A. Contributions to the Study of Leadership in Small Groups

The main theoretical and empirical themes of my research have revolved around, 1) understanding the impact of personality or individual differences on team interactions and performance, 2) managing conflict in teams, both minimizing the potential harm in escalating conflict as well as harnessing the positive impact of debate, and 3) investigating the special problems of leading top management teams. 

1. The Effects of Individual Differences. 
2. Conflict Management.
3. Top Management Teams.

1. The Effects of Individual Differences. This stream of my research is the longest running and dates back to my very earliest research working with Phil Tetlock on how integrative complexity as a personality variable predicts perceptions of managerial competence and leadership potential (Tetlock, Peterson, & Berry, 1994). Having sparked my interest in issues of personality, I began to think about how these differences in personality reflect different approaches to teamwork. That began a collaborative relationship with (then Ph.D. student) Lisa Moynihan that resulted in two theoretical publications promoting what we call the contingent configurational approach (Moynihan & Peterson, 2001, 2004). In those papers, we argue that past research has taken one of three basic theoretical approaches to explaining the nature of member personality effects on group process and team performance, (1) universal -- certain traits always predict teamwork success, (2) contingent -- certain traits predict team performance depending on the task or organizational culture, and (3) configuration -- the mix of traits within a group, or the "fit" of individual members with each other, predicts team performance. Each of these three approaches to personality in groups has received significant empirical support in the literature and yet has some shortcomings. We offer suggestions for improving research on the role of personality in group decision making using each approach, but argue that a full understanding of the role of personality in group processes must integrate all three of these approaches into what we call the contingent configuration approach. Lisa Moynihan and I are now in the process of empirically testing our theoretical model by creating a large database with all of the core teams at London Business School over the past two years including their scores on the NEO personality inventory and their performance on a variety of group tasks (e.g., case writing grades, a creative task, etc.). No one we know of has a comparably large and comprehensive database in order to test these ideas, so we look forward to results very soon. Additionally, this work has had practical implications for pedagogy at London Business School in how we design our MBA curriculum as I will discuss in subsequent parts of this statement.
back to top

2. Conflict Management. Conflict is probably the single most important dynamic for groups to get "right" if they are to achieve peak performance. On the positive side, conflict promises great potential benefits if debate and exchange of ideas is well-managed (Nemeth, 1986; Peterson & Nemeth, 1996). Much of my early work looked at the positive aspects of conflict to document, a) that exposure to different points of view encourages divergent thought-expanding cognitive processes (Peterson & Nemeth, 1996), and b) allowing a full discussion encourages buy-in from the group and better quality decision-making (Peterson, 1997). However, there is a limit to these effects such that a full consensus model where anyone in the group can stop the group from making a decision can actually lead to problems under conditions of persistent or deeply held conflict (Peterson, 1999). This implies some of the downside of conflict suggesting that it is also the source of some of the most poisonous behavior in groups that distracts the group focus from making good decisions, demoralizes group members, and destroys creativity (Jehn, 1995). Here I have also looked at the destructive effects of relationship-based or affective conflict (e.g., Peterson & Behfar, 2003). In particular, I have argued and found support for the notion that the field has causality reversed because interpersonal conflict is a result of performance feedback processes in groups such that poor performance leads to interpersonal conflict, rather than interpersonal conflict leading to poor group performance.

The key leadership challenge in managing conflict is to understand the different types of conflict and to manage each appropriately. The first type is relationship or affective conflict. Relationship conflict is defined as an awareness of interpersonal incompatibilities, including affective components such as feeling tension and friction between group members and is associated with a sense of personal disliking. Empirical studies tend to show that relationship conflict is reliably associated with lower individual and group performance, member satisfaction, and the likelihood the group will work together effectively in the future (De Dreu & Weingart, 2003; Jehn, 1995).

Task or cognitive conflict is an awareness of differences in viewpoints and opinions pertaining to the group's task (Jehn, 1995). Task conflict under certain conditions has historically been considered beneficial to group performance and satisfaction (De Dreu & Weingart, 2003). Task conflict has specifically been proposed as a key source of divergent thinking, encouraging use of unique information, and pooling of resources to create a better solution (Nemeth, 1986; Peterson & Nemeth, 1996).

Process conflict is defined as an awareness of controversies about aspects of how task accomplishment should proceed, how to delegate work assignments, and who has responsibility for different group tasks (Behfar, Peterson, Mannix, & Trochim, 2004; Jehn, 1997). Teams scholars have long argued that process issues such as agreeing on a common approach, developing work plans, and dividing responsibility are essential to a group being able to perform well and avoid dysfunctional conflict (Pondy, 1967; Porter & Lilly, 1996). Despite its theoretical importance, process conflict is the least well examined and understood of the three types of conflict.

Some of my most recent work with my Ph.D. student from Cornell, Kristin Behfar, attempts to refine the construct validity and measurement of the tripartite model of group conflict (Behfar, Mannix, Peterson, & Trochim, 2004). This paper is aimed at better understanding the meaning and measurement of task, relationship, and process conflict, with a particular focus on refining our understanding of process conflict. We present a series of studies, using different methodologies, which indicate that the three types of conflict are more complex in their measurement and explication than was originally proposed by Jehn (1995) (e.g., task and process conflict rarely factor in their original proposed form). This paper presents revised task, relationship, and process conflict scales with more clearly specified definitions and improved divergent validity from each other.
back to top

3. Top Management Teams. The third major stream of my research looks at leadership in small teams, specifically focused on the special challenges of top management teams. This research stream integrates across the other two streams with different studies that look at both conflict management in top management teams (Simons & Peterson, 2000) and examining the role of chief executive officer personality in forming top management team dynamics and firm performance (Peterson, Smith, Martorana, & Owens, 2003). However, this stream makes a distinct additional contribution from my other work because the area of top management teams speaks to different academic audiences (e.g., strategy process) from the other work I do.

I have made two quite distinct contributions to scholarship in the top management teams area. First, in my collaborations with Tony Simons , I looked at the key role of intragroup trust in top management teams (Peterson, Simons, & Rodgers, 2004; Simons & Peterson, 2000). Here the focus was on whether task and relationship conflict in top management teams have the effects found in other kinds of teams. We were also interested in whether task conflict promotes good quality decision-making, and whether relationship conflict is as destructive as hypothesized. Here when looking back at past research on the topic we discovered that the two conflict types are typically positively correlated in ongoing groups, despite their opposite hypothesized relationships with performance. This creates a prescriptive dilemma for top management teams because CEOs intending to harness the positive effects of task conflict may actually trigger the negative effects of relationship conflict. We tested two different hypotheses about the nature of this relationship with survey data from 79 ongoing top management teams. The first hypothesis was that the problem is behavioral such that people behave in inappropriate ways when debating and thus trigger relationship conflict, and the second hypothesis proposed that people often misinterpret task conflict as relationship conflict. We found that high intragroup trust severs the relationship between task conflict and relationship conflict regardless of how top management team members behave, which provides evidence that interpretation processes are key to this phenomenon. Thus trust appears to be a key to gaining the benefits of task conflict without suffering the costs of relationship conflict in top management teams. In a new collaboration (Peterson, Simons, & Rodgers, 2004), we have been looking in greater detail at this prescriptive dilemma - extending our original analysis by testing whether the negative effects typically associated with a lack of trust in top management teams can be attenuated through the use of a consensus decision rule. We hypothesize that the unanimity required in using consensus grants every group member veto power, forcing group members to work from a collaborative mindset by hearing the opinions of all group members (i.e., giving them voice).

The second major contribution I have made to the top management teams literature is to clearly specify the connection between the personality of the chief executive officer and firm performance (Peterson, Smith, Martorana, & Owens, 2003). The literature on leader and chief executive personality has a long and controversial history, with a wide range of opinions represented. My argument, put simply, is that any effects of chief executive officer personality must be mediated through the top management team rather than having a direct unspecified effect on the larger organization. This paper specifically explores the mechanism by which leader personality affects organizational performance. Results from independently observed measures of chief executive officer (CEO) personality and TMT dynamics supported our hypothesized relationships both between CEO personality and TMT group dynamics, as well as between TMT dynamics and organizational performance. This empirical work complements the theoretical work with Lisa Moynihan on the role of personality in teams (e.g., Moynihan & Peterson, 2001).

This integrative work on top management teams (including my early work on elite political groups) has emerged in recent years as the work that has generated the most interest from both the scholarly community (i.e., my most highly cited work) and the practitioner community (e.g., my article, derived from my academic work on this topic, that appeared in the Financial Times has generated considerable executive education interest). I plan to continue this area as a primary area of concentration. I have been focussing on developing a theoretical perspective on groups and top management teams that suggests that establishing legitimacy is the primary driver of organizational team effectiveness (e.g., Peterson, Ronson, Anand, & Rodgers, 2004). In this work, we argue that effective organizational teams are embedded within a social structure that requires them to establish two, sometimes conflicting, purposes. The first and widely studied purpose is in seeing groups as information processors, suggesting the need for cognitive legitimacy (i.e., making a sensible decision). The second less studied purpose is for groups to be seen as embedded political agents in need of establishing socio-political legitimacy (i.e., managing the expectations of others). This perspective suggests, for example, that organizational teams have three distinct audiences to please, a) an internal audience that demands the team persuade members to commit to investing in the group, b) an audience outside the group but internal to the organization who are looking for actions consistent with organizational norms, and c) an audience external to the organization for which cognitive diversity is essential to persuading and influencing them.
back to top