The Team Experiences Survey: a Biodata Test for Team Selection

Team Experiences Survey 1

THE TEAM EXPERIENCES SURVEY: A BIODATA TEST FOR TEAM SELECTION

MICHAEL D. JOHNSON

FREDERICK P. MORGESON

MICHIGAN STATE UNIVERSITY

ABSTRACT

Although many organizations have moved to teams as a way of structuring work, few tools exist for selecting individuals to work in teams. A new biodata selection test is developed that taps into twenty team experience constructs. The Team Experiences Survey is almost wholly independent of cognitive ability, but somewhat related to personality factors.

PRESS PARAGRAPH

Many organizations have moved to teams as a way of structuring work, but little research has been done into how to best hire people to work in teams. Organizations often use standardized tests in making their hiring decisions. Tests based on biographical information (biodata) about the applicant, have been shown to do a good job of predicting job performance, while minimizing discrimination in hiring decisions. This paper outlines the development of a biodata test to select individuals who will work in teams.

Teams have become an increasingly common way to structure work in organizations (Morgeson, Aiman-Smith & Campion, 1997). The types of teams used by organizations vary widely, including self-managing teams, empowered teams, cross-functional teams, crews, and committees, among many others (Guzzo & Dickson, 1996). Two recent surveys of U.S. organizations found that 48% of the organizations utilized at least one type of team in their organization (Devine, Clayton, Philips, Dunford, & Melner, 1999). Although the use of teams is not a new phenomenon (Sundstrom, DeMeuse, & Futrell, 1990), the last ten years have witnessed a considerable growth in research on teams in organizations.

A great deal has been learned about factors affecting team performance and effectiveness, but relatively little work has been done on developing selection systems for work teams (Hough & Oswald, 2000). It is not clear if the vast amount of research on individual personnel selection will transfer to situations where individuals are expected to work in collaborative, highly interdependent work teams. For example, it has been suggested that the skills, knowledge, and motivation needed to function effectively in a team go well beyond the core technical skills often measured in traditional selection contexts (Barrick, Stewart, Neubert, & Mount, 1998; Guion, 1998). Others have noted that selecting individuals for teams requires one to consider problems that are seldom considered when selecting individuals to work by themselves (Jones, Stevens, & Fischer, 2000).

For example, Stevens and Campion (1999) have suggested that a unique set of team-related knowledge, skills, and abilities are needed to effectively function in a team environment. Others have found evidence for the role of personality in jobs requiring teamwork or interpersonal interaction (Barrick et al., 1991). Beyond these efforts, however, there have been few investigations into what other selection tools might be useful in team contexts. In particular, the usefulness of experience-based measures for team settings has not been explored. This is unfortunate, because at the individual level, experience-based measures have shown good validity and low adverse impact.

Experience-based selection measures, such as structured interviews (Campion, Campion, & Hudson, 1994), background investigations (McDaniel, 1989), training and experience evaluations (McDaniel, Schmidt, & Hunter, 1988), and accomplishment records (Hough, 1984) have been shown to have be useful tools. Experience-based measures have several positive features, which include: (a) they provide an operationally well-defined description of prior behavior of the applicant (Mumford & Owens, 1982); (b) they yield standardized life histories of applicants that can be quantitatively analyzed (Owens, 1976); and (c) they are generally unrelated to cognitive ability. One particularly well-researched type of work experience measure is biographical information tests. Also termed biodata tests, they have evidenced good validity (Schmidt & Hunter, 1998), incremental validity beyond cognitive ability measures (Karas & West, 1999; Mael & Ashforth, 1995; Mount, Witt, & Barrick, 2000) and various personality factors (Mount et al., 2000).

Given the usefulness of a biodata approach in individually oriented selection, we decided to develop a biodata test to select individuals in team settings. This involved the development of a model of team performance and the identification of a series of team experience constructs that are relevant for performance in team settings. This informed the development of the Team Experiences Survey (TES), a biodata test measuring past experiences in group and team settings. This paper details the development of the TES and presents data on its psychometric properties as well as relationships with ability and personality measures.

This study contributes to the literature in at least two ways. First, Hough and Oswald (2000) have asserted that advances in team selection research await good taxonomies of “team difference” variables. In this study, we develop a typology of twenty experience-based constructs that are expected to be positively related to team performance. Each of these constructs is defined and operationalized to reflect the individual team member’s impact on team-level variables. Second, we develop an experience-based tool for selection of individuals for work teams. This represents an important extension of biodata measures to the team context.

Team Experience Constructs

Several literature reviews on team composition, performance, and effectiveness have been published within the past ten years (Cohen & Bailey, 1997; Guzzo & Dickson, 1996; Kozlowski & Bell, in press; Morgeson et al., 1997; Paris et al., 2000). A great deal of work in organizational studies has gone into identifying the variables that significantly predict team effectiveness. Using the input-process-output (IPO) framework from McGrath (1984) and Hackman (1987), we identified twenty input and process constructs from the literature that are most likely to predict future team performance based on team member experience. These are summarized in Table 1 and guided the development of the TES.

Inputs

Inputs represent all variables that either the team members bring with them—such as traits, knowledge, and experiences—or that the organization provides for the team—such as the team task, structure, or context. Four broad categories of team inputs have emerged from the literature: design, interdependence, composition, and context.

Design covers all aspects related to the structure of the team. One important design consideration involves the amount of self-management given to the team. Self-managing teams are given a high degree of autonomy in their task design, which has repeatedly shown positive correlations with satisfaction, organizational commitment, and trust in management (Cohen & Bailey, 1997). We expect that experiences with self-management will lead to good team effectiveness and performance in the future, because self-management is presumed to increase the team member’s sense of responsibility and ownership of the team task (Campion, Medsker, & Higgs, 1993).

Interdependence concerns the degree with which team members must rely on each other to achieve the team’s goals or accomplish the team’s task. Wageman (1995) found that highly interdependent groups performed better than moderately interdependent groups, and that highly interdependent groups exhibited “high-quality social processes, extensive mutual learning, and a sense of collective responsibility for performance outcomes” (p. 174). Moreover, Campion et al. (1993) suggest that interdependence increases “the motivational properties of work,” which in turn affects team effectiveness. From an experience perspective, we expect that experiences both with task interdependence—where team members depend on each other to accomplish the work—and goal interdependence—where individual goals are linked to group goals—will lead individuals to be more predisposed to interdependence in future teams.

Composition (also called team characteristics) covers all of the traits or dispositional characteristics that team members bring to the team, and the way in which they are configured in the team. A great deal of recent research has dealt with personality factors and abilities in team composition, but the benefit of these variables is task- or team-specific (Cannon-Bowers, Tannenbaum, Salas, & Volpe, 1995). Therefore, we focus on composition variables which have been shown to be more beneficial for team effectiveness and performance across tasks and contexts. Preference for team work refers to team members’ desire to work collectively, rather than individually. Campion et al. (1993) distinguish it from cohesiveness in that cohesiveness refers to a desire to work with a particular group, but preference for team work refers to a general desire to work in groups. We expect that preference for team work will influence motivation to work in future teams. Studies on heterogeneity in demography, abilities and experiences have found that although homogeneous groups initially performed better than heterogeneous groups, this effect diminished over time, with heterogeneous groups ultimately outperforming homogeneous ones (Watson, Kumar, & Michaelsen, 1993). We expect that experiences in team heterogeneity should be related to future team effectiveness, as they serve to enhance the team member’s ability to positively relate to a broad range of people.

Context covers all situational, organizational, and even extra-organizational variables within which the team operates. Boundary management (or boundary spanning) refers to the way a team manages its interactions with other groups or individuals (Tannenbaum, Beard, & Salas, 1992). From an ecological point of view, boundaries both separate work teams in an organization and serve as the points that connect them (Sundstrom et al., 1990). Good boundary management has been associated with high levels of performance, as the teams continually revise their knowledge of the environment (Ancona, 1990). We expect that positive experiences in boundary management will enable team members to engage in these behaviors effectively in future teams, thus enhancing team performance.

Process

Process constructs in a team setting represent all of the interpersonal transactions between team members (Hackman, 1987). We found three major categories of process constructs in the teams literature: energy, direction, and affect.

Energy includes all of the interactions that directly affect the level of exertion and activity within the team. Effort refers to the conscious exertion of power or hard work on the part of team members. Although effort has often been considered as a dependent variable in group research, we conceptualize effort as a process variable in terms of the amount of energy each member is willing to expend on behalf of the team. Past experiences where team members exert a great deal of effort toward team tasks and goals may predict individual effort on behalf of future teams. In problem solving, the team members first decide on which issues they need to seek information and their approach to the problem (Ilgen, Major, Hollenbeck, & Sego, 1995); then the team gathers and integrates information, generates alternatives, and makes a decision (Cannon-Bowers et al., 1995). Positive experiences in team problem solving are presumed to develop learning among team members about good problem solving behaviors, which should lead in turn to future team effectiveness.

Skill usage refers to the effective deployment of each team member’s unique knowledge, skills, and abilities toward achieving team goals. Past experiences with skill usage should increase both future collective efficacy and future skill usage, leading to increased team performance. Potency refers to “the collective belief of the group that it can be effective” (Shea & Guzzo, 1987, p. 335). Potency has been distinguished from collective efficacy in that collective efficacy is task-specific, while potency refers to a general shared belief that the team can be effective across many tasks (Guzzo, Yost, Campbell, & Shea, 1993). Past experiences with potency should heighten expectancy of future team performance. Perseverance in a team setting refers to the team’s persistence in the face of obstacles, opposition, or discouragement. Individuals who have been in highly persevering teams in the past should be more likely to persevere in future teams, compared to individuals who have been in teams that gave up when they faced obstacles.

Direction includes all of the group interactions affecting the focus of the team’s attention and energy. Accountability refers to “the degree to which the individual feels personally accountable and responsible for the work he or she does” (Hackman & Oldham, 1976, p. 256). In a team setting, the extent to which an individual has had past experiences where the team members have felt a great deal of responsibility for the team should predict a future sense of accountability to teams. Learning orientation refers to a team’s “seeking to develop competence by acquiring new skills and mastering new situations” (VandeWalle & Cummings, 1997, p. 392). In contrast to performance orientation, where teams seek to produce without improving their performance, teams that seek to learn direct attention toward continually refining and improving their skills. We expect that past experiences with teams that have a learning orientation should both enhance the individual’s knowledge of team processes and encourage future learning orientation, leading to future team effectiveness and performance.

Adaptability refers to the ability of teams to adjust the team’s resources and behavior in response to changes in the team’s task or environment (Cannon-Bowers et al., 1995). Individuals who have been in highly adaptable teams will presumably be more flexible in their behavior in future teams, and will be able to adapt to new team tasks and environments more effectively than individuals who have been in less adaptable teams. Communication refers to “the process by which information is clearly and accurately exchanged between two or more team members in the prescribed manner and with proper terminology” (Cannon-Bowers et al., 1995, p. 345). Although frequency of intrateam communication has been associated with lowered performance because it is indicative of high team conflict (Smith et al., 1994), the ability of teams to communicate effectively has long been shown to enhance effectiveness to the extent that effective communication influences cooperation within the team (Campion et al., 1993). Positive communication experiences in teams, then, should lead to future cooperation.

Affect includes all group interactions that are related to the emotional states of the team members. Cooperation refers to “the willful contribution of personal efforts to the completion of interdependent jobs” (Wagner, 1995, p. 192). Numerous studies have shown associations between cooperation and both performance outcomes and psychosocial outcomes (Kozlowski & Bell, in press). Past cooperative experiences in teams are expected to lead individuals to cooperate more willingly in future teams; even more significantly, however, uncooperative past experiences are expected to decrease trust and potency, lowering future team performance. Social support refers to what Gladstein (1984) called “group maintenance behavior;” it is the sense that team members help each other. Because social support is motivating, it may help teams remain on task during mundane activities (Campion et al., 1993). Workload sharing refers to the prevention of social loafing or free-riding. We expect that that having many team experiences where social loafing does not occur will predispose a team member to share workload in future teams; perhaps more significantly, individuals who have been in many teams where social loafing did occur may be more likely to accept social loafing in the future. Conflict resolution refers to the successful working out of both relational and task conflict within the team. Although it has been theorized that task conflict may be beneficial for team performance, most studies have shown detrimental effects from both types of conflict (De Dreu & Weingart, 2002). Therefore, we expect that positive past experiences in conflict resolution will enable a team member to more effectively resolve conflict in future teams. Trust is a psychological state that indicates the team members’ willingness to be vulnerable to each other, and can rely on each other (Rousseau, Sitkin, Burt, & Camerer, 1998). Lack of trust will diminish team effectiveness, and can lead to withdrawal from the team. Past experiences with team trust should predispose a team member to trust more readily in the future.

Method

Participants and Procedures

In exchange for class credit, 525 undergraduate business students from a large Midwestern university wrote “Team Experience Essays” that documented team experiences in their lives that they felt were related to team performance. Similar to a method described by Russell, Mattson, Devlin, and Atwater (1990), we content-coded each essay by team construct, and wrote items based on these experiences. We wrote ten items in each construct based on the team experience essays, intentionally writing more items than we expected to use in the final measure. This yielded 200 total items in the first iteration of the Team Experiences Survey (TES). Additionally, we wrote ten items each to reflect the output constructs of team performance and team satisfaction, for a total survey of 220 items. The items were written such that subjects were asked to aggregate their past experiences in teams and groups in responding to each of the items in the TES. The TES is based on the premise that it is the aggregation of experiences in teams—rather than simply a positive or negative experience in teams—that will significantly impact future team performance and effectiveness. The anchors for the item responses included agreement, extent, and other anchors unique to specific items. Example items in each construct can be found in Table 2. The following semester, the TES was administered to a different group of 536 undergraduate business students. Items corresponding to the various constructs were randomly distributed throughout the TES.