Abstract
Three studies examined the role of leader consideration in diverse teams. Based on the categorization-elaboration model, we argue that leader consideration can address the negative group processes that result from categorization processes in diverse teams as well as influence the perceptions of the team’s diversity. Studies 1a and 1b focus on the leadership preferences of members of heterogeneous versus homogeneous teams. Results show that participants faced with the prospect of working in a diverse team had a higher preference for considerate leadership than participants expecting to work in a homogeneous team. In Study 2, we examined whether diverse teams indeed function better with a leader who scores higher on considerate leadership, and whether this can be explained by changes in the ways in which the leader and followers perceive the team’s diversity. An interaction between group diversity and consideration shows that highly considerate leaders improved team functioning for heterogeneous (but not homogeneous) teams, and that leader individuation, or the ability of the leader to see members as unique individuals, explained the positive effects of considerate leadership on the functioning of diverse teams.
Team diversity can be defined as “differences between individuals on any attribute that may lead to the perception that another person is different from the self” (van Knippenberg, De Dreu, & Homan, 2004, p. 1008). These differences within teams have been found to affect a wide array of important team outcomes, including performance, satisfaction, creativity, and innovation (e.g., Milliken & Martins, 1996; Williams & O’Reilly, 1998). Interestingly, the direction of the effects is often difficult to predict (Milliken & Martins, 1996). On the one hand, diversity can initiate subgroup categorizations (“us–them” distinctions) and intergroup bias that may result in reduced interpersonal liking, low trust, and high levels of conflict (e.g., Earley & Mosakowski, 2000; Thatcher, Jehn, & Zanutto, 2003). This can impede important team outcomes, such as performance and satisfaction. On the other hand, diversity can instigate the exchange and processing of different perspectives and ideas, which can enhance team performance and increase work motivation and satisfaction (e.g., Cox, Lobel, & McLeod, 1991; van Knippenberg et al., 2004).
To account for these inconsistent findings, the focus in diversity research has recently switched to a contingency approach (e.g., van Knippenberg et al., 2004; van Knippenberg & Schippers, 2007), examining likely moderators of the effects of diversity (see e.g., Chatman, Polzer, Barsade, & Neale, 1998; Harrison, Price, & Bell, 1998; Homan et al., 2008; Homan, van Knippenberg, van Kleef, & De Dreu, 2007; Jehn, Northcraft, & Neale, 1999; Pelled, Eisenhardt, & Xin, 1999). A new and growing line of research suggests that leadership (i.e., influencing a group of people to obtain a certain goal; Stogdill & Coons, 1957) may be a critical determinant of the effects of diversity on team outcomes (e.g., Chen & van Velsor, 1996; DiTomaso & Hooijberg, 1996; Eagly & Chin, 2010; Greer, Homan, De Hoogh, & Den Hartog, 2012; Homan & Jehn, 2010; Pittinsky, 2010). Because of their influential positions, leaders might hold the key to successful diversity management in teams (Zaccaro, Rittman, & Marks, 2001).
In the present paper, we adopt a functional leadership approach (McGrath, 1962) and propose that considerate leadership (Bass, 1999; Yukl, 2010) is an effective way to deal with the potential downsides of diversity. Considerate leaders “show concern and respect for followers, look out for their welfare, and express appreciation and support” (Judge, Piccolo, & Ilies, 2004, p. 36; see also Bass, 1990) and should therefore be able to address the negative processes associated with diversity, that is the potential “us–them” distinctions and deteriorated relationships. More specifically, we will argue that considerate leadership limits subgroup categorization processes and smoothes interpersonal processes, both of which are critical to the effective functioning of diverse groups (Mannix & Neale, 2005).
Importantly, within this model, we add to past work on leadership and diversity (see e.g., Greer et al., 2012; Kearney & Gebert, 2009; Klein, Knight, Ziegert, Chong Lim, & Saltz, 2011; Shin & Zhou, 2007; Somech, 2006) by (a) examining followers’ preferences for considerate leadership in diverse teams, and (b) investigating whether the preference for considerate leadership in diverse teams can be explained by the impact of considerate leadership on the ways in which the leader and team members perceive the team’s diversity. This means that the contribution of this paper is twofold. First, we focus on the leadership preferences of members of diverse teams. While past research has examined which type of leader can benefit the performance of diverse groups (e.g., Greer et al., 2012; Kearney & Gebert, 2009), little research has examined which type of leader followers instinctively prefer when called to work in a diverse setting. We will argue that people who are confronted with a diverse team will have a higher preference for considerate leadership because people in diverse teams prefer a leader who is focused on relationships and meeting individual needs, which is especially important in the potentially tense environment of a diverse team. Our focus on leader preference is important because previous research has shown that employees work better when they work for leaders who can fulfill their needs (e.g., Brief, 1998; Eagly & Chaiken, 1993; Ehrhardt & Klein, 2001; Miner, 1978). Second, we examine whether these preferences translate into actual performance benefits for diverse teams. We do this by focusing on the influence of considerate leadership on the ways in which both the team leader and the team members perceive the team’s diversity. We propose that leader consideration makes subgroup formation less likely and makes leaders individuate their team members more, and that this perception of diversity in terms of individual differences (as opposed to subgroups) aids team functioning. This focus on the underlying cognitive, perceptual mechanisms offers a new and theoretically driven explanation of the process by which certain leadership styles may help or hurt the performance of diverse groups.
Team diversity
The broad definition of diversity that was postulated before illustrates that diversity can occur on any possible dimension that can lead people to perceive someone else to be different (see also Jackson & Associates, 1992; Mannix & Neale, 2005; Triandis, Kurowski, & Gelfand, 1994; Williams & O’Reilly, 1998). As stated before, the effects of diversity on team outcomes have been found to range from negative to neutral to positive (van Knippenberg & Schippers, 2007). These inconsistent findings can be explained by the two distinct processes that can be instigated by diversity (Milliken & Martins, 1996). On the one hand, diversity can initiate subgroup categorization (“us–them” distinctions and subgroup formation) and intergroup bias that may result in reduced interpersonal liking, low trust, and conflict (Earley & Mosakowski, 2000; Thatcher et al., 2003). On the other hand, diversity can instigate the thorough exchange and processing of different perspectives and ideas, potentially resulting in enhanced performance (Cox et al., 1991; Homan et al., 2008; Homan et al., 2007).
These two competing processes have been combined in the categorization-elaboration model (van Knippenberg et al., 2004). Based on this model, the authors proposed that in order to predict whether diversity has positive or negative effects, one needs to incorporate crucial moderators. In line with this reasoning, recent research has shown that the effects of diversity depend on contextual factors, such as time, diversity beliefs, organizational culture, and task type (e.g., Chatman et al., 1998; Ely & Thomas, 2001; Homan et al., 2008; Homan et al., 2007; Jackson, Joshi, & Erhardt, 2003; Jehn et al., 1999; Joshi & Roh, 2009; Schippers, Den Hartog, Koopman, & Wienk, 2003).
The theoretical logic behind these contextual determinants of the effects of diversity is twofold. On the one hand, contextual factors that make groups more likely to elaborate upon their divergent information and perspectives (e.g., reflexivity) can be expected to enhance the relationship between diversity and team performance (e.g., Schippers et al., 2003). On the other hand, factors that make it less likely that individuals will categorize their diverse team into demographic subgroups (e.g., prodiversity beliefs), can be expected to improve the relationship between diversity and team performance by limiting the likelihood of conflict and distrust (e.g., Homan et al., 2008). Importantly, van Knippenberg et al. (2004) argued that negative processes between salient subgroups within diverse teams undermine the potential positive effects of diversity on information exchange and processing—that is, teams in which relationships have deteriorated due to subgroup formation will not be able and/or motivated to discuss the available divergent perspectives and ideas (e.g., De Dreu, Nijstad, & van Knippenberg, 2008). Thus, these theoretical and empirical insights indicate that managing the degree to which subgroups are formed is crucial for harvesting the value in diversity. Here we propose that considerate leadership is a way to reduce subgroup formation and enhance the performance of diverse teams.
Leader consideration
Leadership is one of the most extensively studied topics in organizational behavior (e.g., Yukl, 2010) and has been found to have a profound impact on the functioning of teams (Burke et al., 2006). Recently, research has started to examine the moderating role of different leadership styles in diverse teams (e.g., Greer et al., 2012; Klein et al., 2011; Somech, 2006). Most research on the role of leadership in diverse teams has focused on transformational leadership (Avolio, Bass, & Jung, 1999) and leader–member exchange (Dansereau, Graen, & Haga, 1975).
Regarding leader–member exchange (LMX), the better the quality of the relationship between leader and team members, the better diverse teams have been shown to perform, because team members feel accepted and safe. In two studies, teams that were diverse in terms of gender, race, age, and tenure exhibited lower turnover and higher performance with higher aggregate LMX as well as lower LMX differentiation (Nishii & Mayer, 2009; Stewart & Johnson, 2009).
With respect to transformational leadership, transformational leaders, who stress a common goal and vision, intellectually stimulate and empower their followers, and set high performance standards (Avolio et al., 1999), have been suggested to promote the positive effects of diversity by transcending differences and by stimulating the discussion of divergent viewpoints and ideas. Supporting this reasoning, Shin and Zhou (2007) found that the relationship between educational specialization diversity and team creativity is more positive when transformational leadership is high rather than low. In a similar vein, Kearney and Gebert (2009) found an interaction between team diversity and transformational leadership, such that teams with higher levels of educational and nationality diversity (but not those with low diversity) performed better to the degree that the leader scored higher on transformational leadership. This effect was mediated by collective team identification and information elaboration. Finally, Kunze and Bruch (2010) showed that age-based diversity faultlines were only detrimental for perceived productive energy when the leader scored low (rather than high) on transformational leadership.
Here we go beyond past findings on transformational leadership and LMX by focusing on considerate leadership. Although considerate leadership shows similarities to the individualized consideration subcomponent of transformational leadership (see e.g., Judge et al., 2004; Kearney & Gebert, 2009; Seltzer & Bass, 1990; Zohar & Tenne-Gazit, 2008) and LMX (Dansereau et al., 1975; Graen & Uhl-Bien, 1995), the concept of leader consideration utilized in the present paper differs from these related leadership concepts in important ways. Whereas LMX focuses on the quality of the dyadic relationship between leader and follower (Graen & Uhl-Bien, 1995), consideration is aimed at the quality of the relationships within the team or between subordinates. Regarding transformational leadership, we follow Bass (1999), who argued that the Ohio State Consideration factor (which is used in the present paper) should be distinguished from individualized consideration. Individualized consideration is conceptualized as the leaders’ behaviors addressing individual development of the individual team members, whereas Bass’ consideration factor is conceptualized as behaviors aimed at solving problems and providing personal support. Although all these leadership styles are person-directed, the focus of this personalized attention represents a crucial and important difference. That is, LMX is aimed at creating positive relations between a leader and a follower (and within teams, a leader can have different-quality relationships with different team members); individualized consideration is aimed at the development, empowerment, and abilities of the individual; and consideration is aimed at making sure that individuals feel safe, supported, and solve personal problems (Bass, 1999). We choose to focus on leader consideration because of its theoretical linkages with the key problems of diverse teams—ensuring that individuals feel safe and able to resolve interpersonal differences and overcome subgroup differences within the team.
Whereas past studies on transformational leadership and LMX provide valuable insights in the role of different leadership styles and behaviors as moderators of the diversity–outcome link, they remain silent regarding the role of (a) member preferences for certain leadership styles and (b) member and leader diversity perceptions in diverse teams. Previous work has shown that perceptions of diversity might be more important than actual demographic differences (Homan, Greer, Jehn, & Koning, 2010; Zellmer-Bruhn, Maloney, Bhappu, & Salvador, 2008), which makes it highly relevant to examine how leadership styles, and particularly, the theoretically relevant style of leader consideration, influence perceptions that team members as well as leaders have of their team.
Linking team diversity and leader consideration
Diverse teams experience negative effects of diversity due to subgroup categorization and resulting intergroup bias (van Knippenberg et al., 2004). This implies that to the degree that the team members perceive their team in terms of subgroups, they are more likely to experience conflicts, distrust, and deteriorated relationships (Homan et al., 2008; Lau & Murnighan, 1998). We propose that leaders who adapt considerate leadership in diverse teams will be preferred more and will be more effective for two reasons. First, in a diverse team, the issues that should be addressed are subgroup formation and concomitant intergroup biases, such as conflicts, low trust, and deteriorated relationships within the team (van Knippenberg et al., 2004). We argue that leader consideration can limit subgroup formation and concomitant negative processes because these leaders focus on healing relationships, increasing trust, and managing frictions (Bass, 1990; Burke et al., 2006; Kerr, Schriesheim, Murphy, & Stogdill, 1974). Leader consideration will therefore be preferred more by team members of diverse compared to homogeneous teams as well as have positive effects on the functioning of diverse teams.
Second, considerate leadership can be expected to influence the way diversity is perceived. Diverse teams are, by definition, characterized by a greater amount of differences between individuals than are homogenous teams. These differences can however be perceived in different ways (Homan et al., 2010). On the one hand, individuals can see team diversity in terms of subgroups, perceiving a dividing line that sets one subgroup in the team apart from one or more other subgroups (e.g., women vs. men, or older members vs. younger members). On the other hand, individuals may perceive team diversity in terms of the unique qualities that each member brings to the team (e.g., one person has a lot of experience, another is very creative). The greater the diversity in the team, the greater is the potential for both subgroup categorization and perception of unique qualities. We propose that leader consideration shapes the way diversity is perceived. Considerate leaders appreciate and acknowledge individual ideas and feelings (Fleishman & Peters, 1962), which makes them likely to perceive team members as unique and separate individuals rather than in terms of diversity-based subgroups. We refer to this tendency as leader individuation. We argue that a leader’s individuation tendency is likely to be taken over by the team members (Griffin, Parker, & Mason, 2010; Shamir, Arthur, & House, 1994), which should in turn influence the behavior and reactions of the team members toward their diverse team (Greer et al., 2012; Shamir et al., 1994). The team members should therefore be less likely to perceive subgroups to the degree that their leader shows more consideration. Additionally, it has been shown that perceiving differences in individuals is required for the positive effects of diversity to occur (Homan et al., 2010). The value-in-diversity hypothesis suggests that diversity in itself can produce positive outcomes. This implies that differences should be perceived in order to profit from diversity (Homan et al., 2010; Homan et al., 2008). Leader individuation should therefore be positively related to team functioning.
In sum, we propose that leaders who apply considerate leadership in diverse teams will be most preferred and most effective, because consideration will address negative intragroup processes and limit subgroup formation, while at the same time leading to the acknowledgement and appreciation of individual differences and diversity perceptions in terms of individual differences rather than diversity-based subgroups. Based on the explained reasoning, we propose the following hypotheses. First, we hypothesize that (members of) diverse teams are more likely to prefer leader consideration than members of homogeneous teams (H1a). This preference will be explained by (anticipated) subgroup formation within diverse teams (H1b; van Knippenberg et al., 2004; Williams & O’Reilly, 1998). Second, diverse teams (compared to homogeneous teams) will perform better (H2a) and experience less subgroup formation (H2b) to the degree that their leader shows more consideration. Additionally, diversity will be more strongly related to leader perceptions of individual differences (i.e., leader individuation) to the degree that the leader scores higher on consideration (H3). Finally, we hypothesize that the interaction between team diversity and leader consideration on team performance will be mediated by perceived subgroup formation (H4a) and leader individuation (H4b). Our research model is visualized in Figure 1. All directional hypotheses were tested using one-tailed levels of significance.
The proposed relationships between the variables of interest.
Overview of studies
We examined these hypotheses in two different research setups. First, we were interested to see whether people would indeed prefer leader consideration when (faced with the prospect of) working with a diverse (as opposed to a homogeneous) team (Studies 1a and 1b). In Study 1a, our sample consisted mainly of university students who read a scenario about working in a generally diverse or homogenous team. We then asked them for their leadership preference. In this study we utilize an operationalization of diversity that is closely related to our working definition of diversity (i.e., general diversity) to test whether the anticipation of working with “different” others affects people’s preference for considerate leadership. In Study 1b, our sample consisted of employees who read a scenario about working in an educationally diverse or homogenous team and then reported their leadership preference. In Study 1b we also tested for mediation by examining whether the expectations of participants regarding subgroup formation within their team influenced their preference for leader consideration. In this study, we focus on educational diversity, because this diversity dimension has often been connected to the positive processes of diversity, that is, information exchange and processing (e.g., Bantel & Jackson, 1989; Mannix & Neale, 1995), but can—according to the categorization-elaboration model—theoretically also instigate subgroup formation and negative group processes (van Knippenberg et al., 2004). As such, our choice of educational diversity makes for a conservative test of our hypothesis. If we find that educational diversity also leads to preferences for considerate leadership, this would therefore boost our confidence in the proposed relationships.
Second, we tested whether leader consideration would indeed positively affect the functioning of diverse teams in organizations by influencing diversity perceptions of the team members and the team leader (Study 2). We zoom in on tenure diversity in this field study. Specifically, we examine consideration as a moderator of the effects of tenure diversity on team performance, and also investigate diversity perceptions as a potential mediator of the interactive effect of leadership and team diversity on team performance. Tenure is a very interesting diversity dimension to include in a study on a moderator of diversity effects, as it can potentially impede as well as benefit group functioning (Ancona & Caldwell, 1992; Mannix & Neale, 2005). This makes the management of this type of diversity highly relevant.
Study 1a
In this study, 36 students and employees participated in the experiment for course credits or the opportunity to win a gift voucher (28 students, 8 employees; 31 women, 5 men; mean age = 23.70, SD = 7.67). The participants read a scenario in which they had to imagine working in either a heterogeneous (N = 19) or homogeneous team (N = 17)—the specific diversity dimensions were not made explicit. We choose this broad manipulation of diversity as a first test of our hypothesis, in order to connect to the broad definition of diversity (van Knippenberg et al., 2004) which describes that diversity can occur on any dimension. In the heterogeneous team condition, people read that the team members were very different from each other and did not have many similarities. In the homogeneous team condition, people read that the team members were very much alike and had many similarities. The exact text of the scenario including the manipulation is provided in Appendix A. After reading the scenario, the participants indicated their preference for leader consideration. Finally, they answered questions to check the manipulation of diversity. There was no difference between students and employees on any of the measures and including this distinction in the analyses did not alter the pattern of results. We also assessed years of work experience, educational level, and nationality. Including these variables in the analyses also did not change our findings nor did these variables predict leadership preference.
Measures
Preference for leader consideration
The preference for consideration was assessed with four questions based on questionnaires developed by Stogdill and Coons (1957) and Indvik (1985). Example questions are “In this situation I would prefer a leader who helps people in the work group with their personal problems” and “In this situation I would prefer a leader who is friendly and easy to approach” (7-point scale; M = 5.64; SD = 0.70; α = .70).
Manipulation check
The manipulation of diversity was checked using three items. Example items are “The members of my team are very different from each other” and “My team members are very much alike [reverse coded]” (M = 3.66; SD = 1.85, α = .91).
Results
The manipulation check showed that participants in the heterogeneous team condition (M = 5.51; SD = 1.43) indicated that the group was significantly more diverse than did participants in the homogeneous team condition (M = 3.04; SD = 1.34), F(1, 34) = 28,36, p < .001, η2 = 0.46, one-tailed.
In line with Hypothesis 1a, participants who were confronted with a diverse group (M = 5.93; SD = 0.76) expressed a stronger preference for considerate leadership compared to participants who were confronted with a homogeneous group (M = 5.31; SD = 0.45, F(1, 34) = 8.81, p = .003, η2 = 0.21, one-tailed).
Discussion
In this first study, we found that participants who were confronted with the idea of working in a generally diverse team, were more likely to prefer leader consideration than participants who anticipated working in a more homogeneous team. The results of this study provide first evidence for Hypothesis 1a. However, given our small sample size, we wanted to replicate and extend this study by examining the effects of another diversity characteristic (i.e., education diversity) among a bigger nonstudent sample. Additionally, we wanted to address the underlying process of the relationship between team diversity and leadership preference by including anticipated subgroup formation as a mediator.
Study 1b
In this study, 103 people participated (all employees; 61 men, 42 women; mean age = 39.13, SD = 13.24). All participants could win one of five lottery tickets (worth €20.00) if they completed the questionnaire. The participants read a scenario in which they had to imagine that they worked in either an educationally heterogeneous (N = 49) or educationally homogeneous team (N = 54; see Appendix B for the text of the scenario). Educational diversity is an often used diversity characteristic in diversity research (Milliken & Martins, 1996; van Knippenberg & Schippers, 2007) and all participants were likely to have experience with educational diversity. After reading the scenario, the participants indicated their preference for leader consideration. Additionally, they filled out a questionnaire regarding the amount of subgroup formation they anticipated within the team. We expected that participants would anticipate more subgroup formation in heterogeneous teams than in homogeneous teams, and that this would explain their preference for consideration (H1b).
We also assessed months of work experience (N = 5 shorter than 1 year; N = 98 longer than 1 year), type of organization (profit [N = 41] vs. nonprofit [N = 57]; unknown for five participants), educational level (university degree N = 31; college degree N = 36; professional training N = 20; high school N = 16) and nationality (N = 99 Dutch, N = 4 non-Dutch). Including these variables in the analyses did not alter our findings nor did they affect our outcome variables.
Measures
Preference for leader consideration
The preference for consideration was measured with the same questionnaire as in Study 1a (Indvik, 1985; Stogdill & Coons, 1957; all items assessed on a 7-point scale; M = 5.81; SD = 0.71, α = .78).
Anticipated subgroup formation
We measured anticipated subgroup formation in the team with two items adapted from Homan et al. (2010). The items are “Our team will consist of subgroups” and “The team will split up into smaller subgroups” (M = 3.54; SD = 1.20, r = .30, p < .001).
Manipulation check
In this study, we used two questions to check the adequacy of the diversity manipulation. Example items are “The members of my team are very different from each other in terms of educational background,” and “This team consists of people with different educational backgrounds” (M = 3.69; SD = 1.33, r = .44, p < .001).
Results
The manipulation check showed that participants in the educational heterogeneity condition (M = 4.67; SD = 0.88) indicated the group to be significantly more diverse than did participants in the educational homogeneity condition (M = 3.20; SD = 1.02, F(1, 101) = 61.02, p < .001, η2 = 0.38, one-tailed).
In line with Hypothesis 1a, participants faced with the prospect of working in an educationally diverse group (M = 5.97; SD = 0.62) indicated a stronger preference for consideration than participants faced with a group that was homogeneous on education (M = 5.55; SD = 0.78, F(1, 101) = 9.65, p = .002, η2 = 0.09, one-tailed).
Results for anticipated subgroup formation showed that participants who read that they would work in an educationally diverse team (M = 4.28; SD = 0.89) anticipated more subgroup formation than did participants in the homogeneous team condition (M = 3.46; SD = 0.73), F(1, 101) = 26.30, p < .001, η2 = 0.21, one-tailed).
To test for the possible mediating role of anticipated subgroup formation (H1b), we used Baron and Kenny’s three-steps approach (1986). 1 First, regression showed (in line with the ANOVA results) that diversity predicted preference for leader consideration (β = .29, p = .003) and anticipated subgroup formation (β = .17, p = .01). Second, we found that anticipated subgroup formation was related to preference for consideration (β = .46, p < .001). Third, we examined anticipated subgroup formation as a mediator in the relationship between diversity and preference for considerate leadership. The strength of the relationship between team diversity and preference for leader consideration was reduced (β = .23, p = .002) when the mediator, anticipated subgroup formation, was included (β = −.21, p = .005). The Sobel test for the indirect effect was significant (Sobel test z = 2.49, p = .01), indicating partial mediation and providing support for Hypothesis 1b.
Discussion
The results of Study 1b again indicate that people prefer their leader to show more consideration when they anticipate working in a diverse team than when they anticipate working in a homogeneous team (H1a). In this study, however, we obtained the effect for a more specific form of diversity rather than general diversity (Study 1a) and for a diversity dimension that has been associated more with positive rather than negative diversity effects (Horwitz & Horwitz, 2007), lending support for the generalizability of the finding. Moreover, we show that this preference for a considerate leader is driven by anticipated subgroup formation, lending support for Hypothesis 1b. Although these vignette studies provide experimental control in gaining knowledge regarding the preferences of people (both students and employees) who anticipate working in a diverse team, the situations are also quite artificial and do not speak to the actual impact of considerate leadership on the functioning of diverse teams. Therefore, in the next and final study, we examine whether team members’ preferences for a considerate leader were legitimate by examining the moderating influence of consideration on the relationship between diversity, diversity perceptions, and team functioning. More specifically, we investigate in a field study of teams and their leaders whether leader consideration affects the relationship between tenure diversity on the one hand and diversity perceptions and team functioning on the other hand.
Study 2: Effects of leader consideration
The findings of the previous two studies suggest that members of diverse teams prefer a considerate leader who focuses on improving positive relationships and trust. This underlines our theorizing that members of diverse teams want leaders who can manage the potential downsides of diversity (i.e., anticipated subgroup formation). In this study, we investigate whether these preferences for considerate leadership among members of diverse teams are legitimate. Specifically, we investigate whether leaders who show considerate behaviors actually lead diverse teams to function more effectively.
As the quality of group functioning in diverse work groups is likely to be reduced when diverse groups experience subgroup formation (e.g., van Knippenberg et al., 2004), we predict that leaders who exhibit high (compared to low) consideration will buffer the potential negative effects of diversity. This is because considerate leaders may reduce the likelihood of subgroup formation in diverse teams by influencing the perception of diversity within the team. As stated before, we argue that considerate leaders are capable of healing frictions and improving relationships among members of diverse teams, making it likely that diverse teams will experience less subgroup formation when their leader shows more consideration.
Additionally, we propose that diversity will be more positively related to leader individuation to the extent that the leader shows consideration. Considerate leaders acknowledge and appreciate individual feelings and ideas (Fleishman & Peters, 1962). This makes it likely that they will perceive their followers as unique individuals rather than as members of diversity-related subgroups. The tendency of considerate leaders to instigate individuation processes should especially be likely to become manifest in diverse teams, which are characterized by more differences between individual team members and therefore offer greater potential for individuation (Homan et al., 2010). This leader individuation will in turn affect how individuals within the team will behave and respond toward each other (Greer et al., 2012). Following the theoretical arguments regarding the value-in-diversity hypothesis, we propose that this individuation of the team members by the leader (Brewer & Miller, 1984; Wilder, 1978) will create the opportunity for teams to make adequate use of their diversity and to function effectively. In sum, we hypothesize that the relationships between team diversity on the one hand and team performance, experienced subgroup formation, and leader individuation on the other hand will be moderated by leader consideration in such a way that diversity is related to better performance, less subgroup formation, and more leader individuation to the extent that the leader scores higher on consideration (H2 and H3). Additionally, the interaction between leader consideration and team diversity on team performance will be mediated by perceived subgroup formation and leader individuation (H4a and 4b).
Participants
Six hundred and seventeen people in 96 retail outlets of a large services organization with offices in the Netherlands participated in this survey study (61% men; mean age = 34.43, SD = 2.91, 10% of employees were of an ethnic minority). Each retail outlet employed three to nine employees (M = 3.67, SD = 1.29), all of whom worked together closely, had a common goal, and identified themselves as a team. More specifically, the teams worked on complex tasks such as tailor-matching financial products to specific customer needs. We therefore decided to treat each retail outlet as a unique team (Hackman, 1987). Only teams of which we had a response rate of at least two team members were included in the analyses (the organization had 223 branch offices in total; 127 teams were excluded because of having one or less response per team, resulting in our final sample of 96 retail outlets). Within that final sample, we had a response rate of 66% (374 of 569). Teams had on average 5.50 members, and on average we had 3.50 responses per team. The distribution of within-group response was between two and nine responses. We had full response from 10% of teams, whereas 8% of teams only had two responses. This implies that most teams (87%) had on average 3–4 responses per team. Due to some nonresponse of the leaders in our sample, we have a sample of 74 teams for the leader-rated variable individuation, and a sample of 59 teams for leader-rated performance quality. In return for their participation, teams received individualized team reports.
Procedure
To assess the variables in our study, we had access to multisource data, including surveys distributed to both leaders as well as employees (leader consideration and subgroup formation as assessed by team members, individuation of team members and team performance quality as reported by the leader). All questionnaires had response scales ranging from 1 = completely disagree to 7 = completely agree.
Measures
Tenure diversity
In this study, we focused on tenure diversity. As stated before, we focused on tenure diversity because it can potentially be related to positive as well as negative team outcomes (Ancona & Caldwell, 1992; Mannix & Neale, 2005; Milliken & Martins, 1996; Murray, 1989; Pelled et al., 1999; Sessa & Jackson, 1995). Moreover, tenure diversity is interesting in this specific research setting, because it may aid teams to perform their tasks better. People who have been working at the retail outlet for a longer period of time have a lot of experience and can quickly oversee the possible products that would fit a certain client’s wishes. New employees, on the other hand, have more knowledge regarding new techniques and insights in the latest developments. Diversity of the teams on tenure was calculated by means of the coefficient of variation—the standard deviation in the number of years members had worked in the team, divided by the mean (Harrison & Klein, 2007).
Leader consideration
Perceived leader consideration was assessed among the team members with the same questionnaire as in Study 1a and 1b (α = .93).
Subgroup formation
We measured perceived subgroup formation in the team with three items from Jehn and Bezrukova (2010; see also Homan et al., 2010). For example, an item was “During our work, our team splits up into smaller subgroups” (α = .86).
Leader individuation
We asked the team leaders to which degree they perceived their work group to consist of unique individuals using three questions: “This team consists of unique individuals,” “I see this team in terms of different individuals,” and “No one is exactly the same as the others in this team” (α = .77).
Team performance quality
We asked the team leaders to indicate their opinion of their team’s performance quality with four questions: “This team delivers good products/services,” “This team finishes the job on time,” “This team does its very best to obtain and maintain high quality work,” and “I am satisfied with the outcomes of this team” (α = .82).
Data treatment
We controlled for sex and ethnic diversity (both calculated using Blau’s index, which is used when a variable V is dispersed across members who might be in one of k = 1,… K possible categories, which is the case for variables such as ethnicity or sex [Blau, 1977; Bunderson & Sutcliffe, 2002; see Harrison & Klein, 2007, for the specific formula]), mean age of the team members, team size, and educational level. Because the company had been through a merger recently, we also controlled for whether the teams contained employees from only the original company (coded by a 0) or from both of the premerger companies (coded by a 1).
Regression analyses were used to test the interaction between group tenure diversity and leader consideration on perceived subgroup formation, leader-rated individuation, and leader-rated team performance quality. Both predictors were centered and their interaction term was computed based on these centered variables. Because all our variables are team-level variables, and also because our dependent variable was a team-level variable (i.e., leader-rated team performance quality), we needed to analyze our data on the group level. Indices of interrater reliability and within-group agreement for leader consideration (ICC[1] = .14; F(67, 547) = 1.63, p < .001; ICC[2] = .14; rwg(j) = .75) and subgroup formation (ICC[1] = .10; F(67, 547) = 1.78, p = .02; ICC[2] = .44; rwg(j) = .80) were sufficient to support the aggregation of variables measured on the individual level to the group level (Bliese, 2000). 2
Results
The means, standard deviations, and correlations of the variables of interest are given in Table 1.
Means, standard deviations, and correlations of the main variables in Study 2
Note: N = 59 (for leader-rated team performance quality), N = 74 (for leader individuation), and N = 96 (for team tenure diversity, leader consideration, and subgroup formation).
p < .05; **p < .01.
Team performance quality
Tenure diversity alone was not related to perceptions of team performance quality, but leader consideration was positively related to team performance quality (see Table 2). More importantly, we found support for an interaction between consideration and tenure diversity (see Table 2 and Figure 2). In line with Hypothesis 2a, simple slopes analyses (Aiken & West, 1991) showed that tenure diversity was not related to team performance when the leader showed lower levels of consideration (β = −.27; p = .10, one-tailed), but that there was a positive relationship between tenure diversity and team performance when leader consideration was rated higher (β = .38; p = .04, one-tailed).
Results of hierarchical regression analyses of Study 2
Note: N = 96 (subgroup formation), N = 74 (leader individuation), and N = 59 (team performance quality).
p < .05; **p < .01 (one-tailed).
The interactive effect of tenure diversity and leader consideration on team performance quality.
Subgroup formation
Leader consideration and tenure diversity alone were not related to perceptions of subgroup formation among team members (see Table 2). In support of Hypothesis 2b, we found a marginally significant interaction between tenure diversity and leader consideration on subgroup formation. Figure 3 visualizes the interaction. Simple slopes analyses (Aiken & West, 1991) showed that tenure diversity was positively related to subgroup formation when the leader showed lower levels of consideration (β = .27; p = .04, one-tailed), and that there was no significant relationship between tenure diversity and subgroup formation when consideration was rated higher (β = −.17; p = .15, one-tailed).
The interactive effect of tenure diversity and leader consideration on subgroup formation.
Leader individuation
Again, there was no significant relationship between leader consideration or tenure diversity and the degree to which the leader perceived the team to consist of unique individuals (see Table 2). We did find an interaction between consideration and tenure diversity on leader individuation (see Figure 4). Simple slopes analyses showed that tenure diversity was positively related to leader individuation in teams with more considerate leaders (β = .56; p = .01, one-tailed). Tenure diversity was not related to leader individuation in teams with less considerate leaders (β = −.08; p = .35, one-tailed). These results support Hypothesis 3.
The interactive effect of tenure diversity and leader consideration on leader individuation.
Mediation
We predicted that perceived subgroup formation (H4a) and leader individuation (H4b) would mediate the interactive effect of tenure diversity and consideration on team performance—a mediated-moderation model (Muller, Judd, & Yzerbyt, 2005). We found that leader individuation was related to team performance (β = .41, p = .002) and that subgroup formation was not related (β = −.21, p = .12) to team performance. Hypothesis 4a could therefore not be confirmed, as the proposed mediator subgroup formation was not related to the dependent variable team performance quality. We therefore proceeded to test for the mediating role of leader individuation (see Table 2, Step 2a). The strength of the relationship between the interaction between team tenure diversity and consideration on team performance quality was reduced (from β = .30, p = .04 to β = .19, p = .17) when the potential mediator, leader individuation, was included. In line with Hypothesis 4b, we found a significant effect of leader individuation on team performance quality (β = .41, p = .001). A Sobel test indicated that the indirect effect was significant (Sobel’s z = 1.70, p = .004, one-tailed), indicating full mediation. 3 These findings lend support for our Hypothesis 4b that leaders who exhibit considerate leadership individuate their team members in diverse teams more and thereby improve team performance.
Discussion
The results of Study 2 provide partial support for our idea that leader consideration can positively affect the performance of tenure-diverse teams by influencing the way in which diversity is perceived. We found that leaders who were observed to be considerate positively influenced the relationship between tenure diversity on the one hand and subgroup formation and leader individuation on the other hand. Tenure diversity was related to less subgroup formation (H2b) and more individuation of the team members (H3) to the extent that the leader scored higher on consideration. Additionally, we found that tenure diversity was positively related to team performance when leaders were rated higher on consideration (H2a). This relationship was mediated by leader individuation (H4b).
General discussion
Although team diversity has received much research attention in the last decades (see e.g., van Knippenberg & Schippers, 2007), research on the moderating role of leadership in diverse teams is still in its infancy. We add to this new area of research by focusing on the preference for and potential beneficial effects of leader consideration in diverse teams, and how this type of leadership may affect perceptions of diversity within diverse teams. We proposed and found that considerate leadership (which is characterized by healing relationships, solving conflicts, increasing trust) is especially preferred by, and beneficial for, diverse groups. Additionally, we identified a mechanism by which considerate leaders are helpful to diverse groups—namely, heightened leader individuation of team members.
In our first set of studies, we focused on how the anticipation of working in a diverse team would influence leadership preferences. In two scenario studies, people were presented with either a diverse or homogeneous team and indicated their preferences for a leader showing leader consideration. In the first study we operationalized diversity as “general diversity” and in the second study as educational diversity. Both studies showed that people preferred consideration more when they anticipated working with a heterogeneous rather than homogeneous team (H1a). Additionally, in Study 1b, educational diversity (compared to educational homogeneity) was related to more anticipated subgroup formation. Anticipated subgroup formation mediated the relationship between diversity and preference for leader consideration, explaining why members of diverse groups preferred this type of leadership more (H1b). They anticipated difficulties in the relationships within the group, and they desired a leader who would address these problems. The fact that we found similar results for different operationalizations of diversity (i.e., general and educational diversity) boosts our confidence in the generalizability of the effects.
In Study 2, we set out to examine whether this heightened preference for a considerate leader when anticipating working with diverse team members would translate into actual improved team performance when diverse teams are being led by a considerate leader. In other words, we wanted to know whether these leadership preferences actually make sense. The potential beneficial effects of leader consideration were examined by means of a survey filled out by teams and their leaders. To illuminate the potential underlying mechanism of this potential benefit, we examined how consideration influenced the diversity perception of team members and team leaders, and how this perception would in turn be related to team functioning. Results showed that consideration was related to better functioning in tenure-diverse teams as well as related to less subgroup formation within the team, providing support for Hypotheses 2a and 2b. Moreover, we found that considerate leaders perceived the team members more as unique individuals to the degree that the team was more tenure diverse (H3). For Hypothesis 4a, we found no evidence for a mediating role of subgroup formation in Study 2. Although subgroup formation was negatively related to leader-rated performance quality, the correlation failed to reach statistical significance. In the present sample, leader individuation seemed to be more strongly related to team functioning than subgroup formation. Because many diversity researchers have put subgroup categorization and subgroup formation at the center of potential clarifying mediators (e.g., van Knippenberg et al., 2004; Williams & O’Reilly, 1998), future research should set out to examine the mediating role of subgroup formation further. The significant negative correlation between perceived subgroup formation and leader individuation does indicate that subgroup formation was related to the degree to which the leader individuated team members. In line with Hypothesis 4b, we found that the interactive effect of tenure diversity and consideration on performance was mediated by the tendency of the leader to see the team members in terms of unique individuals. This finding is in line with research by Homan et al. (2010) who proposed that perceiving unique individuals rather than subgroups in diverse teams is a prerequisite for obtaining the benefits in diversity. In sum, the two sets of studies paint a complementary picture of the role of considerate leadership in diverse teams. The preference that participants have for a considerate leader when being confronted with a (educationally) diverse team indeed seems to translate into better functioning for tenure-diverse teams that are being led by a considerate leader.
Together, our studies support a functional/contingency approach to leadership by showing that depending on the situation, different leadership behaviors may be required (Fiedler, 1965; Kerr et al., 1974; McGrath, 1962; Morgeson, DeRue, & Karam, 2010). Previous work on diversity has illuminated that diversity can instigate negative group processes within teams due to subgroup formation processes (e.g., van Knippenberg et al., 2004; Wiliams & O’Reilly, 1998). These negative processes can disrupt positive processes that are associated with diversity, such as the elaboration of diverse pieces of information (van Knippenberg et al., 2004). This suggests that diverse teams need leaders who address and manage these categorization processes. The present results indeed indicate that this is the case. Both subjective preferences of participants who anticipate working in a diverse team and actual experiences of members of existing diverse workgroups indicate that consideration is more preferred and more effective in diverse teams. Considerate leadership helps to counteract subgroup formation, a potentially detrimental group process that needs attention in heterogeneous teams. Additionally, we are the first to show that leader consideration affects the way the leader perceives the team’s diversity. Leaders who show more consideration pay more attention to personal ideas and feelings and perceive the team members more in terms of unique individuals. Study 2 showed that such individuation was related to better team performance.
These findings also have relevant practical implications for organizations. Based on our results we argue that diverse teams require leaders to show more considerate leadership, thereby improving relationships within the team, valuing members’ unique differences, and reducing subgroup formation. This will not only help to dampen the potential negative effects of diversity (van Knippenberg et al., 2004), but will also improve employees’ responses to working with these leaders because they fulfill their needs in a diverse team (Ehrhardt & Klein, 2001; see also Brief, 1998; Eagly & Chaiken, 1993). This in turn can lead to more motivated and better performing subordinates (e.g., Miner, 1978).
Future research and limitations
We show across three studies, involving different samples and operationalizations of diversity, that leader consideration is preferred by and beneficial for diverse teams. Although we focused on only three operationalizations of diversity, that is general diversity, educational diversity, and tenure diversity, we do find compatible results over studies, which strengthen our belief that the results are not limited to one specific dimension of diversity. Nevertheless, future research is needed to test our hypotheses with other diversity dimensions.
With this paper we contribute to the growing interest on leadership as a moderator of diversity effects (see e.g., Greer et al., 2012; Kearney & Gebert, 2009; Klein et al., 2011; Shin & Zhou, 2007; Somech, 2006; Stewart & Johnson, 2009). Because we were interested in when leaders can minimize the dark side of diversity, and bring out the positive side, we focused on more complex tasks in our sample, as diversity is most likely to benefit performance on such tasks (including tasks such as tailor-matching financial products to specific customer needs and writing a detailed reorganization plan; e.g., Bowers, Pharmer, & Salas, 2000; Jehn et al., 1999; see also van Knippenberg et al., 2004). In light of previous theorizing (van Knippenberg et al., 2004) as well as research by Homan et al. (2010) indicating that task requirements are a critical moderator of the effects of diversity on group functioning, it would be important for future research to test whether our results generalize to less complex tasks. Indeed, when leaders focus on diversity, individual differences, and relationships and make it possible to discuss differences, they might instigate processes that are harmful for teams that do not need diversity to perform their tasks (e.g., simple, physical tasks; Homan et al., 2010). Therefore, an empirical investigation of potential boundary conditions of the effects of consideration in teams, such as task type, is crucial.
There are some limitations to our findings. One, although our experimental approach in Studies 1a and 1b increases confidence in the internal validity of our conclusions, we cannot be sure that our results are generalizable to preferences of members working in actual diverse and nondiverse teams. Future research could explore leadership preferences in actual teams. Additionally, although our research methodology of Study 2 made use of multiple sources (i.e., objective information regarding tenure diversity based on archival data, team member-rated and leader-rated variables), the relationships are still based on correlational data. Future research could set out to test these relationships in a more controlled, experimental setting in order to speak to causality. Furthermore, in Study 2, our dependent variable, team performance quality, was subjectively determined by the supervisor. Although this is an often used operationalization of team performance (e.g., Kearney & Gebert, 2009) that reduces common-method bias, future research should include more objective performance measures when available, such as the financial performance of teams. Also, Study 2 had a moderately high nonresponse rate, which might have affected the strength and direction of our effects (Maloney, Johnson, & Zellmer-Bruhn, 2010). However, we have no reason to suspect that this nonresponse was systematic as opposed to random (Schafer & Graham, 2002). Indeed, a comparison of our sample with the organization’s population showed that it did not differ in terms of sex, age, ethnicity, educational background, or tenure, which alleviates concerns regarding nonresponse (Newman & Sin, 2009; Schafer & Graham, 2002).
As put forward in the first part of the paper, we believe that the leadership consideration concept utilized in our paper zooms in on a different type of person-focused leadership than individualized consideration or LMX (Bass, 1999). However, as there are also similarities between the different leadership constructs, it would be interesting to examine these in future research. Whether and how the leadership frameworks can be integrated is still open to empirical research (e.g., Graen & Uhl-Bien, 1995; Judge et al., 2004), and this question was beyond the scope of the present paper. Finally, future research could set out to examine whether these constructs show similar or different effects in diverse teams, and whether there is overlap in preferences for the different leadership styles among people working in diverse teams.
Finally, we did not focus on initiating structure (i.e., task-oriented leadership), although several authors have proposed that both types of leader styles can go together (e.g., Judge et al., 2004; Klein et al., 2011; Stogdill & Coons, 1957), meaning that leaders can score high or low on both initiating structure and consideration (Yukl, 2010). Relatedly, our findings do not speak to the question of whether diverse teams might benefit similarly from task-oriented leadership. Interestingly, recent research by Klein et al. (2011) found that diverse teams benefitted from task-focused leadership and that relationship-oriented leadership harmed rather than benefited these diverse groups. They argued that relationship-oriented leadership creates low-structured situations and gives room for discussion and conflicts. One possible explanation for these contrasting effects lies in the task type investigated in the Klein et al. (2011) study, which was perhaps not as conducive to the beneficial effect of diversity (tasks were more routine, including clearing trails in the community and distributing food) as the more complex tasks in our sample were (which include tasks such as tailor-matching financial products to specific customer needs; e.g., Bowers et al., 2000; Jehn et al., 1999; van Knippenberg et al., 2004). Along similar lines, Gratton, Voigt, and Erickson (2007) proposed that leaders show task-oriented leadership in the beginning of a diverse team’s life, because this task-orientation focuses the attention on the performance and requirements of the teams. After some time, leaders should then switch to relationship-focused leadership to address the tensions that arise due to deep-level diversity dimensions. In sum, future research should examine the role of initiating structure separately and in combination with consideration in order to address these issues in greater detail.
Conclusion
In sum, we believe that the finding that considerate leaders are preferred by, and benefit the functioning of, diverse teams might set the stage for future research on the crucial role of leadership in diverse teams. Our findings do not only provide important new insights in the role of leadership in diverse teams, they also provide an important qualification of previous work and suggest that the exploration of the best possible ways to lead diverse teams has only just begun.
Footnotes
Appendix A
Scenario Study 1a: Homogeneous/heterogeneous on general diversity.
Below, you will find a description of a hypothetical situation. Please assume that you are the person at the center of this situation. While reading about the situation, try to imagine yourself in this situation as vividly as you can:
Since a couple of months, you are a member of a team which develops advertising campaigns. The members of the team have to work together a lot. The team consists of 10 members and a supervisor. You experience that the team members are
As a team you are responsible for the communication with customers, developing new campaigns, work on graphics and copyright, and the practical side of the work.
Given the situation just described, please rate each of the statements below [not provided here] as to how much it is true for you being present in this situation.
Appendix B
Scenario Study 1b: Homogeneous/heterogeneous on educational diversity.
Below, you will find a description of a hypothetical situation. Please assume that you are the person at the center of this situation. While reading about the situation, try to imagine yourself in this situation as vividly as you can:
You work at a medium-sized consultancy firm in an urban area. A characteristic of this consultancy firm is that they work with project teams. At this time, the team in which you work has been hired by an external company to help develop and manage their reorganization. This external company is in a crisis at the moment. When they do not thoroughly restructure the organization, they will need to fire people and might even go bankrupt. Outlining a thorough and good reorganization proposal is thus very important.
Your project team consists of eight persons. All team members have a
Given the situation just described, please rate each of the statements below [not provided here] as to how much it is true for you being present in this situation.