Peer control and team performance:
the mediating and moderating role of
team learning orientation and team identification
Department of Human Resource Management & Organizational Behavior Faculty of Economics and Business
University of Groningen
2 Abstract
The aim of this research is to examine the relation between peer control and team performance. It is proposed that team learning orientation mediates the relationship between peer control and team performance. Furthermore, team identification is proposed to play a moderating role in the relationship between peer control and team performance and the relationship between peer control and team learning orientation. The hypotheses are tested on the responses of 191 team members and 33 supervisors by two online-surveys which are conducted among 34 teams in Dutch organizations. The findings suggest that there is no mediation effect of team learning orientation on the relationship between peer control and team performance. In contrast, the proposed moderation effects of team identification on the relationship between peer control and team performance and the relationship between peer control and team learning orientation are found.
3 Introduction
Imagine that you are an employee, who is a member of a team in an organization. You are detecting behavior of a fellow team member which is not in line with the team standards. However, you are hesitating whether you will address your fellow team member about this incorrect behavior, because it could result in different outcomes. It can positively influence the team performance, but it might also negatively influence the relationship between yourself and your fellow team member. Since you speak up to your fellow team member about this incorrect behavior and show him/her the way to correct behavior, he/she could perform better in accordance with the team goals, which has a potentially positive effect towards team performance. However, your criticism could also be interpreted by your fellow team member as an attack on his/her identity or competencies which potentially results in a damaged relationship (Milliken, Morrison, & Hewlin, 2003).
Peer control is the behavior involved in the above-mentioned scenario, because peer control is about detecting incorrect behaviors of coworkers and informing them about it (Loughry & Tosi, 2008). Specifically, the worker who exercises peer control is the controller and the coworker who is exposed to peer control is the controlee (De Jong,
4 to organizational silence (Morrison & Milliken, 2000). Organizational silence is a
phenomenon which entails withholding opinions and concerns about organizational problems. As a result, this potentially valuable information of opinions and concerns about
organizational problems is lost, since the team norms are preventing employees from sharing their opinions and concerns (Morrison & Milliken, 2000). In the end, this loss of valuable information is a missed opportunity to increase team performance, because sharing these opinions and concerns has the potential to exert a positive effect on team performance (Nemeth, 1997; Morrison & Milliken, 2000). Besides, self-managing teams (SMTs) and teams which partially depend on self-management become increasingly important (Loughry, 2010; Druskat & Wheeler, 2003; Roberson & Williamson, 2012; Lanaj & Hollenbeck, 2015; Erez, LePine, & Elms, 2002). Managers are struggling with decision-making processes and wider spans of control (Alper, Tjosvold, & Law, 1998; Loughry, 2010). Therefore, SMTs have been introduced, which have a decreased direct supervision and an increased importance of interpersonal influence and lateral coordination (Loughry, 2010; Stewart, Courtright, & Barrick, 2012; Barker, 1993), and rely more on peer control (Loughry, 2010). Considering the possible negative effects of peer control for an organization (e.g. organizational silence which will have a potentially negative effect on the team performance), it is important to examine in what ways peer control can have a positive contribution rather than a negative contribution to team performance. In this study, I will focus on team conditions which promote the proposed positive effects of peer control on team performance through the mediator team learning orientation (TLO) and the boundary condition team identification.
Prior research showed mixed results on whether organizational control is beneficial or harmful for teams and organizations. Positive effects of organizational control include
compliance with rules (Tyler & Blader, 2005) and increased motivation to achieve
5 organizational control include resistance of employees to change (Ezzamel & Willmott, 1998) and the suffocation of innovation and flexibility (Bachmann, Gillespie, & Priem, 2015). Specifically, there are also mixed results regarding the effects of peer control within teams or organizations. Examples of positive effects of peer control are increased team learning (Edmondson, 2004) and increased information sharing (Welbourne & Ferrante, 2008), while negative effects of peer control encompass for instance that peers can encourage each other to act against organizational interests (Westphal & Khanna, 2003) and workers might contribute to performance problems because they do not know what behavior is appropriate (Welbourne & Ferrante, 2008). Given these mixed results, Loughry & Tosi (2008) and Loughry (2010) proposed to conduct more research into the specific boundary conditions under which peer control is either beneficial or harmful for teams. Stewart et al. (2012) proposed, moreover, to conduct more research on the mediating processes through which peer control could affect team performance, since little is known about such mediating processes. A few studies have examined the relation between peer control and performance (De Jong et al., 2014; Stewart et al., 2012). Therefore, this study will contribute to existing literature by focusing on the effect of a possible mediator team learning orientation (TLO) on the relationship between peer control and team performance. Furthermore, I examine if the relationship between peer control and team performance and the possible relationship between peer control and TLO will be moderated by team identification.
This study uses survey data of 34 teams with each 4 to 14 subordinates and one supervisor. Loughry (2010) speculated in her study that when peers are observing each other, this may lead to learning, which enhances job performance. In this study, I follow the same idea that TLO will be an important mediator in the relationship between peer control and team performance. Hirst, Van Knippenberg & Zhou (2009) pointed out that when peers are
6 learning activities. Such an environment is characterized by peers who appreciate learning from each other and encourage each other to do so (Rosenthal & Zimmerman, 1978). I expect that when team members are detecting inappropriate behavior and speak up to each other about what behavior is desired, this will be a starting point of a learning process from inappropriate to desired behavior which support attaining of the team goals. This learning process will only be started, if the controlee is open to receive criticism from the controller on his/her inappropriate behavior and is willing to improve his/her behavior to contribute to the team goals.
I expect that if there are high levels of team identification, it is more likely that controlees will see the criticism of the controllers on their inappropriate behavior as an opportunity to increase team performance, since controlees will contribute to a positive team identity and maintain it (Tajfel, 1979; Tajfel & Turner, 1979). When team identification is low, it is more likely that controlees will focus on individual interests, such as protecting their personal identity against criticism of controllers and therefore will reject the perspectives of their controllers on their inappropriate behavior (Elsbach & Bhattacharya, 2001; Cameira & Ribeiro, 2014; Scott, 1997). To conclude, I expect that when team identification is high, peer control will be positively related to team performance, and when team identification is low, peer control will be negatively related to team performance.
7 identification is low, controlees will be more focused on their individual interests rather than collective interests (Brewer & Kramer, 1986; Van der Vegt, Van de Vliert, & Oosterhof, 2003; Pearsall & Venkataramani, 2015), resulting in controlees’ tendency to overlook or reject the perspectives of controllers (Scott, 1997). Therefore, I expect that controlees, if they are addressed by peer control, will not be motivated to improve their behavior in a learning-oriented way and thus will be more likely to ignore the criticism of controllers. In conclusion, when team identification is high, I expect that there will be a positive relationship between peer control and TLO, and when team identification is low, I expect that there will be a negative relationship between peer control and TLO.
If the controlees will be open to receive criticism on their inappropriate behavior, I expect that a proactive learning environment – also referred to as TLO (Bunderson & Sutcliffe, 2003) – will be stimulated, because controlees will be willing to adjust their behavior to the shared team goals in a learning-oriented way. A proactive learning
environment is characterized by a team that is highly adaptable to the environment to operate effectively and focuses on continuous improvement (Edmondson, 1999; Argote, Gruenfeld, & Naquin, 2001; Kozlowski, Gully, Nason, & Smith, 1999). Proactive learning is positively related to team performance (Edmondson, 1999), because it helps to discover new and better ways of achieving team objectives (Bunderson & Sutcliffe, 2003). Thus, I expect that peer control will be related to higher levels of team performance through the mediator TLO.
8
Figure 1. Conceptual model
Theory and Hypotheses Definitions
In this chapter, the concepts included in the model are defined. Moreover, the proposed relations of the model are further explained and supportive research is presented. Peer Control
Peer control is a form of organizational control (Loughry, 2010). It happens when workers are at the same organizational level or are working in the same field and have no formal control over coworkers (Loughry, 2010). Peer control can be seen as an interplay between controllers (the workers who exercise peer control) and controlees (the workers who are exposed to peer control) (De Jong et al., 2014). Loughry & Tosi (2008) used the concept of peer monitoring in their study. In general, peer monitoring involves “setting standards,
observing peers’ results, and sanctioning coworkers if their efforts are below the monitor’s standards” (Loughry & Tosi, 2008, p. 877). It consists of two types: direct and indirect peer
monitoring. Direct peer monitoring happens when controllers detect incorrect behavior of others and address controlees about it (Loughry & Tosi, 2008). Indirect peer monitoring happens when controlees are excluded from the group because of gossiping about and
peer-9
based rational control and peer-based normative control. Peer-based rational control is
described as “a motivational state that occurs when team members perceive their rewards as
being dependent on the input and decisions of their teammates” and peer-based normative
control is described as “a motivational state whereby individuals feel influenced by their
teammates through the pull of social inclusion and internalization” (Stewart et al., 2012, p.
436). Thus, Stewart et al. (2012) suggest the existence of two forms of peer control. In this study, this specific focus on one of the two forms of peer control is not applied. De Jong et al. (2014) conceptualize peer control as a combination of controls (p. 1704). In this study, I use direct peer monitoring of Loughry & Tosi (2008) as definition for peer control, because I measure the extent to which controllers are observing controlees and putting pressure on controlees to change their behavior.
Team Performance
There are several levels on which performance can be measured. For example, performance can be measured at individual level, team level or organization level (Nuhn, Heidenreich, & Wald, 2019). In this study, I focus on team performance, because peer control often happens within a team, since controllers and controlees are at the same organizational level (Loughry, 2010). Team performance can be defined as “the extent to which the
productive output of a team meets or exceeds the performance standards of those who review and/or receive the output” (Hackman, 1987, p. 323). Several indicators can be used to
10 Team Learning Orientation
Edmondson (1999) defines learning behavior as “an ongoing process of reflection and
action, characterized by asking questions, seeking feedback, experimenting, reflecting on results, and discussing errors or unexpected outcomes of actions” (p. 353). Learning
behaviors can also be defined as “activities by which team members seek to acquire, share,
refine, or combine task-relevant knowledge through interaction with one another” (Van der
Vegt & Bunderson, 2005, p. 534). Bunderson & Sutcliffe (2002) define team learning
orientation (TLO) as “a team’s emphasis on search (for new practices, knowledge, skills) and
experimentation in an effort to improve performance” (p. 51). In another study of Bunderson
& Sutcliffe (2003), they define TLO as “a shared perception of team goals related to learning
and competence development; goals that guide the extent, scope, and magnitude of learning behaviors pursued within a team” (p. 553). They pointed out that TLO can be seen as a
11 Team Identification
Broadly, social identification involves “the extent to which team members feel
psychologically intertwined with the group’s fate, sharing its common destiny, and experiencing its successes and failures” (Mael & Ashforth, 1995, p. 310). According to
Pearsall & Venkataramani (2015), team identification is closely related to team cohesion and group pride. Team cohesion is a phenomenon that “attracts and holds team members together
and unifies them” (Pearsall & Venkataramani, 2015, p. 737). Group pride arises when “a person attributes progress or success on a task to the joint efforts of the group” (Delvaux,
Meeussen, & Mesquita, 2016, p. 103). Van der Vegt & Bunderson (2005) define team identification as “the emotional significance that members of a given group attach to their
membership in that group” (p. 533). The definition of Van der Vegt & Bunderson (2005) will
be used in this study, because it emphasizes the importance of emotional involvement of the members (in this study the controllers and controlees) in a given group (in this study the team), which also plays an important role in this study.
The Relationship between Peer Control and Team Performance
12 already performing in accordance with the team goals, will control the agent (controlee), who exercises inappropriate behavior. The principal (controller) will encourage and steer the agent (controlee) towards behavior that is in line with the team goals (Loughry & Tosi, 2008). Contingency theory of control describes the alignment of the organization strategy and organizational processes with the behavior of employees (Kirsch et al., 2002; Karim, Carroll, & Long, 2016). The controllers will control the behavior of controlees by for example setting task outputs and measure task performance (Kirsch et al., 2002). In terms of peer control, it means that an organization has a strategy and the teams will act in accordance with this to contribute to this organization strategy. To make this possible, the team will set a task output and measure and control the behavior of each other upon these task outputs. However, De Jong et al. (2014) criticize the limited applicability of these theories to peer control in teams. For example, both theories fail to define the mediating processes through which they affect performance (De Jong et al., 2014). Specifically, the theories do not describe psychological and social factors that may play an important role in the relationship between peer control and team performance. For example, agency theory has a strong focus on rational economic reasoning and it simplifies assumptions about human behavior (Loughry & Tosi, 2008). This point of criticism is important for this study, because this makes it legitimate to examine a possible mediator through which the relationship between peer control and team performance is mediated.
Besides, empirical research also showed that the relationship between peer control and team performance is rather complex as outlined by De Jong et al. (2014), because there are mixed results towards this relationship. To begin with, prior research has found that
13 1978; Sundaramurthy & Lewis, 2003; Ezzamel & Willmott, 1998; Bachmann et al., 2015). Specifically, there are also mixed results regarding the effects of peer control within teams or organizations. Examples of positive results are increased team learning (Edmondson, 2004), increased information sharing (Welbourne & Ferrante, 2008), fewer behavioral problems among front-line workers (Loughry & Tosi, 2008) and higher quality and quantity
information about performance of workers (Fama & Jensen, 1983; Welbourne & Ferrante, 2008). Since controllers often understand how their controlees perform and do not only focus on objective measurements of the job, there is both objective and subjective information available about performance of workers. Furthermore, prior research showed that peer control can detect opportunities to assist, motivate, or encourage poorly performing coworkers or to compensate for team members’ poor performance (LePine & van Dyne, 2001; Sewell, 1998).
Peer control has also been found to enhance work-unit performance (Loughry & Tosi, 2008; Stewart et al., 2012). In contrast, other studies has found that peer control can stiffen
flexibility (Barker, 1993), that peers can encourage each other to act against organizational interests (Westphal & Khanna, 2003), or that workers might contribute to performance problems because they do not know what behavior is appropriate (Welbourne & Ferrante, 2008).
Since there are mixed results in both fields of control in regard to team performance, I will add the boundary condition team identification to examine if high levels of team
identification will interact to let peer control work out positively in regard to team performance. This will be further explained on pages 15, 16 and 17.
The Mediating Role of Team Learning Orientation
14 2009). In such a social context, peers appreciate learning from each other and encourage each other to do so (Rosenthal & Zimmerman, 1978). This will create a social environment
characterized by proactive learning (Edmondson, 1999), which involves improvement of practices, discovery of new ideas and development of competencies (Bunderson & Sutcliffe, 2002). A crucial finding from the study of Edmondson (2002) is that team learning takes place when there are interactive reflections within the team and actions based on these
reflections. For instance, in the studies of Sutcliffe & Obstfeld (1999) and Edmondson (2004) peer control was positively related to respectively organizational learning and team learning. Team learning breaks down when teams fail to reflect on their own actions (Edmondson, 2002). This is exemplified by in the study of Welbourne & Ferrante (2008), where it was the case that workers did not know which behavior was appropriate. Hence, they could not reflect on inappropriate behavior to subsequently learn appropriate behavior. Peer control is
accompanied by such critical interactive reflections on the behavior of controlees and addressing each other about it (Loughry & Tosi, 2008).
I propose that when peers control and reflect on the behavior of each other, this can be a social factor to develop and stimulate a social context of TLO – a climate of proactive learning (Bunderson & Sutcliffe, 2003). I expect that this will happen if controlees are open for peer control given by their controllers and subsequently are willing to improve their behavior in a learning-oriented way to contribute to the shared team goals. Then, peer control will positively influence TLO. This is in line with evidence Edmondson (2002) has found, because team learning takes place under the conditions that peers successfully criticize each other’s behavior in an interactive way and the controlee is willing to take actions to improve
15 identity or competencies. Then, team members will fail to interactively reflect on the behavior of each other, which will lead to a breakdown of team learning (Edmondson, 2002).
Therefore, I expect that team identification will be an important boundary condition in this relationship, which is further explained on pages 17 and 18. A climate of proactive learning (TLO) is recognizable by a team that is highly adaptable to the environment in order to operate effectively and focuses on continuous improvement (Edmondson, 1999; Argote et al., 2001; Kozlowski et al., 1999). Edmondson (1999) found that proactive learning is associated with higher team performance, because the team members are constantly searching for new and better ways of achieving team objectives (Bunderson & Sutcliffe, 2003). A large body of empirical research also showed that TLO is positively related to team performance (Van der Vegt & Bunderson, 2005; Edmondson, 1999; Schippers, Homan, & Van Knippenberg, 2013). According to this reasoning, I expect that the relationship between peer control and team performance is mediated through TLO.
Hypothesis 1: The relationship between peer control and team performance is mediated by team learning orientation.
The Moderating Role of Team Identification
16 (Van Knippenberg, 2000; Ellemers, De Gelder, & Haslam, 2004; Ouwekerk, De Gilder, & De Vries, 2000).A high team identification reflects affective commitment to the team as an entity, their sense of unity, and their desire to see the team succeed as a whole (Pearsall & Venkataramani, 2015)
Considering SIT in the perspectives of both controlees and controllers, I expect that under high levels of team identification controlees see peer control of controllers on their inappropriate behavior as an opportunity to improve team performance. They will be more willing to adjust their behavior towards the team goals, because if controlees contribute to a good team performance, this will create a positive team identity which will positively affect the self-esteem of the controlee (Tajfel, 1979; Tajfel & Turner, 1979). From controllers’ perspective, I expect that high team identification will act as a motivational incentive to put more effort into helping controlees’ understanding of their inappropriate behavior and steer them towards behavior that is in accordance with the team goals. Prior studies indeed showed that peer control can detect opportunities to assist, motivate, or encourage poorly performing coworkers or to compensate for team members’ poor performance (LePine & van Dyne,
2001; Sewell, 1998). In this way, I expect that from both perspectives high team identification will play a positive interacting role in the relationship between peer control and team
performance.
In contrast, when team identification is low, team members will be more focused on their own interests. When peer control is exercised, I expect that controlees will be more likely to interpret it as an attack on their individual competencies and performance. Hence, they will be likely to protect their personal identity against criticism of controllers (Elsbach & Bhattacharya, 2001; Cameira & Ribeiro, 2014), resulting in the disapproval of the
17 ignored by controlees (Scott, 1997) and are afraid that this will negatively influence their relationships with controlees (Morrison, Milliken & Hewlin, 2003). This will facilitate organizational silence, which will have a potentially negative influence on team performance (Nemeth, 1997; Morrison & Milliken, 2000). Therefore, I expect that low levels of team identification will play an interacting role with peer control in such a way that peer control will be negatively related to team performance.
Hypothesis 2: Peer control and team identification interact to influence team performance, in such a way that…
Peer control is positively related to team performance if team identification is high. Peer control is negatively related to team performance if team identification is low.
Based on the same reasoning, I also expect that team identification will also play a moderating role in the relationship between peer control and team learning orientation (TLO). Following from SIT, when team identification is high, I expect that this will work as an important motivational incentive to meet the named conditions on page 14 (controlees are open for peer control by their controllers and are willing to improve their behavior in a learning-oriented way to contribute to the team goals). First of all, on the controllers’ side, high team identification will serve as a motivational incentive to put extra effort into the controlees’ understanding about which behavior is inappropriate and which behavior is
desired (Haslam & Ellemers, 2005; Barker & Tompkins, 1994). On the controlees’ side, high team identification will also serve as a motivational incentive in the way that they are more willing to improve their behavior in a learning-oriented way towards the shared team goals due to higher levels of motivation, effort and commitment to their team members and goals (Rosendaal & Bijlsma-Frankema, 2015; Ellemers et al., 2004; Van Der Vegt & Bunderson).
18 Vliert, & Oosterhof, 2003; Pearsall & Venkataramani, 2015), because team members see each other as separate workers (Van der Vegt & Bunderson, 2005). Therefore, controlees will see peer control from their controllers as criticism on their individual competencies and
performance, resulting in a higher tendency to overlook or reject controllers’ opinions (Scott, 1997). Thus, I expect that controlees will be less motivated to improve their behavior in a learning-oriented way and will be more likely to reject or ignore controllers’ criticism. In this way, peer control will be negatively related to TLO due to low levels of team identification.
Hypothesis 3: Peer control and team identification interact to influence team learning orientation, in such a way that…
Peer control is positively related to team learning orientation if team identification is
high.
Peer control is negatively related to team learning orientation if team identification is
low.
Overall, I suggest that there is a moderated mediation effect between peer control and team performance which is mediated through team learning orientation and moderated by team identification (see figure 1).
Hypothesis 4: The direct relationship between peer control and team performance as mediated by team learning orientation is stronger when team identification is high rather than low.
Method Participants and Procedures
As four students of the MSc-programme Human Resource Management, we conducted a two-wave online research among employees and an online survey-research among their supervisors. The survey-research setting involved 55 teams in Dutch
19 these organizations received an email which included a link to the online questionnaire. The questionnaires for the employees contained the exact same questions about several team subjects, including the variables of my model peer control, team learning orientation (TLO) and team identification (see appendix p. 55/56) and demographic questions. The questionnaire for the supervisors contained questions about the grades they would give for the individual performance of the team members in terms of quantity and quality of work and demographic questions about themselves. Since the analyses of this study were at team level, a cutoff score of 60% response was used to ensure that the teams were sufficiently represented. This meant that if 60% or more respondents in a team answered the questionnaire, this team was included in the final dataset.
The term of the first employee questionnaire ran from February 14, 2020 to March 17, 2020. 302 employees had responded to the first questionnaire, which resulted in a response rate of 72.42%. 14 respondents did not fill in any answer, 16 respondents did not fully answer the questionnaire and 81 employees did not reach the cutoff score of 60%. These employees were excluded from the dataset. To sum up, 191 employees who worked in one of the 34 teams remained.
20 The term of the questionnaire for the supervisors ran from April 20, 2020 to May 1, 2020. 45 supervisors responded to the questionnaire, which resulted in a response rate of 86.54%. Specifically, out of the 34 teams which remained in the dataset, 33 of the 34 supervisors had responded to the questionnaire, which resulted in a response rate of 97.06% (Mage = 43.79, SD = 10.502; 51.50% female).
To conclude, the final dataset consisted of 33 supervisors and 191 employees who belonged to one of the 34 teams. The team sizes of the employees who were working in one of the 34 teams ranged from 4 to 14 employees and the average team size was 6.71 (N = 228,
SD = 2.680). Female respondents represented 51.30% of the total sample, male respondents
represented 48.20% and 0.50% indicated different than female or male (SD = .511). One employee, who filled in to be of the age of 100 years, was excluded when determining the range of age. After excluding this, the age ranged from 18 to 64 years (M = 42.03, SD = 12.242).
Measures
Peer control. Peer control was measured by a single-item network question on a
seven-point Likert scale (1 = “strongly disagree”; 7 = “strongly agree”), developed by Giulio Ockels (ex-PhD student University of Groningen), based on the scale of De Jong et al. (2014). Participants filled in a score in a table which included the names of their fellow team
members. The item was: “If this team member is doing something I perceive as not right, I
will address him/her about it”.
Team performance. Team performance was measured by two items on an
eleven-point Likert scale (1 = “very poor”; 11 = “very well”) of De Jong et al. (2014). The supervisor rated each team member. The two items were indications of: “The amount of work provided
21
Team learning orientation. TLO was measured by Bunderson & Sutcliffe’s (2002)
four items on a seven-point Likert scale (1 = “strongly disagree”; 7 = “strongly agree”), which was adapted from VandeWalle’s (1997) team-level scale. Example items were: “Our team
looks for opportunities to develop new skills and knowledge” and “Our team likes challenging and difficult assignments that teach new things” (α = .858).
Team identification. Team identification was measured by four items on a seven-point
Likert scale (1 = “strongly disagree”; 7 = “strongly agree”). This scale was adapted from Allen & Meyer’s (1990) scale, which was also used in the study of Van der Vegt et al. (2003)
study. Sample items of the four-item scale included: “I strongly identify myself with other
team members” and “I would like to continue working with my team” (α = .869).
Control variables. The age of workers was taken into account as a control variable,
because prior research demonstrated that supervisory ratings show small declines in workers’ performance with increasing age of the workers. This was possibly explained by the tendency experienced by supervisors to bias their appraisals, resulting in lower ratings for older
workers (Waldman & Avolio, 1986). Moreover, team size was included as a control variable, because prior research showed that this could influence performance appraisals given by supervisors (De Jong et al., 2014). This study showed that supervisors were more likely to give higher ratings of team performance when the team size was small. No clear explanation for this effect of team size on team performance had been given (De Jong et al., 2014). However, this could be explained by the fact that a supervisor had more possibilities to compare individual performances of team members within a larger team. If the team size of the sample was relatively small, there were less possibilities for comparison between individual performances, which will likely result in biased performance appraisals of
22 relatively small mean and high variance. Following from this study, this bias of team size on performance appraisals given by supervisorscould be expected.
Results Data Analysis
First, for the peer control question, a network matrix was created. The column and row names were in the same order to ensure that it was an organized matrix. In this way, self-ratings could be taken out. If there was one missing value, the score of the symmetric position was taken for that place, because the assumption was made that there is a high change of reciprocity in peer ratings between two persons (Huisman & Krause, 2017; Huisman, 2009). If there was an entire row missing, the mean of that specific column was taken to solve that problem, and fill in the mean of that column for the missing value (Huisman & Krause, 2017; Huisman, 2009). These two decisions were taken in the light to increase the statistical power of the analyses. After that, the sum of the upper half of the matrix was taken and divided by the denominator. For this sort of network questions, the following formula was used as
denominator: (Robins, 2015). To sum up, the following formula was used to calculate
the team level score of peer control: .
23 performance and the multicollinearity between TLO, team identification and team
performance.
Third, hypotheses 1, 2, 3 and 4 were tested by using the Hayes’ approach. Prior to the use of Process model 4 to test for the mediation effect of TLO (hypothesis 1), it was examined if the conditions of mediation were met, as outlined by Baron & Kenny (1986). One of the advantages of the Hayes’ approach compared to Baron & Kenny’s approach is that regression
analyses could be done based on a non-normal distribution, because the method of
bootstrapping corrects for biased distributions (Hayes, 2009; Shrout & Bolger, 2002). Until recently, Baron & Kenny’s approach was the optimal method to conduct mediation analysis. In this way, it could be tested whether the same results for mediation appear by conducting the method of Baron and Kenny and the method of Hayes. Baron and Kenny (1986)
24 and TLO and the relationship between peer control and team performance. This analysis also showed whether there is a total moderated-mediation effect.
Factor Analysis
The structure of factorability for the variables TLO, team identification and team performance were checked (table 1). A principal component analysis was used with an eigenvalue greater than 1, with Kaiser’s Varimax rotation. I used .40 as the criterion value difference (Stevens, 1992). The factor analysis produced three factors, with for each variable loadings between .787 and .951, which were appropriate factor loadings. There were no cross-loadings. To conclude, all items were maintained.
Table 1. Factor analysis. Factor loadings based on forced principle with Varimax rotation for 10 items.
Team learning orientation (TLO) Team identification (TI) Team performance (TP) TLO 1 .787 TLO 2 .827 TLO 3 .851 TLO 4 .793 TI 1 .837 TI 2 .817 TI 3 .885 TI 4 .790 TP 1 .944 TP 2 .951
25 Normal Distribution and Outliers
It was tested whether the data reflects a normal distribution of team performance. This was done by the use of the Shapiro-Wilk test (table 3), because the Shapiro-Wilk test is the most powerful significance test to check for normality compared to other tests, such as the Kolmogorov-Smirnov test (Razali & Wah, 2011). The descriptive statistics of the distribution of team performance are demonstrated in table 2.
Table 2. Degrees of skewness and kurtosis.
Statistic SE
Degree of skewness -.015 .180
Degree of kurtosis -.393 .358
Note: N = 182. Dependent variable: team performance.
The p-value of the test should be higher than .05 to meet the condition of normality. Considering the significance of the Shapiro-Wilk test, the condition of normality was not met, which could be a problem to perform regression analyses. However, Hayes’ approach
provided the method of bootstrapping, which did not require a normal distribution to perform regression analyses (Efron & Tibshirani, 1993). Bootstrapping produces an estimation of the sampling distribution, then calculates the confidence intervals and then runs the requested samples over the estimated sampling distribution (Efron & Tibshirani, 1993). In this way, regression analyses can still be performed.
Table 3. Shapiro-Wilk test.
M team performance Statistic Degrees of freedom Significance
Shapiro-Wilk test .969 182 .000
26 The presence of possible outliers was also taken into consideration. Since there was a non-normal distribution of team performance, the graphical representations of the expected normal were critically investigated. The common cutoff score for outliers is usually a difference of -3.0 and +3.0 standard deviations from the mean (Ilyas & Chu, 2019).
Considering the non-normality of the team performance scores, this common cutoff score was narrowed down to -2.5 or +2.5 standard deviations from the mean (Rousseeauw & Hubert, 2011). However, the biggest difference of a team performance score below the mean was approximately -.035 standard deviation and above the mean was approximately +.025 standard deviation. Furthermore, the box plot of team performance also did not indicate a score that is classified as an outlier. To conclude, it had been assumed that there is no outlier in the score of team performance. This meant that no score had to be excluded.
Multicollinearity
The variance inflation factors (VIF) were calculated to check whether these factors were under 4.0 and the tolerance is above .2, because otherwise there could be problems with multicollinearity (Hair, Black, Babin, & Anderson, 2010). The outcomes showed that there were no problems with multicollinearity (table 4).
Table 4. Collinearity analysis.
Tolerance VIF
Peer control .978 1.022
Team learning orientation .818 1.223
Team identification .805 1.242
Descriptive Statistics
Table 5. Descriptive statistics and intercorrelations
Variable M SD 1 2 3 4 5 6
1. Age of employees 42.026 12.242 1
2. Team size 6.706 2.680 .188** 1
3. Peer control 5.877 .531 .139+ -.312*** 1
4. Team learning orientation 5.456 .959 .099 -.045 -.018 1
5. Team identification 5.462 1.073 -.013 -.256*** .121+ .435*** 1
6. Team performance 8.921 .703 -.047 .130+ -.284*** .001 -.081 1
Notes. N = 182-191, team size M and SD are calculated over the whole team (N = 228). +p < .10, *p < .05, **p < .01, ***p < .001 (two-tailed). Rounded to max. 3 decimal places. Pearson correlation coefficients are presented.
Peer control: 1 = totally disagree, 7 = totally agree.
In table 5, the means, standard deviations and correlations of the variables in my model are presented. There was a negative correlation between peer control and team performance, which was highly significant (r = -.284, p < .001). Furthermore, peer control showed a marginally positive correlation with team identification (r = .121, p < .10).
Moreover, there was a positive relationship between TLO and team identification (r = .435, p < .001).
Besides, possible correlations between team performance and the control variables age of employees and team size were examined. Age of employees correlated positively with team size (r = .188, p < .01). Age of employees also marginally correlated with peer control (r = .139, p < .10). Team size correlated negatively with peer control (r = -.312, p < .001) and team identification (r = -.256, p < .001) and had a marginally positive correlation with team performance (r = .130, p < .10). The control variables that (marginally) significantly
correlated with team performance were taken into account for further analyses. To conclude, team size is taken into consideration in further analyses, because team size showed a
Preliminary Analysis
Table 6. Simple and multiple regression analyses results. Regression I (DV: team
performance)
Regression II (DV: team learning orientation)
Regression III (DV: team performance) Regression IV (DV: team performance) B SE t p B SE t p B SE t p B SE t p Team size .045 .024 .602 .548 -.056 .025 -.729 .467 .130+ .024 1.750 .082 .027 .024 .598 .550 Peer control -.270*** .075 -3.580 .000 -.036 .077 -.467 .641 - - - - -.270*** .076 -3.569 .000 Team learning orientation - - - .007 .074 .093 .926 -.002 .072 -.026 .980 R2 .082 .003 .017 .082 Adj. R2 .072 -.007 .006 .067
In the preliminary analysis (table 6), it is shown how the regression coefficients were related to the dependent variables and whether the conditions of mediation were met,
mentioned by Baron and Kenny (1986). In the first regression model, peer control was significantly related to team performance (B = -.270, SE = .075, p < .001), which meant that the first requirement of mediation was met (Baron & Kenny, 1986). However, except for a marginally significant result of team size on team performance in the third model (B = .130,
SE = .024, p < .10), there were no significant results in the other models, which meant that no
mediation through TLO could be expected (Baron & Kenny, 1986). Hypotheses Testing
I tested the mediation effect of TLO using model 4 of Hayes first, before the moderated mediation results were presented. In table 7, the regression results for the mediation effect of TLO are presented. These results provided an answer to the first hypothesis: team learning orientation mediates the relationship between peer control and
team performance. On the next page, the conclusion on the first hypothesis is mentioned and
31
Table 7. Regression results for the mediation effect (Hayes model 4).
Mediator variable model DV = team learning orientation
Predictor B SE t p LLCI ULCI
Team size -.037 .029 -1.276 .204 -.093 .020
Peer control -.059 .079 -.745 .458 -.215 .097
Dependent variable model DV = team performance
Predictor B SE t p LLCI ULCI
Team size .009 .028 .341 .734 -.032 .045
Peer control -.268** .076 -3.531 .001 -.293 -.083 Team learning
orientation
-.003 .072 -.037 .971 -.101 .097
Total effect model DV = team performance
B SE t p
Direct effect of IV on DV
-.268** .076 -3.531 .001
Indirect effects (mediation model)
Effect Boot SE BootLLCI BootUCLI IV > M > DV .000 .007 -.017 .014
Note: N = 182. +p < .10, *p < .05, **p < .01, ***p < .001. Mediator variable model R2 = .010; Dependent variable model R2 = .081**. Rounded to max. 3 decimal places. Standardized regression coefficients are presented.
32
Figure 2. Graphical representation of the mediation effect.
In table 8 on the next page, the regression results for moderated mediation are
presented. Conclusions on the hypotheses two, three and four can be drawn from these results. These conclusions can be found below the table. Below the conclusions of the second and third hypothesis, the graphical representations of the moderation effects are presented. In this way, it can be seen which role team identification has in the relationship between peer control and team performance (hypothesis 2) and in the relationship between peer control and TLO (hypothesis 3).
The second, third and fourth hypotheses were formulated as follows:
Hypothesis 2 stated that peer control and team identification interact to influence team performance, in such a way that…
Peer control is positively related to team performance if team identification is high. Peer control is negatively related to team performance if team identification is low.
33 Peer control is positively related to team learning orientation if team identification is
high.
Peer control is negatively related to team learning orientation if team identification is
low.
Hypothesis 4 stated that the direct relationship between peer control and team performance as mediated by team learning orientation is stronger when team identification is high rather than low.
Table 8. Regression results for the moderated mediation effect (Hayes model 8).
Mediator variable model DV = team learning orientation
Predictor B SE t p LLCI UCLI
Team size .006 .027 .239 .812 -.047 .059
Peer control -.057 .071 -.797 .426 -.197 .084 Team identification .487*** .073 6.672 .000 .348 .631 Peer control * Team identification .168* .065 2.524 .013 .036 .292
Dependent variable model DV = team performance
Predictor B SE t p LLCI UCLI
Team size .005 .020 .268 .789 -.034 .045
Peer control -.190*** .053 -3.596 .000 -.295 -.086 Team learning orientation .035 .056 .619 .537 -.076 .145
Team identification -.075 .061 -1.231 .220 -.194 .045 Peer control * Team identification -.105* .049 -2.135 .034 -.202 -.008
Index of moderated mediation Mediator Index BootSE Boot
LLCI Boot UCLI Team learning orientation .006 .009 -.013 .024
Conditional indirect effects of peer control on team performance at values of the moderator Mediator Team identification Effect BootSE Boot
LLCI Boot UCLI Team learning orientation -.897 low -.007 .012 -.033 .015 Team learning orientation .268 medium .000 .005 -.013 .008 Team learning orientation .967 high .004 .008 -.012 .020
34 Hypothesis 2. The results of this moderation effect can be found in table 8. The results showed that there was a significant interaction effect of team identification on peer control to influence team performance (B = -.105, SE = .049, p < .05). The explained variance of this interaction model on team performance was 10.6% (R2 = .106). To conclude, a negative relationship between peer control and team performance is indeed found, when team identification is low. However, this effect is marginal, as can be seen in figure 3.
Unexpectedly, the graphical representation showed that when team identification is high, there is a negative relationship between peer control and team performance. When peer control is high, team identification even has an opposite effect; team performance is lower when team identification is high than when team identification is low. This is graphically represented in figure 3 below. Thus, the second hypothesis is partially supported.
Figure 3. Graphical representation of the moderation effect on team performance.
Hypothesis 3. The results of this moderation effect can be found in table 8. The results showed that there was a significant interaction effect of team identification on peer control to
1 1,5 2 2,5 3 3,5 4 4,5 5
Low peer control High peer control
35 influence TLO (B = .168, SE = .065, p < .05). The explained variance of this interaction model on TLO was 22.1% (R2 = .221). This effect is represented in figure 4 below. As expected, peer control is positively related to TLO if team identification is high and peer control is negatively related to TLO if team identification is low. Thus, the third hypothesis is supported.
Figure 4. Graphical representation of the moderation effect on team learning orientation.
Hypothesis 4. The results of moderated mediation can be found in table 8. Since the relationship between peer control and team performance was not mediated through TLO, the proposed moderated mediation effect was also expected to be non-significant. Indeed, this was clearly demonstrated by the presence of zero in the confidence interval CI95% [-.013, .024]. At all levels of team identification, there were no significant effects. Therefore, there is no support for the fourth hypothesis.
1 1,5 2 2,5 3 3,5 4 4,5 5
Low peer control High peer control
36 Discussion
This study aimed to give a better understanding of the relationship between peer control and team performance by examining the influence of the potential mediator team learning orientation (TLO). Besides, the boundary condition team identification was added to the model, as prior research showed mixed results towards the relationship between peer control and team performance.
The results of this study showed a negative relationship between peer control and team performance, which is inconsistent with earlier studies including the study of Stewart et al. (2012), which revealed a positive relationship between peer control and team performance. However, the difference between this study and the study of Stewart et al. (2012) is that the latter study found a positive relationship between peer-based rational control and team
performance (see definition p. 9), whereas this study does not focus on such a specific form of peer control, which might imply that different forms of peer control could have different effects on team performance. A possible explanation for the negative relationship between peer control and team performance found in this study is that it depends on the pay structure that is used within the company, which was demonstrated by Welbourne & Ferrante (2008). Specifically, when there was a group-based pay structure (if the whole group is performing well, then the whole group will be rewarded) peer control was related to higher individual performance appraisals of employees given by their managers. In contrast, when companies used an individual-based pay structure (if an individual is performing well, then he/she will be rewarded) peer control was related to lower individual performance appraisals of employees given by their managers (Welbourne & Ferrante, 2008). This might be because within a group-based pay structure, controllers and controlees might be more dependent on each other’s performances in relation to the amount of the rewards. In this way, controlees might
37 peers might perceive peer control as a communication instrument that could help to increase the group rewards (Welbourne & Ferrante, 2008; Loughry, 2010). In an individual-based pay structure, there might be no interdependence between employees in relation to the amount of reward. Therefore, peer control might be perceived by controlees as being overruled by controllers, which might have a negative effect on team performance (Loughry, 2010). Following from this, if the teams used in this sample often have an individual-based pay structure instead of a group-based pay structure, this may be a possible explanation for the negative relationship between peer control and team performance, since team performance is also measured in this study by individual performance appraisals of employees given by their managers. Besides, in the introduction section, it was described that self-managing teams (SMTs) rely on peer control (Loughry, 2010). The finding of the negative relationship between peer control and team performance contributes to the ongoing discussion of the effectiveness of SMTs. There are studies which found beneficial effects of SMTs, such as more favorable work attitudes (Cordery, Mueller, & Smith, 1991) and higher levels of job and group satisfaction (Cohen & Ledford, 1994), but other studies found negative effects, such as higher absenteeism and turnover (Cordery, Mueller, & Smith, 1991), reluctance to monitor each other and lower team performance (Langfred, 2004). To conclude, the negative relationship between peer control and team performance might indicate for SMTs that peer control may contribute to the negative effects.
It was suggested beforehand that this negative effect of peer control on team
performance can be turned into a positive relationship when team identification is high. The second hypothesis suggested that there was a moderation effect of team identification on the relationship between peer control and team performance. Indeed, this moderation effect was found. However, this moderation effect was not exactly working in the way it was
38 to team performance. It might indeed be that controlees are more capable to protect their own identity against criticism of controllers, because there is no social connection between the controlees and controllers, as described in the introduction and theory section (Elsbach & Bhattacharya, 2001; Cameira & Ribeiro, 2014). Surprisingly, high team identification may play a counterintuitive role, since the relationship between peer control and team performance was more strongly negative in the context of high team identification, compared to when team identification was low. A possible explanation for this surprising effect might be that when team identification is high, controlees might have the perception that they already behave appropriately, although this does not correspond with the controllers’ perception. Hence, when controllers address controlees about their inappropriate behavior, they may react more shocked, because controlees interpret the criticism as an attack on their personal identity. This is because of a stronger social connection between controllers and controlees resulting from a high team identification (Elsbach & Bhattacharya, 2001; Cameira & Ribeiro, 2014). This may result in an even higher detrimental effect on the controlees’ performances, since the
controlees’ efforts to contribute to a positive team identity are not recognized by their
controllers (Ashforth, Harrison, & Corley, 2008). To conclude, high team identification may have a moderating effect to let peer control work out negatively on team performance.
The proposed mediation effect of TLO was not supported by the results. In the theory section, two conditions were described (controlees are open for peer control by their
controllers and are willing to improve their behavior in a learning-oriented way to contribute to the team goals) to let peer control work out positively on TLO. A possible explanation why there is no mediation through TLO is that at least one of these conditions is not met, which is due to perceptions of controlees. As outlined in the theory section, this may be because
39 breakdown of team learning (Edmondson, 2002). An alternative explanation, which is in the field of controllers’ perceptions, could be that controllers speak up about the behavior of controlees, but that controllers may not specifically refer to team goals when they provide criticism. If they only refer to shared beliefs about task behavior to improve controlees’ behavior, then controlees may not be encouraged by controllers to turn the criticism into a learning process (Stewart et al., 2012).
To examine the relationship between peer control and TLO, a boundary condition in the form of team identification was introduced. It was hypothesized that team identification plays an interacting role in the relationship between peer control and TLO. Indeed, team identification was found to have this moderating effect on the relationship between peer control and TLO. Team identification may therefore serve as a motivational incentive for controllers and controlees, as described in the theory section. Following from this, it is more likely that the two aforementioned conditions (see introduction and theory section) may be met and that controllers may be more open for the controlees’ learning process from inappropriate to more desired behavior, in order to contribute to the shared team goals.
As mentioned before, the proposed mediation effect was not found. Surprisingly, the relationship between TLO and team performance was also not supported, although prior research showed a positive significant relationship between TLO and team performance (Van der Vegt & Bunderson, 2005; Edmondson, 1999; Schippers et al., 2013). A possible
40 explanation for this insignificant result could be found in the field of the methodological differences between this study and other studies. For example, Van der Vegt & Bunderson (2005) used a different scale with different items to measure TLO. The items for the TLO scale they used were developed with an eye on team performance, which is for example shown by the question: “criticize each other's work in order to improve performance”. The scale used in this study does not contain such a specific link to team performance.
Practical Implications
As described in the introduction section, SMTs often use peer control as a
communication system to communicate and give feedback to each other (Loughry, 2010; Stewart et al., 2012; Barker, 1993). The results of this study suggest that there is a negative relationship between peer control and team performance. This suggests that peer control is perhaps not the optimal method of communication to maximize team performance, because it may negatively influence the team performance. To conclude, this study implies to reconsider peer control as a communication instrument in teams.
Furthermore, the results of the moderated effect that team identification may play in the relationship between peer control and TLO and the relationship between peer control and team performance could be important SMTs or teams which partially depend on
self-management, considering the proposed practical implications described in the introduction section. According to the results of team identification in this study, investing in team building activities to increase team identification could be practically beneficial or
41 view, supervisors and team members perhaps have to avoid investments in team building activities to increase team identification.
Limitations and Future Research
There are several limitations which might influence the results that have been found. Therefore, these limitations will also be addressed. Furthermore, suggestions for future research are provided.
The first limitation was the corona crisis (caused by the COVID-19 pandemic) which might have influenced the results of this study. The respondents from this study had to work from home due to governmental measures (Dutch government, 2020), which might have led to biased responses and the exclusion of the second survey-research, since there was a considerable decline in response rates between the first (response rate: 72.42%) and the second questionnaire (response rate: 47.36%). To conclude, the different working
environment (at home instead of the office) and the exclusion of the second survey-research might have harmed the statistical power of this study.
Furthermore, the analyses have been conducted on the dependent variable team performance, which was non-normal distributed. However, the bootstrapping method corrected for the non-normal distribution, which made it possible to conduct the analyses, which was already explained on page 25. The non-normal distribution could be seen as a limitation.
42 underlying reason could be that they were afraid that the supervisors or team members would see their answers. Some of the respondents answered all the questions, but had answered all the network questions with an eight (“this is me”), which made these data unusable. A possible reason for this is that these respondents were willing to fill in all the questions, except the network questions for the same reasons as mentioned above. Furthermore, the variables peer control, TLO and team identification are self-report scales. Considering the analyses which consisted of two or more of these variables, it is possible that these analyses were biased, because of self-report biases (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). To conclude, it might have harmed the statistical power of this study.
Future research should focus on elaborating the relationship of peer control and team performance. In the theory section, it is mentioned that peer control could be connected with socio-/psychological factors, such as social inclusion, internalization and cohesion (Stewart et al., 2012). Since this study showed that team identification might play a moderating role in this relationship, this may indicate that other socio-/psychological variables also influence the relationship between peer control and team performance. Future research could focus on analyzing other socio-/psychological variables to expand our knowledge about the
relationship between peer control and team performance. A possible direction is to examine the influence of psychological safety in the relationship between peer control and team performance, since prior research showed that psychological safety is related to cohesion within teams (Johnson & Avolio, 2019; Stewart et al., 2012).
43 will not affect team performance when team performance was already high in the past. This effect cannot be excluded by this study, because this study only measured team performance once.
Conclusion
Prior research had started to examine the relationship between peer control and team performance (De Jong et al., 2014; Stewart et al., 2012; Loughry & Tosi, 2008). This study had focused on elaborating the relationship between peer control and team performance. This had been done by analyzing the effects of the possible mediator team learning orientation (TLO) and the possible boundary condition team identification. The results from this study showed a negative relationship between peer control and team performance. In addition, it can be concluded from the results of this study that there was no mediating effect through TLO. Furthermore, team identification moderated the relationship between peer control and team performance and the relationship between peer control and TLO. When team
identification was low, peer control was negatively related to team performance, but
44 References
Allen, N. J., & Meyer, J. P. (1990). The measurement and antecedents of affective,
continuance and normative commitment to the organization. Journal of Occupational
Psychology, 63(1), 1-18.
Alper, S., Tjosvold, D., & Law, K. S. (1998). Interdependence and Controversy in Group Decision Making: Antecedents to Effective Self-Managing Teams. Organizational
Behavior and Human Decision Processes, 74(1), 33-52.
Argote, L., Gruenfeld, D., & Naquin, C. (2001). Group learning in organizations. In M. E. Turner, Groups at work: Advances in theory and research (pp. 369–411). Hillsdale, New Jersey: Erlbaum.
Ashforth, B. E., Harrison, S. H., & Corley, K. G. (2008). Identification in Organizations: An Examination of Four Fundamental Questions. Journal of Management, 34(3), 325-374.
Bachmann, R., Gillespie, N., & Priem, R. (2015). Repairing Trust in Organizations and Institutions: Toward a Conceptual Framework. Organization Studies, 36(9), 1123-1142.
Barker, J. R. (1993). Tightening the Iron Cage: Concertive Control in Self-Managing Teams.
Administrative Science Quarterly, 38(3), 408-437.
Barker, J. R., & Tompkins, P. K. (1994). Identification in the self-managing organization: Characteristics of target and tenure. Human Communication Research, 21(2), 223-240.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of
45 Beazley, I., Dougherty, S., Penn, C., Philips, L., & James, C. (2019). Performance
measurement systems in the health sector and their budgetary implications. OECD
Journal on Budgeting, 19(3), 41-66.
Biemann, T., & Kearney, E. (2010). Size Does Matter: How Varying Group Sizes in a Sample Affect the Most Common Measures of Group Diversity. Organizational Research
Methods, 13(3), 582-599.
Brewer, M. B., & Kramer, R. M. (1986). Choice behavior in social dilemmas: Effects of social identity, group size, and decision framing. Journal of Personality and Social
Psychology, 50(3), 543-549.
Bunderson, J. S., & Sutcliffe, K. M. (2002). Why some teams emphasize learning more than others: Evidence from business unit management teams. In H. Sondak, Research on
Managing Groups and Teams, 4, Toward phenomenology of groups and group
membership: an introduction (pp. 49–84). Bingley, United States: Emerald Publishing
Limited.
Bunderson, J. S., & Sutcliffe, K. M. (2003). Management team learning orientation and business unit performance. Journal of Applied Psychology, 88(3), 552-560.
Cameira, M., & Ribeiro, T. A. (2014). Reactions to intragroup deviance: Does disidentification have a role? Journal of Social Psychology, 154(3), 233-250.
Cohen, S. G., & Ledford, G. J. (1994). The Effectiveness of Self-Managing Teams: A Quasi-Experiment . Human Relations, 47(1), 13-43.
Cordery, J. L., Mueller, W. S., & Smith, L. M. (1991). Attitudinal and behavioral effects of autonomous group working: A longitudinal field study. The Academy of Management
46 Cortina, J. M. (1993). What Is Coefficient Alpha? An Examination of Theory and
Applications. Journal of Applied Psychology, 78(1), 98-104.
Cravens, D. W., Lassk, F. G., Low, G. S., Marshall, G. W., & Moncrief, W. C. (2004). Formal and Informal Management Control Combinations in Sales Organizations: The Impact on Salesperson Consequences. Journal of Business Research, 57(3), 241-248.
De Jong, B. A., Bijlsma-Frankema, K. M., & Cardinal, L. B. (2014). Stronger than the sum of its parts? The performance implications of peer control combinations in teams.
Organization Science, 25(6), 1703-1721.
Delvaux, E., Meeussen, L., & Mesquita, B. (2016). Emotions are not always contagious: Longitudinal spreading of self-pride and group pride in homogeneous and status-differentiated groups. Cognition & Emotion, 30(1), 101-116.
Druskat, V. U., & Wheeler, J. V. (2003). Managing from the boundary: The effective
leadership of self-managing work teams. The Academy of Management Journal, 46(4), 435-457.
Dutch government. (2020, March 12). Information about the measures according to the
coronavirus. Retrieved from the website from the Dutch government:
https://www.government.nl/latest/news/2020/03/12/new-measures-to-stop-spread-of-coronavirus-in-the-netherlands
Edmondson, A. C. (1999). Psychological safety and learning behavior in work teams.
Administrative Science Quarterly, 44(2), 350-383.
47 Edmondson, A. C. (2004). Learning From Mistakes Is Easier Said Than Done: Group and
Organizational Influences on the Detection and Correction of Human Error. Journal of
Applied Behavioral Science, 40(1), 66-90.
Efron, B., & Tibshirani, R. J. (1993). An Introduction to the Bootstrap. Boca Rotan, Florida: Chapman & Hall/CRC.
Eisenhardt, K. M. (1989). Agency Theory: An Assessment and Review. The Academy of
Management Review, 14(1), 57-74.
Ellemers, N., De Gelder, D., & Haslam, A. S. (2004). Motivating Individuals and Groups at Work: A Social Identity Perspective on Leadership and Group Performance. The
Academy of Management Review, 29(3), 459-478.
Elsbach, K. D., & Bhattacharya, C. B. (2001). Defining Who You Are By What You're Not: Organizational Disidentification and The National Rifle Association. Organization
Science, 12(4), 393-413.
Ennew, C. T., Reed, G. V., & Binks, M. R. (1993). Importance-Performance Analysis and the Measurement of Service Quality. European Journal of Marketing, 27(2), 59-70.
Erez, A., LePine, J. A., & Elms, H. (2002). Effects of Rotated Leadership and peer Evaluation on the functioning and effectiveness of self-managed teams: A quasi-experiment.
Personnel Psychology, 55(4), 929-948.
Ezzamel, M., & Willmott, H. (1998). Accounting for Teamwork: A Critical Study of Group-Based Systems of Organizational Control. Administrative Science Quarterly, 43(2), 358-396.
Fama, E. F., & Jensen, M. C. (1983). Agency problems and residual claims. Journal of Law
