Effects of Team Personality Composition on Member Performance: A Multilevel Perspective

Abstract

Personality traits are often theorized to affect team performance by predisposing members to perform individual-level behaviors. Yet, member personality traits may also affect team performance by creating contextual influences on member behaviors. As such, the purpose of the present study was to examine the effect of team personality composition on individual-level performance using hierarchical linear modeling. A range of effects for team-level elevation were observed, but few effects emerged for team-level heterogeneity. Main effects from elevation in Extraversion and Openness to Experience were consistently observed across analyses. The main effects from team elevation in Conscientiousness and Agreeableness, however, were only observed prior to controlling for individual-level trait scores or when using a group-mean centering method for individual-level trait scores. In addition, elevation in Conscientiousness and heterogeneity in Emotional Stability moderated the relationships between individual trait scores and performance, such that individual-level relations were stronger when team elevation was higher (Conscientiousness) and heterogeneity was lower (Emotional Stability). These findings provide evidence that team member personality can influence performance through contextual phenomena.

Keywords

personality elevation hierarchical linear modeling personality diversity teamwork team composition

Team-based jobs are common in the workplace, and the essential knowledge, skills, and abilities (KSAs) for effective teamwork are the subject of great interest to researchers and practitioners alike. However, the effects of individual differences on team performance are complicated by the general multilevel nature of teams. That is, the study of team effectiveness involves constructs at multiple levels of analysis, including individual, team, and organizational levels. Furthermore, explanations of many phenomena in team settings require investigations of cross-level relationships: Individual characteristics, such as personality and cognitive ability, are integrated to create phenomena at the group level (Kozlowski & Klein, 2000). Likewise, group-level characteristics influence individual-level cognitions and behaviors (e.g., Forsythe, 2009). In making sense of this complex system, researchers have often focused on a single level of analysis, akin to the proclivity in the physical sciences to reduce phenomena down to its lowest level. Unfortunately, team-level effects cannot be exclusively explained from effects observed at the team level, and member actions and behaviors stem from group phenomena as well as individual characteristics (Hackman, 2003). By focusing only on a single level of analysis, we may be missing the forest for the trees.

This issue is particularly salient when considering the effect of personality composition in work teams. Team personality composition and team performance describe multilevel constructs, in that individual-level traits combine in some form to produce team-level performance. Conversely, team-level personality characteristics may exert top-down influences on the attitudes and behaviors of individual members. As identified by Ployhart and Schneider (2005), top-down effects may constrain or enhance work performance at the individual level as well as the relationships between individual-level predictors and criteria. Yet, the extant literature has predominantly tested either the relationships between individual traits and individual member performance (e.g., Morgeson, Reider, & Campion, 2005) or the relationships between team-level personality scores and team performance (e.g., Prewett, Walvood, Stilson, Rossi, & Brannick, 2009). Reviews on team personality composition have represented the multiple levels that personality and performance may be conceptualized, but these theoretical models typically focus on the aggregation of individual traits and behaviors to team-level phenomena (e.g., LePine, Buckman, Crawford, & Methot, 2011; Salas & Driskell, 2014). The current study expands this theoretical framework by drawing attention to potential cross-level effects that occur from team-level phenomena to individual outcomes. In this manner, the examination of top-down cross-level effects can provide a substantial contribution to our theoretical understanding of team personality composition.

The effect of team composition on individual-level performance also poses a critical issue for the use of personality in team member selection and staffing. Validation studies in the area of personality and performance are typically conducted at the individual level, with consideration given to the job context but not given to the characteristics of the team itself. If the personality traits of other team members also affect an individual member’s performance, the utility of making staffing decision on personality test scores in a team setting could be either under- or over-estimated without considering person–group (P-G) fit. Thus, an examination of both individual- and team-level personality traits in predicting member performance should clarify our understanding of the relationships between personality traits and individual performance in team settings.

With these theoretical and practical benefits in mind, the goal of the present study was to determine the degree to which (a) team-level personality composition affects individual team member performance and (b) team-level personality composition interacts with an individual member’s personality trait to affect his or her performance. Evidence for either type of effect would suggest that the individual-level test validation paradigms typically used (where one person’s score on a trait scale is correlated with that person’s performance) may not accurately estimate the usefulness of personality tests to inform staffing decisions in team settings. As a result, this study provides valuable insight in further developing our understanding of the interplay between individual traits, team composition, and team member performance.

Team Personality Composition and Team Performance

Team personality composition is defined here as a team-level index of the personality traits within the team, reflecting the strength (or elevation) of a given trait within the team and/or the heterogeneity of a trait within a team. A primary reason for creating team-level indices of personality is to relate personality characteristics to team-level criteria without inappropriately crossing levels of analysis. The elevation and diversity approaches are based on the person–environment models of supplementary and complementary fit as proposed by Muchinsky and Monahan (1987). Supplementary fit describes attributes where homogeneity is desired among team members (typically, it is preferred that each team member has a high level of a job-relevant attribute, leading to the term elevation). Complementary fit refers to attributes where heterogeneity among team members is desired. Thus, team personality elevation (TPE) refers to the overall level or strength of the trait within the team and is typically measured using the mean of member scores, whereas team personality diversity (TPD) describes the variability of a given trait within the team, and is typically examined using the variance of member scores.

Personality research has generally centered on the Five-Factor Model (FFM) of personality (Costa & McCrae, 1992). At the individual level, research has documented consistent and positive relationships with performance for the personality factors of Conscientiousness and Emotional Stability (e.g., Barrick, Mount, & Judge, 2001). These factors are theorized to promote work motivation (Conscientiousness) and emotion regulation (Emotional Stability), competencies which are relevant for a wide variety of occupations. Extraversion, Agreeableness, and Openness to Experience are positively related to performance for certain occupations, particularly ones that place a greater emphasis on interpersonal skills or adaptability (Barrick et al., 2001).

In team settings, each of the five factors have shown modest, positive correlations with performance (Bell, 2007; Prewett et al., 2009), with Conscientiousness and Agreeableness showing the strongest effect sizes. Stronger correlations have been observed in studies that used applied work samples, teams with higher interdependence, and behavioral criteria for evaluating performance (Prewett et al., 2009). In examining the effect of different aggregation methods, the meta-analysis by Prewett and colleagues (2009) suggested that TPE was positively correlated with team performance, whereas TPD showed either negative or null relationships.

However, it remains unclear whether the relationship between team personality composition and team performance is reflective of bottom-up processes, top-down processes, or some combination. Prevailing theory posits a bottom-up approach, in which that individual traits lead to individual behaviors that are favorable in a team context, and these behaviors serve as the foundation for better team outcomes (LePine et al., 2011). This framework views individual traits as the drivers of team performance, and team personality composition is merely an analytical tool for relating individual traits to team performance without inappropriately crossing levels of analysis. Although this framework provides logical connections between individual personality traits and team performance, it neglects the possibility that the team-level personality composition exerts top-down effects on team member behaviors. We next discuss the theoretical underpinnings of these potential cross-level effects.

Cross-Level Effects of Team Personality Composition

Although teams perform as a unit, team members are often evaluated as individuals in performance appraisals. Such evaluations may be made in addition to the appraisal of performance of the group as a whole, but the contribution of each individual is generally considered and remains a valuable criterion. Team composition may function as an important contextual variable that affects the performance of individual members (i.e., a cross-level main effect) or affects the relationship between a member’s personality trait and his or her performance (i.e., a cross-level interaction effect).

The presence of group-level effects upon individual behaviors is a long accepted phenomenon in psychology (Forsythe, 2009), which may come from characteristics of the group’s task or from characteristics of the group itself. With regard to task characteristics, empirical work has documented that member behaviors are affected by factors such as reward interdependence (e.g., Pearsall, Christian, & Ellis, 2010), task interdependence (e.g., Saavedra, Earley, & Van Dyne, 1993), and core task demands (Stewart & Barrick, 2000). Regarding characteristics of the team, research in social psychology has illustrated that majority opinion can influence the direction and the extremity of individual member opinions (i.e., polarization; Goethals & Zanna, 1979). As group identification increases, members become more committed to group goals and more hostile toward outside groups (e.g., Hogg, 2005). Finally, member behavior is partially governed by the strength of group norms (e.g., Gelfand, Nishii, & Raver, 2006).

The mixture of personality traits within the team should logically influence individual-level performance. Specifically, the configuration of traits among team members should function to define the climate and constraints within which they work. However, to date there has been little empirical examination of the potential cross-level effects of team personality composition (LePine et al., 2011). In one of the few studies to examine cross-level relationships, Ployhart, Weekley, and Baughman (2006) observed that job- and organizational-level elevations in Agreeableness, Conscientiousness, Emotional Stability, and Extraversion were all positively related to individual-level satisfaction and performance on the job. In another study, similarity in Conscientiousness, Agreeableness, and Emotional Stability helped mitigate relationship conflict and negative team affect, whereas heterogeneity in Extraversion was negatively related to these criteria (Tekleab & Quigley, 2014). Other research has found that individual Conscientiousness had a stronger relationship with performance in football teams with higher elevations of Conscientiousness (Schmidt, Ogunfowora, & Bourdage, 2011). These initial findings suggest that team personality composition can exert meaningful cross-level effects on individual performance, but there is still little empirical research on these effects within organizational team contexts.

Previous research on team performance has suggested that homogeneous teams tend to develop stronger norms than diverse teams (Kirkman, Tesluk, & Rosen, 2004). This finding is further bolstered by the social psychology research showing that groups tend to become more polarized, or more strongly hold certain opinions, when members perceive that they share the same abilities or opinions beforehand (e.g., Goethals & Zanna, 1979). In some cases, even a single, toxic member may lead to dysfunction in the group (Felps, Mitchell, & Byington, 2006). With this perspective, we view team elevation as a key driver of top-down effects because these groups are more likely to develop effective performance norms. TPD should, thus, show weak or null relationships with member performance, as higher TPD tends to occur with middling levels of TPE (i.e., higher or lower TPE must necessarily occur with low TPD). Because the FFM traits discussed here are expected to benefit team performance at higher levels of the trait, we expect that TPE in personality will positively relate to member performance. We next provide a more detailed rationale for hypothesized main effects of TPE on member performance.

TPE in Conscientiousness will create a context where member effort is encouraged and loafing is discouraged, thus improving the work effort from team members. For example, individuals high in Conscientiousness are more likely to engage in “backup behaviors” and other forms of contextual performance (e.g., Morgeson et al., 2005). These behaviors can further reinforce proscriptive performance norms and communicate strong performance expectations to team members (Cialdini & Trost, 1998).

We also expect TPE in Extraversion to positively relate to member performance. We note that this prediction is at odds with much of the theoretical work in this area, which cites Extraversion as a factor where team heterogeneity (TPD) is preferred to homogeneity (e.g., Humphrey, Hollenbeck, Meyer, & Ilgen, 2007). Some research posits that extraverted members within the same team would compete for status, such as leader-oriented roles, and neglect other relevant roles in the team (Barry & Stewart, 1997). However, empirical research has generally not supported TPD in Extraversion as a correlate of team performance (Bell, 2007; Prewett et al., 2009). The generally weak correlations between TPD in Extraversion and team performance suggests that competition for status can be constructive, perhaps by inducing members to give more effort toward the team task to achieve a higher status. In addition, low diversity in team Extraversion may encourage shared or collective leadership in the team, which is associated with positive outcomes (Carson, Tesluk, & Marrone, 2007). Furthermore, Extraversion involves more than dominance or need for power: It also includes sociability and friendliness, which are traits that may lead to cooperativeness and positive affectivity in the team when maximized (e.g., McNeil & Fleeson, 2006). From this perspective, we predict that TPE in Extraversion would positively relate to member performance by promoting information sharing, collaboration, healthy emulation, and achievement striving among members.

TPE in Agreeableness and Emotional Stability will create a context where members are expected to interact constructively and courteously (Ashton & Lee, 2007), which should bolster customer service as well as team member relations. Elevation in these traits is related to team cohesion and the effective management of disagreements between team members (Bradley, Klotz, Postlethwaite, & Brown, 2013). Thus, teams that are generally high on these traits may encourage a positive social and emotional environment. This environment is favorable as it serves to motivate member performance, positive team atmosphere, and team outcomes. However, low elevation in these traits can lead to a negative and tense team climate, harming individual motivation and morale in the process (Driskell, Goodwin, Salas, & O’Shea, 2006).

Finally, TPE in Openness to Experience should facilitate member adaptability and help members resolve unique or complex challenges. For example, LePine, Colquitt, and Erez (2000) found that teams with greater openness were more successful in adapting to changing contexts in a decision-making task. This result likely occurred because member adaptability is predicted by individual-level Openness to Experience, but we argue that member adaptability is also enabled by the flexibility of other team members. That is, individual flexibility and imagination may be constrained if other team members fail to reciprocate in this regard. Thus, we expect that TPE in openness will positively relate to member performance.

To summarize, the hypotheses for the main effects of TPE on member performance are as follows:

Hypothesis 1 (H1): TPA for (a) Agreeableness, (b) Conscientiousness, (c) Emotional Stability, (d) Extraversion, and (e) Openness to Experience will be positively related to individual performance.

As mentioned previously, members in teams with diverse personalities are less likely to have strong norms governing their behavior, but these lack of norms are not necessarily worse for performance than teams who have strongly dysfunctional norms. For example, teams that are homogeneously low in Conscientiousness may have strong norms for loafing and perform worse than teams lacking performance norms in either direction (i.e., weak norms). Thus, low levels of heterogeneity may indicate either a highly positive or highly negative climate, which may effectively suppress the correlation between TPD and individual performance. However, some effects for TPD may emerge after controlling for the effects of TPE. That is, when assuming an average level of TPE, is it better to have more or less heterogeneity in the examined personality trait? This possibility presents the following research question:

Research Question 1(RQ1): After controlling for TPE, does TPD significantly predict team member performance, and for which personality traits?

Team Personality Composition as a Cross-Level Moderator

In addition to main effects on member performance, team personality composition may affect member performance through cross-level interactions with individual-level traits. Previous research in team settings has found that member personality for each of the five factors has a positive correlation with individual performance (e.g., Barrick et al., 2001). We expect that these relationships will be moderated by the elevation of the corresponding trait within the team as a unit, in which the correlations between a member’s personality score and his or her performance are generally stronger in teams with higher elevation in the corresponding trait.

This type of cross-level interaction is supported by trait activation theory (TAT), which contends that the relationships between traits and performance are moderated by the presence of trait-relevant cues (Tett & Burnett, 2003). This theory presents an interactionist view of job performance, suggesting that certain situations at work allow the expression of specific underlying personality traits, which facilitate job performance (Tett & Burnett, 2003). In essence, high TPE creates situations that contain the relevant cues for expressing individual traits, whereas modest or low TPE creates situations that constrain individual trait expression.

Because team homogeneity generally corresponds to stronger norms, it seems reasonable to believe that greater heterogeneity in teams would enable the expression of individual traits within the team. This expectation would assume, however, an equal influence from different member personalities and neglects unique social pressures that may arise in diverse teams. In particular, negative traits and behaviors are often more powerful in group contexts than positive ones (Baumeister, Bratslavsky, Finkenauer, & Vohs, 2001). Because “bad is stronger than good,” it may become increasingly difficult for team members to express traits associated with effective teamwork if other team members score lower on these traits. Stemming from this logic, we argue that TPE is a key moderator for the relationship between individual-level personality and performance, whereby the relationships between individual personality traits and member performance to be stronger when TPE for that trait is high. For example, a conscientious group member should be more likely to put forth effort with other conscientious team members than with members who are low in Conscientiousness, such that individual conscientious has a stronger relationship with performance when TPE in the factor is also high.

The effect of group composition on trait expression is further illustrated in the “bad apple” effect (Felps et al., 2006), in which one or two members who are low on a beneficial personality trait may disproportionally affect the expression of those traits in other members. For example, Camacho and Paulus (1995) found that teams with two socially anxious members and two socially calm members generated almost the same number of ideas as teams composed of all anxious members, but significantly less ideas than teams composed entirely of socially calm members. The authors suggested that socially calm team members feel more comfortable expressing ideas when other members are also socially calm than when other members are anxious. Other research has found that minimum group personality scores can exert undue influence on group performance (e.g., Neuman & Wright, 1999). In short, some research has suggested that homogeneous groups encourage the expression of individual traits in common with the group. However, the expression of personality traits that are adaptive for team performance depends on the elevation of the trait.

With regard to Conscientiousness, a more conscientious team should create a context in which member Conscientiousness is encouraged and rewarded, thus eliciting greater activation in Conscientiousness for those who are predisposed to behave in such a way (Mohammad & Angell, 2003). Conversely, a low conscientious team member is more likely to withhold effort, which will yield feelings of inequity in the other team members. Thus, rather than activate behaviors associated with Conscientiousness, loafing by even one member may result in a desire by other members to reduce their contributions as a way to restore equity. This phenomenon is known as the “sucker effect” (Hart, Bridgett, & Karau, 2001).

We also argue that higher TPE in Extraversion will also form a climate that “activates” individual Extraversion or Introversion, primarily through participation in team discussions and coordinated activities. Extraverted members of the same team will likely “activate” each other’s predisposed behaviors in discussion and action, whereas extraverted members on an otherwise introverted team may feel inhibited in the presence of introverted team members. This dynamic is supported by the attraction paradigm (Byrne, 1971). Introversion, however, is likely to yield poor individual performance on teams, regardless of whether the rest of the team is introverted or extraverted. If these expectations are true, then the relationship between individual-level Extraversion and job performance will be stronger when team-level Extraversion is high and weaker when team-level Extraversion is low, provided that this trait is conducive to performance. That is, teams high on Extraversion will more likely facilitate the expression of Extraversion, whereas teams low in Extraversion will likely inhibit its expression.

As with the main effect hypotheses for TPE, Agreeableness and Emotional Stability are theorized to behave similarly due to their contributions to interpersonal relationships within the team. Empirical work has indicated that negative emotions in one individual are readily adopted by others (Barsade, 2002), and the physiological effects of negative emotions are longer lasting than those of positive emotions (Rein, McCraty, & Atkinson, 1995). Taken together, these results suggest that negative traits associated with Emotional Stability and Agreeableness, such as anger, anxiety, or mistrust, are more likely to be expressed if another team member expresses them. Conversely, teams with strongly supportive interpersonal environments will cue the expression of positive traits associated with Emotional Stability and Agreeableness (e.g., serenity, trust, and optimism). Thus, individual Emotional Stability and Agreeableness should relate to member performance when with high TPE in these corresponding factors.

Similar to the main effects hypothesized previously, team elevation on Openness to Experience should enable the expression of member flexibility. Conversely, members who are otherwise flexible may hesitate to express flexibility if other members act inflexible. If a team member is acting flexible when others are not, this should violate norms for reciprocity and perceptions of fairness (Blau, 1964), thus discouraging member flexibility in future work interactions.

Although the TAT framework for interactions between personality traits defined and measured at multiple levels has been frequently cited, researchers have only begun testing moderating effects for team-level phenomena on individual-level relationships. As such, we could only locate a single study that proposed moderation hypotheses between a trait at the group and individual levels. In this study, the authors found that the positive relationship between individual-level Conscientiousness and performance was stronger for members of more highly conscientious teams (Schmidt et al., 2011). However, these researchers also found that individual-level Extraversion was more strongly related to counterproductive work behaviors in high TPE teams. These initial results are a promising attempt to disentangle the effects of individual and team composition variables, but they also clearly indicate the need for further research. In the current study, TPE is expected to moderate the individual-level relationship between the FFM traits and performance such that higher elevations would lead to stronger effects at the individual level.

Hypothesis 2 (H2): TPA for (a) Agreeableness, (b) Conscientiousness, (c) Emotional Stability, (d) Extraversion, and (e) Openness to Experience will moderate the relationship between individual personality traits and performance such that the relationship between individual-level traits and performance will be stronger within teams with higher elevations.

Given the expected interactions with TPE, the interaction between TPD and member personality is likely confounded by TPE. After controlling for the effect of TPE, TPD may interact with individual personality traits, but the direction of this interaction is debatable. Assuming equivalent levels of TPE, trait diversity may lead to weaker norms for behavior, allowing greater expression of individual personality traits. However, trait diversity may decrease psychological safety and create other social pressures that reduce the expression of individual personality traits, such as impression management or social matching. As TPD decreases at higher levels of TPE, it is even possible that a three-way interaction accounts for variance in individual performance, in which individual personality traits show positive relationships with individual performance when TPD is low and TPE is high. Thus, TPD present a complex system of relationships with individual personality and TPE that is difficult to predict. Nevertheless, we investigate this possibility for these interactions through the following research question:

Research Question 2 (RQ2): After controlling for TPA, is there a cross-level interaction between TPD and the individual personality trait score?

Method

Participants

The sample used in the present study was based on data originally collected by Neuman, Wagner, and Christiansen (1999) from 82 teams working for a large retailing organization with stores located across the United States. Seventy-six percent of the sample were males, who averaged 13.5 years of education, 3.2 years of employment, and 37 years of age. Females had an average of 14.2 years of education, 3.7 years of work experience, and 34 years of age. Each of the 82 teams consisted of four members who were responsible for all tasks within their given department, yielding a total sample of 328 individual workers. There were no missing data in this sample.

These retail teams received goals, performance evaluations, and incentives at both the individual and team level, such that teams were interdependent in goals and outcomes. Teams were also task interdependent, in that the actions of one employee often required an adjustment in the actions of another employee. For example, if one employee was on the phone with a customer, the other employees had to adjust their priorities to handle incoming store customers. Thus, team processes such as goal-setting, performance monitoring, coordination, and backup behaviors are necessary components for effective staff performance. The average department tenure for workers was 3.4 years.

Measures

Personality

Employee personality was assessed using scale composites from the California Personality Inventory (CPI; Gough, 1987) and the Personal Audit (PA; Science Research Associates, 1989). Personality indices were created by standardizing the inventory scores for personality traits relevant to the FFM. Standardized scores for the CPI and PA scales were then averaged together to yield composite scores for each of the five factors, consistent with the approach reported by Neuman and colleagues (1999). Specifically, Extraversion was computed by combining “seriousness” from the PA and “dominance,” “sociability,” and “social presence” from the CPI. Emotional Stability was created by adding “emotionality” from the PA to the “self-control” and “well-being” scores from the CPI. Openness to Experience was assessed by summing “stability” from the PA with “self-acceptance” and “flexibility” from the CPI. Agreeableness was measured by combining “tolerance” from the PA with “psychological mindedness” and “socialization” from the CPI. Finally, Conscientiousness was computed by adding the PA scale of “firmness” to the CPI scales of “achievement via independence,” “achievement via conformance,” “status,” and “responsibility.” A factor analysis of the FFM yielded acceptable confirmatory fit index (CFI = .92) and root mean square error of approximation (RMSEA = .09) values to demonstrate fit, as well as acceptable factor loadings for the individual scales (Neuman et al., 1999).

Because team personality composition variables are constructed using an additive model (Chan, 1998), agreement indices such as r_wg or AD_m were not necessary to justify aggregation for these variables. TPE was assessed by calculating the mean of individual team member trait scores within each team. TPD was calculated by determining the standard deviation of trait scores among the members of each team such that higher scores indicated greater heterogeneity. Because facets from different scales were used to form the personality factors, each facet score had to be standardized prior to computing the factor score. After summing the facet scores, the means for all personality variables were zero, and the standard deviations ranged from 2.66 (Openness) to 4.51 (Conscientiousness). The standard deviations were not equal to 1 because the variables reflected a summation of z scores and not actual z scores. For more information on means, standard deviation, and factor loadings for the personality variables, please refer to the original article (Neuman et al., 1999).

Job performance

Employee job performance was assessed using two performance measures, each with several items on a 10-point rating scale. The first measure was a rating of customer service over a 1-month period, provided by human resources personnel (M = 4.62, SD = 1.75). The second measure was a rating of task performance based on supervisor evaluations on how frequently the employee completed work on time over the same period (M = 4.73, SD = 2.30). These measures were combined as an attempt to minimize the threat of common method bias and to reflect the results of a job analysis that was performed prior to data collection in the original study. The internal consistency coefficient of the combined scale indicated that the performance composite was reliable (α = .88). Unfortunately, the original items of the team member performance were proprietary, such that sample items cannot be reproduced in this article.

Analyses

As individual member performance scores were nested within groups, hierarchical linear modeling (HLM) was used to test study hypotheses. HLM allows for the testing of individual- and group-level relationships within a single model without inflating the Type 1 error rate by ignoring the dependency among team member data points (Hox, 2010). We used grand mean centering for testing main effects (H1 and RQ1) and group-mean centering for testing cross-level interactions (H2 and RQ2), as recommended by several researchers (Aguinis, Gottfredson, & Culpepper, 2013; D. A. Hoffman & Gavin, 1998).

All HLM models were run using full maximum likelihood (FML) estimation. Although restricted maximum likelihood (RML) is more commonly used in HLM, the differences in these estimation methods is generally trivial when using greater than 50 Level 2 units (Browne & Draper, 2000; Hox, 2010). The Level 1 and Level 2 residuals were also approximately normally distributed, satisfying an important assumption for FML estimation. Prior to hypothesis testing, the null model was estimated to observe the degree to which employee performance was nested within and between teams. The results identified a significant amount of variance in performance between teams, τ₀₀ = 0.95, χ²(81) = 201.99, p < .001, with 27% of the variance in individual performance occurring between teams.

After establishing the need for HLM, the next analysis examined extent to which team personality composition accounted for variance in performance beyond individual personality. Researchers have acknowledged the potential for mean scores to confound the relationship between heterogeneity indices and study criteria (Harrison & Klein, 2007). Thus, individual personality scores were first entered into the model without other predictors. Then, TPE and TPD were added into subsequent models both individually and conjointly. The hypothesized main effects for TPE (H1) were conducted using a one-tailed test to reflect the specific direction expected in results, whereas all other analyses, reflecting research questions and/or cross-level interactions, were evaluated using a two-tailed test.

After estimating the main effects at Level 2, cross-level interactions between team personality composition (TPD and TPE) and individual personality were tested. Before entering the interaction terms, we compared fixed slope models with random slope models for each FFM trait. Results of the Wald’s chi-square test indicated significant slope variance across teams for all the personality traits with the exception of Openness to Experience, τ₁₁ = 0.018, χ²(81) = 87.11, p = .30. Therefore, cross-level interactions were not tested for this trait. We reported unstandardized coefficients for the cross-level interactions because standardizing the coefficients from interaction terms requires standardizing the predictor variables prior to analysis (Friedrich, 1982). However, using z scores of the predictor variables for analysis can bias the estimates of slope variance in HLM (Hox, 2010). Although facet scores were already standardized, we did want to risk further biasing the slope variances by standardizing the personality factor scores. Therefore, we reported unstandardized coefficients for the interaction terms and estimated variance accounted for by using ordinary least squares (OLS) regression with a cross-level term (D. A. Hoffman, Morgeson, & Gerras, 2003), as OLS does not bias the effect size with nested data as it does the standard errors and significance test.

Finally, supplemental analyses were performed as a robustness check, so as to ensure that the results from HLM were not biased by the overlap between the individual- and team-level personality scores. These analyses used OLS by regressing member performance onto “self” and “other” team member personality scores. OLS with cluster-adjusted standard errors were also run to account for the effect of nesting in these analyses. Results from these supplemental analyses are provided in the appendix.

Results

Hypothesis Testing

Correlations among all individual- and team-level variables are reported in Table 1. We note that the zero-order correlations between team personality composition and individual performance were computed by standardizing the estimates yielded from HLM (due to the different levels of measurement between the variables). These coefficients should be considered baseline estimates of the cross-level main effects from team personality composition. H1 stated that the TPE for all FFMs (Agreeableness, Conscientiousness, Emotional Stability, Extraversion, and Openness to Experience) will be positively related to performance. Prior to controlling for other predictors, TPE was positively related to performance for all five factors (Table 1). With regard to TPD, only Agreeableness was found to be even marginally related to individual performance (β = –.13, p = .065, two-tailed).

Table 1.

Correlation Matrix for Individual- and Team-Level Variables.

Variables	1	2	3	4	5	6	7	8	9	10
Individual level
1. Agreeableness
2. Conscientiousness	.02
3. Emotional Stability	.01	.14
4. Extraversion	.21*	.11	.05
5. Openness to Experience	.08	.18*	.01	−.01
6. Job performance	.47**	.47**	.21**	.17*	.34**
Team level
1. TPE Agreeableness
2. TPD Agreeableness	−.39**
3. TPE Conscientiousness	−.12	.04
4. TPD Conscientiousness	−.14	.23*	−.39**
5. TPE Emotional Stability	−.06	.11	.47**	−.04
6. TPD Emotional Stability	−.22*	.19	.31**	.10	.11
7. TPE Extraversion	.28*	−.03	−.02	−.03	.19	−.25*
8. TPD Extraversion	−.14	.14	.29**	−.02	.10	.55**	−.26*
9. TPE Openness	.11	−.16	.29**	−.39*	−.01	−.15	.02	.04
10. TPD Openness	−.21	.24*	.07	.31*	.19	.10	−.02	.02	−.27*
11. Member performance^a	.37**	−.13*	.32**	−.10	.16*	.09	.18*	.09	.31**	−.06

Note. n = 328 for Level 1 variables, n = 82 for Level 2 variables. Significance tests for individual-level correlations were conducted using HLM due to data dependency. TPE = team personality elevation; TPD = team personality diversity; HLM = hierarchical linear modeling.

Relationships between team-level personality and individual member performance were estimated using standardized coefficients from HLM, these coefficients represent the relationship between a team-level measure of personality and individual performance, independent of any other variable.

p < .05. **p < .01.

After controlling for individual trait and TPD scores, TPE in Extraversion (β = .17, p = .018, one-tailed) and Openness to Experience (β = .20, p = .005, one-tailed) were both related to individual performance, with greater TPE among team members associated with better individual performance on the job. In contrast, TPE for the other FFM dimensions yielded relationships in the predicted direction, but they were not statistically significant.

Considering these results collectively, H1 was partially supported; TPE in Extraversion and Openness to Experience were positively related to individual performance. However, positive relationships between individual performance and TPE in Agreeableness and Conscientiousness were substantially diminished after controlling for individual-level scores for these traits.

RQ1 investigated whether TPD predicted team member performance after controlling for TPE and for which traits. We found that only TPD in Extraversion (β = .17, p = .031, two-tailed) accounted for incremental variance in performance.

Moderating Effects of Personality Composition

H2 proposed that TPE would moderate the relationships between individual-level trait scores and performance, such that these relationships would be stronger within teams with higher TPE scores of the corresponding trait (Table 2). Results supported this hypothesis only with regard to Conscientiousness (γ = .04, ΔR² = .04, p = .028, two-tailed). As shown in Figure 1, the positive relationship between individual personality and performance was stronger within teams with greater TPE.

Table 2.

Cross-Level Interactions Between Individual and Team-Level Personality.

	Random intercept and slope (Model 4)	Cross-level interactions (Model 5)
Agreeableness
TPE × Individual (γ₁₁)		−0.01 (.03)
TPD × Individual (γ₁₂)		−0.02 (.05)
Slope variance (τ₁₁)	0.05	0.05
ΔR² TPE		0.00
ΔR² TPD		0.00
Conscientiousness
TPE × Individual (γ₁₁)		0.03** (.01)
TPD × Individual (γ₁₂)		−0.01 (.02)
Slope variance (τ₁₁)	0.02	0.01
ΔR² TPE		0.03
ΔR² TPD		0.00
Emotional Stability
TPE × Individual (γ₁₁)		0.06 (.03)
TPD × Individual (γ₁₂)		−0.10* (.04)
Slope variance (τ₁₁)	0.11	0.11
ΔR² TPE		0.00
ΔR² TPD		0.01
Extraversion
TPE × Individual (γ₁₁)		−0.01 (.02)
TPD × Individual (γ₁₂)		−0.01 (.03)
Slope variance (τ₁₁)	0.06	0.06
ΔR² TPE		0.00
ΔR² TPD		0.00

Note. All parameter estimates are unstandardized (standard errors in parentheses); ΔR² was estimated using OLS regression. Levels 1 and 2 main effects (Table 3) were included in the hierarchical model but are not reported here. TPE = team personality elevation; TPD = team personality diversity; OLS = ordinary least squares.

p < .05 (two-tailed). **p < .01 (two-tailed).

Figure 1.

Moderating effect of team elevation in Conscientiousness on the relationship between individual Conscientiousness and job performance.

Finally, RQ2 queried cross-level interactions between TPD and individual personality trait after controlling for TPE. A significant moderating effect was found for TPD in Emotional Stability on the relationship between individual Emotional Stability and performance (γ = –.10, ΔR² = .01, p = .011), though this only accounted for an additional 1% of the variance in performance. Illustrated in Figure 2, the relationship between Emotional Stability and performance was stronger within more homogeneous teams (lower TPD). However, no meaningful cross-level interactions were found for the other personality factors.

Figure 2.

Moderating effect of team diversity in Emotional Stability on the relationship between individual Emotional Stability and job performance.

Supplemental Analyses

Results from the OLS analyses yielded highly similar effect sizes as those from the HLM analyses (see the appendix). Of particular interest, the effect sizes from TPE in Extraversion and Openness to Experience were similar to those produced in HLM, as was the interaction between individual Conscientiousness scores and the Conscientiousness scores from other team members (excluding the self-score). The results from the OLS regression with cluster-adjusted standard errors also replicated results, with one exception. Whereas the effect of “Other SD” in Emotional Stability was significant with standard OLS (p = .031), the effect was not statistically significant with adjusted standard errors (p = .085).

In addition, we suspected that different facets of Extraversion may show differential relationships between TPE and TPD, as some personality researchers have found that the facets of Extraversion show different relationships with work criteria (Hogan & Hogan, 2007). In addition, Extraversion has yielded weak coefficients in past research (e.g., Prewett et al., 2009), which could result from different facets operating differently within the team, similar to the facet-level research at the individual level (Hogan & Hogan, 2007). Thus, supplementary analyses were conducted on specific facets of Extraversion: dominance, sociability, and seriousness. These analyses first examined the effect of TPE and TPD for these facets without other variables in the model, and then with the other variables. Before controlling for other variables, TPE in dominance (β = .13, p = .048, two-tailed) and sociability (β = .35, p < .001, two-tailed) were significant predictors of member performance. After controlling for other variables, none of the TPE composition variables was significant.

For TPD, sociability was found to be negatively related to individual-level performance (β = –.17, p = .016, two-tailed), whereas TPD in dominance (β = .15, p = .042, two-tailed) was positively related to member performance. After controlling for both individual-level and TPE scores for the facet, TPD in dominance (β = .14, p = .050, two-tailed) maintained a significant relationship, but the effect of sociability was weak and not statistically significant.

For seriousness, both TPE (β = .18, p = .035, one-tailed) and TPD (β = .20, p = .048, two-tailed) emerged as a significant, though modest, predictors when entered into the same model. These results suggest that seriousness operated similar to dominance.

To summarize these results, the sociability trait operated quite similar to Conscientiousness and Agreeableness, in which TPE showed generally positive relationships with performance and TPD showed either negative or null relationships with performance. As with the team-level variables in Conscientiousness and Agreeableness, TPE and TPD in sociability did not account for a significant amount of performance variance after controlling for the individual-level scores of the trait. The dominance and seriousness facets, however, showed results that were more in line with those of Extraversion, with positive relationships for TPD and, in the case of dominance, a positive relationship between TPE and performance, even after controlling for individual-level trait scores.

Discussion

This study examined the extent to which an individual’s job performance is influenced by the personality of their teammates. Our findings demonstrated that team member performance is affected not only by one’s own personality but also by the personality composition of the team as a whole, though the nature of these effects appears complex.

Cross-Level Main Effects

When estimating the relationships between composition and performance without controls, we observed strong team-level effects for TPE in Agreeableness and Conscientiousness (Table 1). After controlling for team composition variables, TPE in Agreeableness and Conscientiousness held weak, non-significant relationships with performance, whereas individual-level scores on Agreeableness and Conscientiousness maintained significant relationships with performance. These results suggest that Agreeableness and Conscientiousness primarily operate at the individual level in team-based jobs. That is, the significant zero-order relationships between TPE in Conscientiousness and Agreeableness likely occurred because individuals higher in Agreeableness and Conscientiousness receive higher performance ratings.

By contrast, a significant effect of TPE on individual performance was observed for teams with higher levels of Openness to Experience. The current sample of retail workers may especially benefit from team flexibility, such that more flexible teams promote individual flexibility, encouraging members to adjust their behaviors to appropriately match the different types of interactions with customers, management, and other team members (Paulhus & Martin, 1988). This point is especially relevant, as a portion of our performance outcome was assessed by ratings of customer interactions.

Without controlling for any other variables, the relationship between TPE and member performance was significant, whereas the relationship between TPD in Extraversion and member performance was not (Table 1). However, researchers have previously noted that the variance index is confounded with the mean index due to their nonlinear association (e.g., Harrison & Klein, 2007). Harrison and Klein (2007) suggested that the effect of within-group variance is most meaningful after controlling for mean scores. When estimating a main effect while controlling for another variable, the controlled variable is essentially fixed to its mean score so long as the controlled variable has been centered prior to analysis (Hox, 2010). Thus, results from the HLM analyses suggested that among teams with an average elevation score in Extraversion, heterogeneity among these scores positively related to individual performance (Table 3). It is worth noting that previous research on team personality composition has been predominantly correlational, such that it has typically not controlled for mean scores when estimating the effect of heterogeneity (e.g., Barrick, Stewart, Neubert, & Mount, 1998). Thus, the current study provides some unique evidence that within-team diversity in Extraversion can provide some benefit to job performance.

Table 3.

Incremental Effects of TPE and TPD on Member Performance With Grand Mean Centering.

	Model 1	Model 2	Model 3
Agreeableness
Individual (γ₁₀)	.46**	.43**	.43**
TPE (γ₀₁)		.05	.06
TPD (γ₀₂)			.02
ΔR²	.34*	.00	.00
Conscientiousness
Individual (γ₁₀)	.54**	.51**	.51**
TPE (γ₀₁)		.02	.09
TPD (γ₀₂)			−.01
ΔR²	.34*	.00	.00
Emotional Stability
Individual (γ₁₀)	.23*	.23*	.23*
TPE (γ₀₁)		.07	.03
TPD (γ₀₂)			.09
ΔR²	.16*	.02	.03
Extraversion
Individual (γ₁₀)	.15*	.11	.10
TPE (γ₀₁)		.12	.17*
TPD (γ₀₂)			.17*
ΔR²	.16*	.03	.09*
Openness to Experience
Individual (γ₁₀)	.32**	.26**	.26**
TPE (γ₀₁)		.17**	.17**
TPD (γ₀₂)			.00
ΔR²	.45*	.08*	.01

Note. All coefficients are standardized; ΔR² for Model 1 indicates the intercept (between-teams) variance accounted for beyond the baseline model; ΔR² for Models 2 and 3 indicates the intercept (between-teams) variance accounted for beyond Model 1. TPE = team personality elevation; TPD = team personality diversity.

*p < .05 (one-tailed). **p < .01 (one-tailed).

Although statistically significant, the cross-level main effects from team-level Extraversion were generally modest. The low magnitude of the regression coefficients with Extraversion can be partially explained through our facet-level analysis of this factor. The results of exploratory analyses found a relatively strong association between TPE in sociability and individual-level performance, whereas elevation in other facets showed null to relatively modest relationships with performance. Likewise, heterogeneity in sociability was not related to individual-level performance, but heterogeneity in dominance and seriousness was associated with better performance. It is not surprising, then, that combing these facets into an Extraversion score would yield an overall modest association between TPE/TPD and performance.

The results from this study suggested that facets associated with friendliness provided supplementary fit, whereas those facets more closely associated with dominance yielded complementary fit. This is consistent with past findings that dominance facilitates greater role clarity through the delineation of leadership and followership roles (Bono & Judge, 2004), but supplementary levels of sociability may lead to more effective interpersonal interactions due to the high overall levels of friendliness within the group (Barrick et al., 1998). The results may further explain the weak meta-analytic relationships observed between team-level Extraversion and team performance (e.g., Bell, 2007; Prewett et al., 2009). Researchers and practitioners in the field of team performance management may be better served by breaking Extraversion down into sociability and dominance components for analyses, a practice that is already done by several personality inventories (e.g., Hogan & Hogan, 2007).

Cross-Level Interactions

There was a significant cross-level interaction between individual-level Conscientiousness and TPE in Conscientiousness. Individual-level Conscientiousness exhibited a positive relationship with individual-level performance in teams when other members were also high in Conscientiousness. In teams that were generally low in Conscientiousness, individual-level Conscientiousness held a weaker relationship with member performance. This contextual effect is likely best explained by the sucker and free-riding effects (e.g., Hart et al., 2001; J. R. Hoffman & Rogelberg, 2001). In low conscientious teams, the sucker effect would describe a conscientious member who fears being burdened with most or all the work, or otherwise playing the “sucker.” To avoid being taken advantage of, the conscientious member may actually withhold more effort than is typical for them, thus attenuating the relationship between their Conscientiousness scores and their performance level (Hart et al., 2001). However, the sucker effect does not explain why highly conscientious teams failed to raise the performance of individual members who were low in Conscientiousness, which is apparent from the lack of a main effect from TPE in Conscientiousness. This effect may have resulted from free riding, in which low conscientious members perceive that they can loaf in their work because the more conscientious members will carry the team (J. R. Hoffman & Rogelberg, 2001).

A cross-level interaction was also found for TPD in Emotional Stability, in which individual-level Emotional Stability was more strongly related to performance in teams with less heterogeneity in this trait. These findings are consistent with the notion that team members are linked in affect, similar to the concept of emotional contagion (Barsade, 2002). Longitudinal research has found that individual affect is affected by group affect over time, even after controlling for other factors (Ilies, Wagner, & Morgeson, 2007). These results may have occurred because homogeneity in trait affectivity, as captured by Emotional Stability, likely strengthens the affect linkages among team members. Similarly, Camacho and Paulus (1995) found that participants who scored low on trait anxiety measures were more participative in groups with similarly calm individuals, but less participative when other group members were socially anxious. Taking these findings together, homogeneity in Emotional Stability may encourage the individual expression of that trait, such that it is more strongly related to performance in teams that share a similar version of the trait.

However, the effect of TPD in Emotional Stability and member performance requires further investigation before firm conclusions can be drawn. The interaction between individual-level and TPD scores in Emotional Stability yielded a weak effect size in all analyses (ΔR² = .01) and the interaction was not significant when using OLS regression with cluster-adjusted standard errors. There are different reasons to trust either the cluster-adjusted OLS regression or the HLM results. On one hand, OLS addressed the overlap between the individual score and the SD of the team personality trait by building the SD term around a mean score that did not include the focal individual. On the other hand, team heterogeneity in the OLS regression analyses is only represented by three members instead of four. By removing the score of the individual from the heterogeneity measure, one fails to include the distance between the focal score and the rest of the team, which is an important component when estimating the effect of within-group heterogeneity on individual behavior. In short, one could argue that the metric of within-team diversity lacks construct validity when removing a score from the team.

Implications for Theory

Our findings provide several theoretical implications in the study of personality composition. By modeling personality at both the individual and team levels, we observed that member performance is uniquely influenced by personality factors across both levels. As such, personality traits appeared to influence member performance in a variety of ways: by directly influencing member behavior, by encouraging the development of effective norms for teamwork and production, and even by affecting how team members are evaluated against their peers. These findings suggest that we expand our existing theoretical frameworks of team personality composition to include the different mechanisms by which personality may affect team performance.

The cross-level interactions were consistent with the TAT framework (Tett & Burnett, 2003), in which individual members will express a personality trait if the other group members show a similar level of that trait. We argue that team personality composition provides the basis for the social cues and performance norms to individual members, which in turn affects the expression of individual personality traits. Based on these findings, we can infer that team personality composition can strengthen or weaken the relationship between individual-level personality and performance.

As suggested by Hackman (2003), a more robust explanation for the effects of personality in work teams is achieved when examining the effects at both individual and team levels. Yet, researchers have generally overlooked the potential for cross-level effects (LePine et al., 2011). This lack of research is troubling, as Ployhart and Schneider (2005) stated that “to ignore the nested nature of selection within a multilevel system is to ignore the very basis of organizational science” (p. 513). Our results suggest that both bottom-up and top-down effects are generated from the composition of team member personalities. Thus, a more nuanced understanding of the relationship between personality and behavior can be reached by modeling these effects across multiple levels of measurement (LePine et al., 2011).

Recently, Ployhart and Moliterno (2011) have promoted the use of multilevel approaches, particularly to study the effects of human capital resources at various organizational levels. To incorporate research dispersed at different organizational levels, the authors defined human resource capital as the unit-level emergence of individual-level KSAs. Utilizing this approach, research at the individual and team levels can be consolidated into a unified area of multilevel research. This effort may provide a useful template for future research seeking to integrate multiple levels of personality and performance into a coherent framework.

Implications for Practice

Beyond advancing the literature of personality in work teams, our study also provides several implications for practice. As work teams are commonly utilized in organizations, multilevel approaches will be appropriate in staffing and personnel selection (Ployhart & Schneider, 2005). The results for Conscientiousness and Openness suggest there may be added importance in maximizing these traits in team staffing. Based on the weak effects from TPE in Emotional Stability and the cross-level interaction between TPD and individual-level Emotional Stability, it may be more advisable to compose teams with similar levels of Emotional Stability rather than focusing on maximizing this trait. Regarding Extraversion, our findings suggest that practitioners should first consider the elevation among team members when staffing, particularly sociability. Heterogeneity in Extraversion should be emphasized for dominance or other facets related to leadership roles (to the extent that it is possible). Finally, our study did not uncover any top-down effects from Agreeableness, but this is not to say that Agreeableness is unimportant in team settings. Rather, the effect of Agreeableness appears to operate primarily at the individual level, demonstrating a bottom-up process more so than the other personality factors. Thus, practitioners are still advised to maximize this factor, as Agreeableness showed strong relationships with performance at the individual level.

More broadly speaking, personality is of particular value in the selection and building of teams in that its impact on performance spans multiple organizational levels. Because the personality composition of an existing team may determine which traits would be best suited for an incoming member, team-level personality is a vital consideration when selecting members into teams. Based on the present personality composition of the team, certain traits may be more or less valid in predicting performance of newly selected team members. Taking multiple organizational levels into account should result in more accurate decisions and greater utility from personality assessments in selection and staffing.

Limitations and Future Directions

Although the current study provides valuable insight into the multilevel nature of personality in work teams, there are a few limitations that suggest caution when interpreting results. First, task performance measures were the only criteria available in the data. Even though most research has only examined personality in regard to task outcomes, process behaviors have been supported as a more proximal outcome of personality (Prewett et al., 2009). In addition, more studies are needed to formally test the mediating effects of processes between personality and more results-oriented outcomes. Therefore, the inclusion of teamwork processes (e.g., Marks, Mathieu, & Zaccaro, 2001) and team roles (e.g., Mumford, Van Iddekinge, Morgeson, & Campion, 2008) will be vital in bridging the relationship between personality and performance. In particular, job analysis data should be referenced to identify the processes most relevant to performance and personality at individual and team levels. Future research should also investigate whether personality has similar, multilevel effects across different team and task types, based on their varying job demands and task characteristics

A second limitation stems from the small number of individuals per team, which created substantial overlap between individual- and team-level personality scores. This overlap was addressed through different analytical methods: controlling for individual-level scores when estimating main effects of TPE and TPD, using group-mean centering when computing interactions, and providing supplemental regression analyses using self and other personality scores. However, these different analytical methods each had disadvantages, and the distinction between effects from individual- and team-level personality scores could be more accurately assessed in larger groups. In particular, the overlap between individual- and team-level scores when using grand mean centering inhibited our ability to assess the relationships between TPE and individual performance, most notably for Conscientiousness and Agreeableness.

An additional limitation can be found in the measurement of personality. It is possible that the rescaling procedures used to combine scale scores served to bias the estimates of slope variance. In general, any transformation that changes the scaling of a variable has the potential to affect the estimate of variability between the predictor slopes for different groups (Hox, 2010). Such biases are typically small, but caution is warranted when interpreting the estimates of slope variance due to the rescaling procedures used to create the personality scores.

Given the already large scope of the study, the current study examined each personality trait in isolation, which did not account for the overlap among personality traits. However, the correlations between personality traits were generally low, such that it is unlikely that the relationships with member performance were confounded by other personality variables. One exception is the association between TPE in Extraversion and Conscientiousness were correlated (r = .29). However, when running a model that included both personality traits, the results were highly similar to those reported in the tables. Regardless, one potential avenue for future research in this area is to examine additive and interactive models between different personality traits

Conclusion

Several conclusions can be drawn from this study. First, the inclusion of traits measured at both the team and individual level provides a more complete validation of personality tests with regard to the prediction of job performance. Beyond displaying stronger relationships at the individual level, Extraversion and Openness had significant effects at the team level. The effects from Extraversion were best explained by separately examining the sociability and dominance facets within that factor. We also provided evidence that the relationship between individual-level Agreeableness and Conscientiousness stems not only from direct effects on work-related behaviors but also from comparative effects between team members. Finally, meaningful cross-level interactions were found for Conscientiousness and Emotional Stability. These interactions suggest that, beyond their ability to predict performance at the individual level, these traits also affect the team environment, leading to stronger or weaker trait activation in those contexts. The results of this study confirm that personality is a valuable predictor that can be used for selection and staffing of work teams. Not only do these traits predict performance, but they also appear to contribute to the development of team norms and can influence the validity of trait scores at the individual level. Gaining a more sophisticated understanding of the interplay between these team- and individual-level phenomena equally benefits theory and practice.

Footnotes

Appendix

Supplemental Analyses for TPA Variables.

	r	OLS regression		OLS regression with cluster-adjusted SE
	r	β	SE	β	SE
Agreeableness	.47**	.45**	0.03	.44**	0.04
Other Agreeableness	.26**	.04	0.01	.04	0.01
Interaction		−.05	0.10	−.05	0.11
Conscientiousness	.47**	.47**	0.02	.49**	0.03
Other Conscientiousness	.18**	.02	0.01	−.00	0.01
Interaction		.20**	0.09	.19**	0.12
Extraversion	.17**	.16**	0.03	.16*	0.04
Other Extraversion	.12*	.10*	0.02	.11	0.02
Interaction		−.07	0.10	−.06	0.15
Emotional Stability	.21**	.20**	0.04	.20**	0.05
Other Emotional Stability	.08	.06	0.02	.07	0.02
Interaction with mean of other Emotional Stability		.06	0.10	.05	0.10
Interaction with SD of other Emotional Stability		−.09*	0.02	−.09	0.03
Openness	.34**	.31**	0.04	.31**	0.05
Other Openness	.20**	.14**	0.02	.13*	0.02
Interaction		.07	0.09	.06	0.13

Note. “Other” personality scores were obtained for each team member by computing the mean and standard deviation of scores among the other three members of that person’s team; standard errors and significance tests were calculated using unstandardized variables; standardized coefficients (β) were produced using standardized variables; ΔR² for the Conscientiousness interaction was .04; ΔR² for the Emotional Stability SD interaction was .01. TPA = team personality elevation; OLS = ordinary least squares.

p < .05 (two-tailed). **p < .01 (two-tailed).

Associate Editor: William Gardner

Authors’ Note

A portion of data used in this study were previously published in Neuman, Wagner, and Christiansen (1999). An earlier version of this research was presented in a symposium at the 28th annual conference for the Society for Industrial and Organizational Psychology, April 2013, Houston, Texas.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author Biographies

Matthew S. Prewett is an assistant professor in the department of Psychology at Central Michigan University. He received his PhD in industrial-organizational psychology from the University of South Florida. His research interests include team performance, group dynamics, and teamwork processes.

Matthew I. Brown is a research scientist at Wonderlic, Inc. He received his PhD in Industrial-Organizational Psychology from Central Michigan University. His research interests include personnel selection, testing and assessment, and multilevel modeling.

Ashita Goswami is an assistant professor of psychology in the department of Psychology at Salem State University. She earned her PhD in industrial-organizational psychology from Central Michigan University. Her research interests include followership, managerial cognition, leader behaviors and study of individual differences.

Neil D. Christiansen is a full professor in the department of Psychology at Central Michigan University. He received his PhD in industrial-organizational psychology from Northern Illinois University. His research interests include employment testing, methodological and legal issues in personnel selection, and organzational uses of personality inventories.

References

Aguinis

Gottfredson

R. K.

Culpepper

S. A.

(2013). Best-practice recommendations for estimating cross-level interaction effects using multilevel modeling. Journal of Management, 39, 1490-1528.

Ashton

M. C.

Lee

(2007). Empirical, theoretical, and practical advantages of the HEXACO model of personality structure. Personality and Social Psychology Review, 11, 150-166.

Barrick

M. R.

Mount

M. K.

Judge

T. A.

(2001). Personality and performance at the beginning of the new millennium: What do we know and where do we go next? International Journal of Selection and Assessment, 9, 9-29.

Barrick

M. R.

Stewart

G. L.

Neubert

M. J.

Mount

M. K.

(1998). Relating member ability and personality to work-team processes and team effectiveness. Journal of Applied Psychology, 83, 377-391.

Barry

Stewart

G. L.

(1997). Composition, process, and performance in self-managed groups: The role of personality. Journal of Applied Psychology, 82, 62-78.

Barsade

S. G.

(2002). The ripple effect: Emotional contagion and its influence on group behavior. Administrative Science Quarterly, 47, 644-675.

Baumeister

R. F.

Bratslavsky

Finkenauer

Vohs

K. D.

(2001). Bad is stronger than good. Review of General Psychology, 5, 323-370.

Bell

S. T.

(2007). Deep-level composition variables as predictors of team performance: A meta-analysis. Journal of Applied Psychology, 92, 595-615.

Blau

P. M.

(1964). Exchange and power in social life. New York, NY: John Wiley.

10.

Bono

J. E.

Judge

T. A.

(2004). Personality and transformational and transactional leadership: A meta-analysis. Journal of Applied Psychology, 89, 901-910.

11.

Bradley

B. H.

Klotz

A. C.

Postlethwaite

B. E.

Brown

K. G.

(2013). Ready to rumble: How team personality composition and task conflict interact to improve performance. Journal of Applied Psychology, 98, 385-392.

12.

Browne

W. J.

Draper

(2000). Implementation and performance issues in the Bayesian and likelihood fitting of multilevel models. Computational Statistics, 15, 391-420.

13.

Byrne

(1971). The attraction paradigm. New York, NY: Academia Press.

14.

Camacho

M. L.

Paulus

P. B.

(1995). The role of social anxiousness in group brainstorming. Journal of Personality and Social Psychology, 68, 1071-1080.

15.

Carson

J. B.

Tesluk

P. E.

Marrone

J. A.

(2007). Shared leadership in teams: An investigation of antecedent conditions and performance. Academy of Management Journal, 50, 1217-1234.

16.

Chan

(1998). Functional relations among constructs in the same content domain at different levels of analysis: A typology of composition models. Journal of Applied Psychology, 83, 234-246.

17.

Cialdini

R. B.

Trost

M. R.

(1998). Social influence: Social norms, conformity, and compliance. In Gilbert

D. T.

Fiske

S. T.

Lindzey

(Eds.), The handbook of social psychology (pp. 151-192). New York, NY: McGraw-Hill.

18.

Costa

P. T.

McCrae

R. R.

(1992). Normal personality assessment in clinical practice: The NEO personality inventory. Psychological Assessment, 4, 5-13.

19.

Driskell

J. E.

Goodwin

G. F.

Salas

O’Shea

P. G.

(2006). What makes a good team player? Personality and team effectiveness. Group Dynamics: Theory, Research, and Practice, 10, 249-271.

20.

Felps

Mitchell

T. R.

Byington

(2006). How, when, and why bad apples spoil the barrel: Negative group members and dysfunctional groups. Research in Organizational Behavior, 27, 174-222.

21.

Forsythe

D. R.

(2009). Group dynamics (5th ed.). Belmont, CA: Cengage.

22.

Friedrich

R. J.

(1982). In defense of multiplicative terms in multiple regression equations. American Journal of Political Science, 26, 797-833.

23.

Gelfand

Nishii

L. H.

Raver

J. L.

(2006). On the nature of cultural tightness-looseness. Journal of Applied Psychology, 91, 1225-1244.

24.

Goethals

G. R.

Zanna

M. P.

(1979). The role of social comparison in choice shifts. Journal of Personality and Social Psychology, 37, 1469-1476.

25.

Gough

(1987). California Psychological Inventory administrator’s guide. Palo Alto, CA: Consulting Psychologists Press.

26.

Hackman

J. R.

(2003). Learning more by crossing levels: Evidence from airplanes, hospitals, and orchestras. Journal of Organizational Behavior, 24, 905-922.

27.

Harrison

Klein

(2007). What’s the difference? Diversity constructs as separation, variety, or disparity in organizations. Academy of Management Review, 32, 1199-1228.

28.

Hart

J. W.

Bridgett

D. J.

Karau

S. J.

(2001). Coworker ability and effort as determinants of individual effort on a collective task. Group Dynamics: Theory, Research, and Practice, 5, 181-190.

29.

Hoffman

D. A.

Gavin

M. B.

(1998). Centering decisions in hierarchical linear models: Implications for research in organizations. Journal of Management, 24, 623-641.

30.

Hoffman

D. A.

Morgeson

F. P.

Gerras

S. J.

(2003). Climate as a moderator of the relationship between leader-member exchange and content specific citizenship: Safety climate as an exemplar. Journal of Applied Psychology, 88, 170-178.

31.

Hoffman

J. R.

Rogelberg

S. G.

(2001). All together now? College students’ preferred project group grading procedures. Group Dynamics: Theory, Research, and Practice, 5, 33-40.

32.

Hogan

(2007). Hogan Personality Inventory (3rd ed.). Tulsa, OK: Hogan Assessment Systems.

33.

Hogg

M. A.

(2005). The social identity perspective. In Wheelan

(Ed.), The handbook of group research and practice (pp. 133-157). Thousand Oaks, CA: Sage.

34.

Hox

J. J.

(2010). Multilevel analysis techniques and applications (2nd ed.). New York, NY: Routledge.

35.

Humphrey

S. E.

Hollenbeck

J. R.

Meyer

C. J.

Ilgen

D. R.

(2007). Trait configurations in self-managed teams: A conceptual examination of the use of seeding for maximizing and minimizing trait variance in teams. Journal of Applied Psychology, 92, 885-892.

36.

Ilies

Wagner

D. T.

Morgeson

F. P.

(2007). Explaining affective linkages in teams: Individual differences in susceptibility to contagion and individualism-collectivism. Journal of Applied Psychology, 92, 1140-1148.

37.

Kirkman

B. L.

Tesluk

P. E.

Rosen

(2004). The impact of demographic heterogeneity and team leader-team member demographic fit on team empowerment and effectiveness. Group & Organization Management, 29, 334-368.

38.

Kozlowski

S. W. J.

Klein

K. J.

(2000). A multilevel approach to theory and research in organizations: Contextual, temporal, and emergent processes. In Klein

K. J.

Kozlowski

S. W. J.

(Eds.), Multilevel theory, research, and methods in organizations: Foundations, extensions, and new directions (pp. 3-90). San Francisco, CA: Jossey-Bass.

39.

LePine

J. A.

Buckman

B. R.

Crawford

E. R.

Methot

J. R.

(2011). A review of research on personality in teams: Accounting for pathways spanning levels of theory and analysis. Human Resource Management Review, 21, 311-330.

40.

LePine

J. A.

Colquitt

J. A.

Erez

(2000). Adaptability to changing task contexts: Effects of general cognitive ability, conscientiousness, and openness to experience. Personnel Psychology, 53, 563-593.

41.

Marks

M. A.

Mathieu

J. E.

Zaccaro

S. J.

(2001). A temporally based framework and taxonomy of team processes. Academy of Management Review, 26, 356-376.

42.

McNeil

J. M.

Fleeson

(2006). The causal effect of extraversion on positive affect and neuroticism on negative affectt: Manipulating state extraversion and state neuroticism in an experimental research approach. Journal of Research in Personality, 40, 529-550.

43.

Mohammad

Angell

L. C.

(2003). Personality heterogeneity in teams: Which differences make a difference for team performance? Small Group Research, 34, 651-677.

44.

Morgeson

F. P.

Reider

M. H.

Campion

M. A.

(2005). Selecting individuals in team settings: The importance of social skills, personality characteristics, and teamwork knowledge. Personnel Psychology, 58, 583-611.

45.

Muchinsky

P. M.

Monahan

C. J.

(1987). What is person-environment congruence? Supplementary versus complementary fit. Journal of Vocational Behavior, 31, 268-277.

46.

Mumford

T. V.

Van Iddekinge

C. H.

Morgeson

F. P.

Campion

M. A.

(2008). The team role test: Development and validation of a team role knowledge situational judgment test. Journal of Applied Psychology, 93, 250-267.

47.

Neuman

G. A.

Wagner

S. H.

Christiansen

N. D.

(1999). The relationship between work-team personality composition and the job performance of teams. Group & Organization Management, 24, 28-45.

48.

Neuman

G. A.

Wright

(1999). Team effectiveness: Beyond skills and cognitive ability. Journal of Applied Psychology, 84, 376-389.

49.

Paulhus

D. L.

Martin

C. L.

(1988). Functional flexibility: A new conception of interpersonal flexibility. Journal of Personality and Social Psychology, 55, 88-101.

50.

Pearsall

M. J.

Christian

M. S.

Ellis

A. P.

(2010). Motivating interdependent teams: Individual rewards, shared rewards, or something in between? Journal of Applied Psychology, 95, 183-191.

51.

Ployhart

R. E.

Moliterno

T. P.

(2011). Emergence of the human capital resource: A multilevel model. Academy of Management Review, 36, 127-150.

52.

Ployhart

R. E.

Schneider

(2005). Multilevel selection and prediction: Theories, methods, and models. In Evers

Anderson

Voskuijl

(Eds.), Blackwell handbook of personnel selection (pp. 495-516). Oxford, UK: Blackwell.

53.

Ployhart

R. E.

Weekley

J. A.

Baughman

(2006). The structure and function of human capital emergence: A multilevel examination of the attraction-selection-attrition model. Academy of Management Journal, 49, 661-677.

54.

Prewett

M. S.

Walvood

A. A.

Stilson

F. R. B.

Rossi

M. E.

Brannick

M. T.

(2009). The team personality–team performance relationship revisited: The impact of criterion choice, pattern of workflow, and method of aggregation. Human Performance, 22, 273-296.

55.

Rein

McCraty

Atkinson

(1995). The physiological and psychological effect of compassion and anger. Journal of Advancement in Medicine, 8(2), 87-105.

56.

Saavedra

Earley

P. C.

Van Dyne

(1993). Complex interdependence in task-performing groups. Journal of Applied Psychology, 78, 61-72.

57.

Salas

Driskell

(2014). Personality and teams. In Christiansen

Tett

(Eds.), The handbook of personality at work (pp. 744-771). Danvers, MA: Routledge.

58.

Schmidt

J. A.

Ogunfowora

Bourdage

J. S.

(2011). No person is an island: The effects of group characteristics on individual trait expression. Journal of Organizational Behavior, 30, 863-891.

59.

Science Research Associates. (1989). Personal audit examiner’s manual. Chicago, IL: Author.

60.

Stewart

G. L.

Barrick

M. R.

(2000). Team structure and performance: Assessing the mediating role of intrateam process and the moderating role of task type. Academy of Management Journal, 43, 135-148.

61.

Tekleab

A. G.

Quigley

N. R.

(2014). Team deep-level diversity, relationship conflict, and team members’ affective reactions: A cross-level investigation. Journal of Business Research, 67, 394-402.

62.

Tett

R. P.

Burnett

D. D.

(2003). A personality trait-based interactionist model of job performance. Journal of Applied Psychology, 88, 500-517.