Team Effectiveness

Journal of Applied Psychology 2006, Vol. 91, No. 1, 97–108

Copyright 2006 by the American Psychological Association 0021-9010/06/$12.00 DOI: 10.1037/0021-9010.91.1.97

Empowerment and Team Effectiveness: An Empirical Test of an Integrated Model John E. Mathieu and Lucy L. Gilson

Thomas M. Ruddy

University of Connecticut

Siemens

The authors developed a model of team empowerment as an emergent state linking inputs (I) with processes (P) and, thereby, with outcomes (O) in the context of an expanded team IPO framework. Using survey responses from 452 members of 121 empowered service technician teams, along with archival quantitative performance and customer satisfaction criteria, the authors tested the model using structural equation modeling techniques. The model was generally supported, although areas for improvement were evident. Specifically, empowerment partially mediated the influences of various inputs on team processes, whereas team processes fully mediated the influence of empowerment on outcomes. Directions for future research and application are discussed. Keywords: teams, empowerment, IPO model, processes

Work teams have been described as “integral” to organizational success in our global, fast-paced, customer-driven economy (Sundstrom, 1999). Empowerment has been seen as a powerful mechanism for increasing employee involvement (Lawler, 1986) and, thereby, enabling organizations to be more flexible and responsive (Bowen & Lawler, 1992; Kirkman & Rosen, 1999; Wellins, Byham, & Wilson, 1991). Empowered teams have greater authority and responsibility for their work than do more traditional teams, and their effectiveness is determined, in part, by the receptiveness or supports of the larger organizational system within which they operate (Hyatt & Ruddy, 1997). Although researchers have sought to understand influences on the success of empowered teams (i.e., Cohen & Ledford, 1994; Kirkman & Rosen, 1999), empowerment has not been examined within the context of the classic input–process-outcome (IPO) framework (e.g., Hackman & Morris, 1975; Marks, Mathieu, & Zaccaro, 2001; McGrath, 1984). Our goals in this study are as follows. First, we advance a model that depicts team empowerment as an emergent state linking team inputs and processes in an expanded IPO framework. In so doing, we argue that team processes serve to mediate the influence of team empowerment on team outcomes. Second, we test our model using survey data collected from a sample of customer service technician teams along with lagged archival measures of team performance and external customer satisfaction measures. Third, we conclude with a discussion of the implications of our findings for future research and practice.

Team Empowerment Construct Empowerment is not a new phenomenon and can be traced back to the early work of Lewin (1947) on employee involvement. In the 1990s, empowerment became a rallying cry for reengineering where it was proposed that individuals should have the authority to make their own decisions regarding how their work was to be done (Hammer & Champy, 1993), resulting in more innovative, happier, and productive employees (Peters, 1992). To date, empowerment has been defined in several different ways, although one point of consensus is that it is an isomorphic construct across levels of investigation (e.g., Kirkman & Rosen, 1997, 1999; Spreitzer, 1996). Isomorphic constructs retain the same basic meaning and function across levels of analysis (Klein & Kozlowski, 2000). Although most theorists and researchers have considered empowerment at the individual level, others have applied the concept to teams or work groups (e.g., Hyatt & Ruddy, 1997; Kirkman & Rosen, 1999; Liden & Tewksbury, 1995), as do we in this article. Essentially, there are two different conceptions of empowerment in the extant literature: structural and psychological (cf. Leach, Wall, & Jackson, 2003; Menon, 2001; Spreitzer, 1995).1 Advocates of the structural approach argue that empowerment is a practice or set of practices that involve the delegation of authority and responsibility to employees. Structural empowerment has been examined at the individual (e.g., Perry, Pearce & Sims, 1999; 1

We should note that there is a third conception of empowerment in the literature often referred to as the process approach, which defines empowerment in terms of the relationships between structural antecedents and resulting psychological states. For example, “empowerment is conceptualized here in terms of changes in cognitive variables (called task assessments), which determine motivation in workers” (Thomas & Velthouse, 1990, pp. 666 – 667). In other words, the process approach roots empowerment in the relationships that exist between structural antecedents and psychological reactions. This approach has not generated much research per se, but it is embedded in studies that have sought to associate organizational contexts with psychological states of empowerment (e.g., Kirkman & Rosen, 1999; Spreitzer, 1996).

John E. Mathieu and Lucy L. Gilson, Department of Management, University of Connecticut; Thomas M. Ruddy, Siemens, Iselin, New Jersey. We thank Scott Taylor and Travis Maynard for their efforts and comments on work related to this project. We also thank Bradley Kirkman, Dov Eden, and Katherine Klein for their helpful comments on earlier versions of this article. Correspondence concerning this article should be addressed to John E. Mathieu, Department of Management, University of Connecticut, 2100 Hillside Drive, Storrs, CT 06269-1041. E-mail: john.mathieu@ business.uconn.edu 97

98

MATHIEU, GILSON, AND RUDDY

Pfeffer, Cialdini, Hanna, & Knopof, 1998; Wall, Cordery, & Clegg, 2002) and at the team (e.g., Arnold, Arad, Rhoades, & Drasgow, 2000; Cook & Goff, 2002; Mills & Ungson, 2003) levels of analysis. Conceptually, structural empowerment draws heavily on work on job design (Hackman & Oldham, 1976) and job characteristics (Campion, Medsker, & Higgs, 1993) and, therefore, focuses on practices such as job enrichment that result in selfmanaging teams or autonomous work groups (Leach et al., 2003; Manz & Sims, 1987). In essence, structural empowerment focuses on work arrangements that alter the role of external leadership and shift responsibilities to team members (Alper, Tjosvold, & Law, 2000; Kirkman & Shapiro, 2001; Menon, 2001). Advocates of the psychological approach consider empowerment as a constellation of experienced psychological states or cognitions. One version of the psychological approach was advanced by Hyatt and Ruddy (1997), who submitted that team empowerment is “the extent to which work group members have the ability to make business decisions, are accountable for the outcomes of their decisions, accept responsibility for the outcomes of their decisions, and can solve problems on their own” (p. 562). Hardy and Leiba-O’Sullivan (1998) and Hechanova-Alampay and Beehr (2001) echoed this view and identified empowerment in terms of members’ perceived authority to control how their work is conducted and having responsibility for their work outcomes. The essence of this two-dimensional definition of psychological empowerment is that it is a psychological state defined primarily in terms of members’ experience of authority and responsibility. Another version of psychological empowerment defines it in terms of a four-dimensional framework of employees’ perceptions: (a) competence to perform their tasks well, (b) self-determination or freedom to choose how they carry out their tasks, (c) sense of meaningfulness that their work is important, and (d) belief that their work has an impact on the effectiveness of the larger system (cf. Kirkman & Rosen, 1997, 1999; Spreitzer, 1995, 1996). As with structural empowerment, psychological empowerment has been examined at the individual (e.g., Conger & Kanungo, 1988; Spreitzer, 1995; Thomas & Velthouse, 1990) and team levels (e.g., Hyatt & Ruddy, 1997; Kirkman & Rosen, 1997, 1999; Kirkman, Rosen, Tesluk, & Gibson, 2004). Recently, Menon (2001) submitted that the structural and psychological approaches are not antithetical; rather, they provide a comprehensive perspective of the empowerment phenomenon. He argued that understanding the process by which structuring acts lead to changes in employees’ psychological states represents a unifying view of empowerment. We concur with this position and seek to differentiate antecedent conditions from the resulting psychological state of empowerment, thereby providing a better understanding of the processes by which the two are associated. Accordingly, we assess both perceptions of the work design features (structural) and examine their relationships with team members’ resulting psychological state of empowerment. This approach allows us to examine the influence of organizational contextual factors, beyond the structural design of work, on the resulting psychological state (cf. Kirkman & Rosen, 1999; Spreitzer, 1996). Given the different versions of psychological empowerment, we prefer the two-dimensional approach for a number of reasons. First, the two-dimensional approach has a long research history dating back to Lewin (1947) and is more consistent with practitioner and dictionary definitions of empowerment. For example,

Ford and Fottler (1995) submitted that, “essentially, empowerment involve[s] passing decision-making authority and responsibility from managers to employees” (p. 21). Second, the twodimensional approach better enables the differentiation between the psychological state of empowerment and other variables (see Liden & Tewksbury, 1995; Menon, 2001). For example, the fourdimensional approach uses members’ reports of task autonomy, which are derived from the job characteristics literature (i.e., a structural input), and perceptions of impact, which are similar to self-reports of effectiveness (i.e., an outcome), thus blurring the construct validity of empowerment as a psychological state (Schwab, 1980). Consequently, we define team psychological empowerment as “team members’ collective belief that they have the authority to control their proximal work environment and are responsible for their team’s functioning.”

Theoretical Model and Hypotheses Team effectiveness theories have long advocated IPO frameworks (Guzzo & Dickson, 1996; Hackman & Morris, 1975). Although substantial work has been conducted on the process 3 outcome relationships (cf. Marks et al., 2001; Weingart, 1997, for reviews), relatively less attention has been devoted to the input 3 process relations. Moreover, Ilgen, Hollenbeck, Johnson, and Jundt (2005) have submitted that IPO models should be expanded to consider “the broader range of variables that are important mediational influences with explanatory power for explaining variability in team performance and viability” (p. 520). Marks et al. (2001) sought to differentiate two major classes of mediator variables in IPO frameworks. They defined processes as “interdependent team activities that orchestrate taskwork in employees’ pursuit of goals” (p. 358). In contrast, they referred to emergent states as “cognitive, motivational, and affective states of teams, as opposed to the nature of their member interaction . . . . constructs that characterize properties of the team that are typically dynamic in nature and vary as a function of team context, inputs, processes, and outcomes” (p. 357). Kirkman et al. (2004) specifically referred to team [psychological] empowerment as an emergent state. Further, Marks et al. (2001) argued that team emergent states serve as an influence on team processes. Accordingly, we advance and test a model that depicts team psychological empowerment as a mediator variable linking inputs with processes. Although the IPO framework has guided team researchers for decades, more recently, appreciation of the role of organizational context has been growing (Hackman, 1987; Sundstrom, 1999). For example, Gully (2000) concluded, “organizational contexts can have a strong influence on the behavior and performance of the work team as an aggregate” (p. 32). In the empowerment literature, factors such as organizational context and management practices have been described as playing key roles in the success of intervention efforts (i.e., Koberg, Boss, Senjem, & Goodman, 1999). Kirkman and Rosen (1997, 1999) suggested that a more detailed breakdown of contextual supports would likely yield further insights into the primary drivers of empowerment. Our model closely follows the foundation established by Kirkman and Rosen (1997, 1999), particularly in terms of antecedents of empowerment. Thus, our underlying logic parallels theirs yet extends the inquiry to include the mediating role of team processes. In addition, we sample different types of teams and use different types

EMPOWERMENT AND TEAM EFFECTIVENESS MODEL

and sources of criteria measures. Our hypothesized model is presented in Figure 1.

Antecedents of Team Psychological Empowerment The present investigation stems from an evaluation of the success of a structural empowerment intervention. The intervention was introduced simultaneously across the entire country. Given the large-scale rollout, we anticipated that there might be important variations in the way in which the structural changes were implemented as well as in the extent to which various supports were in place for different teams. Accordingly, we assessed members’ perceptions of the resulting work design and of three different types of general supports and model their influences on team (a) psychological empowerment, (b) processes, and (c) outcomes.

Work Design The structural view of empowerment suggests that the design of work will enhance employees’ psychological empowerment and thereby yield benefits in terms of increased effectiveness. Team empowerment intervention designs have included the delegation of authority and responsibility for certain human resource functions (e.g., hiring), establishing, executing, and monitoring performance strategies, and holding members responsible for their work outcomes. For example, Wageman (2001) found that teams working in empowered designs were better able to self-manage and exhibited higher performance levels than those working in more traditional designs. Moreover, she found that the teams’ self-managing behaviors mediated the relationships between design features and team effectiveness. In short, we contend, as have others (e.g., Kirkman & Rosen, 1999; Leach et al., 2003), that when the structural design of work shifts control and responsibility from external management to teams, members should experience greater team psychological empowerment. Although such work arrangements serve to enhance members’ feelings of empowerment, they are not identical with them. So, for clarity, we refer to structural

Figure 1.

99

empowering acts as work design and team psychological empowerment as team empowerment. We anticipate that team empowerment levels will be a product of work design features in concert with the influences of other aspects of the embedding environment (Kirkman & Rosen, 1997, 1999). Specifically, the design of team activities affects members’ capabilities of working together (Ancona & Caldwell, 1992). Kirkman and Rosen (1999) found support for positive relationships between what they referred to as “team production/service responsibilities” and “team empowerment.” Therefore, work designs that have allocated tasks such as determining workflow, performance strategies, and internal functioning from external managers to team members (i.e., implemented structural empowerment designs) should enhance members’ feelings of team empowerment. This led to the following hypothesis: Hypothesis 1: When work is structured such that team members make their own human resource decisions, develop and execute performance strategies, and coordinate their work (i.e., work is high in structural team empowerment), they will report greater psychological team empowerment.

Organizational Support A supportive organizational context has long been purported as a necessary condition for team effectiveness (Hackman, 1987). Organizational context refers to sources of support that are external to the team yet emanate from within the larger organizational system (except for the influence of their external leader, which we address next). One of the key components of a supportive organization is a climate in which teams can openly communicate with one another and freely share and exchange information. In effect, teams are distinct entities that are “coupled to one another and to the organization as a whole” (Gully, 2000, p. 32). Indeed, Kirkman and Rosen (1999) referred to this as providing teams with a well-developed social structure and sociopolitical support. This type of climate has also been referred to as “psychologically safe”

Hypothesized model. HR ⫽ human resources.

MATHIEU, GILSON, AND RUDDY

100

(Edmondson, 1999) because employees seeking input from others are encouraged and nurtured rather than frowned on. Mathieu, Marks, and Zaccaro (2001) advanced the notion of multiteam systems (MTSs), an organizational arrangement in which teams of teams work collectively to achieve collective goals. Empowered teams, such as the ones examined here, often work in larger organizational contexts that emulate MTSs and therefore place a premium on interteam processes. In other words, in traditionally designed organizations the coordination of actions, sharing of sharing, development of a common vision, and so forth across teams is the province of upper-level management. In empowered designs, the responsibility for cross-team processes shifts to the teams and is enhanced (or constrained) by the extent to which the larger organizational system facilitates such efforts. Indeed, it has been found that effective cross-team processes enhance team and MTSs outcomes when teams work in interdependent settings (Marks, DeChurch, Mathieu, Panzer, & Alonso, 2005). In addition, it has been suggested that for teams to function effectively as self-contained units, they must have strong information networks as well as communication and cooperation channels both within and between teams (Cummings, 1978). A supportive social structure with open communication and cooperation between teams should provide an enabling condition facilitating team empowerment (Beer, Eisenstat, & Spector, 1990; Kirkman & Rosen, 1999). Indeed, Kirkman and Rosen (1999) found a significant positive relationship between team’s social structure and their empowerment. This led to the following hypothesis: Hypothesis 2: Team members’ perceptions of organizational support in terms of open communications and multiteam cooperation will relate positively to their reported levels of team psychological empowerment.

External Team Leadership Initially, it was thought—and perhaps feared—that flatter organizational structures would eliminate the role of external managers. However, the results of self-directed work initiatives have shown that there is a very real, albeit different, need for effective external leaders (Manz & Sims, 1987). Effective external leadership has been shown to be an important driver in the success of empowered organizations (Druskat & Wheeler, 2003; Sims & Manz, 1984). For instance, Arnold et al. (2000) submitted that empowering leadership involves implementing conditions that increase employees’ sense of control, removing conditions that foster a sense of powerlessness and allowing them the freedom to be as flexible as circumstances warrant. A lack of effective external leader support can result in teams that feel “abandoned” by their organizations (Hackman, 1990; Lawler, 1986) and is often cited as one of the main reasons that self-managed teams fail (Manz & Sims, 1987). Whereas empowered teams are more than simply self-managed, the influence of external leadership on either type of team is comparable. However, with empowered teams, the role of external leaders is fundamentally different than it is in traditionally designed organizations in that the key functions shift from focusing on day-to-day work activities to focusing on the procurement of needed resources and team facilitation (Druskat & Wheeler, 2003). Manz and Sims (1987) referred to this style of

leaderships as “leading others to lead themselves” (p. 119). Kirkman and Rosen (1999) found support for the positive influences of such external leadership on team empowerment. This led to the following hypothesis: Hypothesis 3: When team members perceive external leaders as pursuing resource acquisition and playing a supportive team facilitation role, they will report greater team psychological empowerment.

Team-Based Human Resource (HR) Practices Traditionally, HR departments provide support functions to work teams (e.g., hiring, training, disciplinary action), whereas empowered teams often subsume such responsibilities. One function that often remains outside of the team, however, is training. Providing teams with control and responsibility will not result in improved outcomes unless members have the skills and abilities needed to handle the task and decisions at hand (Liden & Tewksbury, 1995). Team training has long been described as a critical component of team performance (Salas & Cannon-Bowers, 2001). In essence, these arguments have suggested that the better-trained teams will be able to exploit the freedom that empowerment affords the best. In addition to training, outcome feedback provision plays a vital role in teams’ ability to learn and develop (Nadler, 1979). Feedback cues can guide teams as to whether they are on course or whether alternative goals and strategies are warranted (Mesch, Farh, & Podsakoff, 1994; Pritchard, Jones, Roth, Stebing, & Ekeberg, 1988). In an empowered environment, teams are responsible for initiating their own corrective actions, which places an even higher premium on accurate and timely performance feedback. To this end, Druskat and Kayes (2000) found that team feedback was significantly related to team learning. Accordingly, we submit that team empowerment will more readily develop in settings that provide high-quality formal training with performance feedback. This led to the following hypothesis: Hypothesis 4: When team members perceive that team-based HR practices in the form of formal training and feedback mechanisms are provided, they will report greater team psychological empowerment.

Consequences of Team Psychological Empowerment Kirkman and Rosen (1999) found significant correlations between team empowerment and external managers’ ratings of team proactivity, productivity, and customer service, as well as of a variety of members’ affective reactions. Yet the question remains, what is the mechanism by which empowerment has such effects? We submit that team processes play a key mediational role in the empowerment 3 outcome relationships. Team processes have been defined as “how” teams get things done (Weingart, 1997). Recently, Marks et al. (2001) identified three superordinate dimensions of team processes: (a) transition, in which members engage in planning, analysis, and goal setting types of activities; (b) action, in which members’ interaction is focused on goal accom-

EMPOWERMENT AND TEAM EFFECTIVENESS MODEL

plishment, coordinating actions, and monitoring team progress; and (c) interpersonal, in which members manage their own conflicts, motivation, and affect. Team processes should provide a link between team empowerment and outcomes. The notion here is that to the extent team members are empowered, they will be liberated to better execute transition and action processes as they see fit. They will be freer to plan and organize their work orders, implement different performance strategies, coordinate their own actions, and otherwise align their collective efforts with work demands to meet their performance goals. Moreover, to the extent that teams are empowered to perform HR functions, they will be better able to manage their own interpersonal processes. For example, empowered teams have the authority to decide what processes to use to handle conflict among members or how to motivate members. Therefore, we submit that team empowerment will facilitate effective team processes. This led to the following hypothesis: Hypothesis 5: Team psychological empowerment will relate positively to team processes. Taken together, Hypotheses 1–5 suggest that team empowerment will fully mediate the relationships between the four antecedents and team processes. However, there could be significant direct effects as well. Therefore, we conducted supplemental analyses in the context of our structural model to test the mediational role of team empowerment. Although Marks et al. (2001) proposed that emergent states such as empowerment are likely to be antecedents of team processes, they also suggested that team processes might impact emergent states. In other words, they left open the possibility of a reciprocal relationship. To test such a possibility, we also examined, in a more exploratory fashion, whether team processes exhibit a reciprocal relationship with team empowerment. Team processes have a long history of relating significantly with team outcomes such as quantitative performance and customer satisfaction (see Guzzo & Dickson, 1996; Ilgen et al., 2005; Kozlowski & Bell, 2001; Marks et al., 2001, for reviews). The notion is that the extent to which team members can effectively execute important team functions, such as planning their work, coordinating their efforts, and managing interpersonal processes, they will be better positioned to respond to the needs of their customers (Hyatt & Ruddy, 1997). Moreover, effective team processes facilitate the accomplishment of team tasks and thereby relate positively to quantitative performance. Therefore, we advance the following hypotheses: Hypothesis 6: Team processes will relate positively to customer satisfaction. Hypothesis 7: Team processes will relate positively to quantitative performance. Finally, we highlight that, taken collectively, Hypotheses 5⫺7 suggest that team processes will fully mediate the relationships between team psychological empowerment and both work outcomes. Accordingly, we conducted supplemental analyses in the context of our structural model to test these mediational relationships.

101 Method

Participants and Setting The participants in this study were Canadian customer service engineers (CSEs) who worked for a major office equipment and technology organization. Their primary job functions included the repair and servicing of office document production systems. This study was conducted after an organizational intervention designed to empower work teams. Prior to the reorganization, CSEs worked primarily as individuals in a traditional top-down organizational structure in which they received job orders from their managers and were compensated on the basis of individual performance. The empowerment intervention created geographical clusters of CSEs who were collectively responsible for planning, organizing, assigning, and completing their work, along with introducing work process improvements. Teams were given greater latitude to make significant financial decisions (e.g., to repair or replace existing equipment), and technology aids were introduced to give the CSEs greater access to customer and equipment histories and better means to adjust and integrate members’ efforts. Moreover, CSE teams became responsible for their HR decisions, including selecting new members, performance monitoring, recognition, and rewards. Finally, the compensation structure was changed to be based, in part, on team performance. Supplemental data and analyses revealed that employees perceived that their work was significantly redesigned as intended and that they felt greater empowerment after the intervention than they did before.2 Individual team members completed surveys, on company time, that assessed perceptions of work design, organizational supports, external leadership, team processes, and empowerment. They were assured of confidentiality and returned their surveys in sealed envelopes to company HR representatives. Although we received 504 surveys in total (57% response rate), we have included only teams for which we received two or more surveys and which had at least a 25% response rate (teams ⫽ 121, respondents ⫽ 452). This represents a participation rate of 51% of the target sample. The number of people in each team ranged from 2 to 11, with a median of 6. To help ensure confidentiality, no demographic information was collected from respondents. However, company records indicate that approximately 90% of this population is male and that company tenure ranged from 4 to 28 years, with a mean of 14.3 years (SD ⫽ 6.67). This sample came from the same organization that sponsored work by Wageman (1995, 2001); Gilson, Shalley, and Blum (2001); and Hyatt and Ruddy (1997). However, our work is entirely separate and does not include any data that have appeared in those publications.

Measures The surveys contained a number of items designed for general organizational assessment purposes. For our use, we concentrated on items and scales that were aligned with the constructs depicted in Figure 1. All items were answered on a 1–5-point Likert-type scale. We used James, Demaree, and Wolf’s (1984) rwg agreement index, with the rectangular response distribution as a comparison, to justify aggregating individual members’ response to the team level. Median rwg values ⱖ .70 are generally considered sufficient to warrant aggregation. In addition, we report intraclass correlations (ICCs), which represent whether measures are sufficiently reliable to model effects at the team level (Bliese, 2000). The ICCs were calculated based on a one-way random-effects analysis of variance model in which team membership serve as levels of an independent variable and members’ scale responses are the dependent variable. In all instances, the analysis of variance F values were significant ( p ⬍ .01). ICC1 represents

2

Further details regarding pre–post intervention comparisons are available from the authors.

102

MATHIEU, GILSON, AND RUDDY

the reliability of a single rating of the team construct, whereas ICC2 represents the reliability of the average of team members’ responses. We calculated team-level internal consistencies for all of our scales using the average item response per team as the input. This strategy aligns the measurement reliability information with the level of analyses used in the substantive tests (Chen, Mathieu, & Bliese, 2004). Alpha is the lower bound of scale internal consistency (Cortina, 1993) and is affected by the number of items on a scale. Whereas some of the subscales we report fall short of the conventional .70 cutoff, this is not that disconcerting for our purposes because we are using the subscales (rather than items) as indicators of the latent constructs depicted in Figure 1 (see Landis, Beal, & Tesluk, 2000). In other words, rather than using less reliable individual items as indicators, we used substantively based item parcels as indicators in the confirmatory factor analytic (CFA) and structural model tests to reduce the number of parameters to be estimated. CSEs rated 17 features of their work design on a 5-point scale that ranged from 1 (manager makes decisions and informs team members) through 5 (team makes decisions and informs manager). The items were designed specifically to address the three primary foci of the redesign intervention: (a) HR decisions included seven aspects (e.g., selection of new hires) and exhibited acceptable reliabilities and agreement (␣ ⫽ .63; rwg ⫽ .93; ICC1 ⫽ .29; ICC2 ⫽ .60); (b) work coordination assessed four facets (e.g., workload distribution among members) and had an acceptable aggregate internal consistency and agreement index (␣ ⫽ .64; rwg ⫽ .88; ICC1 ⫽ .38; ICC2 ⫽ .70); and (c) performance strategies included six items (e.g., “how to deal with poor performing machines”) and evidenced good psychometrics properties (␣ ⫽ .76; rwg ⫽ .92; ICC1 ⫽ .30; ICC2 ⫽ .62). Organizational support describes characteristics of the larger system or context and was assessed with two subscales. Multiteam cooperation was assessed with two items (e.g., “there is good cooperation among teams in my branch”; ␣ ⫽ .73; rwg ⫽ .80; ICC1 ⫽ .20; ICC2 ⫽ .48) and open communications was assessed with three items (e.g., “members of my team feel that we are listened to in this organization”; ␣ ⫽ .86; rwg ⫽ .87; ICC1 ⫽ .13; ICC2 ⫽ .36). External team leadership was measured with two subscales as indicators: (a) resource acquisition, with two items (e.g., “our manager gets us whatever we need to do our job effectively”; ␣ ⫽ .75; rwg ⫽ .90; ICC1 ⫽ .27; ICC2 ⫽ .36) and (b) team facilitation, which was measured with two items (e.g., “our manager does a good job of building teamwork”; ␣ ⫽ .86; rwg ⫽ .91; ICC1 ⫽ .20; ICC2 ⫽ .49). Team-based HR practices was assessed with two scales: (a) formal training, with two items (e.g., “members of my team are well trained on the products we service”; ␣ ⫽ .48; rwg ⫽ .87; ICC1 ⫽ .10; ICC2 ⫽ .30) and (b) feedback mechanisms, a three-item scale (e.g., “members of my team get regular feedback about performance”; ␣ ⫽ .87; rwg ⫽ .89; ICC1 ⫽ .24; ICC2 ⫽ .54). Team empowerment describes the state in which members share a collective sense that they have the responsibility and the authority to control their work. Accordingly, we used two scales: authority, three items (e.g., “my team is empowered to change our work processes in order to improve our performance”; ␣ ⫽ .82; rwg ⫽ .94; ICC1 ⫽ .10; ICC2 ⫽ .29) and responsibility, three items (e.g., “members of my team are responsible for determining the best way to satisfy our customers’ requirements”; ␣ ⫽ .88; rwg ⫽ .94; ICC1 ⫽ .20; ICC2 ⫽ .48). Team processes were indexed using three scales that correspond to Marks et al.’s (2001) superordinate categories, each with three items: transition (i.e., “members of my team discuss” [a] “our performance vision,” [b] “what we can do day to day to make our performance vision a reality,” and [c] “our district’s objectives”; ␣ ⫽ .88; rwg ⫽ .92; ICC1 ⫽ .12; ICC2 ⫽ .34), action (i.e., “members of my team” [a] “take the time we need to share task-related information,” [b] “actively learn from one another,” and [c] “effectively communicate with each other throughout the workday”; ␣ ⫽ .78; rwg ⫽ .90; ICC1 ⫽ .26; ICC2 ⫽ .57), and interpersonal (i.e., “members of my team” [a] “create an environment of openness and

trust,” [b] “really trust each other,” and [c] “think in terms of what is best for the team”; ␣ ⫽ .87; rwg ⫽ .93; ICC1 ⫽ .33; ICC2 ⫽ .65). Quantitative performance was indexed using three archival measures that are routinely gathered by the organization to track team effectiveness and determine team rewards. These measures were tabulated for 3 months following the survey data collection and have been described by Hyatt and Ruddy (1997). Because of their confidential nature, they were converted to z scores. “Machine reliability” refers to the average number of copies made by machines between service calls. “Response time” refers to the average length of time between the customer’s call for service and the arrival of one or more team members, and “parts expense” refers to the percentage of budget associated with replacing parts. Because lower response time and parts expense represent more effective performance, we multiplied their z scores by ⫺1 so that higher values correspond to better performance. Customer satisfaction was measured by the company through surveys sent directly to a random sample of customers to yield a representative sample for each type of machine serviced by each team. As did Hyatt and Ruddy (1997), we used the following item as an overall index of customer satisfaction: “How was your overall experience with the company’s equipment maintenance service?” Responses were made on a 5-point scale that ranged from 1 (very dissatisfied) to 5 (very satisfied). The surveys were keyed according to which team was responsible for the customers’ machines and were returned to corporate headquarters. On average, 76 surveys (range ⫽ 6 –170) per team were returned over a 3-month period following the data collection. An overall index was calculated for each team by averaging their customers’ responses and then converting that average to a z score.

Results Structural Equation Model Tests The psychometric properties and descriptive statistics for all measures used in the structural model tests are presented in Table 1. We used Anderson and Gerbing’s (1988) two-step strategy to test the model depicted in Figure 1. We first fit a CFA model, and then we tested a series of structural models. The survey subscales noted above were used as indicators of the four antecedents, empowerment, and team process latent variables. The three archival scores were used as indicators of quantitative team performance, and the average customer satisfaction survey response was used as a single indicator of its corresponding latent variable. We report the comparative fit index (CFI; Bentler, 1990) and the standardized root-mean-square residual (SRMSR; Hu & Bentler, 1999) to gauge model fit. The CFI has been identified as the best approximation of the population value for a single model, with values ⱖ .90 considered indicative of good fit (Medsker, Williams, & Holahan, 1994). SRMSR is a measure of the average standardized residual per degree of freedom with values ⱕ .08 considered a “relatively good fit for the model” and values ⱕ .10 considered as “fair” (Browne & Cudeck, 1993). We also report chi-square values that provide a statistical basis for comparing the relative fit of nested models. Table 2 presents a summary of the fit indices for the various models that were tested. Confirmatory factor models. Using the covariance matrix, we estimated the CFA model, which yielded good fit indices, ␹2(108, N ⫽ 121) ⫽ 184.23, p ⬍ .01; CFI ⫽ .92; SRMSR ⫽ .08. All indicators exhibited significant ( p ⬍ .05) relationships with their intended latent variable. In contrast, we fit a null latent CFA model (which constrains the correlations among the latent variables to zero) to the data, ␹2(136, N ⫽ 121) ⫽ 492.24, p ⬍ .01; CFI ⫽ .60; SRMSR ⫽ .15, and obtained a significantly worse fit,

EMPOWERMENT AND TEAM EFFECTIVENESS MODEL

103

Table 1 Variable Descriptive Statistics and Correlations Variable

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

1. HR decisions .63 2. Work coordination .67 .64 3. Performance strategies .56 .71 .76 4. Multiteam cooperation .11 .01 .02 .73 5. Open communications .21 .27 .22 .39 .86 6. Manager resource acquisition ⫺.01 ⫺.00 .05 .26 .43 .75 7. Manager team facilitation .04 ⫺.02 .04 .32 .43 .87 .86 8. Team-based HR practices .14 .11 .12 .10 .29 .41 .43 .48 9. Feedback mechanisms .02 ⫺.01 .06 .20 .43 .52 .47 .46 .87 10. Empowerment: authority .19 .34 .28 .21 .54 .34 .31 .36 .38 .82 11. Empowerment: responsibility .26 .43 .42 .15 .58 .37 .34 .39 .43 .55 .88 12. Transition processes .09 .05 .15 .31 .46 .49 .46 .47 .62 .45 .60 .88 13. Action processes .21 .21 .25 .29 .30 .29 .34 .27 .30 .33 .49 .45 .78 14. Interpersonal processes .18 .27 .36 .32 .29 .28 .32 .29 .19 .30 .48 .38 .61 .87 15. Machine reliabilitya .33 .30 .22 .00 .01 .05 .07 .23 .05 .13 .16 .09 .27 .11 1.00 16. Response timea .37 .21 .16 ⫺.05 ⫺.10 .08 .04 .27 ⫺.03 .00 .05 .03 .11 .21 .42 1.00 17. Parts expensea .15 .25 .16 .15 .11 .17 .17 .18 .24 .12 .28 .21 .21 .15 .45 .14 1.00 18. Customer satisfactiona .00 .08 .13 .10 .03 .14 .09 .11 .09 .01 .14 .24 .20 .14 .19 .33 .14 1.00 M 3.39 4.22 3.98 2.82 3.06 3.67 3.73 3.75 3.24 3.86 3.98 3.68 3.89 3.78 .00 .00 .00 .00 SD 0.51 0.59 0.58 0.56 0.57 0.49 0.52 0.44 0.61 0.39 0.50 0.46 0.47 0.51 1.00 1.00 1.00 1.00 Note. Diagonal entries are scale reliabilities. N ⫽ 121 teams. Correlations: .19, p ⬍ .05; 兩.23兩, p ⬍ .01; and 兩.31兩, p ⬍ .01. HR ⫽ human resources. Z scores.

a

␹2difference(28) ⫽ 308.01, p ⬍ .01. These results indicate that the measurement properties fit reasonably well, and there is sufficient covariance among the latent variables to warrant examining structural models. Notably, it is quite possible that although the full CFA model might exhibit acceptable fit indices, portions of it might not exhibit sufficient discriminant validity among a subset of latent variables. Therefore, we fit several “more focused” CFAs, which featured subsets of latent variables and their indicators, to better ensure the discriminant validity of our measures. One such analysis focused strictly upon the nine indicators of the four antecedent variables and evidenced quite good model fit, ␹2(21, N ⫽ 121) ⫽ 19.63, ns; CFI ⫽ 1.00; SRMSR ⫽ 0.0. We then collapsed the indicators of the two highest correlating latent variables (i.e., external team leadership and team-based HR practices, r ⫽ .71, p ⬍ .01) and fit

a three-factor model. The three-factor model exhibited reasonable overall fit, ␹2(24, N ⫽ 121) ⫽ 35.89, p ⬍ .05; CFI ⫽ .97; SRMSR ⫽ .06, but was significantly worse than the four-factor model, ␹2difference(3) ⫽ 16.26, p ⬍ .01. This provides ample evidence for the discriminant validity of the four antecedent variables. Similarly, we fit the five indicators of the work design and empowerment latent variables to both a two-factor and a singlefactor model. Whereas the two-factor model exhibited quite good fit indices, ␹2(4, N ⫽ 121) ⫽ 3.94, ns; CFI ⫽ 1.00; SRMSR ⫽ 0.0, the single-factor model revealed a poor fit, ␹2(5, N ⫽ 121) ⫽ 32.72, p ⬍ .01; CFI ⫽ .88; SRMSR ⫽ .21, that was significantly worse, ␹2difference(1) ⫽ 28.78, p ⬍ .01. This provides support for the contention that employees can perceive the difference between indicators of structural and psychological empowerment. Last, we fit the five indicators of the empowerment and team process latent

Table 2 Model Fit Indices Model

df

␹2

CFI

SRMSR

Measurement Null latent Hypothesized Empowerment 43 processes Work design and supports 3 processes (including empowerment) Work design and supports 3 processes (excluding empowerment paths) Empowerment 3 outcomes (including processes) Empowerment 3 outcomes (excluding processes paths) Revised model

108 136 123 122

184.23** 492.24** 229.82** 228.82**

.92 .60 .88 .88

.08 .15 .09 .09

119

213.49**

.89

.08

124 121

338.30** 225.95**

.76 .88

.12 .09

124 121

293.77** 202.68**

.81 .91

.11 .08

Note. N ⫽ 121 teams. CFI ⫽ comparative fit index, SRMSR ⫽ standardized root-mean-square residual. ** p ⬍ .01.

MATHIEU, GILSON, AND RUDDY

104

variables to both a two-factor and a single-factor model. The two-factor model exhibited quite good fit indices, ␹2(4, N ⫽ 121) ⫽ 17.56, p ⬍ .01; CFI ⫽ .93; SRMSR ⫽ 0.06. Although the single-factor model also exhibited a reasonable fit, ␹2(5, N ⫽ 121) ⫽ 22.34, p ⬍ .01; CFI ⫽ .91; SRMSR ⫽ .07, it was significantly worse, ␹2difference(1) ⫽ 4.78, p ⬍ .05. This analysis provides support for the contention that employees can perceive the difference between an emergent state of empowerment and their team processes. Structural models. We fit the hypothesized structural model and obtained “fair” model fit indices, ␹2(123, N ⫽ 121) ⫽ 229.39, p ⬍ .01; CFI ⫽ .88; SRMSR ⫽ .085, although the model did differ significantly from the measurement model, ␹2difference(14) ⫽ 44.59, p ⬍ .01, and, therefore, has room for improvement. The results of this analysis are presented in Figure 2. As shown, all of the hypothesized paths were significant in the hypothesized model ( p ⬍ .05), with the exception of the unique influence of external team leadership on team empowerment (␤ ⫽ ⫺.09, ns). Therefore, Hypotheses 1, 2, 4, 5, 6, and 7 were supported, whereas Hypothesis 3 was not. Next, we fit a model that added a path from team processes to empowerment to test the exploratory reciprocal relationship. This model did not fit significantly better, ␹2difference(1) ⫽ .57, ns, than the hypothesized model, nor was the path from processes to empowerment significant, ␹2(122, N ⫽ 121) ⫽ 228.82, p ⬍ .01; CFI ⫽ .88; SRMSR ⫽ .085. These results suggest that the hypothesized empowerment 3 processes order is more consistent with the data.

Additional Analyses Mediational tests. We fit additional structural models to specifically test the two implied mediational relationships in our overall model. First, we tested the mediation role of team empowerment linking the antecedent variables and team processes. Second, we tested the mediation role of team processes linking team

Figure 2.

empowerment and the two outcome variables. In both instances, we developed two additional models to provide adequate tests of mediation (James & Brett, 1984). First, we created for comparison purposes a more inclusive baseline model that added direct paths from the exogenous variable(s) to the endogenous variable(s) in each instance. Second, we eliminated all paths leading to and stemming from the mediator variable from these more inclusive baseline models, but we left the mediator latent variable in the model. Nested comparisons between the more inclusive models and the models dropping the mediator paths provide a test of the value of the mediator. Notably, there should be significant relationships between the exogenous and endogenous variables in this latter model to fulfill the x 3 y precondition for tests of mediation (cf. Baron & Kenny, 1986; James & Brett, 1984). The more inclusive model for testing team empowerment as a mediator exhibited good fit indices, ␹2(119, N ⫽ 121) ⫽ 213.49, p ⬍ .01; CFI ⫽ .89; SRMSR ⫽ .081, and, in fact, fit better than did the hypothesized model, ␹2difference(4) ⫽ 15.90, p ⬍ .01. We next tested a model that included the empowerment latent variable but dropped all paths leading to, or coming from, it. This latter model (without empowerment paths), evidenced very poor fit, ␹2(124, N ⫽ 121) ⫽ 338.30, p ⬍ .01; CFI ⫽ .76; SRMSR ⫽ .12, and differed significantly from the inclusive baseline model, ␹2difference(5) ⫽ 124.81, p ⬍ .01. These results provide substantial support for the notion that team empowerment serves as a partial mediating mechanism linking work design, organizational support, team based HR practices, and team processes. The more inclusive baseline model for testing team processes as a mediator exhibited fair fit indices, ␹2(121, N ⫽ 121) ⫽ 225.95, p ⬍ .01; CFI ⫽ .88; SRMSR ⫽ .085, and did not fit better than the hypothesized model, ␹2difference(2) ⫽ 3.87, ns. The second model (without team process paths), evidenced very poor fit, ␹2(124, N ⫽ 121) ⫽ 293.77, p ⬍ .01; CFI ⫽ .81; SRMSR ⫽ .110, and differed significantly from the inclusive baseline model, ␹2difference(3) ⫽

Results of hypothesized model. N ⫽ 121. *p ⬍ .05.

EMPOWERMENT AND TEAM EFFECTIVENESS MODEL

67.82, p ⬍ .01. Notably, team empowerment exhibited a significant direct relationship with quantitative performance (␤ ⫽ .25, p ⬍ .01) in this model, although it did not with customer satisfaction (␤ ⫽ .13, ns). Therefore, the precondition for testing mediated relationships was fulfilled for performance but not for customer satisfaction. In summary, our results provide substantial support for the notion that team processes fully mediate the relationships between empowerment and quantitative performance. In contrast, our results indicate that the relationship between empowerment and customer satisfaction is not mediated yet is linked indirectly through team processes. Model revisions. The results previously chronicled provide substantial support for the hypothesized model, although it still differed significantly from both the measurement model and from the inclusive baseline model used to test the mediational role of team empowerment. Therefore, we developed a revised model to further our understanding of the underlying relationships. Naturally, such revisions are exploratory and should be interpreted cautiously. First, we dropped the nonsignificant path from external team leadership to empowerment. Second, we examined the model modification indices and concluded that the lack of fit stemmed primarily from a missing direct path from team-based HR practices to team processes (␤ ⫽ .59, p ⬍ .01) and missing direct links from organizational support (␤ ⫽ ⫺.35, p ⬍ .01) and work design (␤ ⫽ .43, p ⬍ .01) to quantitative performance. Incorporating these four changes yielded the revised model presented in Figure 3 that exhibited good fit indices, ␹2(121, N ⫽ 121) ⫽ 202.68, p ⬍ .01; CFI ⫽ .91; SRMSR ⫽ .08, and did not differ significantly from the measurement model, ␹2difference(13) ⫽ 18.45, ns. In summary, the revised model results indicated that team empowerment partially mediates the influence of team-based HR practices on team processes. Moreover, the empowerment 3 process mediational path accounts for the influences of organizational support and work design features on quantitative performance. Finally, the negative relationship between organizational supports and quantitative performance is contrary to what we would have anticipated. Unpacking team processes. The model tests above treated team process as a unified omnibus latent variable. Yet Marks et al. (2001) argued that the three superordinate dimensions of team processes (i.e., transition, action, and interpersonal) may exhibit

Figure 3.

Results of revised structural model. N ⫽ 121. *p ⬍ .05.

105

different relationships with outcome variables. To test this possibility, we regressed customer satisfaction and quantitative performance on the three team process subscales.3 We formed a performance composite for these analyses by averaging the standardized scores for the three components, each multiplied by its respective weight from the CFA analysis. Both transition (r ⫽ .24, p ⬍ .05) and action processes (r ⫽ .20, p ⬍ .05) evidenced significant correlations with customer satisfaction, although interpersonal processes did not (r ⫽ .14, ns). Simultaneously regressing customer satisfaction on the three variables yielded R2 ⫽ .066, p ⬍ .05. Inspection of the individual beta weights, however, illustrated that transition processes had a marginal relationship (␤ ⫽ .18, p ⬍ .10), whereas neither action processes (␤ ⫽ .11, ns) nor interpersonal processes (␤ ⫽ .01, ns) were significant. The composite quantitative performance criterion correlated significantly with action (r ⫽ .26, p ⬍ .05) and interpersonal processes (r ⫽ .21, p ⬍ .05), but only marginally with transition processes (r ⫽ .14, p ⬍ .10). Simultaneously regressing the composite on the three variables yielded R2 ⫽ .074, p ⬍ .05. Inspection of the individual beta weights, however, illustrates that action processes had a marginal relationship (␤ ⫽ .21, p ⬍ .10), whereas neither transition processes (␤ ⫽ .02, ns) nor interpersonal processes (␤ ⫽ .08, ns) were significant. In summary, these more detailed analyses are consistent with the idea that the collective aspects of the three team processes are working in concert in this study to drive the overall effects. However, these findings also suggest that there may be some utility in more detailed approaches in the future.

Discussion The aim of this research was to test a model of team psychological empowerment in the context of an integrated IPO framework. Additionally, our goals were to examine whether team processes play a mediational role linking empowerment and team outcomes and to incorporate methodological advances into the study of team empowerment. Using perceptual survey data, archival performance measures, and external customer satisfaction ratings and structural equation modeling, we found that team empowerment is significantly influenced by the embedding organizational environment. Both organizational support and team-based HR practices exhibited significant positive influences on empowerment beyond the influence of work design features. These results are consistent with our hypotheses, as well as with prior research (Kirkman & Rosen, 1999). These findings extend further prior empowerment research in that we found that team processes fully mediated the relationship between empowerment and quantitative performance. Alternatively, although customer satisfaction and empowerment were not correlated significantly, they were associated indirectly through each having significant relationships with team processes. Our data further suggest that empowerment, as a psychological state, influences team process, but the relationship is not reciprocal. Finally, our revised model suggests that empowerment and team processes do not fully mediate the relationships between work design features and organizational supports and quantitative performance. 3

Further details available from John E. Mathieu.

106

MATHIEU, GILSON, AND RUDDY

Theoretical and Practical Implications Our hypothesized model received substantial (although not complete) support. Team empowerment evidenced the anticipated positive relationship with team processes, which in turn, related significantly to both customer satisfaction and quantitative team performance. The more detailed mediational analyses revealed that an empowerment 3 performance relationship was fully mediated by team processes. In contrast, empowerment exhibited no significant relationship with customer satisfaction but was associated indirectly via team processes. Although we hypothesized and found support for parallel relationships linking team processes with customer satisfaction and quantitative performance, the two criteria were substantially different and only marginally correlated. This implies that the effects of empowerment are fairly ubiquitous and relate positively to a number of the different outcomes, but perhaps through different underlying mechanisms. Our supplemental analyses suggest that the indirect relationship between empowerment and customer satisfaction might be most attributable to enhanced transition processes. In contrast, our results suggest that action processes were perhaps serving as the primary mediating mechanism through which empowerment relates to quantitative performance. Whereas our hypothesized model implies that the team empowerment 3 process mechanism fully mediates the relationship between work design and the various supports on team outcomes, our revised model suggests that the underlying relationships were more complex than initially anticipated. Specifically, team-based HR practices not only related significantly to team empowerment but also exhibited a direct positive relationship with team processes. This suggests that the team-based HR functions went beyond facilitating feelings of responsibility and control among members to provide them with the necessary skills to execute effective team processes. Similarly, the work design features not only related positively to team empowerment but also evidenced direct positive relationships with quantitative performance. Therefore, it appears that structural empowerment efforts are not only beneficial in terms of their impact on psychological states but also directly enhance team effectiveness by shifting the nexus of decision making to those in the best position to handle problems. In other words, freeing CSEs to make on-the-spot decisions about how to deal with a machine (e.g., to repair or replace parts) enables them to better manage their environment and enhances quantitative performance. These findings are consistent with prior research that has found that when teams are proactive and free to shape their work environment, they are more effective compared with teams that remain in a reactive mode (i.e., Ancona & Caldwell, 1992; Tesluk & Mathieu, 1999). Two other results warrant special mention. First, although external team leadership evidenced a significant correlation with team empowerment, it failed to exhibit a significant unique influence in the context of the structural model. This finding may be attributable to a statistical by-product of its high covariance with other antecedents, although supplemental analyses indicated that multicollinearity was not a serious problem.4 Alternatively, external leaders may, indeed, influence team empowerment indirectly by shaping aspects of the work environment. That said, our result may be a substantive finding in that the role of external managers was simply overshadowed by the work design and other forms of support. Additional evidence obtained from the organization, however, suggested that whereas the role of teams had changed dra-

matically, external managers had not substantially altered their behaviors. In other words, it could be that the external managers were simply not yet “on board” with the change and had not fully bought into the new work environment. Certainly, empowered designs such as the one we studied pose threats to the traditional role of external leaders. Developing the support of external leaders has been seen as difficult to engender, yet critical to the effectiveness of empowering interventions (Druskat & Wheeler, 2003). In fact, the sponsoring organization focused its attention on developing external management support for the empowerment initiative as a consequence of this study. Nevertheless, this relationship warrants further research. The negative direct influence of organizational support on quantitative performance is a more puzzling finding. One potential explanation could be that teams might hamper their own performance if they allocate substantial resources to cross-team efforts to maximize overall effectiveness (Choi, 2002). However, this finding is likely to be spurious because none of the zero-order correlations between the organizational support subscales and the indices of quantitative performance were significant. In other words, the evidence is more consistent with this being a statistical artifact attributable to a suppressor-type effect than a substantive relationship. Because we did not anticipate this relationship, and it is counterintuitive, it needs to be replicated before we have any confidence in it.

Study Limitations and Future Research Our structural model tests relied heavily on data collected from team members, although the outcomes came from separate sources and were lagged in time. Consequently, our causal ordering of the antecedents, empowerment, and team processes are subject to debate. Although we formulated a model that is consistent with traditional IPO theory and research (e.g., Hackman & Morris, 1975; Marks et al., 2001) and tested whether a reciprocal process 3 empowerment relationship was evident, nevertheless causal inferences would be strengthened by the use of a complete longitudinal design that assesses all substantive variables multiple times. Regarding directions for future research, our supplemental analyses suggest that it may prove informative to examine the dynamics of team processes at a more fine-grained level. Marks et al. (2001) differentiated three different transition, action, and interpersonal bases of team processes and argued that they exhibited differential effects with team outcomes. Of importance, their work was cast in a time-based episodic framework, which was not feasible to test in our design. However, we encourage future researchers to examine the development of team emergent states, such as empowerment, along with team processes and outcomes, to better illuminate the dynamic nature of their relationships. Future investigations could also use more explicit multilevel designs and analyses to incorporate contextual influences on team empowerment effects. For example, we modeled the influences of organizational supports operationalized at the team level of analysis. However, teams are also subject to the influence of the larger embedding organizational environment (Gully, 2000; Guzzo & Dickson, 4

Detailed explanations of the analyses referred to in this section are available from John E. Mathieu.

EMPOWERMENT AND TEAM EFFECTIVENESS MODEL

1996; Klein & Kozlowski, 2000), and it would be valuable to model divisional, organizational, or even national influences on the effectiveness of empowered team designs. Finally, future investigations should seek to expand the array of antecedent variables that might influence the effectiveness of empowered teams. For instance, these could include team composition, task features, and the types of teams that are sampled. Because our teams were all fairly homogeneous and were all service teams, those factors certainly serve as boundaries for generalizations. Replications with more diverse teams performing different tasks are warranted. Finally, we adopted a fairly narrow, albeit traditional, definition and operationalization of team empowerment that focused on collective perceptions of authority and responsibility for work. Other researchers (e.g., Kirkman & Rosen, 1999) have used the four-dimensional approach to the study of team empowerment. Future research should consider the relative advantages and disadvantages of the two approaches and, perhaps, conduct comparison studies.

Conclusion Our results are both consistent with and extend previous research that has shown empowering work designs are correlated with enhanced team effectiveness (Hyatt & Ruddy, 1997; Kirkman & Rosen, 1999). In addition, our study highlights the need to consider, and to adapt, the larger organizational context to support an empowerment initiative. A work redesign effort will not yield optimal levels of empowerment unless it is complemented by various support mechanisms, such as facilitating organizational climate and team-based HR practices. This highlights the fact that for empowerment interventions to be effective, they must be comprehensive and not simply introduce limited changes in the design of work.

References Alper, S., Tjosvold, D., & Law, K. S. (2000). Conflict management, efficacy, and performance in organizational teams. Personnel Psychology, 53, 625– 642. Ancona, D. G., & Caldwell, D. F. (1992). Demography and design: Predictors of new product team performance. Organizational Science, 3, 321–341. Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: A review and recommended two-step approach. Psychological Bulletin, 103, 411– 423. Arnold, J. A. Arad, S., Rhoades, J. A., & Drasgow, F. (2000). The Empowering Leadership Questionnaire: The construction and validation of a new scale for measuring leader behaviors. Journal of Organizational Behavior, 21, 249 –269. Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychology research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1183. Beer, M., Eisenstat, R., & Spector, B. (1990). The critical path to corporate renewal. Boston: Harvard Business School Press. Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107, 238 –246. Bliese, P. D. (2000). Within-group agreement, non-independence, and reliability: Implications for data aggregation and analyses. 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.

107

Bowen, D., & Lawler, E. (1992). The empowerment of service workers: What, why, how, and when? Sloan Management Review, 33, 31–39. Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In. K. A. Bollen & J. S. Long (Eds.), Testing structural equation models (pp. 136 –162). Newbury Park, CA: Sage. Campion, M. A., Medsker, G. J., & Higgs, A. C. (1993). Relations between work group characteristics & effectiveness: Implications for designing effective work groups. Personnel Psychology, 46, 823– 847. Chen, G., Mathieu, J. E., & Bliese, P. D. (2004). A framework for conducting multilevel construct validation. In F. J. Dansereau & F. Yammarino (Eds.), Research in multi-level issues: The many faces of multi-level issues (Vol. 3, pp. 273–303). Oxford, England: Elsevier Science. Choi, J. N. (2002). External activities and team effectiveness—Review and theoretical development. Small Group Research, 33, 181–208. Cohen, S. G., & Ledford, G. E., Jr. (1994). The effectiveness of selfmanaging teams: A quasi-experiment. Human Relations, 47, 13– 43. Conger, J., & Kanungo, R. (1988). The empowerment process: Integrating theory and practice. Academy of Management Review, 13, 471– 482. Cook, R. A., & Goff, J. L. (2002). Coming of age with self-managed teams: Dealing with a problem employee. Journal of Business & Psychology, 16, 485– 496. Cortina, J. M. (1993). What is coefficient alpha? An examination of theory and applications. Journal of Applied Psychology, 78, 98 –104. Cummings, T. G. (1978). Self-regulating work groups: A socio-technical synthesis. Academy of Management Review, 3, 625– 634. Druskat, V. U., & Kayes, D. C. (2000). Learning versus performance in short-term project teams. Small Group Research, 31, 328 –353. Druskat, V. U., & Wheeler, J. V. (2003). Managing from the boundary: The effective leadership of self-managing work teams. Academy of Management Journal, 46, 435– 457. Edmondson, A. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44, 350 –383. Ford, R. C., & Fottler, M. D. (1995). Empowerment: A matter of degree. Academy of Management Executive, 9, 21–29. Gilson, L. L., Shalley, C. E., & Blum, T. C. (2001). Team and organizational attitudes as a lens and mirror impacting customer satisfaction; an empirical test in self-managed teams. Journal of Quality Management, 6, 235–256. Gully, S. M. (2000). Work teams research: Recent findings and future trends. In M. M. Beyerlein (Ed.), Work teams: Past, present and future (pp. 25– 44). Dordrecht, Netherlands: Kluwer Academic. Guzzo, R. A., & Dickson, M. W. (1996). Teams in organizations: Recent research on performance and effectiveness. Annual Review of Psychology, 47, 307–338. Hackman, J. R. (1987). The design of work teams. In J. W. Lorsch (Ed.), Handbook of organizational behavior (pp. 315–342). Englewood Cliffs, NJ: Prentice Hall. Hackman, R. (1990). Groups that work and those that don’t. San Francisco: Jossey-Bass. Hackman, J. R., & Morris, C. G. (1975). Group tasks, group interaction processes, and group performance effectiveness: A review and proposed integration. In L. Berkowitz (Ed.), Advances in experimental social psychology (pp. 45–99). New York: Academic Press. Hackman, J. R., & Oldham, G. R. (1976). Motivation through the design of work: Test of a theory. Organizational Behavior and Human Performance, 16, 250 –279. Hammer, M., & Champy, J. A. (1993). Re-engineering the corporation: A manifesto for business revolution. New York: Harper Collins. Hardy, C., & Leiba-O’Sullivan, S. (1998). The power behind empowerment: Implications for research and practice. Human Relations, 51, 451– 483. Hechanova-Alampay, R., & Beehr, T. A. (2001). Empowerment, span of control, and safety performance in work teams after workforce reduction. Journal of Occupational Health Psychology, 6, 275–282. Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance

108

MATHIEU, GILSON, AND RUDDY

structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55. Hyatt, D. E., & Ruddy, T. M. (1997). An examination of the relationship between work group characteristics and performance: Once more into the breech. Personnel Psychology, 50, 553–585. Ilgen, D. R., Hollenbeck, J. R., Johnson, M., & Jundt, D. (2005). Teams in Organizations: From I-P-O models to IMOI models. Annual Review of Psychology, 56, 517–543. James, L. R., & Brett, J. M. (1984). Mediators, moderators, and tests for mediation. Journal of Applied Psychology, 69, 307–321. James, L. R., Demaree, R. G., & Wolf, G. (1984). Estimating within group interrater reliability with and without response bias. Journal of Applied Psychology, 69, 85–98. Kirkman, B. L., & Rosen, B. (1997). A model of work team empowerment. Research in Organizational Change and Development, 10, 131–167. Kirkman, B. L., & Rosen, B. (1999). Beyond self-management: Antecedents and consequences of team empowerment. Academy of Management Journal, 42, 58 –74. Kirkman, B. L., Rosen, B., Tesluk, T., & Gibson, C. (2004). The impact of team empowerment on virtual team performance: The moderating role of face-to-face interaction. Academy of Management Journal, 47, 175–192. Kirkman, B. L., & Shapiro, D. L. (2001). The impact of team members’ cultural values on productivity, cooperation, and empowerment in selfmanaging work teams. Journal of Cross-Cultural Psychology, 32, 597– 617. Klein, K., & Kozlowski, S. W. J. (2000). Multilevel theory, research and methods in organization. San Francisco: Jossey-Bass. Koberg, C. S., Boss, R. W., Senjem, J. C., & Goodman, E. A. (1999). Antecedents and outcomes of empowerment: Empirical evidence from the healthcare industry. Group and Organizational Management, 24, 71–91. Kozlowski, S. W. J., & Bell, B. S. (2001). Work groups and teams in organizations. In W. Borman, D. Ilgen, & R. Klimoski (Eds.), Comprehensive handbook of psychology: Industrial and organizational psychology (pp. 334 –375). New York: Wiley. Landis, R., Beal, D., & Tesluk, P. (2000). A comparison of approaches to forming composites measures in structural equations models. Organizational Research Methods, 3, 186 –207. Lawler, E. E. (1986). High involvement management, San Francisco: Jossey-Bass. Leach, D. J., Wall, T. D., & Jackson, P. R. (2003). The effects of empowerment on job knowledge: An empirical test involving operators of complex technology. Journal of Occupational Psychology, 76, 27–52. Lewin, K. (1947). Frontiers in group dynamics: II. Channels of group life; social planning and action research. Human Relations, 1, 143–153. Liden, R. C., & Tewksbury, T. W. (1995). Empowerment in work teams. In G. Ferris, S. Rosen, & D. Barnum (Eds.), Handbook of human resources management (pp. 386 – 403). Oxford, England: Blackwell. Manz, C. C., & Sims, H. P. (1987). Leading workers to lead themselves: The external leadership of self-managing work teams. Administrative Science Quarterly, 32, 106 –129. Marks, M. A., DeChurch, L. A., Mathieu, J. E., Panzer, F. J., & Alonso, A. A. (2005). The importance of goal hierarchies and teamwork processes for multi-team effectiveness. Journal of Applied Psychology, 90, 964 –-971. Marks, M. A., Mathieu, J. E., & Zaccaro, S. J. (2001). A temporally based framework and taxonomy of team processes. Academy of Management Review, 26, 356 –376. Mathieu, J. E., Marks, M. A., & Zaccaro, S. J. (2001). Multi-team systems. In N. Anderson, D. S. Ones, H. K. Sinangil, & C. Viswesvaran (Eds.), Handbook of industrial, work and organizational psychology, (Vol. 2, pp. 289 –313), London, Sage. McGrath, J. E. (1984). Groups: Interaction and performance. Englewood Cliffs, NJ: Prentice Hall. Medsker, G. J., Williams, L. J., & Holahan, P. J. (1994). A review of

current practices for evaluating causal models in organizational behavior and human resources management research. Journal of Management, 20, 439 – 464. Menon, S. T. (2001). Employee empowerment: An integrative psychological approach. Applied Psychology: An International Review, 50, 153–180. Mesch, D. J., Farh, J. L., & Podsakoff, P. M. (1994). Effects of feedback sign on group goal setting, strategies, and performance. Group and Organizational Management, 19, 309 –333. Mills, P. K., & Ungson, G. R. (2003). Reassessing the limits of structural empowerment: Organizational constitution and trust as controls. Academy of Management Review, 28, 143–153. Nadler, A. (1979). The effects of feedback on task group behavior: A review of the experimental research. Organizational Behavior and Human Decision Processes, 23, 309 –338. Perry, M. L., Pearce, C. L., & Sims, H. P. (1999). Empowered selling teams: How shared leadership can contribute to selling team outcomes. Journal of Personal Selling & Sales Management, 19(3), 35–51. Peters, T. J. (1992). Liberation management: Necessary disorganization for the nanosecond nineties, New York: Knopf. Pfeffer, J., Cialdini, R. B., Hanna, B., & Knopof, K. (1998). Faith in supervision and the self-enhancement bias: Two psychological reasons why managers don’t empower workers. Basic & Applied Social Psychology, 20, 313–321. Pritchard, R. D., Jones, S. D., Roth, P. L., Stebing, K. K., & Ekeberg, S. E. (1988). Effects of group feedback, goal setting, and incentives, on organizational productivity. Journal of Applied Psychology, 73, 337–358. Salas, E., & Cannon-Bowers, J. A. (2001). The science of training: A decade of progress. Annual Review of Psychology, 52, 471– 499. Schwab, D. P. (1980). Construct validity in organizational behavior. Research in Organizational Behavior, 2, 3– 43. Sims, H. P. J., & Manz, C. C. (1984). Observing leader verbal behavior: Toward reciprocal determinism in leadership theory. Journal of Applied Psychology, 6, 222–232. Spreitzer, G. M. (1995). Psychological empowerment in the workplace: Dimensions, measurement, and validation. Academy of Management Journal, 38, 1442–1465. Spreitzer, G. M. (1996). Social structural characteristics of psychological empowerment. Academy of Management Journal, 39, 483–504. Sundstrom, E. (1999). The challenges of supporting work team effectiveness. In E. Sundstrom (Ed), Supporting work team effectiveness (pp. 3–23). San Francisco: Jossey-Bass. Tesluk, P., & Mathieu, J. E. (1999). Overcoming roadblocks to effectiveness: Incorporating management of performance barriers into models of work group effectiveness. Journal of Applied Psychology, 84, 200 –217. Thomas, K. W., & Velthouse, B. A. (1990). Cognitive elements of empowerment an interpretive model of intrinsic task motivation. Academy of Management Review, 15, 666 – 681. Wageman, R. (1995). Interdependence and group effectiveness. Administrative Science Quarterly, 40, 145–180. Wageman, R. (2001). How leaders foster self-managing team effectiveness: Design choices versus hands-on coaching. Organization Science, 12, 559 –577. Wall, T. D. Cordery, J. L., & Clegg, C. W. (2002). Empowerment, performance and occupational uncertainty: A theoretical integration. Applied Psychology: An International Review, 51, 146 –169. Weingart, L. R. (1997). How did they do that? The ways and means of studying group process. Research in Organizational Behavior, 19, 189 – 239. Wellins, R., Byham, W., & Wilson, J. (1991). Empowered teams. San Francisco: Jossey-Bass.

Received December 24, 2003 Revision received October 13, 2004 Accepted January 24, 2005 䡲