June 11st, 2017
TEAM IDENTIFICATION – DOES IT ENABLE THE PREDICTION OF TEAM EFFORT BY TEAM PEER PRESSURE?
JULIA PRÖMPELER Student number: S3223639 Saffierstraat 210, 9743 LN Groningen
j.b.j.prompeler@student.rug.nl
Supervisor: Dr. K. M. Bijlsma-Frankema
Master thesis - MSc, Human Resource Management Faculty of Economics and Business, University of Groningen
Acknowledgements: I thank Katinka Bijlsma-Frankema whose helpful remarks and passion for research constantly challenged me to enhance the quality of my research and writing. I also thank my friend and fellow student, Kathrin Kappen, for her abiding emotional and cognitive support. Additionally, I wish to thank my family, boyfriend and friends for their support and believe in me during all of my studies including the thesis period.
ABSTRACT
Promoting team effort including in-role and extra-role behaviors is a challenge for organizations. Peer controls can possibly contribute to this dilemma. Thus they gain importance but are insufficiently understood. This paper enhances theoretical understanding by examining whether and how team peer pressure, as a form of peer control, predicts team effort. It assesses whether team identification mediates the team peer pressure-team effort relationship. Additionally, moderating roles of team legitimacy of peer pressure and team trust in supervisor are examined. Based on the literature review, team legitimacy of peer pressure is expected to strengthen the positive team peer pressure-team identification relationship and team trust in supervisor is proposed to strengthen the positive team identification-team effort relationship. Testing the full model with Survey data of 220 teams of a Dutch non- profit organization suggests that team identification fully mediates the positive team peer pressure-team effort relationship. Support for the mediating role of legitimacy of peer pressure cannot be found. Team trust in supervisor marginally moderates the team identification-team effort relationship, however, such that the relationship between team identification and team effort is stronger at low levels of team trust in supervisor. Theoretical and practical implications are discussed.
Keywords: team peer pressure, team identification, team effort, trust in supervisor, legitimacy, peer control
INTRODUCTION
Organizational control is ubiquitous and has been acknowledged as integral to the functioning of all organizations for a long time (Blau & Scott, 1962; Cardinal, 2001;
Tannenbaum, 1968). Organizational control is applied “to align employee capabilities, activities, and performance with the organization’s goals and aspirations” (Cardinal, Sitkin, & Long, 2004: 411); thus to ensure that all teams exert effort, which benefits their organization. In this context, it is desirable that team effort, which is “the extent to which team members devote their resources (i.e., energy, attention, time) to executing team tasks” (De Jong & Elfring, 2010: 537) encompasses both in-role and extra-role behaviors. But which form of organizational control can achieve this goal?
Hierarchical organizational controls can monitor, demand and enhance in-role behaviors, which are predictable, routinized behaviors. However, they are not suitable for demanding extra-role behaviors (Henderson & Lee, 1992). This is problematic, as in todays working world tasks are increasingly variable and the prediction of work sequencing is merely impossible (Singh, 2008). Thus, adaptability and extra-role behaviors become central to success. Extra-role behaviors, which go above and beyond required activities and contribute to the effectiveness of an organization (Organ, 1988; Van Dyne, Graham & Dienesch, 1994), are proposed as even more critical for successful performance than in-role behaviors (Van Dyne & LePine, 1998).
Hence, the effectiveness of conventional control systems decreases (Singh, 2008) and decentralized lateral controls are advocated as more effective (Henderson & Lee, 1992; Loughry, 2010). Peer control, a lateral control in which “workers who are at the same organizational level or in the same field exert lateral control over their peers”
(Loughry, 2010: 324), is thus a highly relevant form of control. Surprisingly, besides the increasing relevance and interest in peer control mechanisms (Loughry, 2010), to date research on peer control is relatively sparse (Loughry & Tosi, 2008) and very little is known about how, why and when peer controls impact organizational outcomes (De Jong, Bijlsma-Frankema, & Cardinal, 2014). To enhance theoretical understanding, the focus of this study is one specific form of peer control called peer pressure, which is “openly and honestly acknowledging dissatisfaction when members behave in ways considered unacceptable by the team” (Druskat & Kayes, 2000: 332).
As team peer pressure is as a form of organizational control, it is inherently about linking team capabilities to team effort. Surprisingly, regardless of this and the strong positive connection between a team’s effort and its performance (De Jong & Elfring 2010; Liden, Wayne, Jaworski, & Bennett, 2004), team peer pressure has only recently been assessed in combination with team effort. De Jong et al. (2014) found support for a prediction of team effort through team peer pressure. However, theoretical understanding would greatly benefit from enlarged understanding about whether, how and when team peer pressure promotes team effort.
To assess this, the current study introduces team identification as a mediator to the team peer pressure-team effort relationship. Team identification is “the emotional significance that members of a given group attach to their membership in that group”
(Van der Vegt & Bunderson, 2005: 533). As a strong motivational source, which can provoke extra-role behavior, team identification is a highly relevant mediation variable (Dukerich, Golden, & Shortell, 2002; Tyler & Blader, 2005). If team peer pressure promotes team effort through team identification, team efforts likely encompass both in-role and extra-role behaviors, which as outlined above is desirable for organizations. Research supports the motivational force of team identification by finding that feeling the sense of belonging to a group fosters the urge to re-affirm membership, contribute to reaching group goals (i.e. Ellemers, De Gilder & Haslam, 2004; Shamir, 1990) and predicts organizational citizenship behavior (van Dick, Stellmacher, & Christ, 2004). In the literature, several lines of exist, which seem nicely to apply to how team peer pressure can foster the development of team pressure.
Hence, it is surprising that to date team identification has not been explored in the team peer pressure context. By assessing the mediating role of team identification in the team peer pressure-team effort relationship, I will answer calls for increased theoretical understanding about peer control consequences (De Jong et al., 2014;
Loughry, 2010; Stewart, Courtright, & Barrick, 2012).
To date, findings regarding the effectiveness and acceptance of peer control and are mixed. Some researchers find it promoting performance, effort, focus on important tasks, coordination, feedback culture, and trust (De Jong et al., 2014; Frink & Klimoski, 1998; Loughry & Tosi, 2008; Marks & Panzer, 2004). However, Loughry (2010) highlights the possibility of teams rejecting to exert peer control. Romeike, Fleschhut,
Nienaber and Schewe, (2017) argue that perceptions of legitimacy are the key to acceptance and outcomes of peer control systems. Legitimacy is the “psychological property of an authority, institution, or social arrangement that leads those connected to it to believe that it is appropriate, proper, and just” (Tyler, 2005: 375). Whether teams perceive a control system as legitimate predicts their reactions and attitudes (Romeike et al., 2017). If teams believe in the legitimacy of a control system, they willingly comply with it and perceive high quality- relationships and trust towards the persons exerting the control (Bijlsma-Frankema & Costa, 2010; McNall & Roch, 2009;
Weibel, 2007; Tyler, 2005). Contrary, inappropriately exercised, control will be rejected and negatively affects attributes (Ferrin, Blight, & Kohles, 2007; Sitkin &
Geroge, 2005). Hence, it is astonishing that team legitimacy has not been studied in combination with team peer pressure. I propose team legitimacy of peer pressure as a moderator on the team peer pressure-team identification relationship. Particularly, I propose that high legitimacy of peer control strengthens the team peer pressure-team effort relationship, while low team legitimacy of peer pressure weakens it.
Another problematic aspect about team peer pressure is that it does not necessarily aligning team efforts to organizational goals (Loughry, 2010). Some research highlights that team peer pressure can direct team efforts against organizational goals (e.g. soldiering behavior: see Loughry, 2010). If teams strongly identify as one group, the organization might be perceived as the out-group, which can provoke behavior that benefits the team instead of the company. I argue that team trusts in the supervisor is crucial for the alignment of team efforts to team goals as defined by the organization. Therefore, I introduce team trust in supervisor, which is the “expectation or belief that the team can rely on the leader's actions or words and that the leader has good intentions toward the team” (Dirks, 2000: 1004), as a moderator on the team identification-team effort relationship. In support of this, it has been argued and found that trust in supervisor helps the alignment of team efforts to organizational interests (Dirks, 2000; Dirks & Ferrin, 2002; Mayer & Gavin, 2005). Furthermore, both team identification and team trust in supervisor have been identified as important motivational sources (Burke, Sims, Lazzara, & Salas, 2007; Ellemers, Kortekaas, &
Ouwerkerk, 1999; Konovsky & Pugh, 1994). Due to support for synergetic effects between motivational sources (Amabile, 1993; Gagné & Deci, 2005; Grant, 2008), I
expect the relationship between team identification and team effort to be stronger when team trust in supervisor is high versus to be weaker when trust in supervisor is low.
In summary, this study addresses the following research questions: Does team peer pressure predict team effort and if so, is the relationship mediated by team identification? Does team legitimacy of peer pressure moderate the team peer pressure – team identification relationship? Does team trust in supervisor moderate the team identification – team effort relationship?
THEORETICAL FRAMEWORK
Team Peer Pressure
As outlined above, peer pressure is part of the broader peer control construct. While peer control refers to the general phenomenon of “workers who are at the same organizational level or in the same field exert[ing] lateral control over their peers”
(Loughry, 2010: 324)”, team peer pressure refers to more specific actions. It is the verbal communication of disapproval of behavior that team members perceive as inappropriate in order to influence others’ behavior (Druskat & Kayes, 2000). Peer pressure can be executed in different forms. Examples include the direct demand for appropriate behavior and high effort, open confrontation of team members who do not conform to team expectations in the presence of others as well as attempts to trigger a feeling of guilt about not being a valuable team member and a bad team player (Hackman, 1992; LePine & Van Dyne, 2001). Noteworthy, this study focuses on direct peer pressure; hence the peer pressure exercised by one or more team members towards one or more team members in a direct relation. Indirect peer pressure activities, such as rejecting, avoiding and gossiping about peers (Loughry, 2010) are not examined. Moreover, peer pressure is a team level construct and will be measured as such in this study. Team members engage in team peer pressure activities in collaboration with other team members and correct poorly performing team members on behalf of the whole team (Barker, 1993; Hackman, 1992). The variable also ought to be measured at the team level because the magnitude of team
peer pressure depends on the degree, to which all team members engage in or refrain from peer pressure activities (De Jong et al., 2014).
To enhance understanding of the variable team peer pressure, it will be distinguished from closely related variables. According to De Jong et al. (2014) an important distinct mechanism from team peer pressure that constitutes peer control are team norms.
Team norms are standards of appropriate behavior. They are a prevailing ambient stimulus to which all team members are constantly exposed. Even though team peer pressure and team norms mostly co-occur within teams, their application and emphasis are distinct (De Jong et al., 2014). Thus, this paper treats them as distinct constructs and only focuses on team peer pressure. Moreover, team peer pressure includes and presumes a different form of control, namely team monitoring. In order to be able to actively address undesired behavior, undesired behavior first has to be detected (Jaworski, 1988). Thus, team peer pressure cannot occur without team monitoring, however team monitoring is not the focus of this study. Having outlined the independent variable of this study, I will delineate the dependent variable, before proposing their relationship.
Team Effort
Team effort is defined as “the extent to which team members devote their resources (i.e., energy, attention, time) to executing team tasks” (De Jong & Elfring, 2010: 537 citing Yeo & Neal, 2004). Team effort constitutes in-role behavior, thus behavior, which meets the organizational requirements, and partly extra-role behavior (Organ, 1988; Van Dyne & LePine, 1998). Extra-role behavior is “behavior which benefits the organization and/or is intended to benefit the organization, which is discretionary and which goes beyond the existing role expectations” (Van Dyne, Cummings, & Parks, 1995: 218). A team, which refrains from extra-role behavior, does not necessarily compromise on their formal duties and responsibilities (Tavares, van Knippenberg, van Dick, 2016). Thus, in contrast to in-role behavior and in-role effort, extra-role behavior is difficult to demand and enforce because it is neither specified in role descriptions, nor recognized by formal reward systems. Thus it is no source of punitive consequences if failed to perform (Van Dyne & LePine, 1998; Organ, 1988). In this study team effort is understood as a combination of in-role and extra-role-
behavior. If directed at organizational goals, team effort is a highly important variable for organizational success and a key determinant for team performance (De Jong &
Elfring, 2010; Liden et al., 2004).
Team Peer Pressure and Team Effort
Several studies find support that peer control can be suitable to regulate effort trough peer control combinations such as peer pressure, monitoring, rational and normative controls (Barron & Gjerdje, 1997; De Jong et al., 2014; Kandel & Lazear, 1992; Loughry
& Tosi, 2008; Stewart et al., 2012). De Jong et al. (2014) argue that peer pressure increases team effort because by enhancing the explicitness and thereby the salience of team expectations. Thus, team peer pressure activities are argued to enhance the psychological pressure of team members to perform expected behavior. This is in line with the finding that norms within a team affect team members through the perceived obligation to behave congruent to team norms (Jasso & Opp, 1997). Every team has certain norms, however only if they are actively upheld, they keep their strength and affect team effort. Team peer pressure induces and strengthens the norm that investing effort is common and expected.
Additionally, peer pressure sends clear signals about the consequentiality of team members’ actions. Being aware of negative consequences for exerting low effort increases the motivation to work productively and reduces the likelihood to rely on one’s own discretion (Cooper & Withey, 2009). Furthermore, team peer pressure is only possible through prior peer monitoring and peer monitoring by itself has been argued to increase effort by creating accountability for enhanced effort and motivating team members to focus on monitored tasks (Frink & Klimoski, 1998). For all above reasons, I expect that team peer pressure promotes team effort.
H1: Team peer pressure promotes team effort.
Having outlined prior support for a positive, direct relationship between team peer pressure and team effort, I introduce the variable team identification. Team identification has been linked to team effort; however, to date it has not been connected to team peer pressure. I will highlight several mechanisms through which
team peer pressure can foster team identification thereby arguing for a mediating role of team identification in the team peer pressure-team effort relationship.
Team Identification
Social identification is the psychological cognitive and affective link between a person and a social entity. It has a deeply self-defining function for every individual (Edwards
& Peccei, 2007). The group with which persons identify gives them a sense of belonging (Tajfel, 1978), reduces uncertainty about oneself and satisfies the basic human need for safety, self-esteem and pride (Hogg & Terry, 2000). To differing extents, the organizations and teams to which people belong, define a part of their identity. These social identities are sometimes more important to a person than his or her ascribed identity based on age, gender, nationality for example (Hogg & Terry, 2000). For this study, team identification is defined as “the emotional significance that members of a given group attach to their membership in that group” (Van der Vegt &
Bunderson, 2005: 533)
Team identification is a multidimensional construct with three empirically distinct components. The cognitive component describes the cognitive awareness of one’s membership of a certain group and allows individuals to categorize themselves. The evaluative component determines group self-esteem by evaluating negative and/or positive connotations, which are linked to the group membership. Thirdly, the emotional component defines affective commitment, as it is an emotional involvement with the group (Tajfel, 1978; Van der Vegt & Bunderson, 2005). The emotional component of team identification most strongly drives motivation and can foster team members’ readiness to engage in interactions, yield benevolent behavior on behalf of team members and can foster extra-role behavior, altruistic acts and sportsmanship (Bergami & Bagozzi, 2000). Thus, I regard the emotional aspect of team identification as most important in this study, and will argue for its effects in the following sections.
Team Identification and Team Effort
In line with the key proposal of social identity theory (i.e. Tajfel, 1978, Tajfel & Turner, 1979), I hypothesize that team identification promotes team effort. Social identity theory suggests that the degree to which people identify with a certain social group
determines their desire and capability to behave congruent to the group membership, thus to conform to group norms and expectations. According to social identity approach (i.e. Tajfel & Turner, 1979), strong team identification lets team members perceive team goals as their own. Through team identification, personal motivations are projected on, informed by and adapted to team goals, needs and expectations.
Thus, team identification arouses internal desire to conform to team expectations (Ellemers et al., 2004). In this context, the emotional component of team identification triggers the internalization of team goals and thereby fosters effort.
In line with social identity theory (Tajfel & Turner, 1986) social identities are very important for an individuals definition of the personal-identity. Identifying with a successful team is thus beneficial for an individual’s personal identity. Ultimately people strive for a positive personal identity (i.e. Hogg & Abrams, 1988; Hogg & Mulin, 1999). Hence, team members who identify with their team will be intrinsically interested in the success of the team. In order to contribute to team success, it is likely that team members exert effort. In support of this, it was found that team identification enhances the willingness to engage in or disengage from interactions (Bergami & Bagozzi, 2000), the willingness to perform activities, which benefit the team (Dutton, Dukerich, & Harquail, 1994) and is linked to organizational citizenship behavior and extra-role behavior (Organ & Ryan, 1995; van Dick et al., 2004). This supports the above-described idea that the emotional aspects teams ascribe to their belongingness in the group have strong motivational power and influence behavior.
On another line of arguing, team identification promotes team members’ desire to engage in team tasks because this expresses and re-affirms their belonging to the team and strengthens their identification with a team (Shamir, 1990). Behaving conform to team expectations, allows team members to be more certain about their belonging to the team. Hence, the urge to strengthen the cognitive component of team identification promotes team effort. Based on all above outlined reasons, I expect that team identification promotes team effort.
H2: Team identification promotes team effort.
Team Peer Pressure and Team Identification
To date no study has examined the direct effect of team peer pressure on team identification. However, several lines of argumentation exist in the literature, which seem nicely to apply to how team peer pressure may foster team identification.
According to Huettermann et al. (2017), team identification occurs when team members feel a sense of belonging and psychological attachment to their team, categorize themselves as a part of the team and describe their team with the same features as they describe themselves. The latter point is important in understanding the proposed relationship, as it highlights the relevance of sharing norms, attitudes and values for the development of team identification. A collective team identity occurs when specific features of a team are shared among most team members (Albert & Whetten, 1985). I argue that team peer pressure will help creating those shared perceptions by creating strong situations. It has been acknowledged that general control mechanisms can lead to strong situations (Chatman, 1989) and also argued that specifically peer control forms, including peer pressure, promote strong situations (Cooper & Whitey, 2009, De Jong et al., 2014). Situational strength theory describes that situations are strong “to the degree that they lead all persons to construe events the same way, induce uniform expectancies regarding the most appropriate response pattern, provide adequate incentives for the performance of that response pattern, and instill the skills necessary for its satisfactory construction and execution.”
(Mischel, 1973: 276). Accordingly, team members perceive and behave more similar in a strong situation (De Jong et al., 2014). Peer control that creates strong situations results in collectively shared norms, attitudes and similar behavior (Meyer, Dalal, &
Herminda, 2010). If peer pressure leads to situations in which team members have a feeling of unity, a necessary condition for developing team identification is met through team peer pressure.
In addition, team peer pressure can promote team identification along a different line.
Huettermann et al. (2017) argue that four consecutive processes are relevant for the emergence of team identification. First, team members enact at least one social identity depending on situational factors including the salience of the team as a social category. I argue that team peer pressure, increases the salience of the team as a social category. Second is the process of sensemaking about the experiences within the team
(Huettermann et al., 2017; Brown, Stacey, & Nandhakumar, 2008). This includes an individual-directed sensemaking episode during which team members evaluate their own situation in the team (Huettermann et al., 2017). I argue that team peer pressure facilitates this process because team peer pressure provides clear cues about team members’ stance in the team and expected behavior. In the collective-directed sensemaking phase, which follows the individual sensemaking phase, team members elaborate their understanding of the whole team (Huettermann et al., 2017). Peer control can facilitate this process because it provides clarity about requirements and responsibilities of the whole team (Bowen & Ostroff, 2004; Meyer et al., 2010). I do not expect the third and fourth processes (evaluation of team outcomes and converging identities) to be influenced by team peer pressure.
In accordance with social categorization theory (Hogg & Abrams, 1988) and social identity theory (Tajfel, 1978) team members identify with teams to satisfy the basic human need for uncertainty reduction, belonging and positive self-esteem (Hogg &
Mullin, 1999, Hogg & Terry, 2000). As team peer pressure reduces ambiguity and highlights the expectations and membership of the team (De Jong et al., 2014), I expect that direct team peer pressure promotes team identification.
Moreover, when team peer pressure is encouraged, all team members share responsibility in monitoring and addressing undesired behavior in order to reach team goals and show high performance. Thus, team members are responsible for each other and interdependent ties emerge. Research on social dilemmas implies that interdependence regarding outcomes will evoke team identification (Brewer &
Kramer, 1986; Chen, 1996; Kramer & Brewer, 1984). Moreover interdependence has been advocated as a strong basis for in-group formation (Flippen, Hornstein, Siegal, &
Weitzmann, 1996), a crucial basis for intragroup attraction and long-term-liking (Duck, 1977; Sherif, 1969), which have all been considered necessary backdrops for team identification (Henry, Arrow, Carini, 1999). Based on all above outlined lines of argumentation, I expect team peer pressure to positively influence the development of team identification.
H3: Team peer pressure promotes team identification.
Team Identification as a Mediator
Based on the argumentations for direct relations between team peer pressure and team identification as well as team identification and team effort, I proceed by connecting those relationships and arguing that team identification helps explain the team peer pressure-team effort relationship. As delineated above, clearly communicating team expectations and sending signals to all team members about the consequences of their actions has been linked to increase psychological pressure to exert effort (Cooper & Withey, 2009; De Jong et al., 2014). Thus, the independent variable team peer pressure is expected to directly promote the dependent variable team effort.
In line with self-determination theory, both intrinsic and extrinsic motivators mostly influence behavior (i.e. Ryan & Deci, 2000). As outlined above, prior research has focused on individual considerations and extrinsic reasons why team peer pressure predicts team effort. I propose that team identification as an intrinsic motivator, is also an underlying mechanism, which explains the prediction of team effort by team peer pressure. I argue that team peer pressure enhances team identification because teams are partially responsible for their performance and behavior, which may trigger feelings of relatedness. Also, through the direct feedback, which all team members provide each other, team peer pressure construes strong situations, which are perceived similar by all team members (De Jong et al., 2014). Shared perceptions and goals foster team identification (Huettermann et al., 2017). Furthermore, team peer pressure decreases uncertainty and the need for certainty is a reason why teams identify as one unit (Huettermann et al., 2017). Based on the above reasons, I expect team peer pressure to strengthen team identification. Team identification predicts team effort because team members are intrinsically motivated to reach team goals because personal motivation is projected on, informed by and adapted to team goals, needs and expectations (Ellemers et al., 2004). Additionally, the wish to re-affirm team membership by behaving congruent to team expectations increases effort (Shamir, 1990). Hence, I propose that team identification as an intrinsic motivator facilitates a better understanding of the relationship between team peer pressure and team effort.
H4: Team identification mediates the relationship between team peer pressure and team effort.
For further investigation of the mediating role of team identification on the team peer pressure-team effort relationship, the relationship between team peer pressure and team identification will be closely examined. Based on prior mixed findings regarding outcomes of team peer pressure, I propose team legitimacy of peer pressure as a moderator on the team peer pressure-team identification relationship. Therefore, first team legitimacy of peer pressure is introduced.
Team Legitimacy of Peer Pressure
Legitimacy is the “psychological property of an authority, institution, or social arrangement that leads those connected to it to believe that it is appropriate, proper, and just.” (Tyler, 2005: 375). Legitimacy lets groups, organizations and societies accept rules and decisions of authorities, institutions and social arrangements voluntarily. It promotes compliance with rules and regulations and allows shaping others’ behaviors. Influence can also be executed by threatening and using punishments to deter undesired behavior or by promising reward to promote desired behavior. However, as this may become costly and inefficient (Tyler, 2005) all authority systems aim at creating a belief in their legitimacy (Zelditch and Walker, 2003). Any form of control can be far more efficient when legitimacy is present.
Studies have shown that inappropriately exercised control systems cause negative employee attitudes and are often rejected (Ferrin et al., 2007; Sitkin & George, 2005).
Contrary, legitimate controls are often accepted and the relationship between controller and controlled are perceived as trustful, high-quality relationships, in which social-exchange relations often develop (Bijlsma-Frankema & Costa, 2010;
McNall & Roch, 2009; Weibel, 2007). Believing in the legitimacy of control systems causes individuals to internalize social values and norms of the controller and to develop an individual desire to behave accordingly (Hoffman, 1977).
Fundamental for legitimacy is the belief that evaluations, rules and decisions made by the other person or authorities are appropriate and thus are entitled to be obeyed (Tyler, 2005). In line with this, team legitimacy of peer pressure is the shared
perception that directly voicing approval and disapproval about desired behavior is appropriate and just.
Team Legitimacy of Peer Pressure as a Moderator
Prior findings regarding the outcomes of team peer pressure are mixed and thus point to the existence of moderators. While several researchers report positive consequences of peer controls (e.g. Bijlsma-Frankema et al., 2008; Marks & Panzer, 2004; Weick, Sutcliffe, & Obstfeld, 2008), Welbourne and Ferrante (2008) find that peer pressure results in team conflicts if peers experience peer pressure activities as overstepping authority and boundaries. This highlights the importance of legitimacy for the acceptance of team peer pressure activities. Therefore, I introduce team legitimacy of peer pressure as a moderator on the team peer pressure–team identification relationship.
Only if legitimacy in a team is high, team members perceive peer pressure activities as appropriate and just. Not needing to worry about the intention of peer pressure activities but assuming positive intentions (instead of possibly the goal to humiliate or correct) reduces uncertainty for both the person who exerts pressure and the person who receives pressure. As reduced uncertainty fosters team identification, legitimacy of peer pressure should strengthen team peer pressure-team identification. Moreover, I argued that exercising team peer pressure activities creates feelings of responsibility for and relatedness with the team. If team peer pressure activities are not perceived as legitimate, teams may refrain from engaging in them and thus fewer feelings of responsibility and relatedness may develop.
Additionally, interpreting team pressure activities as appropriate and thus positive fosters team identification, because another reason for individuals to identify with a social group is the tendency for self-esteem enhancement (Brewer, 1991;
Huettermann et al., 2017). Regarding team peer pressure activities positive, enhances the general evaluation of the team. Team members more likely identify with a team they perceive positively, because identifying with a highly valued team increases their social-identity (for social identity approach see Tajfel, 1978; Tajfel & Turner, 1979), which influences their personal identity (Ellemers et al., 2004). Since everybody strives for an affirmative personal identity (for self-esteem hypothesis see Hogg &
Abrams, 1988; Hogg & Mulin, 1999; Rubin & Hewstone, 1998), I expect that a team legitimacy of peer pressure strengthens the positive team peer pressure-team identification relationship.
Additionally, if teams regard peer pressure feedback as justified all team members are aware of their stance in the team, which reduces ambiguity and thus enhances the team identification process (Huettermann et al., 2017). Contrary, when teams doubt the legitimacy of peer pressure they regard feedback as inappropriate, are insecure about the intention of the feedback and hold negative feelings towards team members that engage in peer pressure. Those negative feelings would reduce the effect of team peer pressure on team identification. Therefore, I advocate team legitimacy of peer pressure as a moderator on the team peer pressure – team identification relationship.
H5: Team legitimacy of peer pressure moderates the relationship between team peer pressure and team identification, such that if legitimacy in peer control is high the relationship is stronger than when legitimacy in team peer pressure is low.
Having proposed the importance of team legitimacy of peer pressure for the development of team identification, I continue by examining the relationship between team identification and team effort more in detail. For this, team trust in supervisor is introduced, which I expected to moderate the team identification-team effort relationship.
Trust in Supervisor
Rousseau, Sitkin, Burt and Camerer (1998: 5) define interpersonal trust as the
„psychological state comprising the intention to accept vulnerability based upon positive expectations of the intentions or behavior of another”. According to McAllister (1995), two principal forms of interpersonal trust exist: cognition-based and affect-based trust. Cognition-based trust is “grounded in individual beliefs about peer reliability and dependability” (McAllister, 1995: 25). This form of trust assumes that everybody consciously chooses whom to trust based on perceptions of competence, responsibility, reliability and dependability. Additionally, cognition-based trust dependents on the success of past interaction, the degree of cultural and ethnical
similarity and professional credentials. In contrast, affect-based trust is “grounded in reciprocated interpersonal care and concern” (McAllister, 1995: 25). Antecedents of affective-trust are personally chosen organizational citizenship behaviors that aim to meet legitimate needs of the other person and frequent interactions. Additionally, some level of cognitive-based trust is argued to be necessary for the development of affective trust (McAllister, 1995).
More specifically, trust in the supervisor can be understood as the “expectation or belief that the team can rely on the leader's actions or words and that the leader has good intentions toward the team” (Dirks, 2000: 1004). Dirks and Ferrin (2002) extended the cognitive/affective trust framework to the context of trust in leadership.
According to them relationship-based and character-based trust in supervisor exists.
Character-based trust focuses on the supervisor’s characteristics such as integrity, dependability, fairness and ability. Thus, relationship-based trust is similar to cognitive trust. In contrast, relationship-based trust is similar to affective trust as consideration and care are central. It operates according to social exchange processes between subordinates and their supervisors. According to social exchange research, teams that believe their supervisor has in the past or will in the future exert effort and perform benevolent acts towards the team perceive the quality of the relationship as high (Dirks & Ferrin, 2002; Konovsky & Pugh, 1994; van Dam, Oreg & Schyns, 2008).
Teams are consequently likely to reciprocate considerations and care and engage in desired behaviors (Dirks & Ferrin, 2002). Social exchange relationships encourage workers to increase the time they spend to accomplish tasks and engage in actions that go beyond in-role behavior (Burke et al. 2007; Konovsky & Pugh, 1994).
Moreover, social exchange theory suggests that teams that trust their supervisor engage in behavior that fosters a positive relationship and helps to reach the goals defined by the supervisor (Dirks, 2000; Konovsky & Pugh, 1994). Team trust in the supervisor is essential for teams to accept the leader’s activities, decisions and goals and as the consecutive step, to work towards reaching the supervisor’s goals as a unit (Dirks, 2000). Contrary, at low levels of trust in supervisor, team effort is directed at self-protection instead of exerting effort towards team objectives (Mayer and Gavin, 2005).
Team Trust in Supervisor as a Moderator
As argued above, team identification is a motivational source for team effort. Team trust in supervisor is a distinct, highly influential motivational source for team effort. I argue that both sources in combination have a stronger impact than one of them individually due to synergetic effects. Teams that exert effort based on strong team identification do so in order to re-affirm their belonging to the team and because they are intrinsically motivated to meet team-goals (Shamir, 1990; Ellemers et al., 1999).
Team effort based on team trust in supervisor is grounded in exchange mechanisms.
Benevolent behavior of the supervisor creates team trust in the supervisor and the wish to reciprocate positive behavior (Dirks & Ferrin, 2002). Therefore, it can be considered a form of prosocial motivation, which is the desire to exert effort in order to benefit others (Batson, 1987). Prosocial motivation in combination with intrinsic motivation has been found to create synergetic effects (Gagné & Deci, 2005; Grant, 2008). In addition, it has been argued that intrinsic motivators create synergetic effects with extrinsic motivators, which provide information about the value of work, increase autonomy or highlights team members’ competence (Amabile, 1993).
Based on all above argumentations, I propose that the team identification-team effort relationship is particular strong when team trust in supervisor is high. Contrary, when teams do not believe that the supervisor has their best interests at heart, teams may doubt the strategies and goals of the supervisor. They will less likely work towards any objectives set by the leader, which will reduce team effort towards performance- related objectives (Dirks, 2000). In this case, strong team identification would still let members exert effort; however at a lower magnitude.
H6: Team trust in supervisor moderates the relationship between team identification and team effort, such that if team trust in supervisor is high the relationship is stronger than when team trust in supervisor is low.
The following conceptual model (figure 1) combines all hypotheses that have been established above based on theoretical background.
FIGURE 1 Conceptual model
METHOD Sample and Procedure
The study was tested with secondary data, which I received from Giulio Ockels, a current PhD student. The original research setting was a Dutch non-profit organization providing social security, with approximately 300 teams working in various locations in the Netherlands. 3790 employees received a questionnaire by email. By the start of this study, responses from 1248 employees were at hand, which formed a response rate of 32.93%. After ensuring that missing data was missing completely at random (Little’s MCAR test: χ2 (534) = 580.59, p = .08), respondents with missing values on the scales and respondents for whom no team member data was present were deleted. All further calculations were made with 1034 respondents working in 220 teams. Team sizes ranged from 2 to 19 with the average team size being M = 4.70. The age within the sample ranged from 16 to 66 (M = 45.52, SD = 10.94). The sample was comprised of 55.30% female and 44.70% male employees.
The tenure of respondents in years ranged from 0 to 46 (M = 15.76, SD = 13.15). The mean amount of working hours in a week was 33.17 (SD = 4.76).
Measures
Team Peer Pressure. Team Peer Pressure was measured by using a four-item scale by De Jong et al. (2014). Example items are „ If a team member behaves in a way we
consider unprofessional, we confront him or her directly “ and „ We openly express our dissatisfaction with team members who behave inappropriately”. The response categories were measured using a 7-point-Likert-scale, ranging from 1 (totally disagree) to 7 (totally agree). Cronbachs alpha was α = .938 (N = 1248), which exceeds the commonly accepted cut-off point for a reliable scale of α = .8 and indicates a highly reliable measure (Field, 2009).
Team Effort. Team effort was measured using a scale derived from De Jong et al.
(2014), which comprises five items from scales of George (1992) and of Mulvey, Bowes-Sperry & Klein (1998). Exemplary items are „I work very hard to perform my duties well“ and „I do everything to achieve team goals, even when others cut corners“. The response categories were measured using a 7-point-Likert-scale, ranging from 1 (totally disagree) to 7 (totally agree). Cronbachs alpha α = .943, (N = 1248) indicates a high scale reliability.
Team trust in Supervisor. Trust in supervisor was measured by using De Jong and Elfring’s scale (2010), which was referent-changed from team members to supervisor.
Exemplary items are „I assume that my direct supervisor takes my interests into account when making decisions“, „I trust that my direct supervisor will inform me concerning issues that are related to my work“. Employees indicated their answers on a 7-point-Likert-scale ranging from 1 (totally disagree) to 7 (totally agree). Scale reliability is given with a Cronbachs alpha of α = .933 (N = 1248).
Team identification. Team identification was measured using a 4-item scale constructed by Van der Vegt, Van de Vliert & Oosterhof (2003). Example items include
“I identify strongly with the other team members”, “I would like to continue working with my team”. On a 7-point-Likert-scale, employees indicated their answers ranging from 1 (totally disagree) to 7 (totally agree). The Cronbachs alpha α = .841 (N = 1248) indicates scale reliability
Team Legitimacy of peer pressure. This construct was measured by using the five- item scale developed by Giulio Ockels, the PhD student who gathered the data. Team members indicated on a 7-point-Likert-scale to which extent they find certain peer control activities appropriate, in which 1 translated as very unjust and 7 as very just.
Exemplary items: “I think it’s good that we talk to each other about the quality of
work” and “I think that we should listen to each other when we address undesirable behavior”. Cronbachs alpha of α = .923 (N = 1248) illustrates high scale reliability.
All measurement scales are provided in Appendix A.
RESULTS Data Analysis
All statistical analyses were conducted with the software package IBM SPSS and excel.
Throughout the entire analysis, a significance level of .05 was used. First, an exploratory factor analysis was conducted to test whether the scales load on different dimensions. The unit of measurement was the individual, however as the analysis is at the team level all data were aggregated. Before doing so, ICC1, ICC2 and r*wg were computed to justify the aggregation. Normality was tested with the Shapiro-Wilk-test as well as measures of skewness and kurtosis. Furthermore, I ensured that multi- collinearity is no concern. Outliers were detected using the interquartile range rule of 3. All Outliers were allowable answers, which I expected to occur due to the negative skewness. As deleting valid responses would falsify the dataset, outliers were included. The analysis was conducted with standardized variables, which were centered on their means to increase comparability of effect sizes between studies (Field, 2009). Respondents without data on the variable scales were deleted.
Additionally, respondents were excluded if no team member data was available and therefore the team size requirement of two was not met. This reduced the data size to 1034 respondents working in 220 teams.
The all analyses were conducted via Hayes PROCESS models. Hayes PROCESS models take all relationships into account and allow bias-corrected bootstrapping.
Bootstrapping was chosen to account for the imperfect normality of the data. B- coefficients were calculated for 5000 bootstrap samples to estimate the data distribution. Based on the bootstrapping standard error, which is calculated with standard deviations of the sampling distribution from the bootstrapping samples, the results of the indirect effects are and bias-corrected (Field, 2009). All models were tested with and without bootstrapping to increase robustness of results. If not reported otherwise, the results were consistent.
For testing hypotheses 1, 2, 3 and 4 Hayes PROCESS model 4 (Hayes, 2013) was used, which tests for mediation. Hypothesis 5 and 6, which are moderated-mediation hypotheses, were tested using Process model 7 and 14. Additionally, to examine the complete conceptual model Process analysis model 21 was used.
Factor Analysis. A principal component analysis (PCA) was conducted on all 24 items with orthogonal rotation (Kaiser-varimax rotation). The Kaiser-Meyer-Olkin measure verified the sampling adequacy of the analysis with KMO = .92 (‘superb’ according to Field, 2009). Bartlett’s test of sphericity χ2 (276) = 21857.51, p < .001 implies that the correlations between items were sufficiently large for conducting PCA. The Scree plot identified five components with eigenvalues over Kaiser’s criterion of 1 that explained 77.05% of the variance in combination. All items showed factor loadings above .7 and each construct significantly differed from the other constructs, indicated by cross loads below .25. Thus, all items were maintained. Table 1 portraits the factor loadings after rotation.
Justification of data aggregation. Because my independent variable team peer pressure concerns a team level phenomenon, I am interested in how this variable relates to other team level variables. Before aggregating the individual data to a team level, I computed ICC1’s, ICC2’s and r*wg’s to justify the aggregation. Table 2 presents the results of the analyses. ICC1 values were below .4 indicating a low level of clinical significance (Cicchetti & Sparrow, 1981). ICC1 values represent the proportion of group-level variance explained by group membership (Bliese, 2000). Low ICC1 values indicate that only a small proportion of the variance within the target variables can be explained by group membership. ICC2 scores are a measure of between-group differences, thus the low ICC2 scores indicate small differences in ratings between groups (Van Mierlo, Vermunt, & Rutte, 2009). Rwg values indicate within-group interrater agreement and imply whether group members provide more similar absolute ratings in the group than under the condition of random response (Bliese, 2000; Klein, Conn, Smith, & Sorra, 2001). For this study, r*wg values were of particular interest as unlike other rwg values, they do not overestimate the true agreement for ratings with many items (Van Mierlo et al., 2009). R*wg values ranged from .81 to .91, which exceeds the common cut-off point for acceptability of
aggregation of .70 (Van Mierlo et al., 2009). Therefore aggregation is justified. The data were aggregated to the team level via the referent-shift consensus model (Van Mierlo et al., 2009) by calculating the means of all team members.
Preliminary Analysis
Multi-collinearity. To test whether the variables met the collinearity assumption, value inflation factors (VIF) were computed. Multi-collinearity is a concern if VIF values are above 10 (Kennedy, 1992; Hair, Anderson, Tatham, Black, 1995). In the present study, VIF’s ranged between 1.08 and 1.39 (see table 3). Therefore, multi collinearity is no concern.
TABLE 1
Factor loadings and communalities
Effort Trust in
supervisor
Legitimacy of peer pressure
Peer pressure
Team
identification
Effort_1 .88
Effort_2 .89
Effort_3 .89
Effort_4 .87
Effort_5 .76 .24
Trust_1 .85
Trust_2 .83
Trust_3 .84
Trust_4 .89
Trust_5 .89
Legitimacy_1 .78
Legitimacy_2 .81
Legitimacy_3 .87
Legitimacy_4 .83
Legitimacy_5 .86
Peer Control_1 .87
Peer Control_2 .89
Peer Control_3 .88
Peer Control_4 .86
Team Id_1 .74
Team Id_2 .71
Team Id_3 .21 .71
Team Id_4 .24 .72
Team Id_5 .21 .82
Eigenvalues 8.99 2.84 2.60 2.27 1.80
% of variance 37.45 11.81 10.85 9.46 7.48
α .94 .93 .92 .94 .84
Note: Factor loadings < .2 were suppressed, table based on a principle components analysis, Kaiser-varimax rotation, 24 items, N = 1248
TABLE 2
Justification of data aggregation via ICC1, ICC2, r*wg
ICC1 ICC2 R*wg
Effort -.04 -.18 .91
Peer Pressure .12* .35 .81
Trust in Supervisor .21* .51 .86
Team identification .08* .27 .87
Legitimacy of peer pressure .01 .03 .90
Note: * ANOVA, p < .05, N=1034
TABLE 3 Collinearity analysis
Tolerance VIF
Team peer pressure .73 1.36
Team trust in supervisor .93 1.08
Team Legitimacy of peer pressure .73 1.39
Team identification .76 1.31
Note: N=220 teams, Dependent variable: Team effort
Correlations and data distribution. Table 3 displays the two-tailed Pearson correlations between the variables. Correlations were calculated to understand how variables are associated and to decide on necessary control variables for the main analysis. Gender, age, work years and hours per week were considered as control variables. However, as none of them significantly correlate with any main variables these variables were excluded from the further analysis. The correlations provide brief information about expected effects. The significant positive correlation between team peer pressure and team effort (r = .29, p < .001), suggests a main effect between the variables. Furthermore, the significant correlation of team identification with both team peer pressure (r = .40, p < .001) and team effort (r = .42, p < .001) indicates the possibility of mediation.
TABLE 3
Pearson correlations
Variables 1 2 3 4 5 6 7 8 9
1. Effort .29** .35** .35** .42** -.05 .09 -.01 .12 2. Team peer pressure .21* .45** .40** -.02 .03 -.09 .08 3. Trust in Supervisor .22 ** .19 ** -.08 -.01 -.01 .08 4. Legitimacy of pp .41** -.01 .01 -.03 .09
5. Team identification .04 .12 .01 .06
6. Age .22** .77** .03
7. Gender .12 .31**
8. Work years .10
9. Hours per week
Note: N=220 teams, *p < .05 (two-tailed), **p < .001 (two-tailed), pp = peer pressure
Table 4 provides information about the distribution of the variables. Means, standard deviations, skewness and kurtosis of the main variables were calculated and the Shapiro-Wilk test for normality was performed. Besides team peer pressure (p = .22) and team identification (p = .94) the Shapiro-Wilk test showed significant results for all variables indicating a lack of statistical symmetry (Field, 2009).
All variables were negatively skewed, meaning that the majority of ratings were at the high end of the scale. Deciding whether the skewness is statistically different from normal took place via testing whether z-scores (skewness value divided by its standard error) exceed the threshold for large sample sizes of 2.58 (Field, 2009). This surfaced that team effort, team trust in supervisor, and legitimacy of peer pressure are significantly skewed. Kurtosis values provide information about the pointiness of the data. The majority of variables had positive kurtosis, indicating a leptokurtic, heavy- tailed distribution. Examining kurtosis z-scores indicated that team trust in supervisor and legitimacy are significantly different from normal.
Noteworthy, if sample size exceeds 200, researchers are advised to rather interpret graphical representations of the distribution of the variables instead of statistical data (Field, 2009). Scale transformation was dismissed, in part because graphical representations allow the interpretation of a fairly normal distribution and in part because of critical voices against normalizing transformations (Glass, Peckham, &
Sanders, 1972; Grayson, 2004). However, it was decided to apply bias-corrected bootstrapping analyses when applicable, as this circumvents the problem of a non- normal distribution (Hayes, 2013).
TABLE 4
Descriptive statistics
Variables M SD Skewness Kurtosis Shapiro-
(SE .164) (SE .327) Wilk
Effort 6.18 .44 -.71* .59 .00
Team peer pressure 4.52 .82 -.12 -.44 .22
Trust in supervisor 5.42 .81 -1.08* 2.01* .00
Legitimacy of pp 5.91 .54 -.90* 1.17* .00
Team identification 5.08 .67 -.02 .04 .94
Note: N=220 teams, pp = peer pressure, * statistically different from normal
Main Analysis
Hypothesis 1. Hypothesis 1 stated that team peer pressure promotes team effort. This was initially tested within the mediation analysis (table 5). Indeed, I found team peer pressure significantly predicting team effort (model 4: b = .13, t (218) = 4.50, p < .001, R2= .09). However, when taking all other variables into account within the moderated-mediation analysis (table 8), the direct effect is not significant anymore (model 21: b = .03, t (215) = 1.45, n.s., R2 = .27). Thus, the hypothesis has to be rejected.
Hypothesis 2. Hypothesis 2 stated that team identification promotes team effort. The results of the mediation analysis (table 5) indeed showed a significant, positive relationship between team identification and team effort (model 4: b = .16, t (217) = 5.32, p < .001, R2 = .19). Testing the complete conceptual model, which includes team trust in supervisor as a moderator (table 8), additionally supports that team identification significantly predicts team effort (model 21: b = .15, t (215) = 4.97, p <
.001, R2 = .27). This supports my hypothesis.
Hypothesis 3. Hypothesis 3 stated that team peer pressure promotes team identification. As expected, the results of the mediation analysis (table 5) show a significant, positive relationship between team peer pressure and team identification (model 4: b = .40, t (218) = 6.48, p < .001, R2 = .19). The test of the complete conceptual model, which includes team legitimacy of peer pressure as a moderator (table 8), additionally supports the prediction of team identification through team peer pressure (model 21: b = .28, t (216) = 3.77, p < .001, R2 = .23) and thus my hypothesis.
Hypothesis 4. Hypothesis 4 stated that team identification mediates the relationship between team peer pressure and team effort. To test the mediation PROCESS bias- corrected bootstrapping was applied. The results of the mediation analysis (table 5) showed a significant indirect effect of team peer pressure on team effort through team identification (b = .06, CI95%[.04, .10]). Without consideration of any other variables team peer pressure explained 9% of the variance in team effort. The model including team identification explained 19% of the variance in team effort. Therefore, as expected, team identification explained variance over and above the direct effect.
When not considering other variables, team identification partially mediated the relationship between team peer pressure and team effort. However, when taking all variables into account, the direct team peer pressure-team effort relationship is not significant anymore (b = .03, t (215) = 1.45, n.s.), which points to a full mediation. In both conditions, my hypothesis is supported.
Hypothesis 5. Hypothesis 5 stated that team legitimacy of peer pressure moderates the relationship between team peer pressure and team identification, such that if team legitimacy of peer pressure is high the relationship is stronger than when team legitimacy of team peer pressure is low. To test this, a bias-corrected bootstrapping moderated-mediation analysis was conducted. The results (table 6) did not support the hypothesis because the interaction term of team legitimacy of peer pressure on the team peer pressure-team identification relationship was insignificant (b = -.05, SE
= .07, n.s., R2 = .23, ΔR2 = .07). The presence of zero in the 95% confidence interval showed this result [-.19, .10]. As visible in figure 2, team legitimacy of peer pressure
did have a significant main effect on team identification (b = .27, SE = .08, p < .01), but did not function as a moderator. Graphing the results (figure 2.1) shows that high legitimacy of peer pressure yields higher team identification values, but that the slopes do not differ significantly. Therefore, H5 must be rejected.
TABLE 5
Regression results for the mediation effect
Mediator variable model (DV=Team identification) R2 = .16, F (1,218) = 42.04
Predictor b p SE t
Team peer pressure .40 .00 .06 6.48
Dependent variable model (DV =Team Effort)
R2 = .19, F (2,217) = 25.52
Predictor b p SE t
Team identification .16 .00 .03 5.32
Team peer pressure .07 .03 .03 2.23
Total effect model (DV=Team Effort)
R2 = .09, F (1,218) = 20.21
Predictor b p SE t
Team peer pressure .13 .00 .03 4.50
Indirect effect
Effect Boot SE BootLLCI BootULCI
IV->M->DV .06 .02 .04 .10
R2 mediation effect size .07 .03 .03 .12
Note: * p < .01, ** p < .001, a standard regression coefficients are presented, Hayes PROCESS model 4, 5000 bootstrapping samples
TABLE 6
Moderated mediation of team legitimacy of peer pressure Mediator variable model (DV = Team identification)
R2 = .23, F(3,216) = 16.78
Predictor b SE t LLCI ULCI
Team peer pressure .28 .08 3.77*** .14 .43
Legitimacy of pp .27 .08 3.45** .12 .43
Interaction -.05 .07 -.64 -.19 .10
(DV = Team effort)
R2 = .19, F(2,217) = 22.03
Predictor b SE t LLCI ULCI
Team identification .16 .03 5.21*** .10 .22
Team peer pressure .07 .03 2.03* .00 .13
Index of moderated mediation
Mediator Index BootSE BootLLCI BootULCI
Team identification -.01 .01 -.03 .01
Conditional indirect effect team peer pressure on team effort at values of moderator
Mediator Legitimacy Effect BootSE BootLLCI BootULCI
Team identification -1.00 .05 .02 .02 .10
Team identification .00 .04 .01 .02 .08
Team identification 1.00 .04 .02 .01 .08
Note: N=220 teams, standardized regression coefficients, * p<.05, ** p<.01, *** p<.001, interaction: team peer pressure x team legitimacy of team peer pressure, Hayes PROCESS model 7, 5000 bootstrapping samples
FIGURE 2
Moderated mediation model with team legitimacy of peer pressure
Note: standardized regression coefficients are reported, *p < .05, **p < .01, *** p < .001, the total direct effect of team peer pressure on team effort with team identification and team legitimacy of peer control in the model is presented in parenthesis.
FIGURE 2.1
Interaction plot of team legitimacy of peer pressure
Note: b low team legitimacy of pp = .05[.02, .10], b high team legitimacy of pp = .03 [.01, .08], moderation on the relationship between team peer pressure and team identification
Hypothesis 6. Hypothesis 6 stated that team trust in supervisor moderates the relationship between team identification and team effort, such that if team trust in supervisor is high the relationship is stronger than when team trust in supervisor is low.
The results of the bias-corrected bootstrapping moderated-mediation analysis (table 7) showed a significant main effect of trust in supervisor on team effort (b = .12, SE = .03, p < .001, R2 = .27, ΔR2 = .08) but a non-significant interaction effect of trust in supervisor and team identification (b = -.05, SE = .03, p < .10). At CI95% zero was present [-.11, .01], therefore the hypothesis must be rejected.
A display of the effects of moderated mediation can be found in Figure 3. Graphing the results (Figure 3.1) shows that team effort reached its peak when both team identification and team trust in supervisor were high, caused by the positive main effects of team identification and team trust in supervisor. When team trust in supervisor was high, the magnitude of team identification only caused small changes in team effort. The slope was less steep as the effect flattens when the ceiling nears.
Contrary, in the low team trust in supervisor condition, the level of team identification strongly positively affected team effort.
TABLE 7
Moderated mediation of trust in supervisor
Note: N=220 teams, standardized regression coefficients, + p < .1, * p < .05, ** p < .01, *** p <
.001, interaction: Team peer pressure x Team legitimacy of team peer pressure, Hayes PROCESS model 14, 5000 bootstrapping samples
Mediator variable model (DV = Team identification)
R2 = .16, F (1, 218 ) = 35.92
Predictor b SE t LLCI ULCI
Team peer pressure .40 .07 5.99*** .270 .534
(DV = Team effort)
R2 = .27, F (4, 215) = 15.90
Predictor b SE t LLCI ULCI
Team identification .15 .03 4.97*** .09 .20
Team peer pressure .05 .03 1.45 -.02 .11
Trust in supervisor .12 .03 3.85*** .06 .18
Interaction -.05 .03 -1.80 -.11 .01
Index of moderated mediation
Mediator Index BootSE BootLLCI BootULCI
Team identification -.02 .01 -.05 -.001
Conditional indirect effect team peer pressure on team effort at values of moderator
Mediator Trust in
supervisor Effect BootSE BootLLCI BootULCI
Team identification -1.00 .08 .02 .04 .13
Team identification .00 .06 .02 .03 .09
Team identification 1.00 .04 .02 .01 .07
Direct effect of team peer pressure on team identification
Effect SE t p LLCI ULCI
.05 .03 1.45 .15 -.02 .11
FIGURE 3
Moderated mediation model with team trust in supervisor
Note: standardized regression coefficients are reported, *p<.05, **p<.01, *** p<.001, the total direct effect of team peer pressure on team effort with team identification and trust in supervisor in the model is presented in parenthesis.
FIGURE 3.1
Interaction plot of team trust in supervisor
Note: b low team trust in supervisor = .08 [.04, .13] , b high team trust in supervisor = .04 [.01, .07], moderation of
team trust in supervisor on the relationship between team identification and team effort controlled for team peer pressure
