The effectiveness of teams has been of interest to researchers and practitioners alike, and a substantial body of research exists on the topic (Mathieu, Maynard, Rapp, & Gilson, 2008). Team effectiveness models suggest that the right mix of team members is an important enabling condition for team effectiveness (Hackman, 2002). Team composition research considers how the configuration (e.g., team-level diversity) of team membersâ attributes (e.g., personality, values, demographics) shapes team dynamics and performance (Bell, 2007).
Teams are not static entities: they are shaped by a number of temporal influences (Mathieu, Kukenberger, & DâInnocenzo, 2014). Researchers consistently call for the integration of temporal influences in team effectiveness research in general, and team composition research more specifically (Mathieu, Tannenbaum, Donsbach, & Alliger, 2014; Mohammed, Hamilton, & Lim, 2009). Mathieu and colleagues (2014a) summarize the implications of time for teamwork: time can serve as a historic context within which teams operate, time relates to cyclical phenomena as teams perform different activities in a performance episode, and time can be a developmental marker signaling how teams move through a lifecycle from birth to death. Each of these has implications for understanding and managing team composition.
First, teams operate within a specific context which, among other things, shapes the extent to which their composition is dynamic over time. Organizations increasingly rely on team-based work structures that are frequently reconfigured, have fluid boundaries, and have members that are assigned to multiple teams (Tannenbaum, Mathieu, Salas, & Cohen, 2012). Second, teams engage in recurring cyclical processes over time related to goal accomplishment (Marks, Mathieu, & Zacarro, 2001). Team composition can shape the team processes, emergence processes, and emergent properties, as well as how team processes and emergent properties relate to performance over time. Third, teams move through a lifecycle from birth to death. Team composition can inform team design (e.g., staffing) and management across the lifecycle (e.g., training, leadership priorities). Indeed, time is important for understanding and managing team composition.
The focus of this chapter is a few key issues related to understanding and managing team composition over time. We briefly describe team composition. Then, we organize our chapter around the implications of time for teamwork as noted by Mathieu et al. (2014a). We discuss how the context shapes team composition with a focus on the three contemporary issues that affect the nature of team composition: membership change, fluid team boundaries, and multiple team membership. Next, we describe empirical research that examined how team composition shapes team processes and emergent properties over time. Finally, we describe how team composition variables can be managed over the lifecycle of a team.
Team Composition
In order to utilize team composition in team design and other interventions (e.g., training), specific attributes (e.g., personality, values, abilities, demographics) need to be identified as well as the specific unit-level configurations (e.g., team-level diversity, average of the strategic core) on the attributes that relate to effectiveness in the specific context (e.g., team performance, time-to-market, sales). Team composition variables can include knowledge, skills, abilities, and other characteristics (KSAOs) of team members. Often the focus of team composition research is on relatively enduring team member characteristics (e.g., demographics, abilities, personality) or knowledge and skill sets that are difficult or time-consuming to train (e.g., professional background). Surface-level composition variables are overt characteristics of a team member that can be reasonably estimated after brief exposure to the team member; examples include age, race, and sex (Bell, 2007; Harrison, Price, & Bell, 1998). Often surface-level composition variables operate through stereotypes, assumptions, or attraction to similar others. The effects of deep-level composition variables emerge as team members interact. Deep-level composition variables are underlying psychological characteristics that shape an individualâs affect, thinking, and characteristic patterns of behavior (Bell, 2007); examples include personality traits, values, work styles, and abilities.
In general, deep-level composition variables such as personality traits and values have stronger effects on performance than surface-level variables (Bell, 2007; Bell, Villado, Lukasik, Belau, & Briggs, 2011). Surface-level composition variables often have small or negligible effects; however, their importance can increase in specific circumstances. As examples, surface-level diversity can be related to performance when faultlines (i.e., hypothetical divides between team members based on one or more attributes) are activated (Lau & Murnighan, 1998), or if the broader organizational or industry context brings emphasis to demographic differences (Joshi & Roh, 2009).
With team composition, the combination of characteristics across team members is of interest. In their review and integrative framework, Mathieu and colleagues (2014b) organize the team composition literature into individual-based composition models and team-based composition models as described next. The different unit-level team composition operationalizations (e.g., team mean, diversity) can be understood within these models. Individual-based models focus on the fit between individualsâ KSAOs and the positions or roles they will occupy (i.e., traditional personnel-position fit models; Mathieu et al., 2014b). The personnel-position fit model can be extended to include teamwork considerations such as role knowledge or generic teamwork skills (called personnel model with teamwork considerations; Mathieu et al., 2014b). With individual-based composition models, teams are expected to be more effective when they are composed of members with higher levels of advantageous KSAOs. As an example, conscientious individuals are described as hardworking, achievement-oriented, and persevering. A team may be better performing when it is composed of conscientious team members. With individual-based models, team composition is operationalized as the aggregate of individual-level attributes (e.g., team mean conscientiousness).
With team-based composition models, the value of a team memberâs standing on a characteristic(s) is relative to: (a) other team membersâ standings on the characteristic(s), or (b) the team memberâs position in the team. For example, a team may be more cohesive when team members are complementary on an attribute (e.g., one team member higher in dominance and the other team members lower on dominance) or similar on an attribute (e.g., shared values). Team-based composition models include team profile models and relative contribution models. With team profile models, the distribution of attributes across the team is important (Mathieu et al., 2014b). Example operationalizations include team diversity on a specific attribute, or the faultlines that develop across multiple attributes (Lau & Murnighan, 1998). With the relative contribution models, characteristics of some team members are thought to be more important than others because of the formal (e.g., strategic core, see Humphrey, Morgeson, & Mannor, 2009) or informal roles (e.g., network position) the team members occupy. This disproportionate influence is accounted for in the team composition operationalization. For example, a team memberâs attribute may be weighted by their position in the social network, and then aggregated across the team (Lim, 2004).
In many circumstances, individual and team-based composition models both contribute to the prediction of valued outcomes. As an example, a new product development team may be best positioned for success when team members prefer to work in teams rather than individually, the team is diverse in terms of functional background, and members in key positions (e.g., boundary spanning roles) have the necessary levels on the attributes needed for the role (e.g., self-monitoring). The different models can be combined via an algorithm such as the one provided by Mathieu et al. (2014b). This algorithm can include a temporal vector to account for changing team composition and outcome relationships over time (Mathieu et al., 2014b). Further, because team composition itself may be dynamic because of membership change, fluid boundaries and multiple team membership, the question becomes to what extent the team has the best combination of member attributes for a particular task or circumstance (Mathieu et al., 2014b).
There are a large number of possible team composition attributes and configurations to consider. In most cases, researchers and practitioners should focus on identifying a few key composition considerations that are important for effectiveness in the specific circumstance. Some team composition considerations are likely to be important for most teams. For example, a highly disagreeable team member may be disruptive to team performance in most organizational circumstances (Bell, 2007). Other key team composition considerations will be highly dependent on the context. For example, self-managing teams with ambiguous leadership structures can thrive when shared leadership emerges (Carson, Tesluk, & Marrone, 2007). Shared leadership is more likely to emerge in teams composed of members that are high on both psychological collectivism and extraversion, or both psychological collectivism and motivation to lead (Chen, 2014).
Analysis of the context within which teams operate and an understanding of the theoretical path through which team composition is expected to relate to valued outcomes can be used to identify important attributes and configurations (Bell & Brown, 2015; Bell, Fisher, Brown, & Mann, 2016). The context can be used to identify important emergent properties (e.g., team cohesion) that contribute to a teamâs human capital. The context also informs how team composition may be most effectively managed (e.g., through staffing, specific leadership behaviors). Team composition is shaped by the context beyond the temporal aspects discussed here (Johns, 2006); however, in this chapter we focus on a few key issues related to team composition over time. We describe: (a) the temporal context and how it informs our understanding of team composition; (b) research that examined how team composition related to team processes, emergence processes, and emergent properties over time; and (c) the management of team composition over time.