The Role of the Star Player in a Cohesive Group

Abstract

Based on social learning theory, we explore when the most competent member in the group leads to high group performance. We argue that the most competent member in the group increases group performance in high cohesive groups where members interact more frequently and maintain closer relationships with one another. To examine this, we used multisource data collected in two waves from an organization in South Korea. The results indicate that the highest competency within the group is positively related to group performance being perceived by the group members and by the upper management in cohesive groups. Moreover, we found different patterns for the group performance rated by different sources.

Keywords

cohesion group process modeling

An increasing number of organizations have begun to structure work around work groups. To complete group tasks, each member’s contribution needs to be coordinated through interpersonal interactions. Moreover, by working in a group, an individual influences the entire group, not just a few members. As determinants of successful group work in organizations, members’ characteristics that contribute to group performance have been of great interest to researchers. Particularly, the configuration of member attributes (e.g., group mean, minimum, or maximum of members’ characteristics) in a group has been thought to have a significant impact on group processes and outcomes (Kozlowski & Bell, 2003). A meta-analysis suggests that some group characteristics composed of members’ individual characteristics such as abilities, personality, and values are significantly related to group performance (e.g., see Bell, 2007 for reviews).

Even with the proliferation of research on groups, we still know little about the influence of an individual group member on group outcomes. Based on the premise that a member’s specific individual characteristic contributes equally to the collective pool of that characteristic, the majority of research has concluded that the average or median of the group describes the group characteristic. However, this stream of research may not capture the fact that social information arising from a member, especially from the most competent group member, may influence group process and outcomes.

Social learning theory (Bandura, 1977) supports the possible role-modeling effect of the best performer within the group on group performance. Individuals learn desirable behaviors in specific contexts and develop a belief about their capabilities to succeed at the task by observing the role model’s behaviors. This means that members are more likely to have an efficacy belief and perform better when the most capable person in the group is a high performer, compared with when the most capable person in the group performs at a mediocre level. Bell’s (2007) meta-analysis empirically offers evidence on this argument. Specifically, he examined the impact of members’ generalized mental ability (GMA) and members’ ability to be effective in different organizations on group performance. He found that the highest level of GMA in the group (i.e., group maximum of GMA) is significantly related to group performance in lab studies but not in field studies in which researchers have minimal latitude for controlling situations. This result indicates that the individual member with the highest potential may influence the group performance, but this may not be always the case in the actual organizational setting. This can be either because GMA is not the best indicator of employees’ work-related ability in an organizational setting or because there are situational factors that differentiate the relationship, or both.

To find clearer answers to these possibilities, we investigate the impact of the most competent member in the group in determining group performance and a boundary condition for this relationship. Competency, an individual’s desirable behavior to achieve specific business objectives, reflects members’ organization-specific ability (Bartram, Robertson, & Callinan, 2002; Stoof, Martens, Merrienboer, & Bastiaens, 2002). Based on a universalistic approach, micro organizational behavior research posits that an individual’s general ability has a positive impact on individual and group outcomes across diverse contexts (Schmidt & Hunter, 1998). In contrast, researchers with macro perspectives call attention to the value of context-specific individual characteristics (e.g., Barney & Wright, 1998). Given diverse situations, different qualities of employees are more or less effective in different organizations. From a strategy point of view, scholars argue that individual competency is a significant predictor of employee performance in the organization (e.g., Bartram et al., 2002; Stoof et al., 2002). Considering that an employee with high competency tends to be perceived as an outstanding and successful person in the organization, the member with the highest competency within the group may have a strong influence on group performance. Thus, we argue that the competency of the most competent member within the group is significantly related to group performance as rated by group members and by upper management. By employing the group maximum for exploring group processes, the current study contributes to filling the gap in the group composition literature where the possibility of a single member’s influence on the whole group has been largely neglected.

Furthermore, we attempt to identify a boundary condition that facilitates the role-modeling effect of the most competent member in the group. Social learning theory (Bandura, 1977) suggests that having close relationships and frequent interactions with the role model is required for a vicarious learning process to occur. That is, members should have a good relationship with the member deemed to be a role model to learn the member’s behaviors and develop an efficacy belief. As such, we argue that group cohesiveness, which captures the extent to which members maintain close interpersonal relationships with each other (Beal, Cohen, Burke, & McLendon, 2003), strengthens the positive influence of the most competent member on other members’ work behaviors.

The Competence of the Most Competent Member and Group Performance

Competency is an organization-specific set of individuals’ characteristics and skills that leads to success in the organization (Bartram et al., 2002; Stoof et al., 2002). The resource-based view (Barney, 1986), which describes four criteria of resources for attaining competitive advantages (i.e., valuable, rare, imperfectly imitable, and not substitutable), suggests that human resources are essential resources for organizational viability. To capture employees’ value in contributing to the organization’s competitive advantages, researchers in human resource management and organizational psychology attempt to develop a general competency model. Although competency is more about the individual’s potential or capabilities rather than behavior, one’s competency would be evaluated based on their attitudes and behaviors. When an employee displays organization-specific desirable behaviors, other people would rate this employee’s competence highly. In this study, we employ supervisor-rated individual competency, which depicts a behaviorally specific description of the skills and traits that employees need to be effective in the organization (Mansfield, 1996).

Social learning theory (Bandura, 1977) supports the idea that an individual with an exceptional level of competency can improve group performance through vicarious learning. The vicarious learning process explains that individuals learn behaviors and form self-efficacy through observation of the role models’ behaviors. Moreover, this process may extend to the group level. When individuals see a competent and successful person, they identify him or her as a role model and learn that person’s behaviors. In a work group, the most competent member, who displays the most desirable behaviors, is apt to emerge as a role model to others. By observing the focal member’s behaviors, other members learn advantageous behaviors in organization-specific contexts and engage in them. Indeed, the members improve their belief in their capabilities to achieve the given task (i.e., collective efficacy) and eventually improve their actual work behaviors (Sundstrom, De Meuse, & Futrell, 1990). As such, we argue that the most competent member’s competency is pivotal for group members’ perceptions of group performance and objective performance rated by upper management. If the most competent member has an exceptionally high competency, he or she tends to be an outstanding and successful person in the organization and to perform as an effective role model. In contrast, if the highest performing member in the group has only a mediocre level of competency, then this member may not significantly enhance other group members’ performance or efficacy beliefs.

The social learning processes that have been frequently studied in leadership literature have demonstrated the role-modeling effect of the leader. However, scholars have suggested that coworkers may play a pivotal role in shaping the individual members’ social environment, just as leaders do (Seers, 1989). Although the leader formally designated by the organization has the formal power, group members may not always believe that the leader’s behaviors are the most desirable ones. For instance, if the leader got the position mainly because he or she has worked for a long time, the group members may not consider the leader to be a good role model. Indeed, a meta-analysis revealed that coworkers are more influential in determining some employee work outcomes (e.g., job involvement) than their leaders (Chiaburu & Harrison, 2008). Therefore, we argue that the person who has the highest competency among all members and the group leader will significantly influence group performance rated by group members and upper management.

Hypothesis 1: The competency of the most competent member within the group is positively related to (a) peer-rated group performance and (b) upper management–rated group performance.

The Moderating Effect of Group Cohesiveness

Social learning theory proposes that the interpersonal relationship with the role model may reinforce the vicarious learning process. When the observers have a close relationship with the role model, they interact more frequently with the role model and are more likely to adopt the role model’s behaviors (Bandura, 1977). Moreover, the members’ close relationship with the role model facilitates the learning process by contributing to their identification with the role model (Kerr & Kaufman-Gilliland, 1994). When this role model is exceptionally competent, the employee can learn how to perform better and therefore feel more competent (Gist, 1989). Extending this rationale to the group level, it is plausible that the vicarious learning process among group members may be more salient in groups where members have close relationships with one another.

Group cohesiveness, which refers to the degree to which members feel attached to their groups and are willing to remain in the group (Price & Muller, 1986), forms a supportive milieu for group members to learn from the most competent member. Indeed, group cohesiveness forms a belief that members share similarities and are close and unified (Carron & Brawley, 2000). In a cohesive group, members achieve social integration and maintain strong interpersonal bonds with other members (Beal et al., 2003). Moreover, a meta-analysis has shown that group cohesiveness is relevant to performance in some situations (Gully, Devine, & Whitney, 1995; Langfred, 1998). Given close relationships with the most competent member in the group, members’ vicarious learning process and identification with the member is reinforced. In other words, members in a cohesive group may learn advantageous behaviors from the role model and share a belief that they can perform as well as their role model can. Thus, they may evaluate their group performance as being higher when they have a high performer in the group. Moreover, they may actually perform better by imitating the most competent member’s behaviors and feeling competent.

Hypothesis 2: Group cohesiveness moderates the positive relationships between the competency of the most competent member within the group and (a) peer-rated group performance and (b) upper management–rated group performance such that the relationships are stronger when group cohesiveness is high than low.

Method

Sample and Data Collection

Data were collected from a manufacturing company in South Korea in two waves. During the first wave, composite scores of a competency rating for each employee and group performance rated by upper management were collected. After 5 months, surveys asking about group cohesiveness and peer-rated group performance were distributed to 462 employees and 52 group leaders. Among these groups, those with more than three people and available performance data were included. Therefore, 427 members of 47 groups were used for analysis. Of the respondents, 96.9% were male and the average age was 32 years. In terms of the highest education level received, 61% reported a bachelor’s degree and 16% reported a master’s degree or higher.

Measures

Competency

The score of each employee’s competency was acquired from the records of the human resource management department. As a company-level project, the company developed a competency model based on Boyatzis (1982) framework. Supervisors assessed their subordinates’ competencies with 5-point scale. The most competent member’s competency is the highest competency score within the group.

Group cohesiveness

Six items from Price and Muller (1986) were used (α = .84). A sample item was “How well do members of your group get along with each other?” Following Chan (1998), we used the group as the referent to measure group-level characteristics based on individual responses and aggregated these to the group level. The r_WG was 0.75, justifying the aggregation of group cohesiveness. Employees responded on 7-point Likert-type scales.

Peer-rated group performance

Four items from Brannick, Salas, and Prince (1997) were employed to measure peer-rated group performance (α = .85). A sample item was “Has our group been a great success so far?” Members’ perceptions of group performance were aggregated to measure peer-rated group performance with a 7-point scale. The r_WG was 0.81.

Upper management–rated group performance

The human resource management department of the company provided the composite score of group performance rated by upper management. The criteria for team performance included fulfillment of tasks, completion of projects, and degree of innovation. A perfect score was 100 points.

Control variables

We controlled for the lowest competency level, and the median of competency for testing the hypotheses. The control variables were standardized to avoid multicollinearity between control variables and predictor variables. Steiner’s (1972) task typology proposes that different operationalization of individual-level variables at group level (i.e., the group median or mean, the group maximum, and the group minimum) may be preferable for predicting different types of tasks. For instance, the group median or mean is thought to be the best operationalization for additive tasks to which each group member contributes equally and for compensatory tasks that allow high-performing members to compensate for the poor performance of low performers. On the other hand, the group maximum can be appropriate for predicting disjunctive tasks that can be completed by a member solving the problem. Finally, for the conjunctive tasks that require a minimally acceptable level of work from every group member, the group minimum may determine the group task. In this study, we did not specify what types of tasks the participants perform. Moreover, in reality, group tasks are combined in complex ways and cannot be categorized into a single type from Steiner’s (1972) taxonomy (McGrath, 1984). To clarify the effect of the best member’s competence on performance, we attempted to rule out the potential influence of other group members’ competencies on group performance by controlling the group median of competency and the lowest competency within the group.

Results

Table 1 presents the means, standard deviations, and correlations among the research variables.

Table 1.

Means, Standard Deviations, and Correlations Among Study Variables.

Variables	M	SD	1	2	3	4	5
1. Group median of competency	3.99	0.18
2. The lowest competency in the group	3.58	0.38	.57**
3. The highest competency in the group	4.27	0.20	.55**	.08
4. Group cohesiveness	5.33	0.53	.19	.24	−.20
5. Peer-rated group performance	5.26	0.66	−.09	.17	−.29*	.50**
6. Upper management–rated group performance	76.83	11.36	−.18	−.02	.08	.36*	.65**

Note. N = 47.

p < .05. **p < .01.

Hypothesis 1 proposes a positive relationship between the best member’s competency and group performance. We conducted a series of hierarchical regression analyses to examine our hypotheses. To reduce the potential problems of multicollinearity (Cohen, Cohen, West, & Aiken, 2003), both independent variables and product terms were standardized before the analyses. As shown in Model 2 in Tables 2 and 3, the highest level of competency within the group was not significantly related to peer-rated group performance (β = −.27, ns) and upper management–rated performance (β =.03, ns). We also created confidence intervals for the coefficients. The 95% confidence intervals for the main effects included zero ([−.67, .10] and [−5.57, .63], for Hypothesis 1a and Hypothesis 1b, respectively), resulting in the lack of support for both hypotheses.

Table 2.

Results of the Relationship Between Competency, Group Cohesiveness, and Peer-Rated Performance.

Variables	Model 1	Model 2	Model 3	Model 4
Step 1: Controls
Group median of competency	−.27	−.08	−.16	−.07
The lowest competency	.33	.24	.16	.07
Step 2: Main effect
The highest competency		−.27	−.21	−.15
Group cohesiveness			.49**	.52**
Step 3: Moderating effect
Group cohesiveness × HC				.30*
Overall F	1.93 (2, 44)	2.07 (3, 43)	5.59** (4, 42)	5.90** (5, 41)
R ²	.08	.13	.35	.42
F change	1.93	2.25	14.25**	4.98*
R² change	.08	.05	.22	.07

Note. N = 47. HC = highest competency within the group.

p < .05. **p < .01.

Table 3.

Results of the Relationship Between Competency, Group Cohesiveness, and Upper Management–Rated Performance.

Variable	Model 1	Model 2	Model 3	Model 4
Step 1: Controls
Group median of competency	−.22	−.24	−.26	−.35
The lowest competency	.09	.10	−.02	−.22
Step 2: Main effect
The highest competency		.03	.15	.11
Group cohesiveness			.42*	.34
Step 3: Moderating effect
Group cohesiveness × HC				.56**
Overall F	.56 (2, 29)	.37 (3, 28)	1.55 (4, 27)	3.62* (5, 26)
R ²	.04	.04	.19	.41
F change	.56	.02	4.96*	9.84*
R² change	.04	.00	.15	.22

Note. N = 32. HC = highest competency within the group.

p < .05. **p < .01.

We suggested the moderating effect of group cohesiveness on the relationship between the most competent member’s competency and performance in Hypothesis 2. Supporting Hypothesis 2a, the results in Model 3 in Table 2 show that the interaction between the highest competency level within the group and group cohesiveness is significant for predicting group performance rated by members (β = .30, p < .05). Moreover, as shown in Model 3 in Table 3, group cohesiveness significantly moderates the relationship between the best member’s competency and performance rated by the upper management (β = .56, p < .01). In addition, we calculated confidence intervals for these coefficients. The 95% confidence intervals for the main effects did not include zero ([.26, 5.20] and [.25, 1.21] for Hypotheses 2a and 2b, respectively); thus, both hypotheses were supported. To illustrate the nature of this interaction, we generated a graph by following the procedure specified by Cohen and colleagues (2003). Figure 1 presents the moderating effect of group cohesiveness on the relationship between the competency of the most competent member within the group and group performance rated by group members. Figure 2 presents the moderating effect of group cohesiveness on the relationship between the best member’s competency and upper management–rated group performance.

Figure 1.

The interaction effect of the highest competency within the group and group cohesiveness on peer-rated group performance.

Figure 2.

The interaction effect of the highest competency within the group and group cohesiveness on upper management–rated group performance.

In addition, we conducted simple slope tests to examine whether the relationships are significant for both high- and low-cohesive groups. The results indicate that while the impact of the highest competency on peer-rated group performance is significant when group cohesiveness is low (t = −2.29, p < .05), it is not significant when group cohesiveness is high (t = .46, ns). Furthermore, for predicting upper management–rated group performance, the highest competency is significant when group cohesiveness is high (t = 2.69, p < .05), whereas it is not significant when group cohesiveness is low (t = −1.63, ns). These results indicate that group cohesiveness interacts with the impact of the highest competency on group performance differently.

Discussion

The current study sheds light on the impact of the best member’s competency on group performance and a group context that strengthens the impact of this competency. We examined the relationships between the competence of the best member and peer-rated and upper management–rated group performance. Unexpectedly, the most competent member does not significantly improve peer-rated group performance or upper management–rated group performance. These findings indicate that the mere presence of a competent member in the group does not guarantee an improvement in group performance.

However, the relationships differ according to the level of group cohesiveness. For upper management–rated group performance, the effect of the highest competency in the group was positive and significant in cohesive groups. In line with social learning theory, this result indicates that close interpersonal relationships with the member with the highest competency, who may be considered a role model, are important for members’ vicarious learning. In a cohesive group where members frequently interact with each other, the members may learn effective work behaviors from the most competent member. When the member is exceptionally competent, the rest of the members learn the behaviors more efficiently and, consequently, the whole group is able to perform better.

Nevertheless, the social learning proposition does not fit with peer-rated group performance. In fact, we found a significant negative relationship between the competency of the most competent member and peer-rated group performance when group cohesiveness is low. A possible explanation can be derived from the contrast effect of upward social comparison in the social comparison literature (see Buunk & Gibbons, 2007, for a review). The theoretical and empirical research has supported a contrast effect, which indicates that people experience negative psychological processes, such as relative deprivation and having a negative self-view when they are with superior others. Moreover, such negative states can be exacerbated when they do not have a close relationship with the superior others (Mussweiler, Rueter, & Epstude, 2004). In non-cohesive groups, the contrast effect is salient because members do not have strong social bonds with one another. When an exceptionally competent member exists in the group, the members are apt to have a negative self-view and may assess themselves (and their group as a whole) as being at a lower level than they would be otherwise. In other words, when group cohesiveness is low, the competency of the most competent member may bring performance down rather than up.

In addition, low group cohesiveness may not only refrain from learning the most competent member’s desirable behaviors but also encourage members to take actions that deviate from the behaviors. Even if the members learned the behaviors from the most competent members, they may not want to imitate the behaviors or may even want to do the opposite because of psychological reactance (Brehm, 1966). In non-cohesive groups, members are apt to experience negative psychological states when an exceptional member exists in the group. Indeed, they may feel relative deprivation and even resentment toward the target of comparison (i.e., the most competent member; Crosby, 1976). As a result, they may deliberately engage in behaviors that may be destructive to group performance.

This study has theoretical implications for the literature on groups, social learning, and competency. First, this study highlights the role of the best member within the group in determining group performance. Despite a proliferation of studies on groups, a majority of them have illustrated group characteristics using the mean or median of the group. Although these approaches are appropriate for understanding aggregated individual characteristics at the group level, they cannot capture the implications of an outstanding member who may influence the whole group process. By considering the best member’s competence, this study broadens our perspective on group processes.

Second, finding different patterns for group performance from different sources, that is, group members and upper management, may add value to the literature on groups. Although group performance has frequently been used for group effectiveness, not many studies have highlighted the differences between these two types of performance. As such, we contend that researchers should choose the appropriate source of rating with a consideration of their research questions.

Third, this study proposes a boundary condition where the potential role-modeling effects become salient. Supporting social learning theory, we found that a close relationship between a role model (i.e., the most competent member in the group) and other members is an essential condition for facilitating the social learning process. Hence, this finding may contribute to the literature on social learning, particularly in group contexts.

Finally, we revisit the value of competency in understanding group phenomena. Although research on competency has been conducted for decades, most of them were theoretical, without empirical support. As such, the findings of this study support the previous emphasis on the importance of individuals’ competency and provide a possible approach to extend the stream of research on competency.

Practical Implications

This study provides some significant practical implications. The primary practical implication comes from a more nuanced understanding of the group process. Given the increasing tendency of having employees work in groups, this understanding is necessary for practitioners when they design groups as well as for group leaders. When managers construct their groups, they need to consider an individual member’s competency and his or her potential influence on the whole group. In particular, the manager needs to carefully assign an exemplary employee who can emerge as a role model for the group. However, to achieve high group effectiveness, the manager needs to understand a subtlety of this potential role-modeling effect. Considering the potential negative impact of a star player in a non-cohesive group, the manager needs to take group cohesiveness into consideration when designing the group. They may expect better performance from a group with a star player when the group is high in cohesiveness. In contrast, they need to be aware that the star player may decrease the group performance when he or she is assigned to a group with low cohesiveness. Moreover, the group leader’s role in improving group cohesiveness is pivotal if there is a member with exceptionally high competency. The group leader’s ability to facilitate interpersonal relationships among the group members will help attain the advantages of the role-modeling effect.

Limitations

First, generalizability may be limited because the competency measure used in this article was developed based on organization-specific information. Although the company developed the measurement based on a general competency model (Boyatzis, 1982), the surveys included organizational-specific items. Although this measurement was effective for capturing group processes within that organization, it may not be replicable in other contexts using the same methods. Therefore, replicating these findings with a more generalizable measure may be helpful to confirm the hypothesized group processes. Second, the mechanisms explaining the relationships between the level of competency and group performance were not measured. Specifically, we cannot conclude whether members actually model the behaviors of the member with the highest level of competency in their groups. Third, although previous research has indicated that the maximum competency levels of the group may influence group performance when group tasks are disjunctive (Steiner, 1972), different types of tasks were not tested in this study. Thus, future research may investigate these mechanisms. Fourth, this study mainly focuses on the role of star players in work groups. However, in some situations, the member who has the lowest level of competency within the group may significantly influence group performance. In particular, when the group task is conjunctive and the existence of a bottleneck slows the whole work process, the role of the bottleneck may become more salient. Therefore, future research may delve into how the least competent member influences group performance. Fifth, given that some variables are from secondary data (i.e., group members’ competency), the researchers did not have full control over the data. Moreover, the measure of competency may have an inherent limitation. Employee competency was measured by managers using the employees’ behaviors and attitudes as proxies of competency. Considering that competency, by definition, is not behavior, but potentiality or capability, managers’ ratings may not fully capture employee competency.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author Biographies

Hae Sang Park is a doctoral candidate in management at the University of Illinois at Chicago, USA. Her research focuses on interpersonal relationships within groups, positive and negative leadership, and employee behaviors.

Soo-Young Shin is an assistant professor in the School of Business at Yeungnam University, Korea. Her research focuses on group dynamics, top management leadership, and self-regulatory perspectives on motivation.

References

Bandura

(1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191-215. doi:10.1037//0033-295X.84.2.191

Barney

J. B.

(1986). Strategic factor markets: Expectations, luck and business strategy. Management Science, 42, 1231-1241. doi: 10.1287/mnsc.32.10.1231

Barney

J. B.

Wright

P. M.

(1998). On becoming a strategic partner: The role of human resources in gaining competitive advantage. Human Resource Management, 37, 31-46. doi:10.1002/(SICI)1099-050X(199821)37:1<31::AID-HRM4>3.0.CO;2-W

Bartram

Robertson

I. T.

Callinan

(2002). Introduction: A framework for examining organizational effectiveness. In Robertson

I. T.

Callinan

Bartram

(Eds.), Organizational effectiveness: The role of psychology (pp. 1-10). Chichester, UK: John Wiley.

Beal

D. J.

Cohen

R. R.

Burke

M. J.

McLendon

C. L.

(2003). Cohesion and performance in groups: A meta-analytic clarification of construct relations. Journal of Applied Psychology, 88, 989-1004. doi:10.1037/0021-9010.88.6.989

Bell

S. T.

(2007). Deep-level composition variables as predictors of team performance: A meta-analysis. Journal of Applied Psychology, 92, 595-615. doi:10.1037/0021-9010.92.3.595

Boyatzis

R. E.

(1982). The competent manager: A model for effective performance. New York, NY: John Wiley.

Brannick

M. T.

Salas

Prince

(1997). Team performance assessment and measurement. Mahwah, NJ: Lawrence Erlbaum.

Brehm

J. W.

(1966). A theory of psychological reactance. New York, NY: Academic Press.

10.

Buunk

A. P.

Gibbons

F. X.

(2007). Social comparison: The end of a theory and the emergence of a field. Organizational Behavior and Human Decision Processes, 102, 3-21. doi:10.1016/j.obhdp.2006.09.007

11.

Carron

A. V.

Brawley

L. R.

(2000). Cohesion: Conceptual and measurement issues. Small Group Research, 31, 89-106. doi:10.1177/104649640003100105

12.

Chan

(1998). Functional relations among constructs in the same content domain at different levels of analysis: A typology of composition models. Journal of Applied Psychology, 83, 234-246. doi:10.1037//0021-9010.83.2.234

13.

Chiaburu

D. S.

Harrison

D. A.

(2008). Do peers make the place? Conceptual synthesis and meta-analysis of coworker effects on perceptions, attitudes, OCBs, and performance. Journal of Applied Psychology, 93, 1082-1103. doi:10.1037/0021-9010.93.5.1082

14.

Cohen

West

S. G.

Aiken

L. S.

(2003). Applied multiple regression/correlation analysis for the behavioral sciences (3rd ed.). Hillsdale, NJ: Lawrence Erlbaum.

15.

Crosby

(1976). A model of egoistical relative deprivation. Psychological Review, 83, 85-113. doi:10.1037//0033-295X.83.2.85

16.

Gist

M. E.

(1989). The influence of training method on self efficacy and idea generation among managers. Personnel Psychology, 42, 787-805. doi:10.1111/j.1744-6570.1989.tb00675.x

17.

Gully

S. M.

Devine

D. J.

Whitney

D. J.

(1995). A meta-analysis of cohesion and performance: Effects of level of analysis and task interdependence. Small Group Research, 26, 497-520. doi:10.1177/1046496412468069

18.

Kerr

N. L.

Kaufman-Gilliland

C. M.

(1994). Communication, commitment, and cooperation in social dilemmas. Journal of Personality and Social Psychology, 48, 349-363. doi:10.1037//0022-3514.66.3.513

19.

Kozlowski

S. W. J.

Bell

B. S.

(2003). Work groups and teams in organizations. In Borman

W. C.

Ilgen

D. R.

Klimoski

R. J.

(Eds.), Handbook of psychology: Industrial and organizational psychology (Vol. 12, pp. 333-375). London, England: John Wiley.

20.

Langfred

(1998). Is group cohesiveness a double edged sword?: An investigation of the effects of cohesiveness on performance. Small Group Research, 29, 124-143. doi:10.1177/1046496498291005

21.

Mansfield

R. S.

(1996). Building competency models: Approaches for HR professionals. Human Resource Management, 35, 7-18. doi:10.1002/(SICI)1099-050X(199621)35:1%3C7::AID-HRM1%3E3.0.CO;2-2

22.

McGrath

J. E.

(1984). Group interaction and performance. Englewood Cliffs, NJ: Prentice-Hall.

23.

Mussweiler

Rueter

Epstude

(2004). The ups and downs of social comparison: Mechanisms of assimilation and contrast. Journal of Personality and Social Psychology, 87, 832-844. doi:10.1037/0022-3514.87.6.832

24.

Price

J. L.

Muller

C. W.

(1986). Handbook of organizational measurement. Marshfield, MA: Pittman. doi:10.1108/01437729710182260

25.

Schmidt

F. L.

Hunter

J. E.

(1998). The validity and utility of selection methods in personnel psychology: Practical and theoretical implications of 85 years of research findings. Psychological Bulletin, 124, 262-274. doi:10.1037//0033-2909.124.2.262

26.

Seers

(1989). Team–member exchange quality: A new construct for role-making research. Organizational Behavior and Human Decision Processes, 43, 118-135. doi:10.1016/0749-5978(89)90060-5

27.

Steiner

I. D.

(1972). Group process and productivity. New York, NY: Academic Press.

28.

Stoof

Martens

R. L.

Merrienboer

J. J. G.

Bastiaens

T. J.

(2002). The boundary approach of competence: A constructivist aid for understanding and using the concept of competence. Human Resource Development Review, 1, 345-365. doi:10.1177/1534484302013005

29.

Sundstrom

De Meuse

K. P.

Futrell

(1990). Work teams: Applications and effectiveness. American Psychologist, 45, 120-133. doi:10.1037//0003-066X.45.2.120