Team Effectiveness 1
Running head: Team effectiveness
Towards a typology of team effectiveness:
A meta-analytic review
Stephen E. Humphrey
PennsylvaniaStateUniversity
Elizabeth P. Karam
Frederick P. Morgeson
Michigan State University
Abstract
Words words words …
Towards a Typology of Team Effectiveness:
A Meta-Analytic Review and Measurement Development
Teams are now ubiquitous in organizations, making it imperative that organizational scholars study teamwork and team effectiveness (Kozlowski & Ilgen, 2006). In this vein, there has been a remarkable increase in empirical and theoretical research on teams over the last several decades, with the published research on teams in the “top quartile” of management and industrial/organizational psychology journals (Podsakoff, Mackenzie, Bachrach, & Podsakoff, 2005) doubling from 1980-1985 to 2000-2005. However, although researchers have focused extensively on testing the influence (e.g., Ilgen, Hollenbeck, Johnson, & Jundt, 2005; Mathieu, Maynard, Rapp, & Gilson, 2008) and structure (e.g., Harrison & Klein, 2007; Marks, Mathieu, & Zaccaro, 2001) of team inputs and processes, there has been only limited research focusing on team outputs.
This is problematic for team research, in part because a lack of clarity concerning the nature of team effectiveness can lead to a lack of precision and potential theoretical misspecification. For example, the lack of a consistent conceptualization of team effectiveness has led scholars to use distinctly different operationalizations (e.g., satisfaction, learning, or performance) of the team effectiveness construct. This lack of consensus can cause confusion in the research literature, making it difficult to interpret conflicting empirical findings. For example, if some studies find a significant relationship with team effectiveness but other studies find no significant relationships, it is unclear whether such mixed findings implicate the underlying theory or whether it reflects differences in construct measurement. As such, the examination and clarificationof the team effectiveness criterion space can result in scholars building more precise theoretical models and testing these models using a more consistent language.
Other research streams have found that a focus on the criterion space propelled the literature forward. For example, an examination of the dimensionality of Organizational Citizenship Behavior (OCB) by LePine, Erez, and Johnson (2002) demonstrated that the five dimensions of OCB proposed by Organ (1988) are essentially equivalent indicators of the latent OCB construct.As another example, Borman and Motowidlo’s (1993) separation of performance into task and contextual dimensions has demonstrated that whereas both dimensions contribute to overall performance, they have different antecedents (Motowidlo & Van Scotter, 1994).By turning our attention to clarifying the criterion space, we hope to aid scholars in building more precise theoretical models that can be tested using a more consistent language, thus pushing the team literature towards a greater consensus on what leads to effective teams.
One problem with clarifying the domain of team effectiveness is that there are many definitions of team effectiveness. Early definitions shaped future discussions by focusing on internal and external criteria. For example, Schein (1970) argued that the function of a team is to meet organizational responsibilities (e.g., getting work out, generating ideas, or serving as liaisons) while simultaneously meeting personal responsibilities (e.g., developing group identity, backing up team members, or providing social support). Nieva, Fleishman, and Rieck (1978) used a motivational approach by defining team effectiveness as “the goal directed behaviors/activities/functions accomplished by the team in performing the task” (p. 52). Hackman and Oldham (1980) expanded on this by defining team effectiveness in terms of the team’s success in meeting (or exceeding) organizational standards of quality and quantity, members’ needs are satisfied, and members want to continue to work together on future tasks.
Several researchers subsequently took a part of these definitions (e.g., Gladstein, 1984, suggested that team effectiveness consisted of performance and satisfaction and Sundstrom, DeMeuse, Futrell, 1990 defined team effectiveness as performance and viability), whereas other researchers changed components of the definitions. For example, Cohen and Bailey (1997) defined team effectiveness as performance outcomes, attitudinal outcomes, and behavioral outcomes;Beal, Cohen, Burke, and McLendon (2003) suggested that team effectiveness as performance behaviors and performance outcomes; and Kozlowski and Bell (2003) argued that team effectiveness was a combination of internal (e.g., satisfaction and viability) and external (e.g., quantity and quality) criteria.
As this selective review makes clear,although these broad conceptualizations of team effectiveness differ, there is even less agreement as to what the specific dimensions of team effectiveness are. For example, there have been inconsistencies on what “productivity” is. Some defined it as the quality or quantity of work (Beersma, Hollenbeck, Humphrey, Moon, Conlon, & Ilgen, 2003), whereas others have considered it the meeting of organizational expectations (Hackman, 1987). Related to this, it is unclear if performance speed is the same as quantity of work accomplished or team productivity; error rates, failure, accuracy, and quality may be synonomous; and “backing up,” assistance, cooperative support, and cooperative behavior may all be forms of helping. In terms of affective reactions, some have considered these as a component of team effectiveness (Guzzo & Dickson, 1996), whereas others have placed them outside the realm of team effectiveness (Beal et al., 2003). As another example, although Schein (1970) discussed helping, most definitions of team effectiveness do not seem to explicitly acknowledge it.
In this manuscript, we attempt to address this confusion in the literature by organizing team effectiveness into three broad domains using self-regulation theory as an overarching theoretical framework. This yields an integrative definition of team effectiveness that draws from traditional views on team effectiveness but also builds on insights gained from self-regulation theory. We then identify the core dimensions of team effectiveness and provide construct definitions. Drawing from self-regulation theory, the team effectiveness dimensionsare organized into behavioral, cognitive, and affective domains. Next, we complete several meta-analyses of the dimensions of team effectiveness in an effort to provide an introductory understanding of the relationships amongst dimensions. We couple this with an analysis of the reliabilities of existing scales in an effort to identify strengths and weaknesses in the literature.Finally, we test develop and test specific hypotheses that provide a more nuanced view of the relationship between teamwork processes and team effectiveness, extending the recent work of LePine, Piccolo, Jackson, Mathieu, and Saul (2008).
ATypology of Team Effectiveness
We developed our typology of team effectiveness in the following manner. First, we reviewed the articles discussing team effectiveness in an effort to assemble a broad set of team effectiveness constructs examined in past research. This included examining definitions of team effectiveness to determine what researchers have considered relevant, as well as empirical studies of team effectiveness. Second, we examined the resulting set of constructs in an effort to determine if different variables represented unique or redundant constructs (e.g., although given different labels, errors and accuracy reflect different ends of a continuum that captures the quality of work). Third, after creating a list of constructs, we organized these constructs into a structure derived from the literature on self-regulation in teams (Hinsz, Tindale, & Vollrath, 1997). This suggested potential gaps and areas needed refinement. Finally, we reflected upon the resulting construct set and added constructs that theoretically made sense.
Teams are self-regulating entities (Hinsz et al., 1997), which means that they set goals, check their progress against the goals, and adjust their behavior in response to deviations from the desired state (c.f., Karoly, 1993; Lord, Diefendorff, Schmidt, & Hall, 2010)Hinsz et al.(1997) suggested that the extent to which information is shared (and what is shared) in a group impacts how groups (as a collective) manage that information to produce group outcomes. Although they present a theoretical model linking the goals of a team, the specific processing undertaken by the team, and the output of this process, they point out how the model retroactively fits existing models and empirical results. Hinsz et al. (1997) go on to suggest that information processing is what teams do (i.e., how they behave), it is a function of what they feel and how much members are attracted to a team (i.e., their affective reactions), and what they think during the process (i.e., their cognitions). Essentially, there are three categories of information processing: behavioral actions, affective reactions, and cognitions.
We utilize these three categories of information processing to organize team effectiveness. It is no surprise that team activities can be grouped into these three domains, as these are the three main research domainsin the social sciences. Classic psychological research focused on the behaviors of people, to the exclusion of all internal processes (Watson, 1913). As noted by Ilgen, Major, and Tower (1994), cognition grew in interest (and applicability) in the social sciences during the 1980’s and early 1990’s. This “cognitive revolution” explicitly acknowledged the thought processes of individuals and teams in developingtheory. The “affective revolution” (Barsade, Brief, & Spataro, 2003) that followed has moved research into the realm of moods, affect, and emotions, all in an effort to better understand the complexity of individuals. Together, these three paradigms nearly encompass the totality of human behavior, and thus serve as a useful organizational scheme for understanding team effectiveness.
With this as an overall theoretical framework, team effectivenesscan be defined by expanding Mathieu et al.’s (2008, p. 412) recent definition as the behavioral, affective, and cognitive products of team activity that are valued by one or more constituencies.Although inclusive, this definition lacks the specificity needed to address the concerns outlined earlier. As such, we seek to articulate the various specific manifestations of behavioral, cognitive, and affective effectiveness indicators.
In order to organize the dimensions of team effectiveness, we placed constructs into one of these domains. Although werecognize that these domainsare not orthogonal (e.g., a behavioral construct may have cognitive undertones), we do believe that each dimension has a primary or predominant characteristic that enables placement within the typology. The final structure and organization of the dimensions is presented in Table 1.
Behavioral Team Effectiveness
Behavioralteam effectivenesscan be defined as “the action tendencies one has to approach or avoid an object or perform some response” (Bagozzi, 1978, p. 10). This is the domainof effectiveness outcomes that is most commonly thought of as “performance” or “effectiveness”because it refers to task related outcomes of specific activities conducted within the team. We identified six behavioral dimensionsof team effectiveness.
Performance qualityis the accuracy or precision of team output (Jenkins, Mitra, Gupta, & Shaw, 1998). It includes both behavioral task performance (e.g., correctly welded joints) and decision-making performance (e.g., jury decisions).In addition to the over 100 years of research on performance quality at the individual level (Elliot, Helsen, & Chua, 2001), team researchers have also frequently focused on quality (Humphrey, Hollenbeck, Meyer, & Ilgen, 2002). For example, in McGrath’s (1984) circumplex of team tasks, both intellective tasks and judgment tasks are defined in terms of finding a correct or preferred answer. Inaccurate decisions would therefore indicate low quality performance and may have serious consequences. Note that performance quality does not need to be bound solely in terms of decision-making. For example, automotive assembly teams can be judged on the number of quality defects per car, and a research team writing a manuscript can be judged on the number of grammatical errors.
Performance quantityreflects the amount of work produced (Jenkins et al., 1998). This dimensionis conceptually similar to performance “speed” (which reflects how quickly work is performed). Quantity has long been considered a core component of performance, such that quantity of performance is often a fundamental component of reward systems (e.g., piece-rate work; Taylor, 1895). In teams, performance quantity can reflect the number of cars produced, the speed of an Olympic relay team, or the number of mortgages approved per day.
Comparing performance quality and quantity, researchers have frequently noted the different antecedents, and often the negative correlation between these two dimensions of effectiveness (Elliot et al, 2001). Because of this, scholars and practitioners alike have often considered both dimensions simultaneously when examining team performance (Beersma et al., 2003).
In-role performance can be thought of as performance on the job-related aspects of work, rather than an overall assessment of effectiveness. For example, this may reflect performance on a simulation (e.g., Hollenbeck et al., 2002) or task performance (e.g., Tjosvold & Yu, 2004). Whereas performance quantity and performance quality reflect facets of behavioral team effectiveness, they do not reflect the totality of in-role behavior. Performance quantity and performance quality do not inherently capture the relative importance of each of these dimensions. That is, on one task, performance quality may be the only thing that matters (e.g., how quickly a jury makes a decision is essentially irrelevant compared to the quality of that decision), whereas on other tasks performance quantity may be significantly more important than performance quality (e.g., the speed of a relay team is more important than the runners’ form). In contrast, in-role performance reflects the performance of the team on a priori defined role behaviors (thus capturing the aggregation of quantity and quality on these specific behaviors).
In contrast, extra-roleperformance reflects team extra-role (rather than in-role) behaviors such as interpersonal facilitation (e.g., helping or backing up team members). As noted by McIntyre and Salas (1995), “this skill is at the heart of teamwork, for it makes the team truly operate as more than the sum of its parts” (p. 26). This echoes the individual level research, which has often argued that interpersonal facilitation shapes the organizational context (and therefore organizational performance) by supporting other’s in-role activities (Borman & Motowidlo, 1993). Extra-role performance may manifest itself as backing up team members who are overloaded or coaching a team member (e.g., providing verbal feedback). At the core of this dimension is the notion that team members are going beyond their defined roles in ways that contribute to the ability of the team to complete its task.
Goal attainment is the extent to which a team reaches or exceeds its goals, where the goals may be set by other organizational actors or the members themselves. As suggested by Hackman (1987), within organizations, the “absolute” level of performance may be less important than the ability to hit goals. That is, organizational evaluations of teams (for compensation, continuation of resource allocation, etc.) are often based upon whether a team has achieved its goals. The focus of team goals may be those dimensions of behavioral performance discussed already (e.g., production levels or error rates) or they may be unique from those already discussed (e.g., customer satisfaction ratings or turnover rates).
Counterproductive workbehaviors(CWBs) reflect voluntary behaviors that are detrimental to the team or organization (Mount, Ilies, & Johnson, 2006). This includes behaviors such as social loafing, tardiness, sabotage, theft, sexual harassment, and verbal (or physical) abuse (Conlon, Meyer, & Nowakowski, 2005; Ones, 1992).The extent to which teams promote or inhibit CWBs is a matter of debate. On the one hand, self-managed teams have been proposed as a way to reduce the incidence of CWBs (Manz & Sims, 1987), as the team context is proposed to connect individual inputs more tightly to outputs. Moreover, teams can be a strong context where pro-organizational norms are reinforced through socialization (Barker, 1993). Yet, that context may also produce CWBs. For example, a team climate that supports CWBs can encourage otherwise compliant team members to damage property, break rules, or hurt coworkers (Robinson & O’Leary-Kelly, 1998). As noted in the Hawthorne Studies (Roethlisberger & Dickson, 1939), workers may feel pressure from the team to avoid being a “rate-buster” by producing too much quantity of output.
Affective Team Effectiveness
Affect is “the positive-negative emotional relationship or feelings one has towards an object or activity” (Bagozzi, 1978, p. 10). In the teams literature, researchers have long focused on the emotional connection members have with each other and have frequently included affective responses as critical outcomes of teams (c.f., Hackman, 1987). Applying Bagozzi’s definition to teams, team affective outcomes can be thought of as the aggregate of the positive-negative emotional feelings individual team member have towards (or relationship with) the remainder of the team.
Teams researchers have frequently focused on four affective dimensions of team effectiveness. Satisfaction with the team represents how positively members feel about the rest of the team. As noted by Hackman (1987), team members should like (rather than be frustrated by) other members or else the emotional cost to frustrated members likely outweighs the benefits of being in a team. Not surprisingly, most models of team effectiveness have included satisfaction with the team.
Viability is similar to satisfaction with the team, such that it represents whether team members would like to continue to work together in the future. Thus, being satisfied with the team is usually sufficient for a team to be considered viable. However, team members may dislike each other and yet see the benefit in continuing to collaborate. For example, professional sports teams frequently encounter this phenomenon, such that team members perceive that the benefit of having a personally distasteful member on the team outweighs the costs associated with having that member. Because the startup costs of a team are high (due to the time and effort necessary to develop norms and determine how to coordinate with each other; Ilgen et al., 2005), having a team that wants to continue working together can be very beneficial for organizations and is thus a desirable component of team effectiveness.
Cohesion reflects the interpersonal attraction towards, or the bond within, a team (Beal et al., 2003). Much like satisfaction with the team and viability, cohesion is primarily affective such that it focuses on how positively team members feel about other members in the team (e.g., are they friends?). Yet, it has a slightly distinctive aspect in that it represents a sense of belongingness and morale associated with being part of the team (Bollen & Hoyle, 1990). As such, cohesion is a dimension of effectiveness that can begin to rise almost immediately after team formation (Turner, 1987).