Can a team have too much cohesion? The dark side to network density

Highlights

• Study empirically finds an inverse curvilinear relationship between team performance and group cohesion.

• Study finds that top performing teams share similar social network topology.

• Study finds that poor performing teams share similar social network topology.

• Study finds vastly different social network topology between poor performing teams and high performing teams.

Abstract

The goal of most work teams is high performance. Prior studies suggest that performance within work groups is influenced by that group’s social network topology. Research has generally revealed to date that group cohesion (i.e., network density) is positively related to team performance under certain conditions. However, more recent research has indicated that this is not the full story. Recent research suggests that an inverse curvilinear relationship exists between social network measures (of which group cohesion is one) and team performance. In response to the need for understanding this relationship more fully, and leveraging the promising new insights that can be garnered with the use of social network analysis (SNA), this study employs SNA as a tool to explore the structural cohesiveness of teams of travel agents. This research extends our understanding of the relationship between intragroup social network relations and team performance by confirming an inverse curvilinear relationship exists between group cohesion and team performance. This paper leverages email communication to determine the social networks of each team, and then examines such in light of team performance. In total, an analysis of more than 7 million emails was undertaken. This study was conducted with work teams within a service organization. Each team in the study carries out the same tasks, i.e., identical task contingency, yet represents a distinct unit of analysis. The study confirms that social network topology is a valuable predictor of team performance and confirms that, like so many other social network measures, group cohesion and team performance share an inverse ‘U’ shaped relationship, not strictly a positive one as previously posited.

Introduction

Despite the increase of teams and work groups within an organization there has been relatively little social network research on the structural properties of work groups and their consequences for team performance (Cummings and Cross, 2003, Lechner et al., 2010). This research attempts to address this deficiency. This research focuses on network topology, in particular the measure of group cohesion, and how such impacts team performance. The Theory of Task Contingency (Donaldson, 2001) postulates that some network topologies are better suited to exploration practices, while others are more suited to exploitation practices. That research finds that social network topology can be optimized to generate performance gains (Donaldson, 2001, p. 2). The Theory of Task Contingency (Donaldson, 2001) was extended by Lechner et al. (2010) to create the Dark Side of Social Capital theory. Under the Dark Side of Social Capital theory (Lechner et al., 2010) too much of any social network measure is as bad as too little. Put more explicitly, the same network structures that help some teams achieve fit with their environment and accomplish goals reach a point of diminishing returns, after which increasing that social measure further leads to negative performance consequences for groups (Lechner et al., 2010). Fig. 1 below illustrates some of the negative and positive influences of intergroup relations on performance.

This research seeks to explore the relationship between a group’s cohesion level and that team’s financial performance, while holding task contingency constant across teams and controlling for team size. By doing so, this research attempts to determine if the relationship between group cohesion and team performance is always positive as found by prior scholars (Beal et al., 2003, Carron et al., 1998, Evans and Dion, 2012) or if the relationship between these variables is actually inversely curvilinear and as found by Hardy, Eys, and Carron (2005) and as first suggested by Carron, Prapavessis, and Grove (1994). Further, this research seeks to affirm that Lechner et al. (2010) Dark Side of Social Capital theory holds true for group cohesion and team performance.