations for the first-order indicators.

In addition, CR and Cronbach’s α were calculated for each construct, as shown in Table 5, the results of > 0.70 for each coefficient confirming the reliability of all three constructs (Hair et al., 2010).

4.4.2.2. Convergent Validity

Convergent validity is the degree to which a measure correlates positively with alternative measures of the same construct (Hair et al., 2013) and is tested by calculating AVE. The recommended value of ≥ 0.50 (Hair et al., 2010) shows that, on average, the construct describes more than half of the variance in its indicators (Hair et al., 2013), which the results in Table 5 satisfy for all constructs.

4.4.2.3. Discriminant Validity

Discriminant validity evaluates the latent constructs for differentiates (Hair et al., 2010) by comparing the AVE with its squared correlation for each construct (Henseler et al., 2014). Table 6 indicates that there were multicollinearity issues with the measurement model, as the diagonal values are higher than those in columns and rows, which are all < 0.85, as recommended by Awang (2015).

Table-6. Discriminant validity index.

Construct
SCMP
IS
HP
SCMP
0.849
IS
0.741
0.850
HP
0.628
0.776
0.867

Notes: Diagonally, figures represent the square root of the AVE; otherwise, they represent the squared correlations.

4.4.3. Normality Test

According to Awang (2015), robust measures of normality are skewness and kurtosis, with recommended values between -1.0 and +1.0. The results for a sufficiently sized sample of 200 in the current study fell within the recommended range, with a critical region (CR) of < 8.0 for skewness and < 3.0 for kurtosis. As the data distribution proved to be normal, it was considered acceptable for SEM (Awang, 2015).

4.5. Structural Equation Modeling (SEM)

4.5.1. Hypothesis Testing: Direct Hypotheses H1, H2, and H3

The direct hypotheses were supported by SEM (Figure 4). Table 7 indicates that SCMP affects IS and HP positively (H1 and H2), and likewise IS on HP (H4). Moreover, the empirical values revealed that IS exerted a stronger effect than SCMP on HP.

Table-7. Structural path analysis results (direct hypotheses).

Hypothesis
Path
Stdβ
SE
CR
Results
H1
IS <-- SCMP
0.742***
0.085
8.055
Supported
H2
HP <-- SCMP
0.613***
0.085
6.226
Supported
H3
HP <-- IS
0.779***
0.110
7.819
Supported

Notes: SE: standard error; CR: critical ratio; ***, p < 0.001.

4.5.2. Mediation Assessment: Hypothesis H4

The method suggested by both Field (2013) and Awang (2014) was followed to assess the mediation effect: path c (the direct effect) was examined before path c’ (the indirect effect). The standardized estimates for paths a, b, and c were statistically significant, as can be seen in Table 8, revealing the existence of a mediation effect. Once the mediator (IS) was introduced the standardized estimate of 0.0613 for path c was abridged to 0.116 for path c′ and was no longer significant, indicating full mediation had occurred. Furthermore, the results fulfilled Awang’s (2014) criterion for a mediation effect: multiplying paths a and b produced a value greater than path c’ that was not significant, confirming not only mediation but full mediation. The indirect hypothesis (H4) was thus supported.

Table-8. Mediation assessment results (IS in SCMP and HP).

Hypothesis
Path
Relationship
Std β
SE
CR
H4
Path c
HP
<--
SCMP
0.613***
0.085
6.226
Path a
IS
<--
SCMP
0.741***
0.082
7.986
Path b
HP
<--
IS
0.690***
0.149
4.924
Path c’
HP
<--
SCMP
0.116 (ns)
0.111
0.979
path a x path b
0.511 (ns)
Full Mediation since direct effect (c′) is not significant
path a x path b > path c’
Mediation occurs
Notes: ***, p < 0.001; (ns): not significant.

4.5.3. Bootstrapping: Direct and Indirect Effects

Bootstrapping is another highly recommended technique to examine the mediation effect (Awang, 2014; Preacher and Hayes, 2004). The bootstrapping results in Table 9 shows the standardized estimate (0.512) of the indirect effect to be statistically significant, while the 95% percent boot CI does not cross 0 (LL = 0.316, UL = 0.926), which demonstrates a mediation effect. In addition, the non-significant standardized estimate of the direct effect illustrates a full mediation effect verifies that H4 is supported.

Table-9. Bootstrapping results.

Hypothesis
Path
Std β (DE)
Std β (IE)
LL
UL
Decision
H4
SCMP --> IS --> HP
0.116 (ns)
0.512***
0.316
0.926
Supported

Notes: DE: direct effect; IE: indirect effect; LL: lower limit of confidence interval; UL: upper limit of confidence interval; ***, p < 0.001.

4.6. Coefficient of Determinants (R²)

The coefficients of determination suggested that 61% of HP could be predicted through SCMP and IS, while 55% of IS could be predicted through SCMP. According to Cohen (1988) and Urbach and Ahlemann (2010), a substantial R² value increases the predictive ability of the model, and as can be seen from Table 10, the R2 values in the current study can be considered substantial based on (Cohen (1988) and Falk and Miller 91992).

Table-10. Recommended substantial values of R2.

 
≥ 0.26
≥ 0.10
≥ 0.65
≥ 0.75

Notes: Most cited substantial R2 values.

4.7. Effect Size (f ²)

Effect size evaluates how adequate the effect of latent exogenous constructs are on a latent endogenous construct by first omitting a relevant exogenous construct and then estimating the path model twice (f2 = (R2 included – R2 excluded) / (1 – R2 included) (Gefen & Rigdon, 2011; Hair et al., 2010). As Cohen (1988) states that effect size ≥ 0.02 is small, ≥ 0.15 is medium, and ≥ 0.35 is large, the current study found a large effect size (0.59) for SCMP on HP.

5. DISCUSSION

Having developed a second-order construct model and validated it for structural equation modeling, the three main latent constructs were examining in terms of the study’s objectives:

Objective 1: This was addressed by testing hypothesis H1. Results from SEM found that SCMP did exert a significant positive effect on IS, as well as proportionate to each other. This finding is consistent with Ondari and Muturi (2016) and Rashid (2016).

Objective 2: This was addressed by testing hypothesis H2. The SEM results found that SCMP did exert a statistically significant positive effect on HP: improving SCMP will enhance HP. This finding is supported by those of Peng, Quan, Zhang, and Dubinsky (2016) and Sutduean, Joemsittiprasert, and Jermsittiparsert (2019).

Objective 3: Testing hypothesis H3 through SEM revealed that IS exerted a statistically significant positive effect on HP: improvements to IS will improve HP. This finding is similar to those of Gurmu and Ibrahim (2017) and Ahmad and Zabri (2018).

Objective 4: Using SEM to test hypothesis H4 determined a significant and full mediating effect on the relationship between SCMP and HP. This confirmed the findings of Rehman, Mohamed, and Ayoup (2019).

6. CONCLUSION

6.1. Research Contribution and Implications

State-funded hospitals play an important role in public health; however, they are susceptible to failure due to inventory maladministration in the pharmacy and laboratory. Thus, relevant research on healthcare facilities in Punjab was required, and SCMP, IS, and HP were integrated into a second-order construct model based on RBV theory.

The findings contribute to both the body of literature and RBV theory to develop a more comprehensive understanding of healthcare performance. The main focus of this study follows the global trend in research: the mediating role played by integrated systems in supply chain management and organizational performance. However, the concept still lacks management support in the public sector, especially the healthcare sector in Pakistan. Consequently, the findings from the current study will assist public healthcare facilities, as well as pharmaceutical companies, multifunctional organizations, and government bodies, to adopt integrated systems that enable the implementation of supply chain management practices to enhance organizational performance.

6.2. Research Limitations and Recommendations for Future Research

Integrated systems act as a mediator between supply chain management practices and healthcare performance to produce positive and significant effects in Punjab, Pakistan; without them, healthcare facilities are unable to maintain accurate inventories of medicines and laboratory equipment, which could lead to higher mortality rates. However, this study had several limitations: only three variables were examined, and future studies should test either other variables or different dimensions of the same variables; the data was based on perceptions rather than absolute values; the study was restricted solely to Punjab; and the researchers worked within a limited time frame and budget, and faced restricted access to relevant departments and poor recordkeeping at healthcare facilities.

Finally, it is recommended that future research studies should consider stakeholders’ input in the decision-making process and include respondents from community healthcare, the pharmaceutical industry, and the private sector for more comprehensive results. In addition, researchers could investigate the relationship between backlog inventories, procurement strategies, and operational performance.

Funding: This study received no specific financial support.  

Competing Interests: The authors declare that they have no competing interests.

Acknowledgement: All authors contributed equally to the conception and design of the study.

REFERENCES

Ahmad, K., & Zabri, S. M. (2018). The mediating effect of knowledge of inventory management in the relationship between inventory management practices and performance: The case of micro retailing enterprises. Journal of Business and Retail Management Research, 12(2), 83-93.

Amirah, N. A., Asma, W. I., Muda, S., Amin, A., & Him, N. F. N. (2019). Analysis of individual factors on employees’ perception towards safety culture in the Malaysian Manufacturing Industry. Paper presented at the In 1st Aceh Global Conference (AGC 2018). Atlantis Press.

Attom, B. E. (2013). The impact of information communication technology (ICT) on business growth strategies of small and medium-scale enterprises (SMEs) in the Awutu-Senya East Municipality of Central Region of Ghana. Asian Journal of Business and Management Sciences, 3(2), 13-28.

Awang, Z. (2014). A handbook of SEM for academicians and practitioners. The step by step practical guides for the beginners. Bangi: MPWS Rich Publication.

Awang, Z. (2015). SEM made simple. Bangi: MPWS Rich Publication.

Bagozzi, R. P., & Yi, Y. (1988). On the evaluation of structural equation models. Journal of the Academy of Marketing Science, 16(1), 74-94.

Baily, P., David, F., David, J., & David, J. (2005). Purchasing principles and management (9th ed.). England: Prentice Hall.

Beavers, A. S., Lounsbury, J. W., Richards, J. K., & Huck, S. W. (2013). Practical considerations for using exploratory factor analysis in educational research. Practical Assessment, Research, and Evaluation, 18(1), 1-13.

Bentler, P. M., & Bonett, D. G. (1980). Significance tests and goodness-of-fit in analysis of covariance structures. Psychological Bulletin, 88(3), 588–606.Available at: https://doi.org/10.1037/0033-2909.88.3.588.

Bollen, K. A. (1990). Overall fit in covariance structure models: Two types of sample size effects. Psychological Bulletin, 107(2), 256-259.Available at: https://doi.org/10.1037/0033-2909.107.2.256.

Byrne, B. M. (2010). Structural equation modeling with AMOS (2nd ed.). New York: Routledge.

Chau, P. Y. K., & Hu, P. J. H. (2001). Information technology acceptance by individual professionals: A model comparison approach. Decision Sciences, 32(4), 699-719.Available at: https://doi.org/10.1111/j.1540-5915.2001.tb00978.x.

Cheon, O. (2016). Management and performance in U.S. healthcare institutions: Do sector-differences matter? (Order No. 10291118). Pro Quest Dissertations & Thesis Global.  

Child, D. (2006). The essentials of factor analysis (3rd ed.). New York: Continuum.

Chin, W. W. (1998). Issues and opinion on structural equation modeling. MIS Quarterly, 22(1), 7-16.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates.

Creswell, J. W. (2014). Research design: Qualitative, quantitative and mixed methods approaches (4th ed.). Thousand Oaks: SAGE Publication Inc.

Dhodi, M. H. (2018). The effect of information technology on inventory management for the manufacturing companies in Mogadishu. European Journal of Purchasing & Supply Chain Management, 6(3), 20-29.

Falk, R. F., & Miller, N. B. (1992). A primer for soft modeling. Akron, OH: University of Akron Press.

Field, A. (2013). Discovering statistics using SPSS (4th ed.). London: SAGE.

Gefen, D., & Rigdon, E. E. (2011). An update and extension to SEM guidelines for administrative and social science research. MIS Quarterly, 35(2), 1-7.

Given, L. M., & Saumure, K. (2008). The Sage encyclopedia of qualitative research methods. Thousand Oaks: Sage Publications.

Gurmu, T. G., & Ibrahim, A. J. (2017). Inventory management performance of key essential medicines in health facilities of east Shewa zone, Oromia regional state, Ethiopia. Cukurovs Med Journal, 42(2), 277-291.

Hair, J. F. J., Black, W. C., Babin, B. J., & Anderson., R. E. (2010). Multivariate data analysis. A global perspective (7th ed.). USA: Prentice Hall.

Hair, J. F., Hult, G. T. M., Ringle, C. M., & Sarstedt, M. (2013). A primer on partial least squares structural equation modeling (PLS-SEM). CA: Sage Publications.

Hair, J. F., Ringle, C. M., & Sarstedt, M. (2011). PLS-SEM: Indeed a silver bullet. Journal of Marketing theory and Practice, 19(2), 139-152.

Henseler, J., Ringle, C. M., & Sarstedt, M. (2014). A new criterion for assessing discriminant validity in variance-based structural equation modeling. Journal of the Academy of Marketing Science, 43(1), 115–135.Available at: http://doi.org/10.1007/s11747-014-0403-8 .

Hsu, C.-C., Tan, K.-C., Kannan, V. R., & Keong Leong, G. (2009). Supply chain management practices as a mediator of the relationship between operations capability and firm performance. International Journal of Production Research, 47(3), 835-855.Available at: https://doi.org/10.1080/00207540701452142.

Jöreskog, K., & Sörbom, D. (1993). LISREL 8: Structural equation modeling with the SIMPLIS command language. Chicago, IL.: Scientific Software International Inc.

Kline, R. B. (2010). Principles and practice of structural equation modelling (4th ed.). New York: The Guilford Press.

Lee, H.-Y., & Hughey, K. F. (2001). The relationship of psychological separation and parental attachment to the career maturity of college. Journal of Career Development, 27(4), 279-293.Available at: https://doi.org/10.1177/089484530102700404.

Lomax, R. G., & Schumacker, R. E. (2004). A beginner's guide to structural equation modeling (3rd ed.). New York: Psychology Press.

McKelvey, F. A. (2013). Efficiency from pharmacy inventory technology: A means to improving quality in non-profit hospitals. Doctoral Dissertation, The University of Texas School of Public Health.  

Ondari, L. M., & Muturi. (2016). Factors affecting the efficiency of inventory management in organizations in Kenya. International Journal of Economics, Commerce and Management United Kingdom, 4(4), 1019-1034.

Peng, J., Quan, J., Zhang, G., & Dubinsky, A. J. (2016). Mediation effect of business process and supply chain management capabilities on the impact of IT on firm performance: Evidence from Chinese firms. International Journal of Information Management, 36(1), 89-96.Available at: https://doi.org/10.1016/j.ijinfomgt.2015.09.006.

Peteraf, M. A., & Barney, J. B. (2003). Unraveling the resource-based tangle. Managerial and Decision Economics, 24(4), 309-323.Available at: https://doi.org/10.1002/mde.1126.

Porter, M. E. (1981). The contributions of industrial organization to strategic management. Academy of Management Review, 6(4), 609-620.Available at: https://doi.org/10.2307/257639.

Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behavior Research Methods, Instruments, & Computers, 36(4), 717-731.Available at: https://doi.org/10.3758/bf03206553.

Pritchard, B., Gracy, C., & Godwin, M. (2010). The impacts of supermarket procurement on farming communities in India: Evidence from rural Karnataka. Development Policy Review, 28(4), 435-456.Available at: https://doi.org/10.1111/j.1467-7679.2010.00491.x.

Rashid, A. (2016). Impact of inventory management in downstream chains on customer satisfaction at manufacturing firms. International Journal of Management, IT and Engineering, 6(6), 1-19.

Rashid, A., & Amirah, N. (2017). Relationship between poor documentation and efficient inventory control at Provincial Ministry of Health, Lahore. American Journal of Innovative Research and Applied Sciences, 5(6), 420-423.

Rashid, A., Amirah, N., & Yusof, Y. (2019). Statistical approach in exploring factors of documentation process and hospital performance: A preliminary study. American Journal of Innovative Research and Applied Sciences, 9(4), 306-310.

Rehman, S.-u., Mohamed, R., & Ayoup, H. (2019). The mediating role of organizational capabilities between organizational performance and its determinants. Journal of Global Entrepreneurship Research, 9(1), 1-23.

Samuel, I. S., & Ondiek, O. (2014). Inventory management automation and the performance of supermarkets in Western Kenya. International Journal of Research in Management & Business Studies, 3(5), 1-12.

Sekaran, U., & Bougie, R. (2010). Research methods for business: A skill building approach (5th ed.). West Sussex: John Wiley & Sons Ltd.

Shah, R., & Shin, H. (2007). Relationships among information technology, inventory, and profitability: An investigation of level invariance using sector level data. Journal of Operations Management, 25(4), 768-784.Available at: https://doi.org/10.1016/j.jom.2007.01.011.

Shibin, K., Gunasekaran, A., & Dubey, R. (2017). Explaining sustainable supply chain performance using a total interpretive structural modeling approach. Sustainable Production and Consumption, 12, 104-118.Available at: https://doi.org/10.1016/j.spc.2017.06.003.

Sutduean, J., Joemsittiprasert, W., & Jermsittiparsert, K. (2019). Supply Chain management and organizational performance: Exploring green marketing as mediator. International Journal of Innovation, Creativity and Change, 5(2), 266-283.

Tabachnick, B. G., & Fidell, L. S. (2014). Using multivariate statistics (6th ed.). Harlow: Pearson.

Tucker, L. R., & Lewis, C. (1973). A reliability coefficient for maximum likelihood factor analysis. Psychometrika, 38(1), 1–10.Available at: https://doi.org/10.1007/bf02291170.

Urbach, N., & Ahlemann, F. (2010). Structural equation modeling in information systems research using partial least squares. Journal of Information Technology Theory and Application, 11(2), 5-40.

Wade, M., & Hulland, J. (2004). The resource-based view and information systems research: Review, extension, and suggestions for future research. MIS Quarterly, 28(1), 107-142.Available at: https://doi.org/10.2307/25148626.

Wernerfelt, B. (1984). A resource-based view of the firm. Strategic Management Journal, 5(2), 171-180.

Worthington, R. L., & Whittaker, T. A. (2006). Scale development research: A content analysis and recommendations for best practices. The counseling Psychologist, 34(6), 806-838.Available at: https://doi.org/10.1177/0011000006288127.

Yusof, Y., Awang, Z., Jusoff, K., & Ibrahim, Y. (2017). The influence of green practices by non-green hotels on customer satisfaction and loyalty in hotel and tourism industry. International Journal of Green Economics, 11(1), 1-14.Available at: https://doi.org/10.1504/ijge.2017.10003675.

Views and opinions expressed in this article are the views and opinions of the author(s), International Journal of Management and Sustainability shall not be responsible or answerable for any loss, damage or liability, etc. caused in relation to/arising out of the use of the content.