University of Groningen Not lean by default Ziengs, Nick

(1)

University of Groningen

Not lean by default Ziengs, Nick

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date: 2018

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Ziengs, N. (2018). Not lean by default: Exploring practices, their design, and underlying mechanisms driving performance. University of Groningen, SOM research school.

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

(2)

517640-L-bw-Ziengs-SOM 517640-L-bw-Ziengs-SOM 517640-L-bw-Ziengs-SOM 517640-L-bw-Ziengs-SOM Processed on: 14-3-2018 Processed on: 14-3-2018 Processed on: 14-3-2018

Processed on: 14-3-2018 PDF page: 33PDF page: 33PDF page: 33PDF page: 33

25

Chapter 2

The Distinctive Roles of Core and

Infrastructural Quality Management

Practices: A Meta-Analytical

Structural Equation Modeling Study

2.1 Introduction

To ensure competitiveness firms need to improve their quality standards relentlessly (Schroeder, Linderman, & Zhang, 2005; Zhang & Xia, 2013). To do so, firms rely heavily on both core and infrastructural quality management practices (Flynn, Sakakibara, et al., 1995; Naor, Goldstein, Linderman, & Schroeder, 2008; Zu, 2009). Core quality management practices refer to the implementation and use of statistical tools and techniques such as control charts, cause-and-effect diagrams, or histograms (Sousa & Voss, 2002). Infrastructural quality management practices, such as quality related training programs and empowering employees to make quality-related decisions, aim to

(3)

Chapter 2 – Core and infrastructural quality management practices

26

2

facilitate the effective use of core quality management practices (Sousa & Voss, 2002). Managers tasked to improve quality face an arduous decision – to prioritize core quality management practices, infrastructural quality management practices, or invest equally in both. Although the distinction between core and infrastructural practices is widely used (Bortolotti, Danese, et al., 2015; Flynn, Sakakibara, et al., 1995; Naor et al., 2008; Zu, 2009), the role of infrastructural quality management practices, in particular, is subject to considerable debate (Sousa & Voss, 2002; Zu, 2009). In short, there are three competing perspectives. Proponents of an indirect view argue that both core and infrastructural practices are necessary to maintain and improve quality performance (Flynn, Schroeder, & Sakakibara, 1995; Zu, 2009). In fact, these authors argue that infrastructural quality management practices support core quality management practices and therefore only indirectly contribute to quality and organizational performance. In contrast, proponents of the direct view argue that infrastructural quality management practices directly contribute to performance and will positively affect performance even in the absence of core quality management practices (Dow, Samson, & Ford, 1999; Powell, 1995; Samson & Terziovski, 1999). Proponents of the first two perspectives subscribe to the idea that core and infrastructural practices can be distinguished which is contested as well by the proponents of a third perspective. Here the argument is that no distinction can be made between the core and infrastructural practices. As such, the debate surrounding the role of infrastructural quality management practices, first voiced by Sousa and Voss (2002), is not yet resolved as indicated

(4)

Chapter 2

27

2

by recent studies in support of both the indirect (Bortolotti, Danese, et al., 2015; Dal Pont et al., 2008; Zu, 2009; Zu, Robbins, & Fredendall, 2010), the direct perspective (Naor et al., 2008), and indistinct perspective (Choi and Eboch, 1998) . In this study, we take the first step to resolve this debate by examining the relationships between core and infrastructural quality management practices and evaluating these alternative perspectives.

To study the relation between core and infrastructural quality management practices and their relation to performance, we use meta-analytical techniques (Borenstein, Hedges, Higgins, & Rothstein, 2009; Cooper, Hedges, & Valentine, 2009; Hunter & Schmidt, 2004) in combination with structural equation modelling (Viswesvaran & Ones, 1995). Meta-analysis is suitable to resolve contradictory findings when a large number of studies have been published and when additional individual studies are not likely to further the debate. Meta-analytical techniques allow us to provide an accurate estimate of the strength of the relationship between infrastructural quality management practices, core quality management practices, and performance. Confirmatory factor analysis and structural equation modeling techniques allow us to assess whether core and infrastructural practices can be considered distinct and to what degree these practices predict performance.

The study contributes to the literature and to manufacturing practice in the following ways. First, we examine the direct relation between core quality management practices and performance and the direct relation between infrastructural quality management practices

(5)

28

2

and performance. In doing so, we update Nair’s (2006) meta-analytical study by adding over a decade’s worth of research on quality management. Second, we examine the relationship between core and infrastructural quality management practices, thereby extending prior meta-analysis which only addressed the relation between quality management practices and performance, but not between the core and infrastructural practices themselves (Nair, 2006; White, 1996). Third, by assessing the relation between infrastructural and core quality management practices we reflect on both the direct and indirect nature of the relationship between core quality management practices, infrastructural quality management practices, and performance. Understanding the role of infrastructural quality management practices will enable manufacturing managers to make an informed decision when investing in quality management practices.

The remainder of this chapter is structured as follows. In section 2.2, we will discuss the three perspectives proposed in the literature and provide three alternative models which fit these perspectives. In section 2.3, we will detail the selection of papers, coding of papers, and the meta-analytical techniques used. The results are presented in section 2.4 and the implications for theory and practice and suggestions for future research are discussed in section 2.5. In the final section, we will conclude.

(6)

Chapter 2

29

2

2.2 Background

According to Dean and Bowen (1994), quality management is best characterized as a managerial philosophy based on the principles of customer centrality, continuous improvement, and organizational-wide effort. These principles provide managers with a set of guidelines for quality improvement initiatives. Quality management’s most noticeable proponents claimed that adherence to these principles would allow any organization to deliver high-quality products and services without compromising cost or delivery performance (e.g. Crosby, 1979; Deming, 1982; Juran, 1988).

Unsurprisingly, claims by quality management proponents and early examples of success gained a foothold with practitioners and led to a surge of quality management initiatives. More often than not, the outcome of these quality management initiatives was disappointing. The apparent contradiction led researchers to investigate quality management in a more rigorous empirical manner to discern what makes quality management initiatives successful (e.g. Flynn, 1994; Krafcik, 1988; Saraph, Benson, & Schroeder, 1989) resulting in the publication of a larger number of studies which address the relationship between quality management practices and performance.

There is considerable agreement in literature as to what the most common or important quality management practices are (Nair, 2006; Sousa & Voss, 2002; Zu et al., 2010), namely (1) process management, (2) product design and management, (3) quality data and analysis, (4) management leadership, (5) customer quality management, (6) supplier

(7)

30

2

quality management, and (7) people management. Table 2.1 provides an overview of these practices and their descriptions as proposed by Zu et al. (2008, p.632) based on their assessment of the literature.

In a further effort to explain the successful and unsuccessful quality management practices, researchers classified quality management practices as either core or infrastructural (Flynn, Schroeder, et al., 1995; Naor et al., 2008; Sousa & Voss, 2002; Zu et al., 2008). To maintain and improve quality, core quality management practices focus on the application of statistical tools and techniques. Infrastructural quality management practices, on the other hand, are intended to support and facilitate the effective use of core quality management practices and emphasize managerial, employee, supplier, and customer involvement. Implementation of core quality management practices without consideration of infrastructural quality management practices has been suggested as an important reason for the failure of many of the early quality management initiatives (Powell, 1995).

The role of infrastructural quality management practices when it comes to improving performance, however, is highly debated. Not all scholars agree on the supporting role of infrastructural quality management practices. In fact, a large number of studies found infrastructural quality management practices to be of greater importance than core quality management practices. These studies go as far as stating that quality improvement can be realized without the use of core quality management practices through reliance on

(8)

31 Chapter 2 T ab le 2. 1. C or e an d in fr astr uct ur al qu alit y m an ag em en t p rac tices Cla ss P ra ct ice 1 Descript io n 2 co re pr oce ss m an ag em en t T her e is an e m ph as is o n m is tak e-pr oo f pr oce ss d esig n. T her e is co ns is ten t us e of statis tical pr oce ss co ntr ol, pr ev en tiv e m ai nte nan ce . Ma nag er s an d em plo yee s m ak e ef fo rt s to m ai ntai n clea n sh op f lo or s an d m ee t s ch ed ules. pr od uct desig n an d m an ag em en t T her e is th or ou gh r ev ie w b ef or e pr od uctio n. Desig n tea m s in vo lv e peo ple fr om dif fer en t fu nctio ns su ch as m an uf ac tu rin g, m ar keti ng , pu rch asi ng dep ar tm en ts . Si m pli fied d esi gn a nd s ta nd ar dizatio n ar e en co ur ag ed f or m an uf ac tu rab ilit y. qu alit y data an al ys is Q ualit y data ar e av ailab le to m an ag er s an d em plo yee s. T her e is an e ff or t to co llect ti m el y qu ali ty d ata. Qu alit y da ta ar e us ed f or im pr ov em en t. in fr astru ct ur al cu sto m er f ocu s C us to m er n ee ds a nd ex pec tat io ns ar e as ses sed . C us to m er s ar e in vo lv ed i n qu ali ty im pr ov em en t p ro jec ts . C us to m er s atis fac tio n is m ea su red . T her e is a clo se co ntact w it h ke y cu sto m er s. m an ag em en t le ad er sh ip T op m an ag em en t ac ce pts r es po ns ib ilit y fo r qu alit y an d is ev alu ated b ased o n qu al it y per fo rm an ce . T op m an ag em en t p ar ticip ates in q ualit y im pr ov em en t e ff or ts a nd m ak es str ateg ie s an d go al s fo r qu alit y im pr ov em en t. peo ple m an ag em en t E m plo yee s ar e in vo lv ed i n qu alit y dec is io ns . E m plo yee s ar e ev al uated b ased o n th eir qu alit y per fo rm an ce a nd t heir co ntr ib utio ns to q ualit y ar e rec og nized a nd r ew ar ded . Ma nag er s en co ur ag e tea m wo rk in g. T her e is tr ain in g on QM f or m an ag er s an d em plo yee s. su pp ly q ua lit y m an ag em en t A s m all nu m ber o f su pp lier s ar e us ed . Su pp lier s ar e in vo lv ed in p ro du ct dev elo pm en t an d qu alit y im pr ov em en t pr oj ec ts . Su pp lier s ar e ev al uat ed b ased o n qu alit y. T he or gan izatio n pr ov id es su pp lier s tr ain in g an d tech nical as si sta nce . 1C on str uct lab els fr om Nair ( 20 06 ) 2C on str uct descr ip tio ns f ro m Z u et al. ( 20 08 , p .6 32 )

(9)

32

2

infrastructural quality management practices alone. For instance, Powell (1995) was the first to argue, based on the resource-based view of the firm, that infrastructural practices are more likely to result in a competitive advantage because of their intangible nature and the belief that infrastructural quality management practices are difficult to imitate. Powell’s results showed infrastructural quality management practices to be positively associated with performance whereas core quality management practices were not. Dow et al. (1999) and Samson et al. (1999) draw similar conclusions based on a larger sample. More recently, Lakhal et al. (2006), Naor et al. (2008), and Fotopoulos (2009) showed that infrastructural practices alone can play a vital and direct role in improving quality performance. The hypothesized direct relation between infrastructural quality management practices and performance suggests that infrastructural practices are not merely supportive. Managers willing to implement quality practices could suffice with just the implementation of infrastructural quality management practices (Powell, 1995) or, at least, infrastructural quality management practices should require greater attention than core quality management practices (e.g. Naor et al., 2008; Samson & Terziovski, 1999). The direct perspective is summarized in Figure 2.1.

In contrast, the second group of scholars argues in support of an indirect relationship between infrastructural quality management practices and performance. For instance, Flynn et al. (1994) were the first to find support for the notion that the success of core quality management practices was likely to depend on infrastructural quality management practices. Later studies, for example, Dal Pont et al.

(10)

Chapter 2

33

2

(2008), Flynn et al. (1995), Lakhal et al. (2006) and Zu et al. (2008), provided additional support for the claim that infrastructural quality management practices are necessary to support core quality management practices. These studies suggest that infrastructural quality management practices serve as a foundation for core quality management practices. The indirect perspective is summarized in Figure 2.2.

Both of the aforementioned perspectives assume that core and infrastructural quality management practices are distinct. Other studies were not able to find support for the distinction between core and infrastructural quality management practices (e.g. Choi & Eboch, 1998). In these studies, the authors consider core and infrastructural quality management practices jointly as a single construct. Sousa and Voss (2002) suggest that the practices themselves are not difficult to imitate, however the combination or integration of infrastructural and core quality management practices might be. This perspective is in line with theory on resource complementarity (Laursen & Foss, 2000) and socio-technical systems theory (Zu, 2009). Figure 2.3 summarizes the final perspective.

Studies that address the relation between quality management practices and performance conceptualized performance in different ways. A number of studies focus on quality performance measured using indicators such as conformance to specifications (Phan et al. 2011a; Phan et al. 2011b), rework (Flynn et al. 1994), or scrap and defect levels (Mahadevappa and Kotreshwar, 2004). Others focus on

(11)

34

2

the impact of quality management practices on a broader range of operational performance dimensions such as cost (Flynn et al. 1994; Noar et al. 2008), flexibility (Swink et al., 2005), dependability (Curkovic et al, 2000), or speed (Lau, 2000). Still others consider the implications of quality management practices on business performance using indicators such as increase in market share (Su et al. 2008), profit (Boyer et al. 1997), or firm performance (Curkovic et al., 2000). In this chapter, we consider quality management practices to influence both quality and operational performance relatively directly whereas business performance is influence indirectly through quality and operational performance. Amongst others, Curkovic et al. (2000) and Kaynak (2003) have argued similarly. In addition, we consider the possibility that quality performance bolsters operational performance or hinders it through potential trade-offs.

To add to the confusion, results from studies using either one of the three theoretical perspectives are difficult to compare because of differences in study design. First, quality management practices are inconsistently categorized as either core or infrastructural. For example, Flynn et al. (1995) classify quality data analysis as an infrastructural practice, whereas Samson and Terziovski (1999), Dow et al. (1999), and Powell (1995) label it as a core quality management practice. Second, different studies use different measures of performance. Some studies use measures of financial performance (Powell, 1995), others use combined measures of operational performance (Samson & Terziovski, 1999), still others focus on quality performance (Forza & Filippini, 1998).

(12)

Chapter 2

35

2

Infrastructural

quality management practices:

1. Management Leadership 2. People Management 3. Supply Quality Management 4. Customer Focus

Core

1. Process Management 2. Quality Data Analysis 3. Product Design and Management

Performance:

1. Quality Performance 2. Operational Performance 3. Business Performance

Figure 2.1. Direct perspective

Infrastructural

1. Management Leadership 2. People Management 3. Supply Quality Management 4. Customer Focus

Core

1. Process Management 2. Quality Data Analysis 3. Product Design and Management

Performance:

Figure 2.2. Indirect perspective

Quality management practices:

1. Management Leadership 2. People Management 3. Supply Quality Management 4. Customer Focus 5. Process Management 6. Quality Data Analysis 7. Product Design and Management

Performance:

(13)

36

2

Third, as Sousa and Voss (2002) indicated, not all methodologies used were suitable to differentiate between direct and indirect effects. As such, Sousa and Voss (2002) argued that results have been extrapolated beyond the range of the methods used. Table 2.2 provides an overview of the methodologies, perspectives, and models implicitly and explicitly used by the studies in our sample.

To summarize, there is a disagreement in the literature about the role of infrastructural quality management practices. A number of studies suggest that both core and infrastructural quality management practices are crucial for improved performance and argue that a mediating role of core quality management practices needs to be considered. Other studies suggest that infrastructural practices alone are sufficient for improved performance. Still, other studies suggest that core and infrastructural quality management practices should not be considered separately. Furthermore, the comparison is difficult due inconsistent classification of practices and differences in study design. To resolve the debate regarding the role of infrastructural quality management practices a combination of meta-analysis and structural equation modeling is used.

(14)

37 Chapter 2 T ab le 2. 2. Mo del class if icatio n Study M et ho d M odel 1 1 A bd ul lah et al. ( 20 09 ) R eg res sio n 2 Ag us et al. ( 20 00 ) Stru ct ur al eq uatio n m od elin g 3 Ag us an d A bd ulla h (2 00 0) Stru ct ur al eq uatio n m od elin g 4 Ah ir e an d Dr ey fu s (2 00 0) Stru ct ur al eq uatio n m od elin g 5 Ah ir e an d O 'S ha ug hn ess y (1 99 8) R eg res sio n 6 Ah ir e et al. ( 19 96 ) C or relatio n an al ys is 7 An der so n et al. ( 19 95 ) P ath an al ys is 8 An h an d Ma ts ui ( 20 06 ) R eg res sio n 9 A ru m ug am e t a l. (2 00 8) R eg res sio n

(15)

38

T ab le 2. 2. Mo del class if icatio n Study M et ho d M odel 1 10 B lack ( 19 95 ) C or relatio n an al ys is 11 B oo n et al. ( 20 07 ) R eg res sio n 12 B oy er et al. ( 19 97 ) R eg res sio n 13 C ho i a nd E bo ch ( 19 98 ) Stru ct ur al eq uatio n m od elin g 14 C ho ng a nd R un du s (2 00 4) R eg res sio n 15 C ho w dh ur y et a l. (2 00 7) R eg res sio n 16 C ur ko vic et al. ( 20 00 ) R eg res sio n 17 Das e t a l. (2 00 8) E xp lo rato ry f ac to r an al ys is 18 De C er io ( 20 03 ) R eg res sio n

(16)

39 Chapter 2 T ab le 2. 2. Mo del class if icatio n Study M et ho d M odel 1 19 Do ug la s an d Ju dg e (2 00 1) R eg res sio n 20 Do w et al. ( 19 99 ) Stru ct ur al eq uatio n m od elin g 21 Dr ey fu s et al. ( 20 04 ) MA N C OV A 22 Fl yn n an d Salad in ( 20 01 ) P ath an al ys is 23 Fl yn n et al. ( 19 94 ) C an on ical co rr elatio n an al ys is 24 Fl yn n et al. ( 19 95 ) R eg res sio n 25 Fl yn n et al. ( 19 99 ) R eg res sio n 26 Fo rza an d Fil ip pin i ( 19 98 ) Stru ct ur al eq uatio n m od elin g 27 Fu en tes et al. ( 20 06 ) C lu ster an al ys is

(17)

40

T ab le 2. 2. Mo del class if icatio n Study M et ho d M odel 1 28 Fu ller to n an d W em pe (2 00 9) Stru ct ur al eq uatio n m od elin g 29 Gr an dzo l a nd Ger sh on ( 19 98 ) C on fir m ato ry f ac to r an al ys is 30 Ho et al. ( 20 01 ) R eg res sio n 31 Hu ng et al. ( 20 10 ) Stru ct ur al eq uatio n m od elin g 32 Ittn er an d L ar ck er ( 19 97 ) R eg res sio n 33 Jay ar am e t a l. (2 01 0) Stru ct ur al eq uatio n m od elin g 34 Ju n et al. ( 20 06 ) Stru ct ur al eq uatio n m od elin g 35 Ka yn ak ( 20 03 ) Stru ct ur al eq uatio n m od elin g 36 Ka yn ak a nd Har tle y (2 00 8) Stru ct ur al eq uatio n m od elin g 37 Kr is tal et al. ( 20 10 ) Stru ct ur al eq uatio n m od elin g

(18)

41 Chapter 2 T ab le 2. 2. Mo del class if icatio n Study M et ho d M odel 1 38 L ai (2 00 3) C or relatio n an al ys is 39 L au ( 20 00 ) C lu ster an al ys is 40 L ee et al. ( 20 10 ) Stru ct ur al eq uatio n m od elin g 41 L o et al. ( 20 07 ) P ath an al ys is 42 L ok e et al. ( 20 12 ) Stru ct ur al eq uatio n m od elin g 43 Ma had ev ap pa an d Ko tr esh w ar ( 20 04 ) R eg res sio n 44 Me llat -P ar ast et al. ( 20 07 ) C or relatio n an al ys is 45 Mo lin a et al. ( 20 07 ) Stru ct ur al eq uatio n m od elin g 46 No ar et al. ( 20 08 ) Stru ct ur al eq uatio n m od elin g

(19)

42

T ab le 2. 2. Mo del class if icatio n Study M et ho d M odel 1 47 Oo i e t a l. (2 00 8) R eg res sio n 48 Oo i e t a l. (2 01 3) Stru ct ur al eq uatio n m od elin g 49 P ar k et al. ( 20 01 ) C or relatio n an al ys is 50 P han et al. ( 20 11 ) C or relatio n an al ys is 51 P ra jo go an d So hal (2 00 6) Stru ct ur al eq uatio n m od elin g 52 R ah m an an d B ullo ck ( 20 05 ) R eg res sio n 53 R ah m an ( 20 01 ) R eg res sio n 54 R un gtu sa nath am et al. ( 19 98 ) P ath an al ys is 55 Sa m so n an d T er zio vs ki ( 19 99 ) R eg res sio n 56 San ch ez -R od rig uez ( 20 04 ) C or relatio n an al ys is

(20)

43 Chapter 2 T ab le 2. 2. Mo del class if icatio n Study M et ho d M odel 1 57 Sh ah an d W ar d (2 00 3) R eg res sio n 58 So lis et al. ( 20 00 ) C or relatio n an al ys is 59 Su et al. ( 20 08 ) Stru ct ur al eq uatio n m od elin g 60 S w in k et al. ( 20 05 ) R eg res sio n 61 T ar i e t a l. (2 00 7) Stru ct ur al eq uatio n m od elin g 62 Ug bo ro an d Ob en g (2 00 0) C or relatio n an al ys is 63 Z han g et al. ( 20 00 ) E xp lo rato ry f ac to r an al ys is 64 Z u et al. ( 20 08 ) Stru ct ur al eq uatio n m od elin g In fr as tr uct ur al pr ac tices ( I) ; c or e pr ac tices ( C ); q ualit y per fo rm an ce ( Q) ; o per atio nal per fo rm an ce ( O) ; a nd bu si ness p er fo rm an ce ( B ) (So us a an d Vo ss , 2 00 2) .

(21)

Chapter 2– Core and infrastructural quality management practices

44

2

2.3 Methodology

To assess the role of infrastructural quality management practices, we use meta-analytical techniques in combination with structural equation modeling (see Landis, 2013; Viswesvaran & Ones, 1995 for an extended discussion). Meta-analytical techniques are used to synthesize previous empirical findings (Borenstein et al., 2009; Cooper et al., 2009; Hunter & Schmidt, 2004). Meta-analysis allows us to provide a more accurate estimate of the population correlation than individual studies can offer and is, therefore, a useful tool to resolve contradictory or conflicting findings in primary studies. In addition, meta-analysis allows us to provide an estimate of the accuracy of the population correlation resulting in a refinement of prior theory through the identification of possible moderating factors. The correlation matrix derived through meta-analysis served as the input for subsequent confirmatory factor analysis and structural equation modeling. Confirmatory factor analysis is needed to assess convergent and discriminant validity. Structural equation modeling is needed to evaluate the relationship between quality management practices and performance.

2.3.1 Sample

A thorough literature search was conducted to obtain a comprehensive sample. The search was conducted in two steps. In the first step, we conducted a computerized search of the EBSCO Business Source Premier database. The database was searched for papers that contained the phrase “quality management” in their abstracts. In addition, to

(22)

Chapter 2

45

2

restrict the search to survey papers, we also required at least one of the following keywords “questionnaire*, “sample*”, or “survey*”. No date restriction was placed on the search. The search was restricted to peer-reviewed journals only as is common in meta-analyses. However, to establish a comprehensive sample of papers we did not differentiate based on indirect study quality measures such as journal ranking or number of time cited. Studies with a comparable research approach were included in our sample.

The papers were carefully examined based on the following inclusion criteria. First, only survey papers were included. Conceptual, analytical, and case study papers were excluded. Second, papers were included if they reported on the relation between at least one quality management practice and performance or between two quality management practices. Papers were excluded if the constructs they reported on did not fit the constructs or definitions listed in Table 2.2. Only papers that relied on samples of manufacturing firms were included. Samples with a large number of service providers were excluded (e.g. Bou-Llusar, Escrig-Tena, Roca-Puig, and Beltrán-Martín, 2009).

In the second step, we conducted forward and backward searches of the papers that fit our inclusion criteria, earlier narrative reviews (Sousa & Voss, 2002), and earlier meta-analytical reviews on or related to quality management (Crook, Ketchen, Combs, & Todd, 2008; Jitpaiboon & Rao, 2007; Nair, 2006; Pereira & Osburn, 2007; Shenawy, Baker, & Lemak, 2007; White, 1996). The two steps yielded

(23)

46

2

63 papers reporting on 65 samples. Table 2.3 provides a list of the included papers and their corresponding journals.

2.3.2 Coding

A predetermined coding scheme was used to code the papers. The coding scheme was based on the constructs detailed by Sousa and Voss (2002), labels assigned by Nair (2006), and classification of infrastructural and core quality management practices by Zu et al. (2008). Effect sizes, sample sizes, reliability measures, item descriptions, construct labels, and definitions were coded for each paper.

Not all papers used the same constructs or the same items to measure those constructs. As such, constructs were classified based on their labels and the items used to measure those constructs. To ensure the items share a common conceptual definition, constructs were discarded if less than seventy-five percent of the items did not closely match the definition provided in Table 1.1. All constructs were coded by the author. An overview of the constructs and associated labels as reported in the primary studies is provided in Appendix A (digital version of the dissertation). The constructs, construct labels and the correlation and reliability coefficients derived from the included studies is provided in Appendix B (digital version of the dissertation).

(24)

47 Chapter 2 T ab le 2. 3. J ou rn als a nd s tu die s Jo urna l Co un t Studies 1 A ca de m y of Ma na ge m en t J ou rn al 2 Do ug la s an d Ju dg e (2 00 1) ; Fly nn et al. ( 19 95 ) 2 Asi an J ou rn al o f T ec hn olo gy In no va tio n 1 L ee et al. ( 20 10 ) 3 B en ch m ar ki ng : A n In ter nat io nal Jo ur nal 1 Das e t a l. (2 00 8) 4 C on fer en ce o f Asi a P ac if ic De cisi on Scien ce s In st itu te 1 An h an d Ma ts ui ( 20 06 ) 5 Dec is io n Sc ien ce s 5 Ah ir e et al. ( 19 96 ); A nd er so n et al. ( 19 95 ); Cu rk ov ic et al. ( 20 00 ); No ar et al. ( 20 08 ); S w in k et al . ( 20 05 ) 6 E ur op ea n Jo ur nal of Op er atio nal R esear ch 2 P ra jo go an d So hal (2 00 6) ; T ar i e t a l. (2 00 7) 7 Glo bal Jo ur nal of Flex ib le S ys te m s Ma nag em en t 1 C ho w dh ur y et a l. (2 00 7) 8 IE E E T ran sac tio ns o n E ng in eer in g Ma nag em en t 1 Dr ey fu s et al . ( 20 04 ) 9 In du str ial Ma na ge m en t & Dat a S ys te m s 1 Oo i e t a l. (2 00 8)

(25)

48

T ab le 2. 3. J ou rn als a nd s tu die s Jo urna l Co un t Studies 10 In ter natio nal Jo ur nal of Op er atio ns & P ro du ctio n Ma na ge m en t 3 Fu ller to n an d W em pe (2 00 9) ; Kr is tal et al. ( 20 10 ); San ch ez -R od rig uez (2 00 4) 11 In ter natio nal Jo ur nal of P ro du ctio n E co no m ics 4 Fo rza an d Fil ip pin i ( 19 98 ); L ai (2 00 3) ; P han et al. ( 20 11 a) ; P han et al. (2 01 1b ) 12 In ter natio nal Jo ur nal of P ro du ctio n R esear ch 4 De C er io ( 20 03 ); L au ( 20 00 ); L o et al. ( 20 07 ), Ho et al. ( 20 01 ) 13 In ter natio nal Jo ur nal of Q ualit y & R eliab ilit y Ma na ge m en t 5 So lis et al. ( 20 00 a) ; So lis et al. ( 20 00 b) ; So lis et al. ( 20 00 c) ; Su et al. (2 00 8) ; Z han g et al. ( 20 00 ) 14 In ter natio nal Jo ur nal of Q ualit y Scien ce 2 Ah ir e an d O 'S ha ug hn ess y (1 99 8) ; G ran dzo l a nd Ger sh on ( 19 98 ) 15 Jo ur nal of B us in es s E co no m ic s an d Ma nag em en t 1 L ok e et al. ( 20 12 ) 16 Jo ur nal of Op er atio ns Ma nag em en t 16 Ah ir e an d Dr ey fu s (2 00 0) ; Bo yer et al. ( 19 97 ); Ch oi a nd E bo ch ( 19 98 ); Fl yn n an d Salad in ( 20 01 ); Fl yn n et al. ( 19 94 ); Fly nn et al. ( 19 99 ); Jay ar am e t a l. (2 01 0) ; J un et al. ( 20 06 ); K ay nak ( 20 03 ); K ay nak a nd Har tle y (2 00 8) ; M olin a et al. ( 20 07 ); P ar k et al. ( 20 01 ); R un gtu sa nath am et al. ( 19 98 ); Sa m so n an d T er zio vs ki ( 19 99 ); Sh ah an d W ar d (2 00 3) ; Z u et al. ( 20 08 ) 17 Jo ur nal of Qu al it y Ma na ge m en t 1 Ug bo ro an d Ob en g (2 00 0) 18 Ma nag em en t Scie nce 1 Ittn er an d L ar ck er ( 19 97 )

(26)

49 Chapter 2 T ab le 2. 3. J ou rn als a nd s tu die s Jo urna l Co un t Studies 19 O m eg a 1 R ah m an an d B ullo ck ( 20 05 ) 20 P er so nn el R ev ie w 1 B oo n et al. ( 20 07 ) 21 P ro du ctio n an d Op er atio ns Ma nag em en t 1 Do w et al. ( 19 99 ) 22 P ro du ctio n P lan nin g & C on tr ol 2 Me llat -P ar ast et al. ( 20 07 ); O oi e t a l. (2 01 3) 23 Sin gap or e Ma nag em en t Re vie w 1 Ag us an d A bd ulla h (2 00 0) 24 T he B ritis h A cc ou nt in g R ev ie w 1 C ho ng a nd R un du s (2 00 4) 25 T he T QM J ou rn al 1 A ru m ug am e t a l. (2 00 8) 26 T otal Qu alit y Ma na ge m en t 3 Ag us et al. ( 20 00 ); B lack ( 19 95 ); Rah m an ( 20 01 ) 27 T otal Qu alit y Ma na ge m en t & B us in ess E xce lle nce 4 A bd ul lah et al. ( 20 09 ); Fu en te s et al. ( 20 06 ); H un g et al. ( 20 10 ); Ma had ev ap pa an d Ko tr esh w ar ( 20 04 )

(27)

50

2

2.3.3 Meta-analysis

The meta-analytical techniques detailed by Hunter and Schmidt (2004) were used. The techniques have been used in comparable studies in operations management (Mackelprang & Nair, 2010; Nair, 2006), supply chain management (Leuschner, Charvet, & Rogers, 2013; Leuschner, Rogers, & Charvet, 2013; Mackelprang, Robinson, Bernardes, & Webb, 2014; Zimmermann & Foerstl, 2014), human resource management (Combs, Liu, Hall, & Ketchen, 2006), and strategic management (Crook et al., 2008). The meta-analytical techniques by Hunter and Schmidt (2004) allow us to correct for two of the most prevalent study imperfections, namely sampling and measurement error.

A composite correlation was calculated and substituted in order not to favor samples which relied on multiple similar constructs. If the reliability was also reported, a composite reliability score was also calculated and substituted. If the reliability of a construct was not reported, the average reliability of the construct across all studies within our sample was substituted.

The heterogeneity was determined using the approach specified by Borenstein et al. (2009). The heterogeneity was assessed using the Q statistic and associated I2 index. The I2 index provides a ratio of total-to-between study variation; a large value for the I2 index, thus, suggests heterogeneity and a value close to zero suggest homogeneity (Huedo-Medina, Sánchez-Meca, Marín-Martínez, & Botella, 2006). Heterogeneity indicates that study characteristics, such as level of

(28)

Chapter 2

51

2

analysis (e.g. firm or plant), industry type (e.g. electronics or metal), industry region (e.g. Asia or Western Europe), or the degree of use of associated practices (e.g. Process Management and Product Design and Management) are likely to moderate the relation under consideration.

The analysis was conducted in three stages at various levels of aggregation (see also Crook et al., 2008; Leuschner, Charvet, et al., 2013; Leuschner, Rogers, et al., 2013; Mackelprang & Nair, 2010; Mackelprang et al., 2014; Nair, 2006; Zimmermann & Foerstl, 2014 for a similar approach). Each successive stage represents a further refinement of the analysis. The first stage examines the correlation between the set of quality management practices as a whole (process management; production design and management; quality data analysis; customer focus; management leadership; people management; and supplier quality management combined) and aggregate performance (business performance; operational performance; and quality performance combined) and individual performance measures (business performance; operational performance; and quality performance). The second stage addresses the correlation between infrastructural quality management practices (customer focus; management leadership; people management; supplier quality management combined), core quality management practices (process management; product design and management; and quality data analysis combined), and aggregate performance. The last stage considers the correlation between individual infrastructural and core quality management practices (process management; production design and management; quality data analysis; customer focus; management

(29)

52

2

leadership; people management; and supplier quality management separately) individual performance measures (business performance; operational performance; and quality performance separately).

2.3.4 Structural equation modeling

To combine meta-analysis and structural equation modeling, the procedures detailed by Viswesvaran and Ones (1995) and Landis (2013) were used. The procedures have been used in comparable studies in, for example, new product development (Sivasubramaniam, Liebowitz, & Lackman, 2012), organizational behavior (Thomas, Whitman, & Viswesvaran, 2010; Hoffman, Blair, Meriac, & Woehr, 2007), human resource management (Meriac, Hoffman, & Woehr, 2014), safety (Clarke, 2013), and strategic management (Li, Eden, Hitt, & Ireland, 2008).

The correlations derived from the meta-analysis served as the input for the confirmatory factor analysis and structural equation model. To complete the required covariance matrix a mean of 0 and a standard deviation of 1 were used. The median of the sample sizes was used as an estimate of the number of observations as the median is less sensitive to extreme values. The median, instead of the harmonic mean, which is often used (Viswesvaran & Ones, 1995), was used because the estimate of the number of observations would otherwise be too conservative due to the inclusion of a few studies which sampled just a few companies (e.g. Boon, Arumugam, Safa, & Bakar, 2007; Ooi, Arumugam, Teh, &

(30)

Chapter 2

53

2

Chong, 2008). Amos 24 IBM® SPSS® was used to conduct the confirmatory factor analysis and structural equation modeling.

Chi-square (x2), the root-mean-square error of approximation (RMSEA), the standardized root-mean-squared residual (RMR), comparative fit index (CFI), Normed fit index (NFI), and the Akaike Information Criterion (AIC) and their corresponding cut-off criteria were used to evaluate model fit (Qu, 2007).

2.4 Results

In the following section, the results of the meta-analysis, confirmatory factor analysis, and structural equation modeling are presented.

2.4.1 Meta-analytical results

The results of the first, second, and third stage are shown in Table 2.4, Table 2.5, and Table 2.6 respectively. Appendix B (digital version of the dissertation) shows the data used for the analysis.

Table 2.4 shows the results of the first stage of the analysis. The results show a strong and significant correlation between quality management practices and aggregate performance (rh = .450; 90% CRh = .194 - .704). The degree of heterogeneity suggests that the correlation is subject to moderating factors (Qrh = 431.642 (51), p < .01; Irh = 88.416). Overall, the first stage of our analysis shows quality management practices to be positively related to aggregate performance.

(31)

54

2

Table 2.5 details the outcome of the second stage. The results show a strong correlation between core quality management practices and aggregate performance (rh = .439; 90% CRh = .136 - .742) and infrastructural quality management practices and aggregate performance (rh = .461; 90% CRh = .212 - .711). Although the correlation between infrastructural quality management practices and aggregate performance is larger than the correlation between core quality management practices and aggregate performance, the overlapping confidence and credibility intervals indicate that there is no significant difference between the two correlations. Again, the degree of heterogeneity suggests that both the correlation between infrastructural quality management practices and aggregate performance (Qrh = 617.674 (45), p < .01; Irh = 87.594) and the correlation between core quality management practices and aggregate performance (Qrh = 454.602 (41), p < .01; Irh = 91.201) are subjected to moderating factors.

Table 2.5 also shows a strong and significant correlation between infrastructural quality management practices and core quality management practices (rh = .523; 90% CRh = .252 - .794). The correlation between core and infrastructural quality management practices is also likely subject to moderating factors (Qrh = 403.025 (51), p < .01; Irh = 92.876). The second stage of the analysis shows both core and infrastructural quality management to be positively related to aggregate performance. In addition, the second stage also shows that the strength of the relationship between core quality management practices and performance is similar to the strength of the relationship

(32)

Chapter 2

55

2

between infrastructural quality management practices and performance. Moreover, partitioning our sample in terms of core and infrastructure does little to mitigate the heterogeneity observed.

Table 2.6 shows the outcome of the final stage. All correlations between individual quality management practices and individual measures of performance are also positive and significant. The correlation between process management and customer service performance is the highest (rh = .705; 90% CRh =.464 - .946) and the correlation between quality data analysis and business performance the smallest (rh = .360; 90% CRh = .168 - .553). The correlation between management leadership and people management is the strongest between practices (rh = .672; 90% CRh = .388 - .957.), whereas the correlation between quality data analysis and customer focus is the weakest (rh = .524; 90% CRh = .139 - .910). Again, even on the level of individual quality management practices and individual measures of performance a high degree of heterogeneity can be observed which indicates that even on the level of individual practices moderating factors might have a large impact on the strength of the relationship under consideration. The analysis in the third stage shows that on the level of individual practices all relations show a strong positive correlation.

(33)

T ab le 2 .4 . M et a-an al yt ic al r esu lts (Q ua lit y man ag eme nt p ra ct ic es a nd a gg re ga te p er fo rman ce ) re la ti on n k nc m ea n [m in m ax ] r 95 % CI r I2 r rh 90 % CI r I2 rh fsN pr ac tic e - pe rf or man ce 10 24 7 52 10 07 0. 37 8 [ -0. 37 9 - 0. 78 1] 0. 37 8 [ 0 .3 61 0 .3 95 ] 88 .7 33 0. 45 [ 0 .1 94 0 .7 06 ] 88 .4 16 14 5 N umb er o f or ga ni za ti on s (n ); n um be r of st ud ie s (k ); n um be r of r ep or te d co rr el at io ns (n c) ; a ve ra ge r ep or te d co rr el at io n (me an ); lo w est r ep or te d co rr el at io n (m in ); h ig he st r ep or te d co rr el at io n (max ); e st im at ed e ff ec t s iz e co rr ec te d fo r samp lin g er ro r; 9 5% co nf id en ce in te rv al e ff ec t si ze e st imat e co rr ec te d fo r samp lin g er ro r (9 5% C Ir ); h et er og en ei ty e ff ec t si ze c or re ct ed f or samp lin g er ro r (I 2r ); e st imat ed e ff ec t si ze c or re ct ed f or sam pl in g er ro r an d me asu re m en t e rr or ( R h) ; 9 0% cr ed ib ili ty in te rv al e ff ec t si ze e st im at e co rr ec te d fo r sa mp lin g er ro r an d me as ur eme nt e rr or ( 90 % C Ir ); h et er og en ei ty e ff ec t si ze e st imat e co rr ec te d fo r samp lin g er ro r an d me asu re me nt e rr or ( I2 rh ); a nd F ai l-Sa fe N ( fsN ). T ab le 2 .5 . M et a-an al yt ic al r esu lts (Q ua lit y man ag eme nt p ra ct ic es a nd a gg re ga te p er fo rman ce ) re la ti on n k nc m ea n [m in m ax ] r 95 % CI r I2 r rh 90 % CI r I2 rh fsN co re in fr ast ru ct ur al 13 01 4 46 46 2 0. 47 6 [ 0. 01 2 - 0. 77 9] 0. 42 7 [ 0 .4 13 0 .4 42 ] 92 .9 97 0. 52 3 [ 0 .2 52 0 .7 94 ] 92 .8 76 15 1 co re p er fo rman ce 88 42 42 31 2 0. 33 7 [ -0. 35 9 - 0. 68 7] 0. 35 6 [ 0 .3 38 0 .3 75 ] 91 .4 4 0. 43 9 [ 0 .1 36 0 .7 42 ] 91 .2 01 10 8 in fr as tr uc tu ra l p er fo rman ce 10 24 7 52 69 5 0. 38 6 [ -0. 39 5 - 0. 79 7] 0. 38 4 [ 0 .3 67 0 .4 ] 88 .1 59 0. 46 1 [ 0 .2 12 0 .7 11 ] 87 .5 94 14 8 N umb er o f or ga ni za ti on s (n ); n um be r of st ud ie s (k ); n um be r of r ep or te d co rr el at io ns (n c) ; a ve ra ge r ep or te d co rr el at io n (me an ); lo w est r ep or te d co rr el at io n (m in ); h ig he st r ep or te d co rr el at io n (max ); e st im at ed e ff ec t s iz e co rr ec te d fo r samp lin g er ro r; 9 5% co nf id en ce in te rv al e ff ec t si ze e st imat e co rr ec te d fo r samp lin g er ro r (9 5% C Ir ); h et er og en ei ty e ff ec t si ze c or re ct ed f or samp lin g er ro r (I 2r ); e st imat ed e ff ec t si ze c or re ct ed f or sam pl in g er ro r an d me asu re m en t e rr or ( R h) ; 9 0% cr ed ib ili ty in te rv al e ff ec t s iz e est ima te c or re ct ed f or sa mp lin g er ro r an d me as ur eme nt e rr or ( 90 % C Ir ); h et er og en ei ty e ff ec t si ze e st imat e co rr ec te d fo r samp lin g er ro r an d me asu re me nt e rr or ( I2 rh ); a nd F ai l-Sa fe N ( fsN ).

(34)

T ab le 2 .6 . M et a-an al yt ic al r esu lts (I nd iv id ua l c or e an d in fr as tr uc tu ra l q ua lit y man ag eme nt p ra ct ic es an d pe rf or man ce ) re la ti on n k nc m ea n [m in m ax ] r 95 % CI r I2 r rh 90 % CI r I2 rh fsN pr oc ess ma na ge me nt pr od uc t d esi gn a nd man ag eme nt 49 48 16 23 0. 50 7 [ 0. 23 0 - 0. 77 5] 0. 51 5 [ 0. 49 5 - 0. 53 6] 93 .305 0. 62 7 [ 0. 41 1 - 0. 84 2] 85 .887 67 qu al it y da ta a na ly si s 79 45 24 37 0. 56 6 [ 0. 22 0 - 0. 86 0] 0. 46 2 [ 0. 44 5 - 0. 48 0] 96 .406 0. 55 2 [ 0. 22 6 - 0. 87 8] 95 .338 87 cu st ome r fo cu s 71 72 24 30 0. 49 5 [ -0 .0 30 0 .8 58 ] 0. 49 1 [ 0. 47 3 - 0. 50 8] 94 .212 0. 61 4 [ 0. 34 8 - 0. 87 9] 90 .114 94 man ag eme nt le ad er sh ip 68 69 28 42 0. 50 8 [ -0 .1 80 0 .7 68 ] 0. 50 9 [ 0. 49 2 - 0. 52 7] 93 .481 0. 61 9 [ 0. 35 5 - 0. 88 2] 90 .243 11 5 pe op le man ag eme nt 10 44 4 34 73 0. 53 6 [ 0. 16 1 - 0. 78 0] 0. 47 8 [ 0. 46 3 - 0. 49 2] 94 .849 0. 58 2 [ 0. 29 9 - 0. 86 5] 93 .002 12 9 su pp lie r qu al ity man ag eme nt 53 47 20 28 0. 45 0 [ -0 .1 00 0 .7 58 ] 0. 46 7 [ 0. 44 6 - 0. 48 8] 92 .960 0. 58 9 [ 0. 32 8 - 0. 84 9] 89 .525 74 bu si ne ss p er fo rman ce 25 96 11 15 0. 27 3 [ 0. 08 6 - 0. 47 5] 0. 30 8 [ 0. 27 3 - 0. 34 3] 82 .423 0. 36 7 [ 0. 19 6 - 0. 53 7] 81 .053 23 op er at io na l p er fo rman ce 38 77 21 70 0. 30 5 [ -0 .3 85 0 .6 62 ] 0. 38 8 [ 0. 36 1 - 0. 41 5] 88 .314 0. 51 4 [ 0. 25 0 - 0. 77 8] 81 .167 61 qu al it y pe rf or man ce 41 98 19 30 0. 31 6 [ -0 .3 62 0 .7 00 ] 0. 43 4 [ 0. 41 0 - 0. 45 9] 94 .545 0. 51 4 [ 0. 18 8 - 0. 84 0] 94 .073 64 pr od uc t d esi gn a nd m an ag eme nt qu al it y da ta a na ly si s 35 54 14 21 0. 49 5 [ 0. 14 0 - 0. 76 4] 0. 51 [ 0. 48 5 - 0. 53 4] 93 .982 0. 60 6 [ 0. 35 3 - 0. 85 8] 90 .553 58 cu st ome r fo cu s 45 53 17 17 0. 40 2 [ 0. 09 0 - 0. 73 0] 0. 43 9 [ 0. 41 5 - 0. 46 2] 94 .017 0. 57 1 [ 0. 27 3 - 0. 86 9] 89 .848 58 man ag eme nt le ad er sh ip 38 76 15 19 0. 49 7 [ 0. 09 0 - 0. 74 0] 0. 54 3 [ 0. 52 1 - 0. 56 5] 94 .367 0. 65 7 [ 0. 40 5 - 0. 90 9] 90 .439 67 pe op le man ag eme nt 53 76 18 36 0. 49 5 [ 0. 11 5 - 0. 75 0] 0. 45 5 [ 0. 43 4 - 0. 47 6] 93 .085 0. 56 9 [ 0. 32 3 - 0. 81 5] 88 .180 64 su pp lie r qu al ity man ag eme nt 49 03 18 22 0. 48 0 [ 0. 11 0 - 0. 81 0] 0. 43 4 [ 0. 41 1 - 0. 45 7] 91 .964 0. 55 4 [ 0. 30 8 - 0. 80 0] 85 .984 61 bu si ne ss p er fo rman ce 19 05 6 6 0. 40 3 [ 0. 20 6 - 0. 59 0] 0. 40 2 [ 0. 36 4 - 0. 44 0] 92 .084 0. 47 5 [ 0. 28 4 - 0. 66 5] 89 .868 19 op er at io na l p er fo rman ce 27 78 13 35 0. 34 9 [ 0. 03 3 - 0. 56 0] 0. 40 7 [ 0. 37 6 - 0. 43 8] 86 .075 0. 52 8 [ 0. 32 3 - 0. 73 3] 80 .057 40 qu al it y pe rf or man ce 36 73 15 18 0. 29 1 [ -0 .2 83 0 .6 30 ] 0. 38 [ 0. 35 2 - 0. 40 8] 91 .004 0. 47 3 [ 0. 22 1 - 0. 72 5] 87 .835 43

(35)

T ab le 2 .6 . M et a-an al yt ic al r esu lts (I nd iv id ua l c or e an d in fr as tr uc tu ra l q ua lit y man ag eme nt p ra ct ic es an d pe rf or man ce ) re la ti on n k nc m ea n [m in m ax ] r 95 % CI r I2 r rh 90 % CI r I2 rh fsN qu al it y da ta a na ly si s cu st ome r fo cu s 69 37 26 33 0. 46 7 [ -0 .2 29 0 .8 85 ] 0. 43 3 [ 0. 41 3 - 0. 45 2] 96 .455 0. 52 4 [ 0. 13 9 - 0. 91 0] 95 .534 87 man ag eme nt le ad er sh ip 77 11 30 51 0. 51 7 [ -0 .0 60 0 .8 05 ] 0. 48 7 [ 0. 47 0 - 0. 50 4] 93 .622 0. 57 7 [ 0. 31 4 - 0. 84 0] 90 .749 11 7 pe op le man ag eme nt 97 17 32 80 0. 51 4 [ 0. 13 0 - 0. 86 0] 0. 42 7 [ 0. 41 1 - 0. 44 4] 94 .605 0. 50 9 [ 0. 22 1 - 0. 79 7] 93 .337 10 5 su pp lie r qu al ity man ag eme nt 53 90 21 31 0. 45 0 [ 0. 20 5 - 0. 78 8] 0. 42 [ 0. 39 8 - 0. 44 2] 89 .283 0. 51 [ 0. 29 4 - 0. 72 5] 87 .224 68 bu si ne ss p er fo rman ce 22 54 8 17 0. 27 5 [ 0. 05 1 - 0. 51 7] 0. 31 5 [ 0. 27 8 - 0. 35 3] 89 .373 0. 36 [ 0. 16 8 - 0. 55 3] 88 .651 18 op er at io na l p er fo rman ce 38 38 19 48 0. 33 9 [ -0 .0 85 0 .6 62 ] 0. 37 3 [ 0. 34 5 - 0. 40 0] 87 .049 0. 46 5 [ 0. 24 1 - 0. 68 8] 78 .898 52 qu al it y pe rf or man ce 54 82 26 39 0. 34 6 [ -0 .2 10 0 .8 56 ] 0. 34 [ 0. 31 6 - 0. 36 3] 93 .484 0. 40 6 [ 0. 06 7 - 0. 74 4] 93 .020 63 cu st ome r fo cu s man ag eme nt le ad er sh ip 80 50 34 46 0. 52 2 [ -0 .2 94 0 .8 85 ] 0. 54 4 [ 0. 52 8 - 0. 55 9] 94 .456 0. 67 2 [ 0. 38 8 - 0. 95 7] 90 .796 15 1 pe op le man ag eme nt 92 72 36 71 0. 50 8 [ -0 .3 21 0 .8 85 ] 0. 49 7 [ 0. 48 2 - 0. 51 3] 95 .474 0. 60 9 [ 0. 28 3 - 0. 93 5] 93 .815 14 4 su pp lie r qu al ity man ag eme nt 68 59 26 33 0. 42 4 [ -0 .3 08 0 .8 85 ] 0. 47 2 [ 0. 45 3 - 0. 49 0] 94 .726 0. 6 [ 0. 28 7 - 0. 91 2] 91 .711 97 bu si ne ss p er fo rman ce 21 46 8 12 0. 28 4 [ 0. 04 6 - 0. 61 0] 0. 32 8 [ 0. 29 0 - 0. 36 6] 93 .336 0. 37 9 [ 0. 12 1 - 0. 63 8] 93 .130 19 op er at io na l p er fo rman ce 44 61 23 52 0. 38 6 [ 0. 05 8 - 0. 88 5] 0. 41 7 [ 0. 39 3 - 0. 44 1] 90 .788 0. 52 3 [ 0. 24 1 - 0. 80 5] 88 .913 73 qu al it y pe rf or man ce 55 19 26 28 0. 42 7 [ -0 .2 41 0 .8 56 ] 0. 45 2 [ 0. 43 1 - 0. 47 3] 93 .131 0. 56 6 [ 0. 25 9 - 0. 87 2] 91 .235 92 man ag eme nt le ad er sh ip pe op le man ag eme nt 98 05 44 10 5 0. 53 4 [ 0. 04 5 - 0. 77 0] 0. 57 1 [ 0. 55 8 - 0. 58 5] 91 .340 0. 68 6 [ 0. 47 0 - 0. 90 1] 87 .665 20 8 su pp lie r qu al ity man ag eme nt 57 23 24 31 0. 48 3 [ 0. 09 0 - 0. 80 2] 0. 50 6 [ 0. 48 6 - 0. 52 5] 90 .370 0. 60 3 [ 0. 39 2 - 0. 81 4] 86 .978 98 bu si ne ss p er fo rman ce 31 11 11 19 0. 27 2 [ 0. 08 8 - 0. 60 0] 0. 30 6 [ 0. 27 4 - 0. 33 8] 90 .133 0. 36 1 [ 0. 14 3 - 0. 57 8] 88 .412 23 op er at io na l p er fo rman ce 59 37 31 84 0. 41 6 [ 0. 07 6 - 0. 76 4] 0. 41 4 [ 0. 39 3 - 0. 43 5] 88 .110 0. 51 7 [ 0. 27 7 - 0. 75 8] 80 .984 98 qu al it y pe rf or man ce 63 14 29 44 0. 38 3 [ -0 .3 44 0 .8 56 ] 0. 39 9 [ 0. 37 9 - 0. 42 0] 91 .410 0. 47 6 [ 0. 20 5 - 0. 74 7] 89 .692 87

(36)

T ab le 2 .6 . M et a-an al yt ic al r esu lts (I nd iv id ua l c or e an d in fr as tr uc tu ra l q ua lit y man ag eme nt p ra ct ic es an d pe rf or man ce ) re la ti on n k nc m ea n [m in m ax ] r 95 % CI r I2 r rh 90 % CI r I2 rh fsN pe op le man ag eme nt su pp lie r qu al ity man ag eme nt 70 02 27 60 0. 46 1 [ 0 .0 80 0 .8 30 ] 0. 43 6 [ 0. 41 7 - 0. 45 5] 90 .752 0. 53 9 [ 0. 30 3 - 0. 77 4] 87 .833 91 bu si ne ss p er fo rman ce 34 86 14 43 0. 30 0 [ 0. 13 9 - 0. 64 0] 0. 29 9 [ 0. 26 9 - 0. 33 ] 84 .384 0. 36 [ 0. 17 4 - 0. 54 7] 82 .770 28 op er at io na l p er fo rman ce 62 30 33 15 4 0. 39 1 [ -0 .0 36 0 .7 89 ] 0. 41 [ 0. 38 9 - 0. 43 0] 88 .891 0. 53 6 [ 0. 27 8 - 0. 79 3] 77 .201 10 3 qu al it y pe rf or man ce 65 68 30 69 0. 35 2 [ -0 .3 02 0 .8 56 ] 0. 35 6 [ 0. 33 5 - 0. 37 7] 92 .971 0. 42 3 [ 0. 10 8 - 0. 73 7] 92 .382 77 su pp lie r qu al ity man ag eme nt bu si ne ss p er fo rman ce 20 83 8 12 0. 29 3 [ 0. 07 1 - 0. 46 0] 0. 31 8 [ 0. 28 0 - 0. 35 7] 85 .966 0. 38 4 [ 0. 20 6 - 0. 56 2] 84 .699 18 op er at io na l p er fo rman ce 38 71 20 49 0. 33 1 [ -0 .1 33 0 .6 87 ] 0. 34 [ 0. 31 2 - 0. 36 8] 86 .333 0. 42 [ 0. 18 3 - 0. 65 8] 85 .260 48 qu al it y pe rf or man ce 51 77 25 34 0. 37 4 [ -0 .3 65 0 .8 56 ] 0. 37 [ 0. 34 7 - 0. 39 4] 93 .018 0. 45 8 [ 0. 12 7 - 0. 78 9] 92 .207 68 bu si ne ss pe rf or man ce op er at io na l p er fo rman ce 27 76 11 22 0. 38 0 [ 0. 15 0 - 0. 57 4] 0. 41 7 [ 0. 38 6 - 0. 44 8] 89 .059 0. 52 9 [ 0. 31 6 - 0. 74 2] 87 .208 35 qu al it y pe rf or man ce 16 39 6 14 0. 45 4 [ 0. 25 5 - 0. 70 5] 0. 52 8 [ 0. 49 3 - 0. 56 3] 94 .303 0. 64 6 [ 0. 43 1 - 0. 86 1] 89 .420 26 op er at io na l p er fo rman ce qu al it y pe rf or man ce 30 09 14 33 0. 41 0 [ -0 .2 01 0 .8 56 ] 0. 46 9 [ 0. 44 1 - 0. 49 7] 91 .898 0. 55 1 [ 0. 30 5 - 0. 79 7] 89 .809 52 N umb er o f or ga ni za ti on s (n ); n um be r of st ud ie s (k ); n um be r of r ep or te d co rr el at io ns (n c) ; a ve ra ge r ep or te d co rr el at io n (me an ); lo w est r ep or te d co rr el at io n (m in ); h ig he st r ep or te d co rr el at io n (max ); e st im at ed e ff ec t s iz e co rr ec te d fo r samp lin g er ro r; 9 5% co nf id en ce in te rv al e ff ec t si ze e st imat e co rr ec te d fo r samp lin g er ro r (9 5% C Ir ); h et er og en ei ty e ff ec t si ze c or re ct ed f or samp lin g er ro r (I 2r ); e st imat ed e ff ec t si ze c or re ct ed f or sam pl in g er ro r an d me asu re m en t e rr or ( R h) ; 9 0% cr ed ib ili ty in te rv al e ff ec t s iz e est ima te c or re ct ed f or sa mp lin g er ro r an d me as ur eme nt e rr or ( 90 % C Ir ); h et er og en ei ty e ff ec t si ze e st imat e co rr ec te d fo r samp lin g er ro r an d me asu re me nt e rr or ( I2 rh ); a nd F ai l-Sa fe N ( fsN ).

(37)

60

2

2.4.2 Confirmatory factor analysis results

Table 2.7 shows the fit indices associated with the two measurement models. The first measurement model consists of two latent variables, namely core and infrastructural quality management practices. Figure 2.4 shows the first measurement model. The decomposition of quality management practices into core and infrastructural quality management practices is a prerequisite for both the indirect and direct effects model. The second measurement model consists of a single latent variable, namely quality management practices which includes both core and infrastructural quality management practices. Figure 2.5 shows the second measurement model. The second measurement model disregards the decomposition of quality management practices into core and infrastructural quality management practices. Both measurement models exhibit adequate fit. A chi-square difference test suggests that model fit is not significantly different.

Figure 2.5 and Figure 2.6 suggest convergent validity to be acceptable for both models. All standardized regression coefficients are larger than 0.5. The first measurement model, however, exhibits poor discriminant validity as the average variance extracted (AVE) is smaller than Maximum Shared Variance (MSV) and the square root of AVA is greater than the any of the inter-construct correlations (Qu, 2007). As such, the second measurement model is preferred over the first. Core and infrastructural quality management practices do not load on two distinct core and infrastructural quality management practice latent factors, rather they load on a single quality management practice latent

(38)

Processed on: 14-3-2018 PDF page: 69PDF page: 69PDF page: 69PDF page: 69 Chapter 2

61

2

factor. In short, the results favor the single-factor model. The range of the correlation coefficients between core and infrastructural quality management practices already suggested a high degree of interrelatedness.

2.4.3 Structural equation modeling results

Table 2.8 shows the fit-indices for the structural model. The proposed model does not distinguish between core and infrastructural quality management practices because the single factor model was preferred over the two-factor model. The proposed model exhibits adequate fit. All fit indices are above the proposed cut-off values. Figure 2.6 shows the proposed structural model. All paths are significant with the exception of the path between quality performance and operational performance. The structural model suggests quality management practices to predict quality and operational performance. Quality and operational performance, in turn, predict business performance. However, quality performance does not directly predict operational performance.

(39)

62

Fig ur e 2. 4. C on fir m ato ry f ac to r an al ys is ; T w o-fac to r m od el

(40)

63 Chapter 2 Fig ur e 2. 5. C on fir m ato ry f ac to r an al ys is ; Si ng le fac to r m od el

(41)

64

T ab le 2. 7. C on fir m ato ry f ac to r an al ys is ; F it in dice s X2 df Δ X 2 AIC CF I NF I G F I RM SE A RM R Sin gle -f ac to r m od el 10 .7 98 13 1. 32 7 40 .7 98 1. 00 0 .986 .9 85 .0 00 .0 17 T w o-fac to r m od el 9 .4 71 13 39 .4 71 1. 00 0 .9 88 .9 86 .0 00 .0 17 ** * p<. 01 ** p <. 05 * p<. 10 T ab le 2. 8. Stru ctu ral eq uatio n m od eli ng an al ys is ; F it in dice s X2 df AIC CF I NF I G F I RM SE A RM R Stru ct ur al m od el 45 .9 92 31 93 .9 92 .987 .9 62 .9 59 .0 49 .0 32 ** * p<. 01 ** p <. 05 * p<. 10