for magz64 help, from Tina Li

http://jom.sagepub.com/ Journal of Management http://jom.sagepub.com/content/39/2/366 The online version of this article can be found at:   DOI: 10.1177/0149206310365901 2013 39: 366 originally published online 8 April 2010 Journal of Management Rebecca R. Kehoe and Patrick M. Wright Attitudes and Behaviors The Impact of High-Performance Human Resource Practices on Employees'     Published by: http://www.sagepublications.com On behalf of:     Southern Management Association can be found at:

Journal of Management Additional services and information for         http://jom.sagepub.com/cgi/alerts Email Alerts:   http://jom.sagepub.com/subscriptions Subscriptions:   http://www.sagepub.com/journalsReprints.nav Reprints:   http://www.sagepub.com/journalsPermissions.nav Permissions:   What is This?   - Apr 8, 2010 OnlineFirst Version of Record  - Jan 23, 2013 Version of Record >> at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from 366 The Impact of High-Performance Human Resource Practices on Employees’ Attitudes and Behaviors Rebecca R. KehoePatrick M. Wright Cornell University Although strategic human resource (HR) management research has established a significant relationship between high-performance HR practices and firm-level financi\ al and market out- comes, few studies have considered the important role of employees’ perceptions of HR practice use or examined the more proximal outcomes of high-performance HR practices that may play mediating roles in the HR practice–performance relationship. To address recent calls in the literature for an investigation of this nature, this study examined the relationships between employees’ perceptions of high-performance HR practice use in their job groups and employee absenteeism, intent to remain with the organization, and organizational citizenship behavior, dedicating a focus to the possible mediating role of affective organizational commitment in these relationships. Data in this study were collected from surveys of employees at a large multiunit food service organization. The model was tested with CWC(M) mediation analysis (i.e.\ , centered within context with reintroduction of the subtracted means at Level 2), which accounted for the multilevel structure of the data. Results indicate that employees’ perceptions of high-perfor - mance HR practice use at the job group level positively related to all dependent variables and that affective organizational commitment partially mediated the relationship between HR prac- tice perceptions and organizational citizenship behavior and fully mediated the relationship Acknowledgments: This article was accepted under the editorship of Talya N. Bauer. We would like to thank the editor and two reviewers as well as Chris Collins for their comments on earlier versions\ of this article.

Corresponding author: Rebecca R. Kehoe, Cornell University, ILR School, 309 Ives Hall, Ithaca, NY 14853 E-mail: [email protected] Journal of Management Vol. 39 No. 2, February 2013 366-391 DOI: 10.1177/0149206310365901 © 2010 Southern Management Association. Reprints and permission: http://www.

sagepub.com/journalsPermissions.nav at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from Kehoe, Wright / Impact of High-Performance Human Resource Practices 367 between HR practice perceptions and intent to remain with the organization. The discussion reviews the implications of these results and suggests future directions for research in this vein.

Keywords: strategic human resour ce management; commitment In recent years, scholars and practitioners have increasingly recognized\ the impor - tance of effective human capital management for organizational performance (Grant, 1996; Hitt, Biermant, Shimizu, & Kochhar, 2001). In particular, strategic human resource management (SHRM) scholars have suggested that organizations can use per - formance- and commitment-oriented human resource (HR) practices to dri\ ve organiza- tional effectiveness (Becker & Gerhart, 1996; Dyer & Reeves, 1995; Wright, Dunford, & Snell, 2001)—a claim now supported by a large body of field research (e.g., Delery & Doty, 1996; Huselid, 1995; MacDuffie, 1995), as well as subsequent confirmatory reviews and meta-analyses of this literature (e.g., Combs, Liu, Hall, &\ Ketchen, 2006; Wright, Gardner, Moynihan, & Allen, 2005). However, studies of the HR–performance relationship have provided limited in\ sight into the effects of high-performance HR systems on the more proximal employee outcomes that they are likely to affect most directly (Dyer & Reeves, 1995)—thereby leading to gaps in\ the field’s understanding of the mechanisms linking HR practices to performance an\ d resulting in calls in the SHRM literature for research attention in this area (e.\ g., Becker & Gerhart, 1996; Takeuchi, Chen, & Lepak, 2009; Wright & Gardner, 2003). Although a few recent empirical studies have provided support for the claim that high-performa\ nce HR practices work most immediately through employee attitudes and behaviors—such a\ s job satisfaction, affective commitment (Gong, Law, Chang, & Xin, 2009; Takeuchi et al., 2009), service- oriented citizenship behaviors, turnover (Sun, Aryee, & Law, 2007), and social exchange (Takeuchi, Lepak, Wang, & Takeuchi, 2007)—recent theoretical work on the HR–performance causal chain suggest that these studies may oversimplify the relationships between HR prac- tices and employee outcomes (e.g., Bowen & Ostroff, 2004; Nishii & Wright, 2008). In particular, some SHRM scholars have argued that HR practices are likely to have desired consequences on employees’ attitudes and behaviors only to the extent that they are consis- tently experienced and perceived by employees in intended ways (Bowen &\ Ostroff, 2004).

Furthermore, empirical work has demonstrated that employees’ perceptions of HR practices significantly vary from managerial reports of the HR practices in use (Liao, Toya, Lepak, & Hong, 2009). Thus, whereas recent empirical studies linking managerial reports of high- performance HR practices to employee outcomes are methodologically consistent with previous SHRM research and valuable for their contribution to our understanding of the beginning steps of the HR–performance causal chain, additional work i\ s still needed assess- ing the role of employees’ perceptions of HR practices in determining their attitudinal and behavioral outcomes (Nishii & Wright, 2008; for a recent empirical article examining out- comes of employees’ attributions of HR practices, see Nishii, Lepak, & Schneider, 2008).

Such an investigation would allow us to determine whether employees’ collective subjective experiences with HR practices are a critical mechanism through which an HR system affects outcomes of interest. at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from 368 Journal of Management / February 2013 In light of these extant research needs, the purpose of the present arti\ cle is threefold: first, to propose a model explicating the effects of employees’ collective perceptions of high- performance HR system use within their job groups on important employee \ attitudinal and behavioral outcomes; second, to provide theoretical and empirical ration\ ales for aggregating employees’ HR practice perceptions to the job group level; and, third, to present results from a multilevel mediation analysis in which relationships at the job group \ and individual levels are computed separately but simultaneously in a single empirical model. \ We propose that support for our model would point to the theoretical and practical importance of employees’ reflections on their job groups’ collective experiences with HR practices—thereby suggest- ing a need to extend the SHRM focus beyond HR system design, to consiste\ ncy and effec- tiveness in HR practice employment. We begin by briefly examining the SHRM approach; we then proceed to underscore the need for research on the relationship \ between employees’ perceptions of high-performance HR practice use and attitudinal and beha\ vioral outcomes (highlighting, as key outcomes of interest, affective organizational commitment, organiza- tional citizenship behavior, intent to remain with the organization, and absenteeism); and, finally, we discuss the significance of considering employees’ perceptions of HR practices in use at the job group level. From here, we present our theoretical model, discuss the employed methodology in testing our predictions, and report our empirical results\ . We conclude by discussing the implications of our findings and by proposing possible di\ rections for future research in this vein. Background and Hypotheses The SHRM Approach to HR Practices Two key themes characterize the SHRM literature and dominate the focus of\ work in this area: First, coherent systems of mutually reinforcing HR practices are l\ ikely to better support sustainable performance outcomes than are any individual practices (Del\ ery & Shaw, 2001; Dyer & Reeves, 1995); second, all HR systems are not equally effective (Arthur, 1994; MacDuffie, 1995). Thus, substantial research in the SHRM realm has sought to pinpoint characteristics of an optimal HR system for attaining competitive advant\ age, with support for a high-performance approach to HR management emerging from this stream. Although the specific HR practices included in high-performance HR syste\ ms have var - ied across studies, a commonality across practices in any high-performan\ ce approach is a focus on promoting workforce ability, motivation, and opportunity (see Applebaum, Bailey, Berg, & Kalleberg, 2000; Combs et al., 2006) to perform behaviors consistent with organi- zational goals. Given these commonalities and our review of HR practices\ examined in previ- ous research in this area, we constructed a list of 15 HR practices reflec\ ting a high-performance HR approach for this study. In particular, we included ability-enhancing practices, such as formal selection tests, structured interviews, hiring selectivity, high pay, and training oppor - tunities; motivation-enhancing practices, such as rewards based on individual and group performance outcomes, formal performance evaluation mechanisms, and meri\ t-based pro- motion systems; and opportunity-enhancing practices, such as formal participation processes, at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from Kehoe, Wright / Impact of High-Performance Human Resource Practices 369 regular communication and information-sharing efforts, and autonomy in work-related deci- sion making.(e.g., Combs et al., 2006; Delery & Shaw, 2001; Huselid, 1995; Sun et al., 2007; Way, 2002).

Studying Employees’ Perceptions of HR Practices and Assessing Proximal Outcomes Dyer and Reeves (1995) posited four sequential levels of impact of HR \ practices—HR (or employee), organizational, financial, and market—suggesting that HR practices are likely to work outward (and upward) through these levels and hence, most immediately, through their effects on employee attitudes and behaviors. Similarly, Becker, Huselid, Pinckus, and Spratt (1997) suggested that HR practices influence the behaviors of employee\ s, which then affect operational, financial, and share price performance outcomes. Thus, both sets of authors suggested that a thorough understanding of the relationships between HR \ practices and employee outcomes is critical to our ability to draw logical inferences \ concerning the HR– performance causal chain as a whole. We seek to make our primary contribution in this area. In particular, as noted above, employees’ attitudinal and behavioral responses to an HR system depend on the HR practices that employees perceive to exist in their work context (Bowen & Ostroff, 2004). Because employees’ perceptions of HR practices necessarily fol- low managers’ HR practice implementation (Nishii & Wright, 2008), employees’ HR prac- tice perceptions are temporally closer to, and consequently likely to be\ more predictive of, their attitudinal and behavioral outcomes than are HR practice ratings a\ s provided by manag- ers. Furthermore, as we discuss, employees’ perceptions of HR practices are likely to influence and be influenced by the experiences and perceptions of their coworkers. Thus, in this article, we consider the proximal effects of employees’ perceptions of HR practice use in the aggregate (at the job group level). Although recent studies have considered various employee attitudi- nal and behavioral outcomes of high-performance HR systems, research rem\ ains scarce explicitly examining, in the same study, the impact of HR practices on both these outcomes.

Thus, the causal sequence by which high-performance HR practices are likely to affect key employee attitudes and then employee behaviors (likely then, as we argue) requires further theoretical and empirical attention, which we provide. Specifically, we consider the attitudinal outcome of affective organizational commitment and the behavioral (or behavior-oriented) outcomes of organizational citizenship behavior, intent to remain with the organization, and absenteeism, for a number of reasons—namely, their relevance, importance, and generaliz- ability. First, high-performance HR systems rely on the creation of a mutual in\ vestment- based employment relationship, wherein an organization invests in workforce skills and opportunities and, in turn, expects employees to be qualified and motiva\ ted to make valuable work-related investments in the organization (Huselid, 1995). Affective commitment reflects a likely attitudinal consequence of this type of positive mutual social \ exchange from the per - spective of an employee. Organizational citizenship behavior, intent to remain with the organization, and attendance at work (the opposite of absenteeism) likely represent behavioral manifestations of employees’ affective commitment to the organization, thereby suggesting relevance of these outcomes to a high-performance HR approach. Second, t\ hese behavioral at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from 370 Journal of Management / February 2013 outcomes have been linked to harder measures of performance at higher or\ ganizational levels— for example, organizational citizenship behavior (for a review, see Podsakoff, MacKensie, Paine, & Bachrach, 2000), intent to remain and turnover (Shaw, Gupta, & Delery, 2005), and absenteeism (Mason & Griffin, 2003)—thereby suggesting their importance and value in organizations, beyond the individual employee level. Finally, these outcomes represent employee behaviors that are likely to affect performance across a variety of organizational settings; that is, high levels of employee engagement in extra-role beha\ viors, regular employee attendance, and low voluntary turnover intentions based on affective commitment to an organization are likely to benefit performance outcomes across industries,\ business strate- gies, and cultures—all of which points to the generalizability of the\ se outcomes in terms of importance and relevance across many contexts.Although perceptions are formed and necessarily assessed at the individu\ al level, we focus our theoretical predictions and empirical analyses on the impact o\ f aggregated percep- tions of HR practice use throughout a job group on employees’ attitudinal and behavioral outcomes, for several reasons. First, some level of consensus in percept\ ions of HR practices in a collective (e.g., a job group) is likely to reflect a common set \ of beliefs concerning the nature of the exchange relationship and the cause-and-effect principles governing that col- lective and is thus likely to be important in eliciting desired outcomes\ in employees working within that context (Bowen & Ostroff, 2004). Furthermore, the concern in SHRM theory and research with higher-level performance outcomes requires some consistency in employees’ perceptions of and reactions to HR practices at a group level; otherwise\ , aggregate perfor - mance effects of any significance would fail to emerge. Second, given the assessment of perceptions, attitudes, and behaviors at the employee level, a strictly \ individual-level analy- sis could lead to a risk of common method variance and thus potentially \ result in artificially inflated observed relationships among variables; aggregating HR practice\ perceptions to the job group level using a split-sample analytical approach mitigates this \ concern (Gerhart, 2008; Ostroff, Kinicki, & Clark, 2002). Finally, employee perceptions of HR practices are likely to vary at the job gro\ up level as a result of job group–level variance in intended, actual, and perceiv\ ed HR practice employ- ment (Nishii & Wright, 2008). Nishii and Wright (2008) argued that intended HR policy is likely to exist at the job group level—reflecting the selection of di\ fferent HR practices for implementation across job groups (Wright et al., 2001); for example, an upper-level manager might determine that salespeople, but not administrative staff, should receive performance- based compensation. Then, given intended HR policy, supervisors who manage at the job group level are likely to differ in whether and how they implement intended practices— thereby reflecting variation in actual HR practice employment (Nishii &\ Wright, 2008); that is, supervisors may fail to implement intended practices or may employ intended practices in a manner that is inconsistent with the intent of the underlying polic\ y. Finally, employees are likely to differ in how they experience and interpret the HR practices with which the\ y and their coworkers are managed (Nishii et al., 2008), thereby reflect\ ing variance in per - ceived HR practice employment. Although individual employees form these practice perceptions, a social information–processing approach suggests that work-relate\ d percep- tions of this nature are filtered through the contextual influences and \ collective sense-making efforts of the group of employees with whom an individual most often works\ and interacts at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from Kehoe, Wright / Impact of High-Performance Human Resource Practices 371 (Bowen & Ostroff, 2004; Jacofsky & Slocum, 1988), which is more likely to include employees in the individual’s immediate job group than employees in other job groups or employees who work in other areas of the organization or unit (Salancik & Pfeffer, 1978; Schneider & Reichers, 1983). Furthermore, in cases in which employees have not had or cannot recall personal experiences with a particular practice, they are \ likely to rely on the apparent experiences of their coworkers in forming judgments concerning \ that aspect of their employment relationship. Thus, we focus on the relationships between employees’ aggregated perceptions of the extent of high-performance HR practice use\ in their job groups and individual employee outcomes, while assessing deviations in individu\ als’ attitudinal and behavioral outcomes from job group means in the same model.As noted, we focus on the employee outcomes of affective commitment, organizational citizenship behavior, intent to remain with the organization, and absenteeism. Specifically, given the principles of social exchange theory (Blau, 1983), we propose a model in which affective commitment mediates the relationships between employees’ perceptions of high- performance HR practice use and employees’ organizational citizenship behavior, intent to remain with the organization, and absenteeism. Although managers may have multiple approaches to HR management from which to choose, we are interested here in employees’ perceptions of the use of high-performance HR practices across job group\ s in an organiza- tion, given our interest in the present outcomes and our belief that hig\ h-performance HR practices are likely to be effective in eliciting these responses from employees; that is, an interest in a different set of employee outcomes would have led us to consider employees’\ perceptions of other HR approaches as appropriate. Thus, as with approaches in previous research assessing HR practice use (e.g., MacDuffie, 1995; Youndt, Snell, Dean, & Lepak, 1996), we operationalize HR practice perceptions along a continuum, and we present and test predictions indicating that the employee outcomes of interest will \ vary with the extent to which employees perceive that high-performance HR practices are emplo\ yed in their job groups.

High-Performance HR Practices and Affective Commitment Allen and Meyer’s three-component model of organizational commitment (1990) con- ceptualizes commitment as consisting of three dimensions: affective, normative, and con - tinuance commitment. The model has been supported by factor analyses and by research confirming that the three components relate differently to various antecedents and conse- quences (cf. Meyer, Stanley, Herscovitch, & Topolnytsky, 2002). In this study, we focus on affective commitment—or positive affection toward the organization, as reflected in a desire to see the organization succeed in its goals and a feeling of pride at being part of the organization (Cohen, 2003)—and we do so for several reasons. First, a\ ffective com- mitment has been shown to be affected by employees’ work and organizational experiences (e.g., HR practices; Meyer et al., 2002). Second, affective commitment has been demon- strated to relate strongly and consistently to desired work outcomes, su\ ch as low absentee- ism and organizational citizenship behaviors (Meyer et al., 2002), and its role \ as a key linking mechanism between high-performance HR practices and higher level performance at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from 372 Journal of Management / February 2013 outcomes has been supported by previous research, pointing to its establ\ ished importance for the SHRM approach (e.g., Gong et al., 2009).Specifically, social exchange theory suggests that individuals are drawn to particip\ ate and invest in rewarding relationships, after which they become bound to retu\ rn benefits or favors to their partners in exchange (Blau, 1983). High-performance HR practi\ ces, if implemented effectively, are likely to cause employees to perceive that their exchange relation\ ship with the organization is characterized by a supportive environment based on investm\ ents in employee skills, regular unbiased performance feedback, availability of \ fair and attractive rewards for performance—including compensation and advancement opportunities—and mutual efforts toward meaningful goals (which employees may have helped in devel\ oping; Wright, Gardner, & Moynihan, 2003). In return, employees are likely to feel an obligat\ ion to the organization’s goals and so develop an affective bond with the organization itself— which may be expressed as affective commitment (Cohen, 2003). Hypothesis 1: Employees’ perceptions of high-performance HR practice use will be positively rela\ ted to affective commitment. The Mediated Effect of High-Performance HR Practices on Behavioral Outcomes We suggest that employees who perceive to be managed with high-performanc\ e HR practices are likely to express increased affective commitment toward the organization based on a sort of obligatory reciprocation in their exchange relationship (s\ ocial exchange theory); employees’ increased levels of commitment are then likely to affect important work behav- iors, for at least two reasons. First, the attitudinal response of increased affective commit- ment alone is not likely to provide a balance in the benefits received b\ y each party in the exchange relationship (i.e., the employees and the organization). Some further contribution on the part of employees would be required to level the field, given the relatively large orga- nizational investment and great number of employee benefits associated w\ ith a high-perfor - mance HR system. Second, employees who are committed to an organization are likely to more naturally behave in ways that reflect this affective bond. In particular, affectively com - mitted employees are likely to act in ways that are in the best interest\ of the employer—spe- cifically, through demonstrating, in their work behaviors, a personal connection \ and devotion to the organization’s activities and goals (Mayer & Schoorman, 1992; Mowday, Porter, & Steers, 1982).

Organizational citizenship behavior. Organizational citizenship behavior has been defined as “individual behavior that is discretionary, not directly or explicitly recognized by the formal reward system, and in the aggregate promotes the efficient and effective functioning of the organization” (Organ, 1988, p. 4). This is distinguished from employees’ task performance in that it reflects work activities that do not fall within the realm of\ formal job requirements (Mayer & Schoorman, 1992) and are not performed on the basis of a moti\ vation to obtain immediate rewards or avoid punishment (Shore & Wayne, 1993). Thus, whereas previous at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from Kehoe, Wright / Impact of High-Performance Human Resource Practices 373 research has linked high-performance HR practices and organizational citizenship behavior (e.g., Sun et al., 2007), this relationship likely relies on some inte\ rvening attitudinal explanation that does not directly depend on immediate rewards from the organization—such as high- performance HR practices (e.g., extensive training, performance-based i\ ncentives)—to increase extra-role behavior. In particular, a high-performance HR–organizational citizenship behavior linkage is likely better explained indirectly, through the impact of high-performance HR practices on employees’ affective commitment predicted by social exchange theory. Specifically, as Scholl (1981) suggested, commitment serves to maintain behavioral d\ irection in the absence of rewards. Thus, employees who are committed to the organization are likely to be eager to contribute to the organization’s goals and are more likely than others to extend their efforts beyond their required task performance to further the effectiveness of the organization, even if they do not expect to be directly rewarded for this behavior on the b\ asis of formal HR practices. Organizational citizenship behaviors provide a straightforward means for c\ ommitted employees to make such optional contributions to the firm. For these rea\ sons—in addition to support from previous research demonstrating a positive relationship bet\ ween affective commitment and organizational citizenship behavior (for a review, see Podsakoff et al., 2000)—we make the following prediction: Hypothesis 2a: Affective commitment will mediate the positive relationship between high-p\ erfor- mance HR practice perceptions and organizational citizenship behavior.

Intent to remain with the organization. Sun et al. (2007) noted that although high- performance HR practices have been empirically linked to retention and turnove\ r (e.g., Batt, 2002; Guthrie, 2001; Huselid, 1995), the underlying mechanisms of this \ relationship remain uncertain. Sun et al., relying on social exchange theory, examined service-oriented organi- zational citizenship behavior as a mediator of the high-performance HR–turnove\ r relationship.

However, the authors’ explanation relied less on the organizational citizenship behaviors per se and more on employees’ affective attachments that led to these behaviors. Thus, relying again on the social exchange perspective, we argue that affective commitment better explains (indirectly) the relationship between high-performance HR practices and employees’\ intent to remain with the organization. Given that affectively committed employees under a high-performance HR system are likely to hold deep bonds with the organization and feel both eager and obliged to contribute to organizational goals (Blau, 1983; Cohen, 2003), they are likely to expre\ ss an intent to remain that is greater than that of less committed employees, for a few reasons\ . First, employees can contribute to organizational effectiveness only as they continue their employment relation- ship; so, a desire to help achieve organizational goals aligns with an intent to remain with the organization. Second, employees are likely to choose to leave an organization when they think that incentives to stay no longer match the contributions they make (March & Simon, 1958). The desire to leave thus represents a form of withdrawal (Sheridan, 1985) and lik\ ely reflects a lack of emotional attachment to the organization and its goals. Thus, we make the following prediction based on the perceived obligations and affective attachments of commit- ted employees in the high-performance HR context, in addition to empirical\ support for a sig- nificant relationship between affective commitment and both intent to remain and turnover at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from 374 Journal of Management / February 2013 (e.g., Angle & Perry, 1981; Mayer & Schoorman, 1992; Ostroff, 1992; for a related meta- analysis, see Harter, Schmidt, & Hayes, 2002): Hypothesis 2b: Affective commitment will mediate the positive relationship between high-p\ erfor- mance HR practice perceptions and intent to remain with the organization.

Absenteeism. Similar to turnover, absenteeism has been proposed as a form of withdrawal behavior (Sheridan, 1985). Whereas individual HR practices, such as bonuses for perfect attendance, might directly affect absenteeism, a system of high-performance HR practices more likely influences absenteeism indirectly through affective commitment. For instance, under a high-performance HR system, extremely high levels of absenteeism\ may directly affect the benefits that employees obtain, but employees who are absent to\ a moderate extent will still likely be eligible to participate in decision making, complet\ e required tasks, attend available training, and enjoy other benefits of high-performance HR prac\ tices. Thus, the immediate rewards provided by a high-performance HR system are not likel\ y to directly account for significant reductions in absenteeism (other than perhaps a\ t extreme levels). Instead, as consistent with the affective bonds and perceived obligations predicted by the social exchange perspective, increased affective commitment is more likely to account for lower absenteeism in high-performance HR contexts. Specifically, employees can contribute to organizational goals only when they are present at work. Furthermore, they\ can provide their organization benefits (which are likely to be perceived as owed; Blau, 19\ 83) only through their activities on the job. Thus, affectively committed employees’ perceived obligations and desires to work toward organizational goals in the high-performance HR context lead us to the following prediction:

Hypothesis 2c: Affective commitment will mediate the negative relationship between high- performance HR practice perceptions and reported absences in the previou\ s year. Method Sample and Research Procedures Data were collected in a large food service organization from employees working in each of 56 self-contained business units, which consisted of up to 10 jo\ b categories:

administrative, day warehouse, drivers, marketing, marketing associates,\ merchandising, night warehouse, administrative supervisors, operational supervisors, an\ d sales supervi- sors. HR directors were asked to administer surveys to a randomly select\ ed group of 20% or more of the employees in their unit. The corporation from which the sample was drawn provides the opportunity for considerable variance with regard to actual\ HR practice implementation across job groups and business units based on its “ear\ ned autonomy” principle, which allows managers within each business unit to manage the\ ir various employees as they deem most appropriate. Consequently, employees who work in the same job category but in different business units may be managed with different HR at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from Kehoe, Wright / Impact of High-Performance Human Resource Practices 375 practices, as may those who are working in different job categories in the same business unit.

Measures High-performance HR practice perceptions. We assessed employees’ perceptions of high- performance HR practices using 15 items compiled from previous empirical\ research in the SHRM literature aimed at improving employees’ KSAs (knowledge, skills, and abilities) and motivation and opportunity to perform. Specifically, we tapped into perceptions concerning selectivity in staffing procedures by asking employees about the use of formal selection te\ sts (Huselid, 1995; Way, 2001), the use of structured employment interviews (Huselid, 1995),\ and the general quality of selected job candidates (Combs et al., 2006)\ for employees in their jobs, given that selectivity in staffing is likely to increase the average level of knowledge and skills held by employees. We assessed opportunities for employee participation by asking about the extent to which employees perceived the existence of formal participation pro- cesses (Sun et al., 2007), implementation of changes based on employee\ suggestions, fair complaint procedures, and employee autonomy in job design modifications \ (Huselid, 1995), namely, because employees who believe that their voice counts in the organization and that they can make a difference with their work are likely to perceive greater opportunity and motivation to make meaningful contributions in the employment context. We measured the use of high-performance compensation practices by asking employees to ra\ te the availability of bonuses based on group and individual performance outcomes (Sun et al., 2007), merit- based pay raises, and overall high-pay levels for the type of work completed in their jobs (Way, 2001) because (a) high pay is likely to attract competent employees and (b) performance- based rewards are likely to increase motivation to perform. Additional practices that we assessed included formal performance evaluations (to increase motivatio\ n and aid in develop - ment; Huselid, 1995), regular information-sharing communication (to in\ crease employees’ sense of involvement and importance with respect to organizational issues; Combs et al., 2006), merit-based promotion opportunities (to increase motivation to perform; Huselid, 1995; Sun et al., 2007), and extensive formal training (to improve skills and ab\ ilities and to demonstrate the commitment of the organization to its employees; Combs et al., 2006; Huselid, 1995). Twelve items required yes/no responses and asked whether or not each practice existed in an employee’s job group (e.g., “Pay raises for associates in this job are based \ on job perfor - mance”); three additional questions were recoded for consistency wit\ h this format. A com- plete list of items used for each measure appears in the appendix. We created an additive index of HR practice items, resulting in a continu\ ous measure reflecting each employee’s perceptions that particular practices from a high-performance HR system were used to manage employees in his or her job group. Then, as noted, we aggregated individuals’ HR practice perception indices to the split job group level; that is, w\ e computed the mean of these indices for half the employees in each job group. The use of an additive index is consistent with previous research (e.g., MacDuffie, 1995; Youndt et al., 1996) and with our conceptual arguments regarding the additive nature of the effects of the high-performance HR system (Delery, 1998). In particular, we contend that each high-performance HR practice at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from 376 Journal of Management / February 2013 is likely to uniquely contribute to the facilitation of a supportive mut\ ual investment-based exchange relationship between a group of employees and their organization. For example, regarding the multiple effects of high-performance compensation practices, an organization’s provision of high pay is likely to enhance the skill levels of its workforce, but the provision of pay based on group performance outcomes is likely to improve cooperation\ in the workplace— thus reflecting two unique contributions of pay practices to a supportiv\ e exchange relationship.We employed a referent–shift composition model for the HR practice per\ ception measure, for two reasons. First, we expected HR practice implementation \ to occur at the job group level rather than the individual level, thereby making the job\ group (as opposed to the individual) the appropriate referent of the HR construct\ (Chen, Mathieu, & Bliese, 2004). Relatedly, in cases in which an employee does not recall a personal experience with a particular HR practice, he or she is likely to rely on\ the apparent experiences of his or her coworkers when making inferences related to th\ at aspect of the employment relationship. Second, this composition model and the correspo\ nding mea- surement technique allowed us to test our expectation that employees in \ a job group share consistent perceptions of HR practice employment (Kozlowski & Kle\ in, 2000).

The HR practice perception index showed good internal consistency reliability (a = .80). Thus, a high score on the HR practice perception index indicates employee percep- tions of extensive use of high-performance HR practices. We calculated intraclass cor - relation coefficients (ICCs) for high-performance HR practice perceptions at the split job group level, given that this was the level to which we sought to agg\ regate the data.

ICCs for the HR practice perception index (ICC 1 = .30, ICC 2 = .86) provided support for our aggregation of employee responses to the split job group level.

Affective commitment. We measured affective commitment using five items that were consistent with items used in previous research (e.g., Meyer & Allen, 1997; Porter, Steers, Mowday, & Boulean, 1974). Sample items include “I feel a strong sense of belonging to this organization” and “I am willing to work harder to help this company \ succeed.” All the affec- tive commitment items appear in the appendix. Employees were asked to re\ port, on a 5-point scale, the extent to which they agreed with each commitment statement; w\ e then computed a mean commitment scale for each individual by averaging the individual’\ s responses to the five commitment items. Support for combining the commitment items into a\ single scale is evident in the high level of internal consistency reliability (a = .89).

Organizational citizenship behavior. Six items, which appear in the appendix, were used to assess organizational citizenship behavior. These items were drawn from a large number of potential organizational citizenship behavior items (Podsakoff et al., 2000) based on col- laboration on the employee survey between the researchers and the organization. Sample items included “I ‘touch-base’ with my coworkers before initiating actions that might affect them” and “I willingly share my expertise with my coworkers.” Employees were asked to indicate, on a scale from 1 to 5, the extent to which they agreed with e\ ach statement. Again, individual scales for organizational citizenship behavior were computed with each employ- ee’s mean score on the six organizational citizenship behavior items, as supported by good internal consistency reliability (a = .79). at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from Kehoe, Wright / Impact of High-Performance Human Resource Practices 377 Intent to remain with the organization. We relied on employees’ reported intentions to remain with the organization. Although we realize this is not a true behavioral outcome measure, it does measure an intention to engage in a behavior. In addition, intentions to turnover (or, in this case, the inverse) have been shown to be a strong predictor of act\ ual turnover (Van Breukelen, Van der Vlist, & Steensma, 2004). Employee respondents were asked to indicate, on a scale of 1 to 5, the extent to \ which they agreed with four statements regarding their intent to remain with t\ he organization.

Sample representative items included “I plan to spend my career at [t\ his organization]” and “I intend to stay at [this organization] for at least the next 12 months.” Item scores were again averaged at the individual level to calculate a single “intent to remain” scale, which was again supported by high internal consistency reliability (a = .83). The four items for this scale appear in the appendix.

Absenteeism. To assess absenteeism, we asked employee respondents to report the number\ of days they had missed work in the last calendar year.

Control variables. Because different types of jobs vary in their relative desirability and likelihood of attracting committed, motivated, and diligent employees, w\ e controlled for job category as a fixed effect in our analyses, with sales supervisors as the baseline category.

Note that employees are grouped and supervised on the basis of their job categories in this organization such that an employee’s job group includes all the employees in his or her job category who also work in his or her business unit. We included business unit and job group as random covariates in our analy\ ses to account for unmeasured influences on the dependent variables that may be present\ at those levels. Analyses and Results Table 1 presents means, standard deviations, intercorrelations, and alpha\ s among all the variables examined. Given recent recommendations made by Zhang, Zyphur, and Preacher (2009), we tested our model using CWC(M) mediation analysis—that \ is, centered within context with reintroduction of the subtracted means at Level 2 (Kreft &\ de Leeuw, 1998)— based on a 2–1–1 model. Specifically, we accounted for the multilevel nature of our data and model using a group mean–centering approach (as opposed to a grand m\ ean–centering approach) in which group-level means are included in the estimation of \ the between-group components of the relationships between variables; note that the group-l\ evel means are removed from the centered variables that reflect lower-level (individual-level) deviations from group means. The 2–1–1 specification reflects the group-level nature of the ind\ epen- dent variable (i.e., aggregated HR practice perceptions) and the indiv\ idual-level nature of the mediator and dependent variables. We refer readers to Zhang et al. (2009) for a thorough explanation of the CWC(M) approach, including its details and advantag\ es. Briefly, though, a primary benefit of this analysis over grand mean centering is that the\ CWC(M) method provides unique estimates for the within- and between-group coefficients of the mediator (i.e., the effects), whereas grand mean centering holds these two coefficients equal, thus at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from 378 Journal of Management / February 2013 reflecting an unnecessary constraint and yielding confounded estimates of the mediator’s effect on the dependent variable (Zhang et al., 2009).Figure 1 depicts the general CWC(M) 2–1–1 model (ignoring, for \ the moment, the As and the B above the boxes). In the present study, Level 2 represents the job group level, and Level 1 represents the individual level. As Zhang et al. (2009) noted, because variance in the inde- pendent variable exists solely at Level 2 in a 2–1–1 model, the in\ dependent variable can account only for between-group variation in the mediator; thus, we can meaningfully focus on only the between-group mediation effect (i.e., the extent to which job group–level percep- tions of high-performance HR practices create variation across job group\ s in the behavioral and behavior-oriented outcomes of individual employees, through their effect on affective commitment). Note, however, that it is still important to estimate the within-group effect of the mediator on the dependent variables to ensure precision of the overa\ ll model estimation (Zhang et al., 2009). Empirical support for this model would indicate that the Level 2 compone\ nt of affective commitment mediates the relationship between the Level 2 independent var\ iable (i.e., HR practice perceptions) and the Level 1 dependent variable, even after th\ e effect of the Level 1 component of affective commitment is taken into account. The A and B designations above the boxes in Figure 1 reflect the split-sample\ analytical approach that we employed to mitigate concerns associated with common method bias. In particular, boxes designated A reflect variables that we drew from employees in one half of each job group, and boxes designated B reflect variables that we drew from employees in the other half of the job group. We chose this pattern of sample splitting to stack the odds against finding support for our predicted mediated model. Specifically, by drawing the group-centered mediator from the same half of each job group as the dependent variable \ while drawing the Level 2 mean of the mediator from the other half of each job group, we m\ aximized the variation in the dependent variable likely to be explained by group-cent\ ered commitment while minimizing the proportion of variance likely to be explained by th\ e job group mean of commitment (i.e., the operational mediator in this study), thus provid\ ing a conservative test of the predicted mediation effect. Table 1 Means, Standard Deviations, Reliabilities, and Correlations Variables M SD 1 2 3 4 1. HR practice perceptions 0.57 0.16 (.80) 2. Affective commitment 4.20 0.41 .62** (.89) 3. Organizational citizenship behavior 3.97 0.54 .23** .26** (.79) 4. Intent to remain 3.80 0.84 .33** .36** .43** 5. Absenteeism 2.16 4.88 –.22 ** –.20 ** –.07 ** –.13 ** Notes: Scale reliabilities are reported on the diagonal. HR (human resource) \ practice perceptions and affective commitment descriptives and the correlations between them were computed \ at the job group level; all other descrip- tives and correlations were computed at the individual level. Consistent\ with subsequent analyses, affective com- mitment values for the computation of descriptives and correlations were\ drawn from employees in Split B; values for all other variables were drawn from employees in Split A.

*p < .05. **p < .01. ***p < .001. at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from Kehoe, Wright / Impact of High-Performance Human Resource Practices 379 Hypothesis 1 predicted that employees’ perceptions of high-performance HR practices would be positively related to affective commitment. The bivariate correlation in Table 1 shows a correlation of .62 between HR and commitment in the A-B analysis. Table 2 dis- plays the results of the mixed-model analyses in which affective commitment is the depen- dent variable. The baseline model, Model 1, includes only the job category controls and\ the job group and business unit covariates. High-performance HR practice per\ ceptions were included as a fixed effect in Model 2. Consistent with the bivariate correlation between high-\ performance HR practices and affective commitment, high-performance HR practice per - ceptions were a significant predictor of affective commitment (b = 0.40, p < .001), thereby providing strong support for Hypothesis 1. Hypotheses 2a and 2b predicted that affective commitment would mediate the positive relationships between high-performance HR practice perceptions and (a) organizational citizenship behavior and (b) intent to remain with the organization, respectively. To test these hypotheses, we used the method suggested by Baron and Kenny (1987), wh\ o suggested that if the independent variable (HR practices) exhibits (a) a significan\ t coefficient when the mediator (affective commitment) is not included in the equation and (b) a nonsignificant coefficient after controlling for the mediator, then full mediation exists. If the regression coefficient maintains significance but is reliably reduced (reduction of th\ e coefficient can be tested with the Sobel [1982] test), then a partial mediating relationsh\ ip exists. We used a mixed-model analysis to test Hypotheses 2 and 3, including only\ job category controls and job group and business unit covariates in Model 1, adding i\ n high-performance HR practice perceptions in Model 2, and estimating the full model with these predictors as well as the Level 2 mean of affective commitment and group mean centered affective com- mitment in Model 3. As displayed in Tables 3 and 4, Model 2 presents results of the tests of the relationships between high-performance HR practice perceptions and b\ oth organizational Figure 1 CWC(M) Mediation Analysis Based on a 2–1–1 Estimated Model Independent Va riable Mean of Mediator at Level 2 Group-centered Mediator Dependent Variable Le vel 2 Le vel 1 AA AB Note:

CWC(M) = centered within context with reintroduction of the subtracted means at \ Level 2. at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from 380 Journal of Management / February 2013 citizenship behavior and intent to remain with the organization, respectively. As can be seen in the tables, significant variance in both these variables is explained\ by high-performance HR practice perceptions (b = 0.24, p < .001; β = 0.28, p < .001, respectively), thus satisfying the first requirement for mediation for these two dependent variables. As already noted, high-performance HR practice perceptions were significantly related to a\ ffective commit- ment, satisfying the mediation requirement that the independent variable predict the mediator. In Tables 3 and 4, Model 3 presents the results of the final step of our med\ iation analysis.

As reflected by the significance of the HR practice perception coefficient in Model 3 in Table 3 (b = 0.05, p < .05), affective commitment does not completely mediate the relationship between HR practice perceptions and organizational citizenship behavior. However, a Sobel test demonstrates that partial mediation does occur (t = 6.81, p < .001), with affective com- mitment mediating 68% of the total effect of HR practice perceptions on organizational citi- zenship behavior, thus reflecting partial support for Hypothesis 2a. As the lack of significance in the HR practice perception coefficient (b = 0.00) in Model 3 of Table 4 demonstrates, affec- tive commitment completely mediates the relationship between HR practice\ perceptions and intent to remain with the organization, thus providing full support for Hypothesis 2b. Hypothesis 2c predicted that affective commitment would mediate the negative relationship between HR practice perceptions and absenteeism. Because absenteeism was\ an overdispersed count variable, we employed negative binomial regression with a log link\ to test predictions concerning this outcome. To test Hypothesis 2c, we included job category controls, in addition to job group and business unit as subject identifiers (clustering), to\ account for the multilevel Table 2 Results of Mixed-Model Analyses Predicting Affective Commitment Model 1 Model 2 Parameter Estimate 95% CI Estimate 95% CI Estimates of fixed effects Intercept 1.06 *** (0.12) 0.81 1.30 0.58 *** (0.12) 0.34 0.82 Job Category 1 –1.00 *** (0.15) –1.28 –0.71 –0.21 (0.17) –0.54 0.13 Job Category 2 –2.01 *** (0.15) –2.30 –1.72 –1.13 *** (0.18) –1.48 –0.79 Job Category 3 –1.33 *** (0.15) –1.63 –1.04 –0.61 *** (0.17) –0.94 –0.28 Job Category 4 –0.46 ** (0.15) –0.76 –0.16 –0.18 (0.15) –0.48 0.12 Job Category 5 –0.53 *** (0.15) –0.82 –0.24 –0.43 ** (0.14) –0.71 –0.15 Job Category 6 –0.71 *** (0.15) –1.00 –0.42 –0.19 (0.16) –0.49 0.12 Job Category 7 –1.87 *** (0.15) –2.16 –1.58 –1.20 *** (0.16) –1.52 –0.88 Job Category 8 –0.24 (0.17) –0.57 0.08 0.08 (0.16) –0.25 0.40 Job Category 9 –0.28 (0.15) –0.57 0.02 –0.05 (0.15) –0.35 0.24 Human resource practice perceptions 0.40 *** (0.05) 0.30 0.49 Estimates of covariance parameters Residual 0.55 *** (0.04) 0.48 0.63 0.52 *** (0.04) 0.45 0.59 Business unit 0.23 *** (0.06) 0.14 0.38 0.10 (0.03) 0.05 0.20 Model fit statistics Schwarz’s Bayesian criterion 1,218.66 1,163.05 Notes: Standardized coefficients are reported (standard errors in parentheses). CI = confidence interval; HR = human resource. *p < .05. **p < .01. ***p < .001. at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from 381 Table 3 Results of Mixed Models Analyses Predicting Organizational Citizenship Behavior Model 1 Model 2 Model 3 Parameter Estimate 95% CI Estimate 95% CI Estimate 95% CI Estimates of fixed effects Intercept 0.51 *** (0.06) 0.39 0.63 0.22 *** (0.06) 0.10 0.34 0.19 ** (0.06) 0.08 0.30 Job Category 1 –0.42 *** (0.07) –0.57 –0.28 0.07 (0.08) –0.09 0.23 –0.05 (0.07) –0.19 0.09 Job Category 2 –0.77 *** (0.08) –0.93 –0.62 –0.25 ** (0.09) –0.41 –0.08 –0.12 (0.08) –0.28 0.03 Job Category 3 –0.61 *** (0.07) –0.74 –0.47 –0.19 * (0.07) –0.33 –0.04 –0.19 ** (0.07) –0.32 –0.06 Job Category 4 –0.23 * (0.09) –0.41 –0.05 –0.02 (0.09) –0.19 0.16 –0.06 (0.08) –0.21 0.10 Job Category 5 –0.40 *** (0.07) –0.54 –0.27 –0.35 *** (0.06) –0.48 –0.23 –0.29 *** (0.06) –0.40 –0.18 Job Category 6 –0.58 *** (0.08) –0.75 –0.42 –0.25 ** (0.08) –0.42 –0.08 –0.29 *** (0.08) –0.44 –0.14 Job Category 7 –0.83 *** (0.07) –0.97 –0.69 –0.44 *** (0.08) –0.59 –0.29 –0.25 *** (0.07) –0.39 –0.11 Job Category 8 –0.01 (0.10) –0.21 0.19 0.18 (0.10) –0.02 0.37 0.12 (0.09) –0.05 0.30 Job Category 9 0.01 (0.09) –0.16 0.18 0.15 (0.08) –0.01 0.31 0.11 (0.08) –0.03 0.26 Human resource practice perceptions 0.24 *** (0.02) 0.22 0.28 0.05 * (0.02) 0.01 0.10 Level 2 affective commitment 0.26 *** (0.02) 0.22 0.29 Group-centered affective commitment 0.42 *** (0.01) 0.40 0.44 Estimates of covariance parameters Residual 0.90 *** (0.02) 0.88 0.93 0.89 *** (0.02) 0.86 0.92 0.71 *** (0.01) 0.69 0.74 Business unit 0.02 ** (0.01) 0.01 0.04 0.01 (0.00) 0.00 0.02 0.00 (0.00) 0.00 0.01 Job group 0.02 *** (0.01) 0.02 0.04 0.01 ** (0.00) 0.01 0.03 0.01 ** (0.01) 0.01 0.02 Model fit statistics Schwarz’s Bayesian criterion 19,509.05 19,009.67 17,481.83 Notes: Standardized coefficients are reported (standard errors in parentheses). CI = confidence interval. *p < .05.

**p < .01.

***p < .001. at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from 382 Table 4 Results of Mixed Models Analyses Predicting Intent to Remain with the Organization Model 1 Model 2 Model 3 Parameter Estimate 95% CI Estimate 95% CI Estimate 95% CI Estimates of fixed effects Intercept 0.54 *** (0.07) 0.41 0.67 0.22 ** (0.07) 0.09 0.35 0.16 *** (0.04) 0.07 0.25 Job Category 1 –0.65 *** (0.08) –0.80 –0.50 –0.10 (0.08) –0.26 0.07 –0.27 *** (0.06) –0.38 –0.16 Job Category 2 –0.81 *** (0.08) –0.96 –0.66 –0.21 * (0.09) –0.39 –0.04 –0.02 (0.06) –0.14 0.11 Job Category 3 –0.60 *** (0.07) –0.75 –0.46 –0.12 (0.08) –0.27 0.03 –0.10 * (0.05) –0.21 0.00 Job Category 4 –0.26 ** (0.09) –0.44 –0.08 –0.06 (0.09) –0.24 0.11 –0.12 (0.06) –0.24 0.00 Job Category 5 –0.26 *** (0.07) –0.40 –0.13 –0.21 ** (0.07) –0.33 –0.08 –0.11 * (0.04) –0.20 –0.02 Job Category 6 –0.50 *** (0.08) –0.67 –0.34 –0.12 (0.09) –0.29 0.04 –0.18 ** (0.06) –0.29 –0.06 Job Category 7 –1.07 *** (0.07) –1.21 –0.93 –0.62 *** (0.08) –0.78 –0.47 –0.32 *** (0.06) –0.43 –0.21 Job Category 8 –0.12 (0.10) –0.31 0.08 0.10 (0.10) –0.09 0.29 0.02 (0.07) –0.11 0.16 Job Category 9 –0.16 (0.09) –0.33 0.02 0.00 (0.08) –0.16 0.17 –0.05 (0.06) –0.16 0.07 Human resource practice perceptions 0.28 *** (0.02) 0.23 0.33 0.00 (0.02) –0.04 0.03 Level 2 affective commitment 0.39 *** (0.02) 0.36 0.42 Group-centered affective commitment 0.63 ** (0.01) 0.62 0.65 Estimates of covariance parameters Residual 0.82 *** (0.01) 0.79 0.85 0.82 *** (0.01) 0.79 0.85 0.41 *** (0.01) 0.4 0.43 Business unit 0.05 *** (0.01) 0.03 0.08 0.02 ** (0.01) 0.01 0.04 0.01 ** (0.00) 0 0.01 Job group 0.04 *** (0.01) 0.03 0.06 0.02 *** (0.01) 0.02 0.04 0.01 ** (0.00) 0 0.02 Model fit statistics Schwarz’s Bayesian criterion 18,909.34 18,497.29 13,733.47 Notes: Standardized coefficients are reported (standard errors in parentheses). CI = confidence interval. *p < .05.

**p < .01.

***p < .001. at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from Kehoe, Wright / Impact of High-Performance Human Resource Practices 383 data structure in our baseline model (Model 1); we also included HR pr\ actice perceptions as a predictor in Model 2, and we added Level 2 affective commitment means and group mean– centered affective commitment in Model 3. The results for these analyses appear in Table 5.

As can be seen in Models 2 and 3 in this table, HR practice perceptions \ are a significant predic- tor of absenteeism both before and after affective commitment is included in the estimated model (b = –0.19, p < .001; b = –0.13, p < .05, respectively), reflecting a lack of support for full mediation of HR perception–absenteeism relationship. Furthermore\ , affective commitment is not a significant predictor of absenteeism in Model 3 (i.e., when included in the model with HR practice perceptions), thereby suggesting that partial mediation doe\ s not occur. Thus, we find no support for Hypothesis 2c. Discussion The primary purpose of this study was to explore the relationship betwee\ n employees’ perceptions of the use of high-performance HR practices at the job group level and important attitudinal and behavioral employee outcomes. Specifically, substantial evidence based on previous empirical work in the SHRM realm supports the relationship betw\ een high- performance HR practices and organizational performance, and a smaller amount of evi- dence based on more recent studies has begun to point to a link between \ high-performance HR practices and employee outcomes; however, both these streams of work have largely depended on the use of managerial reports of HR practice use. As discussed earlier, a variety of reasons explain why managers’ reported HR practice use may not relate in expected ways to desired employee outcomes (Bowen & Ostroff, 2004; Nishii & Wright, 2008). As such, an examination of the role of employees’ collective perceptions of HR practice use in determin- ing positive outcomes represented an important need in the SHRM literatu\ re and, thus, a key contribution of the present study. To address this need, as well as recent calls in the literature, for studies assessing outcomes of employees’ HR practice perceptions and for additional empirical work assessing the proximal outcomes of high-performance HR practices, we developed and tested a model based on social exchange theory in which af\ fective commit- ment mediates the relationship between employees’ collective perceptions of high-perfor - mance HR practice use and organizational citizenship behavior, intent to remain with the organization, and absenteeism. In general, our findings support our predictions: Employees’ collective perceptions of high-performance HR practice use are positively related to affective commitment, organiza- tional citizenship behavior, and intent to remain with the organization and negatively related to absenteeism. Furthermore, affective commitment partially mediates the relationship between high-performance HR practice perceptions and organizational citizenship behavior, and it completely mediates the relationship between high-performance HR practic\ e perceptions and intent to remain with the organization. Contrary to our expectations, affective commit- ment does not mediate the HR practice perception–absenteeism relation\ ship, nor is it a sig- nificant predictor of absenteeism when HR practice perceptions are inclu\ ded in the model. In addition, the empirical support for our model reflects the importance\ of employees’ aggregated perceptions of HR practices as they are employed throughout a\ job group above and beyond the individual-level mechanisms through which HR practices ma\ y affect at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from 384 Table 5 Results of Negative Binomial Regression Analyses Predicting Absenteeism Model 1 Model 2 Model 3 Parameter Estimate 95% CI Estimate 95% CI Estimate 95% CI Estimates of fixed effects Intercept –0.13 (0.21) –0.53 0.27 0.08 (0.21) –0.34 0.50 0.10 (0.21) –0.32 0.52 Job Category 1 1.23 *** (0.22) 0.80 1.66 0.86 ** (0.25) 0.37 1.35 0.89 *** (0.25) 0.40 1.38 Job Category 2 1.36 *** (0.23) 0.91 1.80 0.92 *** (0.25) 0.43 1.41 0.86 ** (0.25) 0.37 1.36 Job Category 3 1.07 *** (0.23) 0.62 1.51 0.73 ** (0.24) 0.26 1.20 0.72 ** (0.24) 0.25 1.19 Job Category 4 0.70 ** (0.24) 0.22 1.17 0.57 * (0.24) 0.10 1.04 0.57 * (0.24) 0.11 1.04 Job Category 5 –0.15 (0.24) –0.62 0.32 –0.18 (0.24) –0.64 0.29 –0.20 (0.23) –0.66 0.26 Job Category 6 0.79 ** (0.26) 0.28 1.29 0.54 * (0.27) 0.01 1.06 0.56 * (0.27) 0.03 1.08 Job Category 7 1.53 *** (0.22) 1.10 1.95 1.21 *** (0.23) 0.75 1.67 1.13 *** (0.23) 0.67 1.59 Job Category 8 0.93 *** (0.27) 0.41 1.45 0.75 ** (0.27) 0.23 1.27 0.77 ** (0.26) 0.25 1.29 Job Category 9 0.38 (0.25) –0.10 0.87 0.31 (0.25) –0.19 0.80 0.32 (0.25) –0.17 0.82 Human resource practice perceptions –0.19 *** (0.05) –0.29 –0.09 –0.13 * (0.06) –0.25 –0.01 Level 2 affective commitment –0.10 (0.05) –0.2 –0.01 Group-centered affective commitment –0.04 (0.03) –0.1 0.02 Model fit statistics:

Quasi-likelihood under the independence 8,531.99 8,368.41 8,355.45 model criterion Notes: Predictors were standardized to allow for comparison, but the dependent \ variable was not standardized, given the need for positive integers in negative binomial regression analysis. Standard errors are in parentheses. CI = confidence interval. *p < .05.

**p < .01.

***p < .001. at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from Kehoe, Wright / Impact of High-Performance Human Resource Practices 385 individuals’ outcomes. In particular, aggregate affective commitment mediates the impact of aggregate HR practice perceptions on individual-level organizational citizenship behavior and intent to remain with the organization—even after the effects of individual-level devia- tions in affective commitment on these dependent variables are taken into account, \ thus pointing to important implications for research and practice.From a theoretical perspective, our findings represent a departure from \ most existing SHRM research in this vein. Specifically, empirical work in this area has tended to treat the employee outcomes of HR practices as employees’ responses to consistently implemented, predetermined management techniques (for an exception, see Liao et al.,\ 2009). However, the current findings suggest that employees’ perceptions of and reactions to HR practices may in fact be influenced by the perceptions and apparent experiences of\ their coworkers as well—thereby pointing to a need for SHRM scholars to focus on employe\ es’ aggregate per - ceptions of HR practice use throughout a job group. Consequently, whereas a reader of tra- ditional SHRM research might conclude that effective HR policy is essential in eliciting desired aggregate outcomes, the current findings suggest that a group of employees’ indi- vidual outcomes are likely affected by the way that the individuals in the group perceive the group to be managed as a whole. From a practical standpoint, our findings point to the need for organizations to move beyond a focus on the effective design or selection of an HR system, to include an emphasis on consistent implementation of and communication about HR practices. The results of our study indicate that employees’ perceptions about the management of their job group likely affect individual-level attitudinal and behavioral outcomes, thereby sugge\ sting that to maxi- mize the positive effects of an HR system, managers should employ HR practices consis- tently across a job group and ensure that all employees are aware of the\ practices in use. Several strengths of our study enhance the contribution of this research\ to SHRM literature and practice. First, we included attitudinal and behavioral outcomes of \ high-performance HR practice perceptions in a single model. This component of our study extends the SHRM literature in two ways. First, the examination of these employee outcome\ s addresses recent calls in the literature for additional empirical work on more proximal relationships in the HR–performance chain (e.g., Becker & Gerhart, 1996; Takeuchi et al., 2009; Wright & Gardner, 2003). In particular, we demonstrated significant relationships between employees’ HR practice perceptions and affective commitment, as well as three important behavioral outcomes—namely, organizational citizenship behavior, intent to remain with the organiza- tion, and absenteeism—which are likely to have important performance \ implications across a variety of organizational contexts. Second, by examining the relationships between at\ titu- dinal and behavioral outcomes, our study informs the field’s understanding of the causal chain by which high-performance HR practices are likely to have an impact. Bas\ ed on a sequence consistent with social exchange theory, our findings suggest that employees’ perceptions of high-performance HR practice use likely affect employees’ behaviors to at least some extent through their effect on an important attitudinal outcome—that is, affective commitment.

This is significant: Whereas previous work in this area has acknowledged the importance of social exchange processes in the sequence by which HR practices affect employee behaviors, to date, little if any work has theoretically and empirically addressed \ the role of attitudes in creating the context necessary for desirable behaviors to consistently e\ merge from the implementation of a high-performance HR system. at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from 386 Journal of Management / February 2013 In light of this consideration, our findings point to an interesting que\ stion with regard to organizational citizenship behavior and absenteeism. As noted, affective commitment par - tially mediated the relationship between HR practice perceptions and organizational citizen- ship behavior, although it mediated a large portion of this relationship (68%); it did not mediate the relationship between HR practice perceptions and absenteeism\ . Thus, it is likely that one or more alternative explanations, independent of social exchang\ e theory, provide the basis for a portion of the influence of HR practice perceptions on organizational citizenship behavior and for the entire impact of HR practice perceptions on employe\ es’ absenteeism.

Although employees’ perceptions that high-performance HR practices are in use could trans- late directly into beliefs that performing extra-role behaviors and limi\ ting absences from work will lead to better rewards, an interesting avenue for future resea\ rch would be to inves- tigate whether this is the case or if perceptions of HR practices work more indirectly, through some other attitudinal mechanisms (other than affective commitment) to affect organiza- tional citizenship behavior and absenteeism. Attitudes can be conceived as having cognitive, affective, and intentional components, and it may be that the cognitive mechanisms account for variance beyond the affective components. For example, practices such as gain sharing (which we could not measure) may not increase affective commitment, but they still might incent employees to exhibit citizenship behavior. In addition, bonuses for attendance could similarly affect attendance behavior without doing so through commitment. The importance of this question is based on the rationale that prompted this study: tha\ t attitudes are likely to precede and predict behaviors (Ajzen & Fishbien, 1980), thereby sugges\ ting that behaviors’ attitudinal antecedents are important points of focus and measurement. Another critical strength of the present study was the use of a multilev\ el approach with consideration of relationships at the job group and individual levels. S\ pecifically, by consid- ering HR practice perceptions at the job group level, we captured import\ ant variance in this variable based on a relevant organizational boundary that acknowledged the contextual influences affecting the formation of perceptions at the individual level. Furthermor\ e, by considering proximal outcomes of employees’ aggregate HR perceptions, we were able to demonstrate the following: Although individuals develop their own exchange relationships with an organization, choosing and enacting their work behaviors accordingly, they develop these relationships and enact these behaviors in the presence of coworke\ rs with whom they are likely to share frequent experiences and interactions—thus, the interpretations, attitudes, and decisions among employees in the same job group are likely shaped by\ similar influ- ences and thus exhibit some level of consistency. Our findings reflected this logic: Although individuals’ differences in affective commitment significantly predicted their behavioral outcomes, the job group mean of affective commitment fully mediated the relationship between aggregate HR perceptions and both individual organizational citizenship behavior and intent to remain with the organization, even after accounting for these effects—thus demonstrating the importance of identifying and accounting for the appropriate contextual effects in the study of HR practices and outcomes. As with most research, the results of this study should be interpreted w\ ith consideration of several limitations. First, the cross-sectional nature of the study d\ oes not allow for any conclusions regarding causal relationships. Employees were asked to report their percep- tions of current HR practices and present levels of affective commitment, intent to remain with the organization, and organizational citizenship behavior. Absenteeism measures were at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from Kehoe, Wright / Impact of High-Performance Human Resource Practices 387 based on employees’ self-reported number of missed workdays in the previous calendar year.

Although the analyses show that our proposed relationships exist with th\ e cross-sectional data set, we are not able to claim a causal relationship from these resu\ lts. Future research would benefit from testing the present study’s model within a longitudinal design.Second, the study relied on employee self-reports of employee outcomes. More accurate reports of organizational citizenship behavior and absenteeism might have been obtain\ ed through departmental records or managers; however, we were unable to obtain departmental or managerial reports for these variables. Third, we did not empirically examine the antecedents of variability in \ employees’ per - ceptions of HR practices. Although we emphasize the likely importance of consistent and effective HR practice implementation in shaping employees’ HR practice perceptions, empirical work is needed examining the relative significance of potentia\ l antecedents. Finally, the high-performance HR practices that we included in the study were n\ ot entirely representative of previously studied high-performance HR systems. For ex\ ample, because of restrictions imposed by the organization, we were unable to question employee respondents about gain sharing or profit sharing, given that such questions could prime employees to later question why such practices did not exist within the organization. However, as previously noted, the HR practices that we did include were consistent with the cor\ e elements underly- ing a high-performance HR system and with our review of previous researc\ h in this area. Overall, we believe that this study’s strengths and contribution to the SHRM literature and practice outweigh its limitations. First, we demonstrated the importance\ of employees’ per - ceptions of the HR practices used to manage their job groups in predicti\ ng attitudinal and behavioral outcomes—an oversight in the majority of previous SHRM res\ earch. Second, we shed light on the relationships between attitudinal and behavioral outcomes in the high-performance HR context—a set of relationships that previous work in this area had\ left vague or unspeci- fied. Third, we presented findings that suggest a variety of fruitful areas fo\ r future research in the SHRM realm, including the controllable and uncontrollable determinan\ ts of employees’ HR practice perceptions, the affective mechanisms by which high-performance HR practices affect absenteeism, and other attitudinal and behavioral outcomes potentia\ lly affected by HR practice perceptions and important to higher-level organizational performance outcomes. We are hopeful that future SHRM inquiry will acknowledge and attempt to inf\ orm the gaps between organizations’ HR policies and employees’ HR practice experiences and perceptions; employees cannot willfully respond to practices they do not perceive. Appendix: Items Used for All Examined Variables High-Performance Human Resource Practice Perceptions 1. Applicants for this job take formal tests (paper and pencil or work sam\ ple) before being hired. 2. Applicants for this job undergo structured interviews (job related questions, same questions asked for all applicants) before being hired. 3. Associates in this job are involved in formal participation processes su\ ch as quality improve- ment groups, problem solving groups, or roundtable discussions. (continued) at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from 388 Journal of Management / February 2013 4. Associates in this job have a reasonable and fair complaint process. 5. Associates in this job have the opportunity to earn group bonuses for pr\ oductivity, perfor- mance, or other group performance outcomes. 6. Associates in this job have the opportunity to earn individual bonuses (or commissions) for productivity, performance, or other individual performance outcomes.

7. At least once a year associates in this job receive a formal evaluation \ of their performance. 8. Associates in this job regularly receive formal communication regarding \ company goals and objectives. 9. In the last 4 months, the company has made a change in how work is compl\ eted in my depart- ment based on the suggestion(s) of an associate or group of associates\ .

10.

Pay raises for associates in this job are based on job performance.

11.

Qualified associates in this job have the opportunity to be promoted to positions of greater pay and/or responsibility within the company.

12. Associates in this job are allowed to make important work related decisi\ ons such as how the work is done or implement new ideas.

13.

The company hires only the very best people for this job.

14.

Total pay for this job is the highest for the type of work in the area.

15.

On average, how many hours of formal training do associates in this job \ receive each year? Affective Commitment 1. I am willing to work harder to help this company succeed.

2. I am proud to work for [this organization].

3. I feel a strong sense of belonging to this organization.

4. I would refer a friend to come work at [this organization].

5. Overall, I am satisfied working at [this organization]. Organizational Citizenship Behavior 1. I provide constructive suggestions about how my department can improve i\ ts effectiveness.

2. For issues that may have serious consequences, I express my opinions hon\ estly even when others may disagree.

3. I “touch-base” with my coworkers before initiating actions that mi\ ght affect them.

4. I encourage others to try new and effective ways of doing their job.

5. I help others who have large amounts of work.

6. I willingly share my expertise with my coworkers. Intention to Remain With the Organization 1. I would turn down a job with more pay in order to stay with [this organization].

2. I plan to spend my career at [this organization].

3. I intend to stay at [this organization] for at least the next 12 months.

4. I do not plan to look for a job outside of this company in the next 6 mo\ nths. Absenteeism 1. How many days did you miss from work in the last 12 months (excludin\ g vacation)? Appendix (continued) at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from Kehoe, Wright / Impact of High-Performance Human Resource Practices 389 References Ajzen, I., & Fishbien, M. 1980. Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ: Prentice-Hall.

Allen, N. J., & Meyer, J. P. 1990. The measurement and antecedents of affective, continuance and normative com- mitment to the organization. Journal of Occupational Psychology, 63: 1-18.

Angle, H. L., & Perry, J. L. 1981. An empirical assessment of organizational commitment and organizational effec- tiveness. Administrative Science Quarterly, 26: 1-14.

Applebaum, E., Bailey, T., Berg, P., & Kalleberg, A. L. 2000. Manufacturing advantage: Why high-performance work systems pay off. Ithaca, NY: Cornell University Press.

Arthur, J. B. 1994. Effects of human resource systems on manufacturing performance and turnove\ r. Academy of Management Journal, 37: 670-687.

Baron, R. M., & Kenny, D. A. 1986. The moderator–mediator variable distinction in social psychological r\ esearch: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51: 1173-1182.

Batt, R. 2002. Managing customer services: Human resource practices, qui\ t rates, and sales growth. Academy of Management Journal, 45: 587-597.

Becker, B. E., & Gerhart, B. 1996. The impact of human resource management on organizational performance: Progress and prospects. Academy of Management Journal, 39: 779-801.

Becker, B. E., Huselid, M. A., Pinckus, P. S., & Spratt, M. 1997. HR as a source of shareholder value: Research a\ nd recommendations. Human Resource Management Journal, 31: 39-47.

Blau, P. M. 1983. On the nature of organizations. Malabar, FL: Krieger.

Bowen, D. E., & Ostroff, C. 2004. Understanding HRM–firm performance linkages: The role of the “strength” of the HRM system. Academy of Management Review, 29: 203-221.

Chen, G., Mathieu, J. E., & Bliese, P. D. 2004. A framework for conducting multi-level construct validation. Research in Multi-Level Issues, 3: 273.

Cohen, A. 2003. Multiple commitments in the workplace an integrative approach. Mahwah, NJ: Erlbaum.

Combs, J., Liu, Y., Hall, A., & Ketchen, D. 2006. How much do high-performance work practices matter? A meta- analysis of their effects on organizational performance. Personnel Psychology, 59: 501.

Delery, J. E. 1998. Issues of fit in strategic human resource management: Impl\ ications for research. Human Resource Management Review, 8: 289-309.

Delery, J. E., & Doty, D. H. 1996. Modes of theorizing in strategic human resource management\ : Tests of universal- istic, contingency, and configurational performance predictions. Academy of Management Journal, 39: 802-835.

Delery, J. E., & Shaw, J. D. 2001. The strategic management of people in work organizations: Review, synthesis, and extension. Research in Personnel and Human Resources Management, 20: 165-197.

Dyer, L., & Reeves, T. 1995. Human resource strategies and firm performance: What do we know and where do we need to go? International Journal of Human Resource Management, 6: 656-670.

Gerhart, B. 2008. Research on human resources and effectiveness: Some methodological challenges. In D. Guest, J. Paauwe, & P. M. Wright (Eds.), Human resource management and performance: Progress and prospects.

Hoboken, NJ: Blackwell.

Gong, Y., Law, K. S., Chang, S., & Xin, K. R. 2009. Human resource management and fir\ m performance in China: The different role of managerial affective and continuance commitment. Journal of Applied Psychology, 94: 263-275.

Grant, R. M. 1996. Toward a knowledge-based theory of the firm. Strategic Management Journal, 17: 109.

Guthrie, J. P. 2001. High-involvement work practices, turnover, and productivity: Evidence from New Zealand. Academy of Management Journal, 44: 180-190.

Harter, J. K., Schmidt, F. L., & Hayes, T. L. 2002. Business-unit-level relationship between employee satisfactio\ n, employee engagement, and business outcomes: A meta-analysis. Journal of Applied Psychology, 87: 268.

Hitt, M. A., Biermant, L., Shimizu, K., & Kochhar, R. 2001. Direct and moderating effects of human capital on strategy and performance in professional service firms: A resource-based perspective. Academy of Management Journal, 44: 13-28.

Huselid, M. A. 1995. The impact of human resource management practices on turnover, productivity, and corporate financial performance. Academy of Management Journal, 38: 635-672. at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from 390 Journal of Management / February 2013 Jacofsky, E. F ., & Slocum, J. W., Jr. 1988. A longitudinal study of climates. Journal of Or ganizational Behavior, 9:

319-34.

Kozlowski, S. W. J., & Klein, K. J. 2000. A multilevel approach to theory and research in organizations: Contextual, temporal, and emergent processes. In K. J. Klein & S. W. J. Kozlowski (Eds.), Multilevel theory, research, and methods in organizations: Foundations, extensions, and new directions (pp. 349-381). San Francisco: Jossey-Bass.

Kreft, I., & de Leeuw, J. 1998. Introducing multilevel modeling. Thousand Oaks, CA: Sage.

Liao, H., Toya, K., Lepak, D. P., & Hong, Y. 2009. Do they see eye to eye? Management and employee perspectives of high performance work systems and influence processes on service qual\ ity. Journal of Applied Psychology, 94: 371-391.

MacDuffie, J. P. 1995. Human resource bundles and manufacturing performance. Industrial & Labor Relations Review, 48: 197.

March, J. G., & Simon, H. A. 1958. Organizations. New York: Wiley.

Mason, C. M., & Griffin, M. A. 2003. Group absenteeism and positive affective tone: A longitudinal study. Journal of Organizational Behavior, 24: 667-687.

Mayer, R. C., & Schoorman, F. D. 1992. Predicting participation and production outcomes through a tw\ o- dimensional model of organizational commitment. Academy of Management Journal, 35: 671-684.

Meyer, J. P., & Allen, N. J. 1997. Commitment in the workplace: Theory, research, and application. Thousand Oaks, CA: Sage.

Meyer, J. P., Stanley, D. J., Herscovitch, L., & Topolnytsky, L. 2002. Affective, continuance, and normative commit- ment to the organization: A meta-analysis of antecedents, correlates, and consequences. Journal of Vocational Behavior, 61: 20-52.

Mowday, R. T., Porter, L. W., & Steers, R. M. 1982. Employee–organization linkages: The psychology of commit- ment, absenteeism, and turnover. New York: Academic Press.

Nishii, L. H., Lepak, D. P., & Schneider, B. 2008. Employee attributions of the “why” of HR practices: Their effects on employee attitudes and behaviors, and customer satisfaction. Personnel Psychology, 61: 503.

Nishii, L. H., & Wright, P. M. 2008. Variability within organizations: Implications for strategic human resources management. In D. B. Smith (Ed.), The people make the place: Dynamic linkages between individuals and organizations (pp. 225-248). Mahwah, NJ: Erlbaum.

Organ, D. W. 1988. Organizational citizenship behavior: The good soldier syndrome. Lexington, MA: Lexington Books.

Ostroff, C. 1992. The relationship between satisfaction, attitudes, and performance: An organizational level analy- sis. Journal of Applied Psychology, 77: 963.

Ostroff, C., Kinicki, A. J., & Clark, M. A. 2002. Substantive and operational issues of response bias across leve\ ls of analysis: An example of climate-satisfaction relationships. Journal of Applied Psychology, 87: 355-368.

Podsakoff, P. M., MacKenzie, S. C., Paine, J. B., & Bachrach, D. G. 2000. Organizational citizenship behaviors: A critical review of the theoretical and empirical literature and suggesti\ ons for future research. Journal of Management, 26: 513.

Porter, L. W., Steers, R., Mowday, R., & Boulean, P. 1974. Organizational commitment, job-satisfaction, and turn- over among psychiatric technicians. Journal of Applied Psychology, 59: 603.

Salancik, G. R., & Pfeffer, J. 1978. A social information processing approach to job attitudes and task design\ . Administrative Science Quarterly, 23: 224-253.

Schneider, B., & Reichers, A. 1983. On the etiology of climates. Personnel Psychology, 36, 19.

Scholl, R. W. (1981). Differentiating organizational commitment from expectancy as a motivating force. Academy of Management Review, 6: 589-599.

Shaw, J. D., Gupta, N., & Delery, J. E. 2005. Alternative conceptualizations of the relationship between voluntary turnover and organizational performance. Academy of Management Journal, 48: 50.

Sheridan, J. E. 1985. A catastrophe model of employee withdrawal leading to low job performance\ , high absentee- ism, and job turnover during the first year of employment. Academy of Management Journal, 28: 83.

Shore, L. M., & Wayne, S. J. 1993. Commitment and employee behavior: Comparison of affective commitment and continuance commitment with perceived organizational support. Journal of Applied Psychology, 78: 774.

Sobel, M. E. 1982. Asymptotic confidence intervals for indirect effects in structural equation models. Sociological Methodology, 13: 290-312. at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from Kehoe, Wright / Impact of High-Performance Human Resource Practices 391 Sun, L., Aryee, S., & Law , K. S. 2007. High-performance human resource practices, citizenship beh\ avior, and organizational performance: A relational perspective. Academy of Management Journal, 5: 558-577.

Takeuchi, R., Chen, G., & Lepak, D. P. 2009. Through the looking glass of a social system: Cross-level effects of high performance work systems on employee attitudes. Personnel Psychology, 62: 1.

Takeuchi, R., Lepak, D. P., Wang, H., & Takeuchi, K. 2007. An empirical examination of the mechanisms mediating between high performance work systems and the performance of Japanese or\ ganizations. Journal of Applied Psychology, 92: 1069-1083.

Van Breukelen, W., Van der Vlist, R., & Steensma, H. 2004. Voluntary employee turnover: Combining variables from the “traditional” turnover literature with the theory of plan\ ned behavior. Journal of Organizational Behavior, 25: 893-914.

Way, S. A. 2002. High performance work systems and intermediate indicators of fi\ rm performance within the US small business sector. Journal of Management, 28: 765.

Wright, P. M., Dunford, B. B., & Snell, S. A. 2001. Human resources and the resource based view of the firm. Journal of Management, 27: 701.

Wright, P. M., & Gardner, T. M. 2003. The human resource–firm performance relationship: Methodological and theoretical challenges. In D. Holman, T. D. Wall, C. W. Clegg, P. Sparrow, & A. Howard (Eds.), The new work- place: A guide to the human impact of modern working practices (pp. 311-328). New York: Wiley.

Wright, P. M., Gardner, T. M., & Moynihan, L. M. 2003. The impact of HR practices on the performance of business units. Human Resource Management Journal, 13: 21.

Wright, P. M., Gardner, T. M., Moynihan, L. M., & Allen, M. R. 2005. The relationship between HR practices and firm performance: Examining causal order. Personnel Psychology, 58: 409.

Youndt, M. A., Snell, S. A., Dean, J. W., Jr., & Lepak, D. P. 1996. Human resource management, manufacturing strategy, and firm performance. Academy of Management Journal, 39: 836-866.

Zhang, Z., Zyphur, M. J., & Preacher, K. J. 2009. Testing multilevel mediation using hierarchical linear models: Problems and solutions. Organizational Research Methods 12: 695-719. at Glasgow University Library on August 19, 2013 jom.sagepub.com Downloaded from