Team-level KSAOs

Research pertaining to compositions of KSAOs in teams has progressed significantly in the last decade (see Table 15.2 for a summary of meta-analytic results on team-level KSAOs and team performance). As a result, group composition (i.e., the configuration of group member attributes) has become one of the most commonly studied variables in the team literature (Guzzo & Dickson, 1996; Hollenbeck, DeRue & Guzzo, 2004), and different configurations of KSAOs have been used to predict a variety of team-related outcomes, such as team performance (Bell, 2007), collective turnover (Hausknecht & Trevor, 2011), OCB (Arthaud-Day et al., 2012) and CWB (Schmidt, Ogunfowora & Bourdage, 2012).

Although many of the individual-level KSAOs discussed above can be applied at the team level, team composition can be far more complex than merely totalling individual attributes (see Chan, 1998; Kozlowski & Klein, 2000). The most popular type of aggregation method has been mean aggregation, followed by diversity (e.g., dispersion, homogeneity) and extreme scores (maximum and minimum), and the results concerning the same type of team-level predictors can vary drastically depending on the type of aggregation method used.

In order to determine the appropriate type of aggregation for team research, Steiner (1972) proposed a task typology for aggregation method which has been considered helpful (Bell, 2007; Cannon-Bowers & Bowers, 2011; Mohammed et al., 2010). According to Steiner’s typology, the mean or sum of individual scores is most appropriate where additive or compensatory tasks (i.e., team performance is the sum of individual performance) are

Table 15.2 Summary of recent meta-analyses on team attributes and team performance.


Validity Estimates from Meta-analyses



General mental ability (GMA)

0.27 (k = 42, n = 2,995)

Bell (2007)

Emotional intelligence

0.18 (k = 6, n = 304)

Bell (2007)



0.11 (k = 39, n = 2205)

Bell (2007)


0.12 (k = 29, n = 1692)

Bell (2007)


0.09 (k = 38, n = 224з)

Bell (2007)

Emotional stability

0.04 (k = 22, n = 1439)

Bell (2007)

Openness to experience

0.05 (k=25, n = 1697)

Bell (2007)

Values and Attitudes


0.25 (k = 14, n = 1299)

Bell (2007)

Preference for teamwork

0.18 (k = 10, n = 490)

Bell (2007)

Collective efficacy

0.38 (s= 64, k = 78, n = 3738,

Stajkovic, Lee, and Nyberg

N = 16009)


Group potency

0.34 (s= 29, k = 32, n = 1613, N=9699)

Stajkovic et al. (2009)


Task-related diversity

0.13 (k = 15, N = 1209) for quality of team performance

Horwitz and Horwitz (2007)

0.07 (k = 9, N = 704) for quantity of team performance

Horwitz and Horwitz (2007)



-0.01 (k = 14, N = 1093) for quality of team performance

Horwitz and Horwitz (2007)

-0.02 (k = 3, N = 182) for quantity of team performance

Horwitz and Horwitz (2007)

Functional background variety

0.10 (k = 31, n = 3726)

Bell, Villado, Lukasik, Belau, and Briggs (2011)

-0.45 (s = 11, n = 3062)

Thatcher and Patel (2011)

Educational background variety

0.01 (k = 13, n = 2629)

Bell et al. (2011)

Education level diversity

-0.01 (k = 14, n = 3914)

Bell et al. (2011)

-0.46 (s= 8, n = 1859)

Thatcher and Patel (2011)

Organizational tenure diversity

0.04 (k = 24, n = 4259)

Bell et al. (2011)

Team tenure diversity

-0.04 (k = 12, n = 2124) -0.41 (s = 8, n = 980)

Bell et al. (2011)

Thatcher and Patel (2011)

Race diversity

-0.11 (k = 31, n = 5298) -0.35 (s = 23, n = 1890)

Bell et al. (2011)

Thatcher and Patel (2011)

Sex diversity

-0.06 (k = 38, n = 6186) -0.47 (s= 19, n = 1620)

Bell et al. (2011)

Thatcher and Patel (2011)

Age diversity

-0.03 (k = 40, n = 10953) -0.35 (s = 22, n = 1584)

Bell et al. (2011)

Thatcher and Patel (2011)

Cultural diversity

-0.02 (k = 42, n = 7184)

Stahl, Maznevsi, Voigt, and Jonsen (2010)


Table 15.2 (Continued)


Validity Estimates from Meta-analyses



Team cohesion

0.32 (s = 16, n = 1460)

Thatcher and Patel (2011)

Social cohesion

0.20 (k = 4, n = 206) for outcome performance

Chiocchio and Essiembre (2009)

Task cohesion

0.35 (k = 4, n = 206) for outcome performance

Chiocchio and Essiembre (2009)

Task conflict

-0.18 (s= 16, n = 1832)

Thatcher and Patel (2011)

Relationship conflict

-0.18 (s= 16, n = 202l)

Thatcher and Patel (2011)

s=number of studies; k = number of correlations; n = number of teams; N = total sample size.

concerned. As for extreme scores, the team’s minimum score is best when the team is performing conjunctive tasks (weakest individual performance determines team performance), whereas maximum score is the most useful aggregation strategy when disjunctive tasks (strongest individual performance determines team performance) are the criterion. However, empirical evidence from meta-analyses suggests that the validity of aggregation method depends not only on the criterion but also the type of predictor (e.g., Bell, 2007). Given that most of the studies on team-level KSAOs were devoted to dispositional traits, we first discuss research findings regarding the team composition of dispositional traits, followed by other attributes (knowledge, abilities, values and needs).

Team composition of dispositional traits Since 2000 there has been an increasing research focus on team-level KSAOs which has been devoted to examining the composition of personality traits in teams. Meta-analytic results from Bell (2007) highlight the importance of team personality operationalization (e.g., average, dispersion, minimum, maximum) in understanding team composition. Based on the FFM of personality (McCrae & Costa, 1985), Bell found that team minimum and average levels of agreeableness were the strongest predictors of team performance in field studies. Although team average conscientiousness, openness to experience, emotional stability and extraversion were shown to be positive predictors of team effectiveness in field studies, the extreme levels of these four personality traits contribute little to team performance.

On a similar note, Peeters, Van Tuijl Rutte and Reymen (2006) found that team average conscientiousness was positively related to team performance in student design teams. On the other hand, team average agreeableness, variability in agreeableness and variability in conscientiousness did not predict team performance. Using helping-norm emergence as the criterion, Raver, Ehrhart and Chadwick (2012) found that the maximum and minimum levels of agreeableness in student project teams were associated with team helping behaviour. Raver and colleagues’ finding was in line with the ‘sucker aversion’ effect (Chen & Bachrach, 2003; Jackson & Harkins, 1985; Schroeder et al., 2003), which arises when team members experience a sense of inequity when one disagreeable person refuses to help others. The positive effect of team agreeableness was also found in project teams, where team average agreeableness was shown to be related to better team communication and cohesion and subsequently performance over time, though this relationship was evident only when team members were interacting face-to-face as opposed to virtually (Bradley, Baur, Banford & Postlethwaite, 2013). In the context of top management teams (TMTs), Colbert, Barrick and Bradley (2014) demonstrated a positive relationship between team average conscientiousness and organizational performance. Turning to dark personality traits, Baysinger, Scherer and LeBreton (2014) found that team average levels of psychopathy and implicit aggression were positively linked to both dysfunctional interactions and negative perceptions of the group, and these links were mediated by task participation and negative socioemotional behaviours.

At the team level, team self-efficacy can be broken down into collective efficacy (i.e., shared beliefs in the collective capabilities to perform specific tasks; Bandura, 1997, p. 447) and team potency (i.e., shared beliefs in the collective capabilities to perform a wide range of tasks across situations; Gully, Incalcaterra, Joshi & Beaubien, 2002; Guzzo, Yost, Campbell & Shea, 1993; Zaccaro, Blair, Peterson & Zazanis, 1995). Collectively, a group’s belief that it is capable of performing tasks can promote the initiation of actions and boost the collective effort towards a common goal. Stajkovic, Lee and Nyberg (2009) showed a positive relationship between collective efficacy and team performance and between group potency and team performance. In addition, collective efficacy was found to fully mediate the effect of group potency on team performance. In a longitudinal study, Goncalo, Polman and Maslach (2010) demonstrated the different effects of team efficacy in the early versus late stages of team functioning. In particular, a high level of team efficacy in the early stages was associated with fewer perceived process conflicts but not with overall performance, whereas teams experiencing more process conflicts early on had higher team efficacy during later stages and better overall performance. Taking a unique approach of viewing team efficacy as either physical efficacy or mental efficacy, Hirschfeld and Bernerth (2008) showed team size increased both types of team efficacy; more team members might offer greater resources for task accomplishments (note that a curvilinear relationship between team size and team efficacy could not be tested due to restriction on team size, which ranged from 12 to 15 members per team). Team mental efficacy also predicted team internal social cohesion, problem solving and teamwork effectiveness, whereas team physical efficacy predicted team cohesion only.

In addition to personality traits and team self-efficacy, team collective motivational traits, such as collective learning and performance orientation, can influence team adaptability and performance, both independently and interactively (Porter, Webb & Gogus, 2010). Relating a team’s affective makeup to team effectiveness, Kaplan, Laport and Waller (2013) uncovered a positive effect of team homogeneity in positive affect (PA), but not average PA, on team effectiveness in nuclear power plant crews during crises, and this effect was carried through by a reduction in negative emotions. Despite the positive links shown between individual-level CSE and team effectiveness, research has shown that CSE does provide incremental validity above and beyond the Big Five in predicting team performance (Haynie, 2012). In addition, the positive relationship between team CSE and team performance was only evident when the team also exhibited a high level of LMX.

Team composition of other attributes Focusing on team knowledge, Wildman and colleagues (2012) suggested that team average level of knowledge is positively linked to team processes and outcomes. Shamsie and Mannor (2013) demonstrated a positive link between tacit knowledge, a form of organizational resource, and sports team performance in a sample of Major League Baseball teams. In terms of team-level abilities, meta-analyses have demonstrated a strong association between general mental ability (GMA) and team performance (Bell, 2007; Devine & Phillips, 2001; Stewart, 2006). Randall, Resick and DeChurch (2011) demonstrated that teams with higher mean levels of cognitive ability were more likely to develop adaptive strategies in decision-making tasks. In addition, a high degree of team-level EI has been shown to be beneficial in teams, but this effect seems to be more robust in laboratory settings than in field settings (Bell, 2007).

Team shared values can function as key motivational components to team functioning. Research has shown that teams that value equality had higher team performance and importance placed on the value of equality played a bigger role in predicting team performance than past performance (Glew, 2009). Interest in the role values and beliefs play in helping norm emergence in teams, Raver and colleagues (2012) found that a team’s minimum, but not maximum, levels of other-oriented values and personal helping beliefs were positively related to the emergence of a helping norm.

Focusing on cultural values and team-related phenomena, a meta-analysis by Taras and colleagues (2010) showed that cooperation in groups was positively associated with team- level power distance and uncertainty avoidance, as well as negatively related to individualism and masculinity. Similarly, Bell’s (2007) meta-analytic evidence demonstrated a positive link between team average collectivism and team performance, and between team average preference for teamwork and team performance in field studies. Research has shown that the positive effect of team average psychological collectivism on team effectiveness was mediated by its influence on information-sharing (Randall et al., 2011). To further examine the different facets of team psychological collectivism (preference, reliance, concern, norm acceptance and goal priority), Dierdorff, Bell and Belohlav (2011) demonstrated that the relationship between team psychological collectivism facets and performance was moderated by performance stage and LMX.

In the context of multicultural teams, Cheng, Chua, Morris and Lee (2012) investigated the relationship between team composition of cultural values and performance in self-managing teams over time. Results from a sample of MBA student teams suggested that cultural value makeup has a differential impact on team performance at various stages of team formation: although teams with a low mean level and a moderate dispersion level in uncertainty avoidance had better performance early on, teams with a high mean level of leadership orientation as well as a moderate dispersion of relationship orientation worked better in the long run.

Shifting the focus from ‘what people believe in’ to ‘what people want’, researchers have investigated the influence of team composition of members’ psychological needs on team effectiveness. Chun and Choi (2014) found that members’ need for achievement was positively linked to task conflict when operationalized as the group mean and negatively related to task conflict when operationalized as dispersion. In addition, the team mean need for affiliation was shown to correlate negatively with relationship conflict, whereas the team mean and dispersion of need for power were positively and negatively related to status conflict, respectively.

Interaction between individual- and team-level KSAOs In addition to studying individual- and team-level KSAOs independently, some researchers have started to examine the joint effects of individual- and team-level constructs in predicting team effectiveness and performance. Building on the main effects of individual conscientiousness and extraversion on team outcomes, Schmidt and colleagues (2012) showed that team compositions of conscientiousness and extraversion, as well as core group evaluations (i.e., a group-level construct that represents ‘fundamental assessments that individuals make about their worth, competence, and capability’; Kacmar, Collins, Harris & Judge, 2009, p. 1572), moderated the effects of individual-level traits on performance and counterproductive behaviours in university football teams. Similarly, team collective efficacy has been shown to moderate some of the positive effects individual-level traits (i.e., CSE, conscientiousness and agreeableness) had on teamwork behaviours (i.e., interpersonal teamwork behaviour and performance management teamwork behaviours; Tasa, Sears & Schat, 2011).

In an attempt to tease out the process of team collective efficacy formation, Tasa, Taggar and Seijts (2007) tested a longitudinal, multilevel model and found that individual teamwork self-efficacy, individual task-relevant knowledge and team collective efficacy each predicted individual teamwork behaviour. In turn, team-level teamwork behaviour (aggregated from individual teamwork behaviour) was positively related to subsequent collective efficacy, which then predicted final team performance.

Taking individual and team efficacy into a global virtual team context, Hardin, Fuller and Davison (2007) examined virtual and generalized team efficacy beliefs in a sample of computer-mediated teams from the US and Hong Kong and found that virtuality negatively affected team members’ perceptions of team efficacy. However, this effect was buffered by individualism, such that team members from the US, an individualistic culture, perceived higher levels of group self-efficacy and virtual team self-efficacy compared to those from Hong Kong, a collectivistic culture. Given that team generalized and virtual efficacy was linked to team outcomes (satisfaction and performance), this research highlights the importance of understanding efficacy beliefs and cultural values in virtual teams.

Linking dispositional goal orientations to self-regulated learning in simulation-based team training, Dierdorff and Ellington (2012) uncovered the interaction between individual- and group-level learning and performance goal orientations in predicting leaning outcomes. In addition, Maynard, Mathieu, Marsh and Ruddy (2007) found that both individual-level and team-level resistance to empowerment climate negatively predicted individual job satisfaction, and the effect of team resistance to empowerment climate on individual job dissatisfaction was partly due to its influence on team interpersonal processes (e.g., conflict and affect management).

Using Schwartz’s (1992, 1994) values theory, Arthaud-Day and colleagues (2012) tested and demonstrated the interactive effect of individual- and group-level power and self-direction on OCB. In particular, group mean power weakened the association between individual power and OCB-I and OCB-O, whereas group mean self-direction strengthened the positive effect of self-direction on OCB-I.

Person-group fit Person-group (P-G) or person-team fit is a type of person-environment (P-E) fit that describes the degree of compatibility between individuals and their teams (Edwards, 1991; Kristof-Brown, Zimmerman & Johnson, 2005). This notion of matching individuals with teams is aligned with Schneider’s (1987) attraction-selection-attrition (ASA) framework, which suggests that individuals are more likely to be attracted to, be selected into and remain in teams that are compatible with their own attributes (Dickson, Resick & Goldstein, 2008). Applying the two most commonly adopted conceptualizations of P-E fit, P-G fit can also be categorized into complementary fit (i.e., individual attributes compensate the weaknesses or the needs of the team, and vice versa) and supplementary fit (i.e., individuals’ attributes replicate the strengths or characteristics possessed by the team; Muchinsky & Monahan, 1987).

Oh and colleagues (2014), in a meta-analysis, showed that P-G fit is positively related to organizational commitment, job satisfaction and job performance, and negatively related to intent to quit in both North America and East Asia. The effects of P-G fit on outcomes were stronger in East Asia than in North America. Results from other analyses suggest that these differential relationships can be explained by cultural values (i.e., collectivism and power distance), which influence how individuals view and value their compatibility with teams. Focusing on P-G fit along the trait of efficacy, Litrico and Choi (2013) demonstrated that individuals who perceived congruence between their self-efficacy, reflected efficacy (i.e., efficacy as perceived by team members) and team efficacy had higher levels of work collaboration engagement. In a sample of manufacturing teams in

Korea, Seong and Kristof-Brown (2012) discovered that distinct dimensions of P-G fit had differential impact on individual behaviour. In particular, KSA-based fit was positively related to knowledge-sharing, personality-based fit positively predicted voice behaviours and values-based fit was positively associated with team commitment. In addition, each dimension of P-G fit was positively linked to performance in the team.

Team diversity Due to the changing nature of the workforce, team researchers have paid increasing attention to the topic of team diversity and its impact on team effectiveness. Overall, results from this area of research have remained largely inconclusive, suggesting that team diversity might be a doubled-edged sword.

The most popular typologies of diversity differentiate between bio-demographic diversity (i.e., team heterogeneity in age, gender and race/ethnicity) that is more observable and less job-relevant, and job-relevant diversity (i.e., team heterogeneity in function, education, knowledge and skills; van Knippenberg & Schippers, 2007). It has, however, been suggested that the effect of bio-demographic diversity diminishes over time as a result of increased interactions among team members, whereas the effect of deep-level diversity amplifies over time (Bell et al., 2011; Korsgaard, Jeong, Mahony & Pitariu, 2008). On the one hand, meta-analytic results have not demonstrated a significant relationship between bio-demographic diversity and team effectiveness, whether the criterion concerns team innovation (Hulsheger, Anderson & Salgado, 2009) or team performance (Horwitz & Horwitz, 2007). On the other hand, job-relevant diversity has been shown to benefit team creativity and innovation (Bell et al., 2011; Hulsheger et al., 2009) and team performance (Horwitz & Horwitz, 2007). Although it seems reasonable to believe that job-relatedness might explain the differential effects of diversity on team performance, an extensive review of team diversity suggests otherwise (van Knippenberg & Schippers, 2007). In order to illuminate the inconsistent findings regarding team diversity, researchers have started to explore and investigate moderators in the team diversity-performance relationship. In a sample of 68 teams from China, Shin, Kim, Lee and Bian (2012) demonstrated the moderating effects of member creative self-efficacy and transformational leadership, such that cognitive team diversity benefited individual creativity only when members of the team had high creative self-efficacy or perceived their leaders as transformational.

Viewing diversity dimensions as interactive, some researchers have studied the relationship between team diversity, demographic fault-lines (i.e., ‘hypothetical dividing lines that split a team into subgroups based on one or more attributes’; Lau & Murnighan, 1998, p. 328) and team effectiveness. Meta-analytic results showed that demographic diversity (age, race, sex, tenure, functional background and education) is significantly related to demographic fault-line strength, which in turn relates to decreased team cohesion, team performance and team satisfaction, as well as increased conflict (Thatcher & Patel, 2011).

In the context of cross-cultural teams, cultural diversity has also been linked to team performance. Results from two experimental studies showed that cultural diversity had a negative main effect on dyadic performance (i.e., joint task performance by a group of two individuals working as a team) even after controlling for team average cultural intelligence, English proficiency and other types of diversity (age, gender and function; Nouri et al., 2013). But more importantly, task structure and task type moderated the relationship between cultural diversity and performance, such that the negative influence of heterogeneity in members’ cultural background diminished when the dyads performed convergent tasks (i.e., tasks that require cooperation and interdependence) with high levels of task specificity and divergent tasks (i.e., tasks that do not require high levels of cooperation or interdependence) with low levels of task specificity. In a meta-analysis, Stahl, Maznevski, Voigt and Jonsen (2010) did not find a direct link between cultural diversity and team performance in multicultural teams. However, cultural diversity negatively predicted social integration and positively predicted creativity, conflict and satisfaction, and these effects were moderated by team tenure, dispersion, size and task complexity. In another field study, the effect of cultural diversity on team performance was also found to be moderated by team members’ goal orientation (Pieterse, van Knippenberg & van Dierendonck, 2012).

In the context of virtual teams, bio-demographic diversity, particularly diversity in age and nationality, was shown to interact with process conflict and technical experience in predicting team creativity (Martins & Shalley, 2011). In addition, nationality diversity also negatively predicted team creativity, whereas diversity in sex and race was not associated with team creativity.

In sum, we have discussed the major research findings regarding team member individual and team compositions of KSAOs that facilitate or hinder team effectiveness. The vast li terature suggests that team selection should strike a balance between seeking the best individuals and the best combination of individuals with regard to their KSAOs. In the next section, we consider assessment of team member candidates based on the KSAOs discussed above.

< Prev   CONTENTS   Source   Next >