ArticlePDF Available

Getting Specific about Demographic Diversity Variable and Team Performance Relationships: A Meta-Analysis

Authors:

Abstract and Figures

The authors revisited the demographic diversity variable and team performance relationship using meta-analysis and took a significant departure from previous meta-analyses by focusing on specific demographic variables (e.g., functional background, organizational tenure) rather than broad categories (e.g., highly job related, less job related). They integrated different conceptualizations of diversity (i.e., separation, variety, disparity) into the development of their rationale and hypotheses for specific demographic diversity variable—team performance relationships. Furthermore, they contrasted diversity with the team mean on continuous demographic variables when elevated levels of a variable, as opposed to differences, were more logically related to team performance. Functional background variety diversity had a small positive relationship with general team performance as well as with team creativity and innovation. The relationship was strongest for design and product development teams. Educational background variety diversity was related to team creativity and innovation and to team performance for top management teams. Other variables generally thought to increase task-relevant knowledge (e.g., organizational tenure) and team performance were unrelated to team performance, although these variables were almost never studied as the variety conceptualization (i.e., the conceptualization that can reflect the breadth of knowledge that can be applied to the task). Team mean organizational tenure was related to team performance in terms of efficiency. Race and sex variety diversity had small negative relationships with team performance, whereas age diversity was unrelated to team performance regardless of diversity conceptualization. Implications for staffing teams and future research are discussed.
Content may be subject to copyright.
http://jom.sagepub.com/
Journal of Management
http://jom.sagepub.com/content/37/3/709
The online version of this article can be found at:
DOI: 10.1177/0149206310365001
2011 37: 709 originally published online 1 September 2010Journal of Management
Suzanne T. Bell, Anton J. Villado, Marc A. Lukasik, Larisa Belau and Andrea L. Briggs
Relationships: A Meta-Analysis
Getting Specific about Demographic Diversity Variable and Team Performance
Published by:
http://www.sagepublications.com
On behalf of:
Southern Management Association
can be found at:Journal of ManagementAdditional services and information for
http://jom.sagepub.com/cgi/alertsEmail Alerts:
http://jom.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
http://jom.sagepub.com/content/37/3/709.refs.htmlCitations:
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Getting Specific about Demographic
Diversity Variable and Team Performance
Relationships: A Meta-Analysis
Suzanne T. Bell
DePaul University
Anton J. Villado
Rice University
Marc A. Lukasik, Larisa Belau, and Andrea L. Briggs
DePaul University
The authors revisited the demographic diversity variable and team performance relationship
using meta-analysis and took a significant departure from previous meta-analyses by focusing
on specific demographic variables (e.g., functional background, organizational tenure) rather
than broad categories (e.g., highly job related, less job related). They integrated different con-
ceptualizations of diversity (i.e., separation, variety, disparity) into the development of their
rationale and hypotheses for specific demographic diversity variable–team performance rela-
tionships. Furthermore, they contrasted diversity with the team mean on continuous demographic
variables when elevated levels of a variable, as opposed to differences, were more logically related
to team performance. Functional background variety diversity had a small positive relationship
with general team performance as well as with team creativity and innovation. The relationship
was strongest for design and product development teams. Educational background variety diver-
sity was related to team creativity and innovation and to team performance for top management
709
Acknowledgments: This project was assisted by a grant awarded to Suzanne T. Bell from the Faculty Research and
Development Program, College of Liberal Arts and Sciences, DePaul University. Thanks to Bethany Denning and
Cort Rudolph, who assisted on earlier versions of this project, and David Fisher for his comments on previous drafts.
Corresponding author: Suzanne T. Bell, Department of Psychology, DePaul University, 2219 North Kenmore
Avenue, Chicago, IL 60614
E-mail: SBELL11@depaul.edu
Journal of Management
Vol. 37 No. 3, May 2011 709-743
DOI: 10.1177/0149206310365001
© The Author(s) 2011
Reprints and permissions: http://www.
sagepub.com/journalsPermissions.nav
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
710 Journal of Management / May 2011
teams. Other variables generally thought to increase task-relevant knowledge (e.g., organizational
tenure) and team performance were unrelated to team performance, although these variables
were almost never studied as the variety conceptualization (i.e., the conceptualization that can
reflect the breadth of knowledge that can be applied to the task). Team mean organizational
tenure was related to team performance in terms of efficiency. Race and sex variety diversity
had small negative relationships with team performance, whereas age diversity was unrelated
to team performance regardless of diversity conceptualization. Implications for staffing teams
and future research are discussed.
Keywords: teams; diversity; demographic diversity; team diversity; meta-analysis
Characteristics of team members that influence team performance are of interest to
rese archers and practitioners (e.g., Bell, 2007; Carpenter, Geletkanycz, & Sanders, 2004). Of
particular interest is how diversity on team member demographic variables (e.g., race, age,
educational background) is related to team performance (e.g., Ancona & Caldwell, 1992;
Kochan et al., 2003; Mannix & Neale, 2005; Milliken & Martins, 1996; Pelled, 1996). The
increased attention given to demographic diversity is primarily due to the changing nature
of the workforce and to social policy concerns surrounding diversity issues (Jackson, May, &
Whitney, 1995).
Despite the quantity and quality of existing team diversity research based on sound psy-
chological theories and paradigms of team behavior (e.g., Byrne, 1971; McGrath, Berdahl, &
Arrow, 1995; Tajfel, 1969), the effects of demographic diversity on team performance are
not clear (e.g., Horwitz & Horwitz, 2007; Webber & Donahue, 2001). Mixed results pervade the
team diversity literature, which offers limited direction to practitioners and scientists alike.
We believe that a primary source of confusion is the oversimplification of team diversity—
an inherently complex construct. Making general statements about the “good” or “bad” effects
of diversity in teams is a flawed approach, and we believe that future research must be guided
by a more nuanced view of diversity itself. Accordingly, we conducted a meta-analysis of
the demographic diversity and team performance relationship using the most current frame-
works for understanding and conceptualizing diversity. The primary interest of our investi-
gation was the extent to which specific demographic diversity variables (e.g., functional
background, organizational tenure, race) are related to team performance, with special con-
sideration for the conceptualization of diversity—that is, separation, variety, and disparity
(Harrison & Klein, 2007). Second, we integrated the team composition and team diversity
research by hypothesizing and testing when the team mean of continuous demographic vari-
ables should have a stronger relationship with team performance than diversity on the vari-
ables. Finally, within the context of the specific diversity variable and conceptualization, we
investigated additional moderators of the demographic diversity and team performance rela-
tionship as suggested by Argote and McGrath (1993). Specifically, we examined the type of
performance outcome (i.e., efficiency, general performance, creativity, and innovation) and
the nature of the tasks (studied as team type) as moderators of the relationships between
demographic diversity variables and team performance, in an effort to demonstrate how
expected relationships may emerge once the specific diversity variable and the conceptualization
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Bell et al. / Getting Specific 711
of diversity are taken into account. Given the pronounced effect of study setting in studies of
team composition and team performance (Bell, 2007) and the likelihood that the study of
some demographic variables (e.g., organizational tenure) but not others (e.g., race) may be
limited to field settings, we also examined study setting as a potential moderator.
Team Diversity Research
Team diversity refers to the distributional differences among members of a team with
respect to a common attribute (Harrison & Klein, 2007). Researchers have suggested that
differences on demographic variables can be related to team performance both positively and
negatively (see Tsui & Gutek, 1999; van Knippenberg, De Dreu, & Homan, 2004). The idea
that demographic diversity improves team performance is based on the informational diversity
cognitive resource perspective (e.g., Cox & Blake, 1991; Williams & O’Reilly, 1998), which
suggests that distributional differences can serve as indicators of available knowledge and
differing perspectives. A team that is more diverse in terms of demographic variables related
to the task may be more successful than a homogeneous team because the former team can draw
on a greater pool of knowledge and different perspectives. Based on this notion, diversity of
attributes that are “highly job related” (e.g., educational background, functional background)
are thought to be positively related to team performance, whereas those that are “less job related”
(i.e., age, sex, race) are not (Pelled, 1996).
Despite the potential positive effects for team diversity on some attributes, several theories
suggest that increased diversity can lead to decreased cooperation, coordination, and cohesion
among team members and, ultimately, decreased team performance (Milliken & Martins,
1996). For example, the similarity–attraction paradigm (Byrne, 1971) suggests that homo-
geneous teams should be more productive than diverse teams because of the mutual attrac-
tion shared among team members with similar attributes. This mutual attraction can result in
more efficient team processes, such as communication, thereby leading homogeneous teams
to outperform diverse teams (Wiersema & Bantel, 1992). Similarly, social categorization
theory suggests that team members categorize other team members into subgroups (Tajfel,
1969; Tajfel & Turner, 1979), which can form the basis for an in-group–out-group distinction.
Team members may develop an intergroup bias (Brewer, 1979) in some conditions (van
Knippenberg et al., 2004) and favor and cooperate with members of their in-group more than
with members of an out-group. As such, team members with similar demographic attributes,
as opposed to differing demographic attributes, may be more attracted to and may cooperate
more with one another, which suggests that homogeneous teams should outperform hetero-
geneous teams. The expectations model (McGrath et al., 1995), which is based on social
categorization theory, suggests an indirect link between demographic diversity and team
performance through expectations based on the target member’s social category. Team mem-
bers make assumptions about other team members based on the target team member’s demo-
graphic status (e.g., female), and they interact with the target team member in a manner
consistent with their expectations. Demographic characteristics that are easily observable—
surface-level variables such as age, sex, and race—are more likely to evoke responses that
result from basic social categorization. Thus, although surface-level variables are typically
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
712 Journal of Management / May 2011
considered less job related, they are still thought to influence team performance, albeit nega-
tively, through social processes.
Although these theories are intuitively appealing, meta-analytic investigations examining
the relationship between demographic diversity and team performance indicate mixed results.
Specifically, Webber and Donahue (2001) found no support for a demographic diversity–
team performance relationship for highly job-related or less job-related diversities. In a
follow-up to Webber and Donahue, Horwitz and Horwitz (2007) found that task-related (i.e.,
highly job-related) demographic diversity is positively related to the quality and quantity of
team performance, whereas biodemographic (i.e., less job-related) diversity has no relation-
ship with team performance. As a result of such inconclusive findings, researchers have
become discouraged with examining main effect relationships between demographic diver-
sity and team performance (van Knippenberg & Schippers, 2007), and they have begun to
expand in directions that include the search for mediators or moderators (e.g., Kearney &
Gebert, 2009; van Knippenberg et al., 2004). For example, in a recent meta-analysis, Joshi
and Roh (2009) abandoned the notion of determining whether diversity attributes have a
positive or negative effect on team performance; instead, they sought to understand how
contextual factors (e.g., occupational demography) shape these relationships. Their results
indicate that accounting for contextual moderators increases the size of the relationship between
team performance and relations- and task-oriented diversity. Although research efforts in
search of moderators and mediators of the demographic diversity–team performance rela-
tionship are undoubtedly valuable, we believe that they must be conducted in conjunction
with a more nuanced approach to diversity.
What Is Meant by Diversity
Clarification of diversity is paramount to a discussion of how differences on demographic
variables may influence team performance. Harrison and Klein (2007) presented a framework
suggesting that diversity is best conceptualized in three ways—separation, variety, disparity—
which vary in terms of their substance, pattern, and operationalization and, ultimately, their
consequences. Although homogeneous teams are equivalent across conceptualizations, the
differences among the conceptualizations become apparent in diverse teams.
Separation refers to differences among team members in their lateral position on a con-
tinuum, such as a value, attitude, or belief (Harrison & Klein, 2007). With separation, diver-
sity effects are thought to be symmetrical. In other words, it is the extent to which team
members are similar or different that is thought to influence team processes and outcomes;
whether team members are high or low on the construct of interest does not matter. Take,
for example, the continuum of educational experience. A team solely composed of high
school graduates and a team solely composed of members with professional degrees would
be considered equally homogeneous when educational-level diversity is conceptualized as
separation. A team half composed of high school graduates and half composed of members
with professional degrees would represent the maximum amount of diversity in terms
of separation. Hypothesized relationships between demographic variables and team perfor-
mance that are based on the theories of similarity and attraction (Byrne, 1971), social
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Bell et al. / Getting Specific 713
identity and self-categorization (Tajfel & Turner, 1969), and attraction, selection, and attri-
tion (Schneider, Goldstein, & Smith, 1995) conceptualize diversity as separation. These
theories suggest that greater similarity (reduced separation) yields positive outcomes that
ultimately lead to increased team performance.
Variety refers to categorical differences among team members wherein the number of
represented categories contributes to team diversity (Harrison & Klein, 2007). For example,
a team with a maximum amount of functional background diversity (conceptualized as variety)
would consist of every member of the team having a functional background different from
the others (e.g., sales, marketing). Having greater variety captures the essence of the infor-
mational diversity–cognitive resource perspective, which suggests that diversity is beneficial
to performance because diverse teams can draw from different pools of information or resources.
These differing perspectives can lead to debate and a broader understanding of the task, ulti-
mately resulting in increased team performance, especially for tasks requiring creativity or
innovation.
Finally, disparity represents differences in the concentration of valued assets or desirable
resources (Harrison & Klein, 2007). Disparity captures the extent to which an inequality
is present; that is, the team displays vertical differences on a resource between a few privi-
leged team members and the rest of the team (Harrison & Klein, 2007). Disparity differs
from separation in that the direction of the difference between a team member and all other
team members matters in terms of predicting the effect of diversity on team outcomes.
Whereas maximum separation manifests when two opposing camps form at opposing ends
of a horizontal continuum (without regard to which camp is high on the variable and which
camp is low), maximum disparity is reached when one team member is high on a dimension
and separated on a continuum from all other team members. For example, organizational
tenure can be viewed as a proxy for access to resources; that is, more tenure is associated
with more privilege. A maximum amount of organizational tenure disparity would consist of
one team member having been with the organization for 20 years and all other team
members’ being fairly new to the organization. If high levels of a variable are associated
with status or power (e.g., tenure with organization, education level), diversity in terms of
disparity might foster conformity and silence and suppress creativity within the team
(Harrison & Klein, 2007).
Separation, variety, and disparity each represent a unique pattern of differences among team
members. Although differences on a given variable (e.g., education level) may be conceptu-
alized in different ways (e.g., separation, disparity), it is important that the choice of the
conceptualization and related operationalization be theoretically driven (Harrison & Klein,
2007). The pattern of differences must be considered when articulating how each demographic
diversity variable of interest is related to team performance. For example, when diversity on
a demographic variable is thought to positively benefit team performance because of an
increased number of perspectives or task-relevant information (e.g., cognitive resource
perspective), diversity is conceptualized as variety. Separation and disparity conceptual-
izations are generally consistent with theories suggesting that diversity on a demographic
variable leads to negative outcomes such as misunderstandings and a lack of cohesion
(e.g., similarity–attraction theory, social identity theory, and social categorization theory;
Harrison & Klein, 2007).
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
714 Journal of Management / May 2011
Despite the potential for diversity effects in the opposite direction (depending on the
conceptualization of diversity), it is not clear what differences were included in previous
meta-analytic estimates that aggregated not only different demographic variables (e.g., func-
tional background, educational background) but also different conceptualizations of diversity
(e.g., functional background variety with organizational tenure disparity) into one overarch-
ing category (e.g., highly job related). This is problematic for one reason: When researchers
made claims that diversity (whether highly job related or less job related) had no relationship
with team performance, they did not make clear whether included estimates measured team
member differences with operationalizations that were able to capture the spirit of the theo-
retical justification. Accordingly, we present hypothesized relationships between specific
demographic diversity variables and team performance based on theoretically derived con-
ceptualizations of diversity—namely, separation, variety, and disparity. Given the recency of
Harrison and Klein’s typology (2007), authors of previous diversity research did not likely
base their theoretical arguments on the different conceptualizations. However, diversity
conceptualizations are expressed empirically through specific operationalizations of diver-
sity variables (e.g., coefficient of variation, standard deviation; Harrison & Klein, 2007),
thereby allowing for conceptualization to be examined as a moderator when authors reported
the operationalization used. Summarizing the extent to which previous research has used
operationalizations consistent with prevailing theories of how demographic diversity vari-
ables are related to team performance is important for clarifying observed meta-analytic
effects and critical for identifying gaps in the literature in need of additional research.
Elevated Levels of Demographic Variables in a Team
Regarding the effect of team member demographics on team performance, a potential
problem in the literature is its singular focus on how differences (i.e., diversity) on demo-
graphic variables affect team performance, with little regard for other team-level representations
of the demographic variables. Even if weak or no effects are observed for the relationship
between diversity on a demographic variable and team performance, the demographic vari-
able may still be an important predictor of team performance. Demographic diversity in
teams has typically been examined in isolation from the team composition literature on abil-
ity, personality, and values. Team composition is concerned with the configurations of attri-
butes in teams (Levine & Moreland, 1990), and team heterogeneity (diversity) represents
only one possible configuration of team members. Although early diversity research was caged
within the context of the larger composition research (Jackson et al., 1995), later research
developed relatively distinct from it (e.g., Horwitz & Horwitz, 2007; Webber & Donahue,
2001) with a few exceptions (e.g., Jackson, Joshi, & Erhardt, 2003).
Team composition research on the relationship between deep-level variables (e.g., per-
sonality, values) and team performance has consistently found larger effects for team mean
operationalizations of the composition variables as compared to those of heterogeneity (Bell,
2007). Some diversity researchers have argued the importance of accounting for the mean
when testing diversity effects (Harrison & Klein, 2007) and that measures of central ten-
dency of the attributes cannot be ignored (Jackson, Joshi, & Erhardt, 2003). Compared with
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Bell et al. / Getting Specific 715
measures of diversity, the team mean of the demographic variable may have a stronger rela-
tionship with team performance when teams with elevated levels of the demographic vari-
able reflect more task-relevant knowledge to be applied to the task. Although our primary
focus was on demographic diversity, we explored the extent to which elevated team levels
of continuous variables may be related to team performance, when elevated levels, rather than
differences, are more logically connected to team performance.
Specific Demographic Attribute of Interest
Because diversity represents distributional differences among members of a team with
respect to a common attribute, diversity is attribute specific (Harrison & Klein, 2007). A
team in not simply diverse; rather, a team is diverse with respect to specific attributes. The
attribute of interest should influence the demographic diversity and team performance rela-
tionship (Argote & McGrath, 1993; Jackson et al., 2003). Previous meta-analyses (e.g.,
Horwitz & Horwitz, 2007; Webber & Donahue, 2001) examined the relationship between
demographic diversity and team performance at the aggregate level, reporting only estimates
for all highly job-related or task-related demographic variables together and all less job-
related or biodemographic variables together. In the ensuing paragraphs, we outline why
particular demographic variables should be related to team performance and how differences
among team members on these variables might affect team performance. Because we used
meta-analysis, we focused on the demographic diversity variables most commonly studied
in the literature: functional background, educational background (major or degree), educa-
tional level, organizational and team tenure, age, sex, and race/ethnicity (Harrison & Klein,
2007). This focus was justified because a recent meta-analysis examined the relationships
between deep-level team composition variables (e.g., personality variables, values) and team
performance at the specific variable level (Bell, 2007).
Functional and Educational Background
Functional background diversity refers to the distribution of work history across the dif-
ferent functional specializations that exist within an organization (e.g., finance, marketing,
research and development; Bunderson, 2003). Functional background is thought to be impor-
tant in terms of reflecting a team members type of knowledge, as well as shaping a team
members attitude and perspective (Bantel & Jackson, 1989; Dearborn & Simon, 1958;
Hambrick & Mason, 1984). Schemas are thought to develop through experiences (Fiske &
Taylor, 2007), and they are further ingrained by goals and rewards relevant to those experi-
ences (Locke & Latham, 2002). Employees who spend their time in a functional division
of an organization should be exposed to and be influenced by information relevant to those
functional areas, and they should develop beliefs consistent with their functional roles
(Chattopadhyay, Glick, Miller, & Huber, 1999). A team composed of members from diverse
functional backgrounds should have a broader range of perspectives and knowledge to draw
on, and they should be able to outperform teams with members from homogeneous backgrounds.
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
716 Journal of Management / May 2011
At maximum levels of functional background variety diversity, a team would have members
spread across different functions, thereby suggesting more information to apply to the task.
Thus, functional background variety diversity should be positively related to team performance,
consistent with the core argument of the informational diversity–cognitive resource perspective.
Similar to functional background, educational background, in terms of major or content
area, has the potential to influence the knowledge, attitude, and perspective that a team mem-
ber brings to the task. Educational background may be directly related to a team members
background knowledge, as compared to his or her current attitude or perspective, given that
a team member may be years removed from when he or she received a degree. Despite this,
it may be beneficial for team performance to have teams composed of members with a variety
of educational backgrounds. Arguing that increased educational background variety diversity
should allow for access to more task-relevant knowledge is also consistent with the informa-
tional diversity–cognitive resource perspective.
Hypothesis 1: There will be a positive relationship between functional background diversity in
terms of variety and team performance.
Hypothesis 2: There will be a positive relationship between educational background diversity in
terms of variety and team performance.
Research Question 1: To what extent has functional background diversity been operationalized as an
index consistent with a conceptualization other than variety, and what is the nature of the effects?
Research Question 2: To what extent has educational background diversity been operationalized as an
index consistent with a conceptualization other than variety, and what is the nature of the effects?
Functional background variety diversity and educational background variety diversity are
argued to be important because they allow for a broader scope of task-relevant perspectives
to be applied to the task. It is therefore important to identify the context (i.e., type of team,
type of performance) within which diversity of functional background and educational back-
ground may be most task relevant. Diversity in knowledge and information is related to
increases in team innovation (Bantel & Jackson, 1989) and is thought to be important for
team creativity (Milliken, Bartel, & Kurtzberg, 2003). Accordingly, teams that are diverse
(in terms of variety) on functional background and educational background should have
multiple perspectives to apply to the task and be divergent in their thinking. When perfor-
mance criteria such as creativity and innovation are of interest (e.g., when performance is
highly based on divergence), functional background and educational background variety
diversity may contribute to performance. Divergent thinking is thought to promote the cre-
ative process (Milliken et al., 2003), although the creative process requires convergence for
later idea evaluation and implementation stages. However, convergence may be of primary
importance when costs or inputs are factored into how well the team is performing. Thus,
when the performance metric is efficiency (e.g., performance is highly based on convergence),
variety on functional background and educational background may be less helpful.
Similarly, executive teams, as well as product development and design teams, are likely to
benefit from team members from a variety of functional and educational backgrounds. Design
teams create and develop new products and services (Devine, 2002). Having a greater scope of
skills and technical influence is thought to be beneficial for teams such as product development
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Bell et al. / Getting Specific 717
or design teams because of the direct access to expertise relevant to the creation of the
product, as well as the facilitation of product transfer back to the various departments when
the product is complete (Ancona & Caldwell, 1992). Finally, executive teams (i.e., top
management teams; TMTs) engage in a variety of ambiguous, ill-defined tasks that influ-
ence the organization’s direction as a whole (Devine, 2002). TMT members typically rep-
resent different functional units within an organization, and their unique information gained
from working or training in various functional areas should help them make the best deci-
sions for the broader organization.
Hypothesis 3: The positive relationship between functional background diversity in terms of variety
and team performance will be stronger when the team performance criterion is creativity or
innovation rather than efficiency.
Hypothesis 4: The positive relationship between functional background diversity in terms of variety
and team performance will be stronger when the team is a design team or TMT as compared to
another team type.
Hypothesis 5: The positive relationship between educational background diversity in terms of vari-
ety and team performance will be stronger when the team performance criterion is creativity or
innovation rather than efficiency.
Hypothesis 6: The positive relationship between educational background diversity in terms of vari-
ety and team performance will be stronger when the team is a design team or TMT as compared
to another team type.
The above arguments suggest that functional and educational background diversity influ-
ence team performance because variety on these variables represents a larger base of knowl-
edge that what can be drawn on to complete the task. Arguments are less compelling when
suggesting that educational level, organizational tenure, and team tenure improve perfor-
mance because variety on these attributes within a team leads to a greater breadth of perspec-
tives and more task-relevant information. Accordingly, we investigated the extent to which
the team mean on these continuous variables are positively related to team performance.
Educational Level
Educational level pertains to an individual’s highest educational achievement. Although
educational level is often investigated as a diversity variable (e.g., Amason, Shrader, &
Tompson, 2006; Jehn & Bezrukova, 2004), having members spread across different education
levels (i.e., variety) is not likely to increase the breadth of perspectives needed to increase
performance on most tasks. Bantel and Jackson’s early work (1989) included educational level
as a predictor of innovation but not in terms of educational-level diversity. Instead, the authors
proposed that education level influences innovation through an additive combination of team
members’ education levels. Indeed, to the extent that educational level is related to general
mental ability (Sewell & Shah, 1967), teams composed of members higher in educational level
should outperform teams composed of members with lower levels of education. Likewise,
previous meta-analyses showed a relationship between general mental ability and team perfor-
mance, with team mean general mental ability a better predictor than heterogeneity (Bell,
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
718 Journal of Management / May 2011
2007; Devine & Phillips, 2001). Although team member education level may be reflective of
knowledge and information relevant to the task, a team should increase the amount of knowl-
edge and information relevant to the task by having team members with higher levels of educa-
tion rather than diverse levels of education.
Hypothesis 7: There will be a positive relationship between team mean educational level and team
performance.
Hypothesis 8: The positive relationship between team mean education level and team performance will
be stronger than the relationship between education-level variety diversity and team performance.
The team type may provide an important context in determining the strength of the rela-
tionship between team mean educational level and team performance. Devine’s team typol-
ogy (2002) distinguishes several types of teams (e.g., advisory, TMT, service) that fall under
the broader distinction of being primarily engaged in intellectual work or physical work. Team
mean education level should be more important for intellectual teams (i.e., advisory,
design, commission, TMT, command, negotiation) than for physical teams (e.g., production,
service) because of the type of functions that intellectual teams perform (e.g., planning,
integrating information, directing). Although team mean education level may be related to
performance in physical teams, it may be less important for the types of functions that
physical teams complete (e.g., building, repairing, assembling).
Hypothesis 9: The positive relationship between team mean educational level and team perfor-
mance will be stronger for intellectual teams as compared to physical teams.
Organizational Tenure
Organizational tenure is the amount of time that a team member has worked with the
organization. Team members’ organizational tenure may influence performance through its
ties with organizational socialization—the process through which an individual comes to
understand the social knowledge, values, and expected behaviors necessary to assume an
organizational role (Chatman, 1991; Sturman, 2003; Van Maanen & Schein, 1979). A team
composed of members with long organizational tenure may have a greater understanding of
how to successfully operate within the organizational system. For example, members of a
research and development team with long organizational tenure might have a better under-
standing of how to access valued organizational resources (e.g., money, upper management
support) needed for team performance. In addition, members of organizations develop a com-
mon unique language that facilitates transmission of work-related information, which should
make communication among team members with greater organizational tenure more efficient.
Despite the apparent advantage of having teams composed of members with long organi-
zational tenure, the attraction–selection–attrition framework (Schneider, 1987) suggests that
organizational members become homogeneous over time, which might have negative conse-
quences in terms of dealing with an uncertain environment or unexpected change. As such,
composing teams with members who all have long organizational tenure might not lead to
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Bell et al. / Getting Specific 719
increased team performance. The type of performance must be considered to understand
when organizational tenure diversity, as compared to elevated tenure levels (i.e., high team
mean), should be most predictive of team performance. Specifically, when team members
need to converge (e.g., efficiency is the primary performance goal or typical tasks are rou-
tine), greater average organizational tenure might benefit team performance. However, when
innovation is the criterion or the team is completing tasks that require access to a greater
variety of perspectives, variety diversity on organizational tenure might benefit team perfor-
mance because members of the team will have been socialized into the organization at dif-
ferent times and will bring unique perspectives to the team in terms of organizational
know-how (Jackson et al., 1995).
Hypothesis 10: There will be a positive relationship between team mean organizational tenure and
team performance when efficiency is the criterion.
Hypothesis 11: There will be a positive relationship for organizational tenure diversity in terms of
variety and team performance when innovation is the criterion.
Research Question 3: To what extent has organizational tenure diversity been operationalized as a
diversity index consistent with a conceptualization other than variety, and what is the nature of
the effects?
Team Tenure
Team tenure is defined as the length of time that team members have interacted with one
another (Katz, 1982). Team tenure affects project performance by reducing the communication
among team members to a particular information domain. Over time, team members become
cohesive and can become increasingly isolated from important sources that provide evaluation,
information, and feedback (Katz, 1982). As such, differences in team tenure among team
members—that is, having a mix of experienced and newer team members—might benefit team
performance. If new team members are integrated into the team over time, new team members
can provide fresh ideas and approaches, and challenge existing methods, while more tenured
team members can offer information about the team’s existing structure and responsibilities.
This team tenure variety diversity should be important for team innovation. Furthermore, team
tenure effects have been observed beyond those of age and organizational tenure (Katz, 1982).
Although team tenure variety diversity may be important for team innovation, it may be
less important for general performance and efficiency. Kozlowski, Gully, Nason, and Smith
(1999) proposed a dynamic theory of team development and team performance wherein teams
navigate through four important phases of team development: team formation, task compila-
tion, role compilation, and team compilation. When a team forms, individuals come together
and seek information about one another and the nature of the team (e.g., its purpose). Teams
next enter a task compilation phase wherein team members demonstrate their task competen-
cies to one another and focus on what they need from one another. In the third phase, role
compilation, team members connect with one another and figure out how their actions affect
other team members. It is during the role compilation phase that team members also focus
on having their needs met and figure out what they need to do to help others. During the final
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
720 Journal of Management / May 2011
phase (i.e., team compilation), team members learn how to improve their network of roles
and how to deal with routine and normative situations. Teams with members who have been
part of the team compilation process should have a better understanding of how the team will
approach a typical team task. Accordingly, teams with longer average tenure should have bet-
ter performance in general and in terms of efficiency in particular.
Hypothesis 12: Team tenure diversity in terms of variety will be positively related to team perfor-
mance when innovation is the criterion.
Hypothesis 13: Team mean tenure will be positively related to team performance when efficiency
is the criterion.
Research Question 4: To what extent has team tenure diversity been operationalized as a diversity
index consistent with a conceptualization other than variety, and what is the nature of the effects?
Race, Sex, and Age
In their categorization elaboration model, Van Knippenberg et al. (2004) argued that diver-
sity research has taken an overly simplified conceptualization of the social categorization
process. They suggested that the salience of various demographic variables contributes to social
categorization and that demographic diversity is negatively related to team performance only
in situations where social categorization results in intergroup bias. The circumstances under
which social categorization leads to intergroup bias (e.g., threats and challenges to sub-
group’s identity; van Knippenberg et al., 2004) have not been examined in the team demo-
graphic diversity and team performance literature; however, whether differences are
perceived is fundamental to whether intergroup bias can occur.
Social psychology research supports the notion that people form first impressions and cat-
egorize one another on easily observable characteristics such as age, sex, and race (Fiske &
Neuberg, 1990; Messick & Mackie, 1989; Stangor, Lynch, Duan, & Glas, 1992). The initial
categorization tends to be tied to immediate physical features thought to be informative
about another person’s disposition. These social categories are so frequently activated in
daily social perception that they are chronically accessible and habitual in all situations
(Fiske & Neuberg, 1990). Furthermore, categorizing sex and race has shown to be fairly
consistent and resistant to short-term manipulations designed to decrease social categorization
on the basis of these variables (Hewstone, Hantzi, & Johnston, 1991; Stangor et al., 1992).
In sum, the research is clear that individuals take surface-level information (e.g., sex, race,
age) into account when categorizing others.
There is evidence that the saliences of race, sex, and age are not necessarily equal. Harrison,
Price, Gavin, and Florey (2002) found that surface-level aspects of diversity (i.e., age, sex,
race) were differentially related to team members’ perceptions of similarity. Specifically, race
diversity was most predictive of team members’ ratings of how similar they were to one
another (r = .52). In comparison, sex diversity had the least influence on perceptions of
surface-level diversity (r = .17), and age diversity had a moderate effect (r = .30). Similarly
strong relationships between race diversity and perceptions of surface-level differences were
observed in MBA project teams (Zellmer-Bruhn, Maloney, Bhappu, & Salvador, 2008). In
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Bell et al. / Getting Specific 721
addition to influencing perceived social similarity to other team members, race diversity influ-
enced early estimates of perceived work style similarity. Taken together, these results suggest
that age, sex, and race diversity might be related to team performance through the activation
of social categorization, each to a different extent. Diversity is conceptualized as separation
when differences are thought to be related to team performance through mechanisms such
as social categorization theory and intergroup bias as well as similarity–attraction theory.
Thus, the negative relationships between age, sex, and race diversity and team performance
are most likely to emerge when age, sex, and race diversity are operationalized as a diversity
index consistent with separation. Separation describes a situation when two subgroups are
formed at separate ends of a continuum. Because sex and race are categorical variables,
operationalizing diversity with indices associated with separation diversity is not appropri-
ate. Therefore, researchers likely captured race and sex diversity in terms of variety. Doing
so, however, presents another problem. Variety—specifically, maximum variety—reflects a
situation where each member within the team comes from a unique category (e.g., a different
race), precluding subgroups defined by the variable in question. As such, maximum variety
is not reflective of differences among team members for which social categorization theory
and intergroup bias would predict negative consequences on team performance. Given these
potential difficulties, although we hypothesized race, sex, and age separation diversity would
be negatively related to team performance, we also investigated the relationship between
race and sex variety diversity via a research question.
Hypothesis 14: There will be a negative relationship between race diversity in terms of separation
and team performance.
Hypothesis 15: There will be a negative relationship between sex diversity in terms of separation
and team performance.
Hypothesis 16: There will be a negative relationship between age diversity in terms of separation
and team performance.
Research Question 5: To what extent has race diversity been operationalized as a diversity index
consistent with variety, and what is the nature of the effects?
Research Question 6: To what extent has sex diversity each been operationalized as a diversity
index consistent with variety, and what is the nature of the effects?
Research Question 7: To what extent has age diversity been operationalized with a diversity index
consistent with a conceptualization other than separation, and what is the nature of the effects?
Study Setting as a Moderator
Finally, a potential confounding variable in a meta-analysis of relationships between demo-
graphic diversity variables and team performance is whether the study was conducted in a
lab or field setting. Study setting is a potential confound because the examination of some
variables might be mostly limited to field studies (i.e., organizational tenure), whereas others
(e.g., race, sex, age) might lend themselves to examination in field or laboratory settings. Study
setting is also likely to be highly correlated with the length of time that a team has been
together. Although the correlation between study setting and time suggests that weaker effects
would be observed between surface-level variables and team performance in field studies as
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
722 Journal of Management / May 2011
compared to lab studies (e.g., Harrison et al., 2002; Watson, Kumar, & Michaelsen, 1993),
other features of the study setting suggest the converse. Demographic diversity may be most
related to team performance when team members are working on projects that they believe
are relevant to the organization’s functioning (e.g., Jackson et al., 1995). Features of the set-
ting (e.g., fidelity) may affect the relationships between demographic diversity variables and
team performance, especially for surface-level variables thought to be related to perfor-
mance via social mechanisms such as intergroup bias. Relationships between surface-level
variables and team performance may not emerge in artificial lab settings, because they require
the investment and concern for the outcome more readily experienced by team members in
organizational settings. Accordingly, we examined the moderating effect of research setting
on the relationship between specific demographic variables and team performance. We expected
that the relationship between surface-level variables (race, sex, and age) and team performance
would be stronger in field studies than in lab studies.
Method
Literature Search
The present study included the demographic diversity and team performance literature
from 1980 to November 2009. The process to obtain relevant studies included electronic
searches of PsycInfo, ABI/Inform, and ProQuest Digital Dissertations, using keyword com-
binations of the specific diversity variable (e.g., age, functional background), diversity and
variants of the word (e.g., heterogeneity or homogeneity), and team (or group). The database
searches were supplemented with manual searches of reference lists from reviews and meta-
analyses of team diversity research (e.g., Bowers, Pharmer, & Salas, 2000; Horwitz &
Horwitz, 2007; Joshi & Roh, 2009; Webber & Donahue, 2001), searches of “in press” articles
available at relevant journals (e.g., Academy of Management Journal, Small Group Research),
and other articles we knew of. We reviewed articles for potential inclusion—for which, stud-
ies had to report sample sizes and information that allowed for the computation of a correla-
tion that represented the relationship between the demographic variable and performance at
the team level. Studies conducted in lab and field settings were included. Self-report mea-
sures of team members’ perceptions of team performance were not included (e.g., Schippers,
Hartog, Koopman, & Wienk, 2003), and sports teams were not included. Because of the
inappropriateness of mixing levels of analyses when calculating sample-weighted effects
(Gully, Devine, & Whitney, 1995), articles that reported only individual-level performance
data were excluded, even if the individual performed in the context of a team. For TMTs, the
performance criteria were often organizational-level performance metrics such as return on
assets. We included these correlations because TMT functioning and goals are closely tied
to these performance metrics and because studies of TMTs tie one team to one performance
metric, thereby allowing the correlations to contribute to the sample-weighted effects at the
team level. Furthermore, because we examined type of team as a moderator, including stud-
ies on TMTs allowed us to be more comprehensive. Measures of relational demography
were excluded when they focused on an individual’s experience as being demographically
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Bell et al. / Getting Specific 723
similar or different from the remainder of the team and how that difference affects individual
outcomes (for discussion of the relational demography and demographic diversity distinc-
tion, see Harrison & Klein, 2007; Tsui & Gutek, 1999; Tsui & O’Reilly, 1989). Finally, only
English-language articles were included. Note that our inclusion criteria are not the same as
previous meta-analyses in the same content area (Horwitz & Horwitz, 2007; Joshi & Roh,
2009; Webber & Donahue, 2001). For example, Horwitz and Horwitz (2007) included studies
with self-report team performance measures and did not include TMTs; Webber and Donahue
(2001) did not include the same set of demographic variables examined here (e.g., they included
occupational background but not organizational tenure).
We considered correlations from the same group of participants to be dependent if they
contributed to the same demographic variable–team performance relationship for a particu-
lar moderator level. To create an independent data set, we computed linear composites for
dependent correlations when intercorrelations were available. When intercorrelations were
not available, we used the mean of the dependent effect sizes. The majority of the articles
included intercorrelations. The final data set was based on 92 sources (e.g., journal articles,
dissertations), and it included 274 independent correlations for analyses that examined diver-
sity conceptualization as a moderator and 323 independent correlations for analyses that exa-
mined criterion type as a moderator.
Coding of Variables
Two authors independently coded each article and discrepancies were resolved via con-
sensus. When consensus could not be reached between the two coders, the first author was
brought in to discuss the coding question. Disagreement between coders rarely occurred
(< 10% of the time). Demographic variables were coded as educational background, functional
background, educational level, organizational tenure, team tenure, age, sex, and race. Study
setting was coded as lab setting when data were collected from teams that were in an artificial
or classroom situation or when student teams were the sample. Study setting was coded as
field setting when data were collected from an organizational setting using real teams. Using
Devine’s typology (2002), we coded type of team if it was explicitly stated or described in
enough detail to make a reasonable judgment. Devine’s team typology includes two major
team categories: those involved in physical work and those involved in intellectual work.
Intellectual teams are further divided as follows: advisory teams, which address workflow
problems and organizational improvement; design teams, which design new products, goods,
and services (we included product development teams and similar cross-functional teams in
this category); commission teams, which handle special and nonroutine decision making that
requires extensive acquisition and integration of information; executive teams (TMTs),
which coordinate work of functions, departments, and organizations as a whole; command
teams, which make organizational-level decisions in real time; and negotiation teams, which
represent larger entities and attempt to maximize the outcomes for their constituents. Physical
teams include production teams, which build, assemble, and harvest; service teams, which
process orders and requests from customers as quickly as possible; and an action/performance/
work team category, which includes performance, medical response, sports, transportation, and
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
724 Journal of Management / May 2011
military combat teams. Team performance was coded as efficiency, general performance, and
creativity or innovation. Efficiency was coded for performance measures where the outputs
were adjusted for inputs (e.g., productivity, efficiency; Beal, Cohen, Burke, & McLendon,
2003). We coded general performance for performance measures that reflected the extent
that the team met its overall objectives or goals, without consideration of the costs or inputs
needed for achieving the results. We coded creativity for performance measures that cap-
tured the uniqueness of an output compared to other outputs, and we coded innovation for
performance measures that captured the development and application of ideas to processes,
products, and procedures that were new to the unit of adoption and were designed to benefit
the recipient of interest (West & Farr, 1990). Finally, we coded diversity conceptualization
using Harrison and Klein’s typology (2007), linking the diversity operationalization to the
different conceptualizations. We coded standard deviation and mean Euclidean distance as
operationalizations consistent with separation. We coded Blau’s index and Techman’s
entropy as operationalizations consistent with variety. We coded coefficient of variation and
Gini coefficient as operationalizations consistent with disparity. We coded all other opera-
tionalizations of diversity (e.g., diversity composition was experimentally manipulated or
the operationalization was not reported) as other unless the manipulation could be mapped
onto the separation, variety, or disparity conceptualizations.
Meta-Analysis of Correlations
We used Arthur, Bennett, and Huffcutt’s SAS PROC MEANS meta-analysis program
(2001) to conduct a random effects model meta-analysis of correlations, using procedures
recommended by Hunter and Schmidt (2004). We estimated a sample-weighted mean cor-
relation (SWMr) between the demographic variable and criterion and calculated 95% confi-
dence intervals around the SWMr as a measure of accuracy of the effect size (Whitener,
1990). We corrected SWMrs for unreliability of the criterion using an artifact distribution.
Because of the nature of the predictor (i.e., demographics), only criterion unreliability was
corrected. We calculated the standard deviation of the population correlation (SDr) and the
percentage of variance attributed to sampling error and artifact corrections, and we report
them as indicators of the presence of moderators. Finally, we calculated a fail-safe k (n) using
procedures outlined by Rosenthal (1979) and Orwin (1983). The fail-safe k indicates the
number of studies in file drawers with null effects needed to reduce the observed effect down
to .05. We chose .05 with the logic that observed effects between demographic diversity vari-
ables and team performance should generally be small (r = .10; see Cohen, 1992) and that
when reduced to .05, the effect would be as close to zero as to a small effect. We tested
moderators using Hunter and Schmidt’s subgroup analysis (2004) in which a meta-analysis
is conducted at each moderator level of the relationship of interest.
Results
First, we examined the influence of study setting on the demographic variable–team perfor-
mance relationship. Whereas 24% of the correlations (32 out of 133) between surface-level
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Bell et al. / Getting Specific 725
diversity variables (race, sex, age) and team performance were examined in lab settings, less
than 3% of the correlations (4 of 139) between the other demographic variables (e.g., func-
tional background, educational background) and team performance were examined in lab
settings. Study setting appeared to be a consistent moderator of the size and direction of the
relationships observed. Given the moderating effect of study setting on the relationships
between surface-level variable and team performance and the extremely limited number of lab
studies examining other demographic variables and team performance, we interpreted our
results for surface-level variables (race, sex, age) within the context of the study setting and
excluded correlations from lab settings for all other demographic variables.
Table 1 presents results for Hypotheses 1 through 6. Hypothesis 1 predicted a positive
relationship between functional background diversity in terms of variety and team perfor-
mance. In support of Hypothesis 1, we observed a small positive relationship between team
performance and functional background variety diversity (r = .11), and the 95% confidence
interval around the SWMr did not include zero. Research Question 1 asked to what extent
functional background diversity was operationalized as an index consistent with a conceptu-
alization other than variety. Almost all studies (30 of 31) reported using a diversity index
consistent with a variety conceptualization; one study did not report enough information to
code conceptualization. We predicted that the relationship between functional background
variety diversity would be stronger when the team performance criterion was creativity or
innovation rather than efficiency (Hypothesis 3), and when the team was a design or TMT
(Hypothesis 4). Consistent with Hypothesis 3, the relationship between functional back-
ground variety diversity and team performance was stronger when the criterion was cre-
ativity or innovation (r = .18) as compared to efficiency (r = .03). We also observed a small
positive effect for the relationship between functional background variety diversity and gen-
eral performance (r = .12). Finally, in partial support of Hypothesis 4, there was a stronger
relationship between functional background variety diversity and team performance for
design teams (r = .16) but not for TMTs (r = .07; the 95% confidence interval around SWMr
included zero), as compared with other team types (r = –.01).
Hypothesis 2 predicted a positive relationship between educational background diversity
in terms of variety and team performance. Educational background variety diversity was
unrelated to team performance (r = .01). Research Question 2 asked to what extent educa-
tional background diversity was operationalized as an index consistent with a conceptualiza-
tion other than variety. All studies reported using a diversity index consistent with a variety
conceptualization. We predicted that the relationship between educational background vari-
ety diversity and team performance would be stronger when the criterion was creativity or
innovation rather than efficiency (Hypothesis 5) and when the team was a design team or
TMT (Hypothesis 6). Consistent with Hypothesis 5, there was a positive relationship
between educational background variety diversity and team performance when the team
performance criterion was creativity or innovation (r = .23) rather than efficiency (r = –.02),
although the creativity and innovation estimate was based on only three correlations and
should be interpreted with caution. There was no relationship between educational back-
ground variety diversity and general team performance (r = –.03). Finally, in partial support
of Hypothesis 6, there was a stronger relationship between educational background variety
diversity and team performance for TMTs (r = .13) but not for design teams (r = .07), as
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
726 Journal of Management / May 2011
compared with other team types (r = –.05). A limited number of correlations were available
for non-TMT teams, so the effect sizes for design teams and other team types should be
interpreted with caution. In sum, educational background variety diversity was unrelated to
team performance, except when creativity or innovation was the criterion of interest or when
the team was a TMT.
Table 2 presents results for Hypotheses 7 through 13. We predicted a positive relationship
between mean educational level and team performance (Hypothesis 7) that would be stron-
ger than the relationship between educational level variety diversity and team performance
(Hypothesis 8). There was no support for Hypothesis 7 or 8. Team mean education level was
not related to team performance (r = .01), but there was also no relationship observed between
education level variety diversity and team performance (r = –.01). Finally, Hypothesis 9
predicted a stronger relationship between team mean education level and team performance
for intellectual teams than for physical teams. Although the relationship between mean edu-
cation level and team performance for intellectual teams suggested a small effect (r = .11),
Hypothesis 9 was not supported in that the 95% confidence interval around the SWMr included
Table 1
Relationship Between Team Performance and Functional/
Educational Background Diversity
Variable k n SWMr SWSD VAR % 95% CI r SDr VAR A % k
fs
Functional background 31 3,726 .09 .15 35.63 .04 .15 .10 .13 35.74 25
Variety 30 3,653 .10 .15 37.57 .05 .15 .11 .13 37.70 30
Efficiency 17 1,338 .03 .17 46.82 -.05 .11 .03 .13 46.83
General performance 12 2,267 .11 .13 30.09 .03 .18 .12 .12 30.27 15
Creativity and innovation 5 493 .16 .16 37.52 .02 .30 .18 .14 37.79 11
Design/cross-functional 6 1,816 .14 .07 58.67 .08 .20 .16 .05 59.72 11
Top management team 16 1,373 .07 .20 32.14 -.03 .16 .07 .17 32.18
Other/mixed team type 9 537 -.01 .17 59.67 -.12 .10 -.01 .12 59.67
Educational background
Variety 13 2,629 .01 .12 34.63 -.05 .08 .01 .11 34.63
Efficiency 5 1,855 -.02 .03 100.00 -.04 .01 -.02 .00 100.00
General performance 5 1,832 -.03 .06 82.35 -.08 .03 -.03 .03 82.41
Creativity and innovation 3 317 .21 .11 72.40 .08 .33 .23 .06 73.34 10
Design/cross-functional 3 291 .06 .11 93.12 -.06 .18 .07 .03 93.22
Top management team 6 696 .12 .11 71.25 .03 .21 .13 .06 71.59 9
Other/mixed team type 4 1,642 -.04 .09 31.31 -.13 .04 -.05 .08 31.37
Notes: Only results for teams from field settings are reported. The number of correlations at a moderator level (e.g.,
different conceptualizations) may not sum to the overall number of correlations for a specific variable (e.g., func-
tional background) if the moderator information was not reported or if a correlation could not be categorized into the
level of the moderator (e.g., included efficiency and innovation in one measure). Results are corrected for criterion
unreliability. k = number of correlations; n = number of teams; SWMr = sample-weighted mean correlation; SWSD =
sample-weighted standard deviation of the SWMr; VAR % = percentage of variance attributed to sampling error; 95%
CI = 95% confidence interval; r = corrected population correlation; SDr = standard deviation of the corrected popu-
lation correlation; VAR A % = percentage of variance attributed to sampling error and artifact corrections; k
fs
= fail-
safe k. Fail-safe k indicates the number of studies in “file drawers” with a mean r = .00, which would reduce the
results in our meta-analysis to a trivial effect size of SWMr = .05. Fail-safe k is reported only for effects that have
more than one study, are greater than SWMr = .05, and have a confidence interval that did not include zero.
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Bell et al. / Getting Specific 727
zero. Only one correlation examined the relationship between team mean education level
and team performance in physical teams, and it indicated a weaker effect in the opposite direc-
tion of the effect observed for intellectual teams (SWMr = –.07).
Hypothesis 10 suggested that team mean organizational tenure would be positively related
to team performance when efficiency was the criterion. Consistent with Hypothesis 10, results
indicated a small positive relationship between team mean organizational tenure and effi-
ciency (r = .14), and the 95% confidence interval around the SWMr did not include zero.
Hypothesis 11 suggested that organizational tenure variety diversity would be positively
rel ated to team performance when innovation was the criterion. Only two studies examined
Table 2
Relationship Between Team Performance and Education Level, Organizational
Tenure, or Team Tenure
Variable k n SWMr SWSD VAR % 95% CI r SDr VAR A % k
fs
Education level
Mean 9 2,571 .01 .13 19.43 –.08 .10 .01 .13 19.43
Intellectual teams 7 1,077 .10 .16 26.13 –.02 .21 .11 .15 26.24
Physical teams 1 1,401 –.07
Diversity 14 3,914 –.01 .10 40.57 –.05 .04 –.01 .08 41.25
Variety 4 244 –.01 .09 100.00 –.09 .08 –.01 .00 100.00
Other conceptualization 9 3,597 –.01 .09 28.14 –.07 .06 –.01 .09 31.26
Organizational tenure
Mean 17 4,039 .07 .13 26.40 .01 .13 .08 .12 26.50 7
Efficiency 9 2,524 .13 .07 81.38 .09 .17 .14 .03 82.44 15
General performance 7 2,685 .00 .15 11.97 –.11 .11 .00 .15 11.97
Creativity and innovation 1 199 –.27
Diversity 24 4,259 .04 .12 38.13 –.01 .08 .04 .11 38.16
Separation 4 296 –.02 .06 100.00 –.08 .04 –.03 .00 100.00
Variety 2 115 .05 .07 100.00 –.05 .16 .06 .00 100.00
Disparity 18 3,848 .04 .13 29.92 –.01 .10 .04 .12 29.96
Team tenure
Mean 15 867 .08 .19 47.40 –.01 .18 .09 .15 47.45
Efficiency 4 238 .10 .12 100.00 –.01 .22 .11 .00 100.00
General performance 8 436 .02 .18 59.30 –.10 .14 .02 .12 59.31
Creativity and innovation 2 116 .09 .22 34.24 –.23 .40 .10 .20 34.28
Diversity 12 2,124 –.04 .10 58.98 –.10 .02 –.04 .07 59.03
Disparity 10 1,986 –.04 .09 69.37 –.10 .01 –.04 .05 50.17
Notes: Only results for teams from field settings are reported. The number of correlations at a moderator level (e.g.,
different conceptualizations) may not sum to the overall number of correlations for a specific variable (e.g., mean
team tenure) if the moderator information was not reported or if a correlation could not be categorized into the level
of the moderator (e.g., included efficiency and innovation in one measure). Results are corrected for criterion unre-
liability. k = number of correlations; n = number of teams; SWMr = sample-weighted mean correlation; SWSD =
sample-weighted standard deviation of the SWMr; VAR % = percentage of variance attributed to sampling error;
95% CI = 95% confidence interval; r = corrected population correlation; SDr = standard deviation of the corrected
population correlation; VAR A % = percentage of variance attributed to sampling error and artifact corrections; k
fs
=
fail-safe k. Fail-safe k indicates the number of studies in “file drawers” with a mean r = .00, which would reduce
the results in our meta-analysis to a trivial effect size of SWMr = .05. Fail-safe k is reported only for effects that
have more than one study, are greater than SWMr = .05, and have a confidence interval that did not include zero.
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
728 Journal of Management / May 2011
organizational tenure variety diversity, and only one study examined innovation as the criterion,
and it was not supportive of our hypothesis (SWMr =.06). Research Question 3 asked to what
extent organizational tenure diversity was operationalized as an index consistent with a con-
ceptualization other than variety and what the nature of the effect was. The majority of cor-
relations that examined the relationship between organizational tenure diversity and team
performance operationalized diversity as an index consistent with disparity (k = 18). These
correlations also failed to support a relationship between organizational tenure diversity and
team performance (r = .04).
Hypothesis 12 predicted a positive relationship between team tenure diversity in terms of
variety and team performance when creativity or innovation was the criterion. No studies
reported the relationship between team tenure variety diversity and team performance, so
Hypothesis 12 was not tested. Hypothesis 13 predicted that mean team tenure would be positively
related to team performance when efficiency was the criterion. Only four studies (k = 4) investi-
gated this relationship. The direction of the effect supported our hypothesis (r = .11), but the 95%
confidence interval around the SWMr included zero. Research Question 4 asked to what extent
team tenure diversity was operationalized as a diversity index consistent with a conceptualization
other than variety and what the nature of the effect was. Ten of the 12 studies (k = 12) that
reported the relationship between team tenure diversity and team performance conceptualized
diversity as disparity and suggested only a negligible relationship (r = –.04). The other 2 studies
did not report enough information for conceptualization to be coded.
Hypotheses 14 through 16 focused on the relationships between team performance and
race, sex, and age diversity. Table 3 presents the results. Study setting moderated the relation-
ships between team performance and race, sex, and age; as such, results are interpreted within
the context of the study setting. Hypothesis 14 predicted a negative relationship between
race diversity in terms of separation and team performance. As expected, no correlations in
either lab or field settings examined the relationship for race diversity in terms of separation
and team performance, so the hypothesis could not be tested. Research Question 5 asked to
what extent race diversity was operationalized as a diversity index consistent with variety and
what the nature of the effect was. Sixteen correlations from field studies examined race diver-
sity consistent with the variety conceptualization. Results supported a small negative rela-
tionship between race variety diversity and team performance (r = –.13), and the 95%
confidence interval around the SWMr did not include zero. Race variety diversity and team
performance were unrelated in lab settings (r = .00).
Hypothesis 15 predicted that sex diversity in terms of separation would be negatively
related to team performance. Only one lab study investigated sex diversity in terms of separa-
tion and team performance, and it reported a negative effect. Three studies investigated the
relationship between sex separation diversity and team performance in field settings and
suggested no effect (r = –.01). Research Question 6 asked to what extent sex diversity was
operationalized as a diversity index consistent with variety and what the nature of the effect
was. In field studies, the majority of studies investigated the relationship between sex variety
diversity and team performance (k = 23). Results supported a small negative relationship
between sex variety diversity and team performance (r = –.09), and the 95% confidence
interval around the SWMr did not include zero. In lab studies, sex variety diversity was
unrelated to team performance (r = .07; the 95% confidence interval around the SWMr
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Bell et al. / Getting Specific 729
included zero). Although we did not make any predictions across criterion type, it is worth not-
ing that the negative effects for race and sex variety diversity and team performance were simi-
lar in that they were amplified when team performance was a measure of creativity or innovation.
Table 3
Relationship Between Team Performance and Race, Sex, or
Age Diversity
Variable k n SWMr SWSD VAR % 95% CI r SDr VAR A % k
fs
Race 31 5,298 –.10 .13 34.70 –.14 –.05 –.11 .12 34.28 31
Lab 15 886 .02 .11 100.00 –.03 .08 .02 .00 100.00
Variety 9 693 .00 .11 100.00 –.06 .07 .00 .00 100.00
Field 16
a
4,412 –.12 .12 25.25 –.18 –.06 –.13 .11 24.56 23
Variety 16 4,412 –.12 .12 24.57 –.18 –.06 –.13 .11 24.86 23
Efficiency 2 1,428 –.04 .02 100.00 –.07 –.01 –.04 .00 100.00
General 9 3,994 –.12 .11 16.74 –.20 –.05 –.14 .12 17.06 13
performance
Creativity and 3 205 –.17 .16 55.09 –.34 .01 –.18 .12 55.38
innovation
Sex 38 6,186 –.06 .11 47.14 –.09 –.02 –.06 .09 47.57 8
Lab 11 644 .01 .12 100.00 –.06 .09 .02 .00 100.00
Variety 6 365 .06 .07 100.00 .00 .11 .07 .00 100.00
Field 27 5,542 –.07 .11 40.90 –.11 –.02 –.07 .09 41.01 11
Separation 3 279 –.01 .14 57.26 –.17 .14 –.01 .10 57.26
Variety 23 5,155 –.08 .10 44.50 –.12 –.04 –.09 .08 44.69 14
Efficiency 4 1,689 –.08 .05 83.29 –.14 –.03 –.09 .02 83.99 2
General 12 4,354 –.05 .10 27.57 –.11 .00 –.06 .09 27.65
performance
Creativity and 5 380 –.15 .16 47.88 –.29 .00 –.16 .13 48.10
innovation
Age 40 10,953 –.02 .12 25.43 –.06 .01 –.03 .11 25.44
Lab 5 307 .06 .13 88.70 –.06 .18 .07 .05 88.76
Field 35 10,646 –.02 .12 23.44 –.07 .01 –.03 .11 23.45
Separation 7 688 .04 .19 28.56 –.10 .18 .04 .18 28.57
Variety 7 321 .01 .08 100.00 –.06 .07 .01 .00 100.00
Disparity 20 9,562 –.03 .11 16.78 –.08 .02 –.04 .11 16.80
Notes: The number of correlations at a moderator level (e.g., different conceptualizations) may not sum to the
overall number of correlations for a specific variable (e.g., age) if the moderator information was not reported or if
a correlation could not be categorized into the level of the moderator (e.g., included efficiency and innovation in
one measure). Results are corrected for criterion unreliability. k = number of correlations; n = number of teams;
SWMr = sample-weighted mean correlation; SWSD = sample-weighted standard deviation of the SWMr; VAR % =
percentage of variance attributed to sampling error; 95% CI = 95% confidence interval; r = corrected population
correlation; SDr = standard deviation of the corrected population correlation; VAR A % = percentage of variance
attributed to sampling error and artifact corrections; k
fs
= fail-safe k. Fail-safe k indicates the number of studies in
“file drawers” with a mean r = .00, which would reduce the results in our meta-analysis to a trivial effect size of
SWMr = .05. Fail-safe k is reported only for effects that have more than one study, are greater than SWMr = .05,
and have a confidence interval that did not include zero.
a
The results for overall race diversity in field studies were not the same as the results for race variety diversity,
because the overall result included a correlation that represented the average effect of two correlations from two
different conceptualizations generated from the same sample (e.g., variety and not coded).
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
730 Journal of Management / May 2011
Finally, Hypothesis 16 predicted that age diversity in terms of separation would have a
negative relationship with team performance. No effect was observed for age separation
diversity and team performance in field settings (r = .04). In fact, the relationship between
age diversity and team performance was negligible (r = –.04 to .07) across study settings and
the diversity conceptualizations (Research Question 7).
Discussion
Although team diversity research is thriving, unclear results and mixed conclusions are
pervasive. We believe that the lack of clarity may be attributed to a consistent oversimplifi-
cation of diversity. Our results support several thematic conclusions. First and most impor-
tant, the strength and direction of the relationship between diversity and team performance
were dependent on the specific demographic variable. Second, diversity on several variables
was primarily operationalized as an index inconsistent with the conceptualization that we
believed would have the strongest relationship with team performance, thereby suggesting
room for additional research before the specific variables are abandoned as predictors of
team performance. Finally, the team mean of organizational tenure had a stronger relation-
ship with team performance compared to that of diversity operationalizations, suggesting that
alternative team-level representations of the demographic variables may be more predictive
of team performance for some variables. We expand on these themes and indicate their impor-
tance for practitioners and future team demographic diversity research.
The Specific Demographic Diversity Variable Matters
Meta-analytic researchers have historically grouped demographic variables into catego-
ries such as highly job-related and less job-related diversity or task oriented and relations
oriented relations oriented (e.g., Joshi & Roh, 2009; Webber & Donahue, 2001). Our results
indicate differential effects for demographic variables on team performance historically
grouped within these categories, ranging from r = .23 to r = –.14. These results underscore
the importance of using precision when discussing the demographic diversity–team perfor-
mance relationship. Making statements that suggest diversity is “good,” “bad,” or unrelated
to team performance without specifying the variable of interest and the way in which diver-
sity is conceptualized, is a flawed approach.
In terms of specific results for the variables historically grouped into highly job-related
diversity, functional background variety diversity was consistently and positively related to
team performance, whereas the relationship between team performance and other variables
historically grouped into highly job-related variables was nonexistent or situationally spe-
cific. There was a small effect between functional background variety and team performance
(r = .11), which was further strengthened in situations where team member differences in
functional background would be expected to have a relationship with team performance—
that is, when team performance was general team performance (r = .12) or innovation (r = .18)
or for design and cross-functional teams (r = .16). Functional background variety diversity
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Bell et al. / Getting Specific 731
was unrelated to team performance in only limited situations—when efficiency was the team
performance metric (r = .03), for example. Educational background variety diversity was
related to team creativity or innovation (r = .23) and to team performance for TMTs (r = .13).
Education-level diversity, team tenure diversity, and organizational tenure diversity were
consistently unrelated to team performance.
We offer three potential explanations for the variation in the relationships observed
between these variables historically grouped into highly job-related diversity and team
performance. First, a team members functional background may influence a team mem-
ber’s perspective more strongly than other variables. The predominant rationale explain-
ing why diversity on highly job-related variables should be related to team performance is
the informational diversity–cognitive resource perspective (e.g., Cox & Blake, 1991;
Williams & O’Reilly, 1998). Team member differences on highly job-related variables are
important because distributional differences serve as indicators of available knowledge
and differing perspectives, which can be beneficial for task completion. It is possible that
team members’ behavior and thinking are more consistent with and shaped by their prox-
imal functional roles (Chattopadhyay et al., 1999) than by their distal developmental
experiences such as educational background, thereby making differences on functional
background more reflective of task-relevant information compared to other variables (e.g.,
educational background). Our results indicate that some highly job-related demographic
variables may be more task related than others. This finding suggests that the lack of sup-
port from meta-analyses for seemingly relevant theories (e.g., informational diversity–
cognitive resource perspective) in the context of demographic diversity and team
performance may be the result of the improper application of the theory (e.g., investigat-
ing diversity on highly job-related variables that are not actually job or task related) rather
than any problems with the general tenants of the theories.
Second, team performance is related to the extent that teams recognize the importance of
elaboration in terms of decision-relevant information (van Ginkel & van Knippenberg, 2008).
The type of team may serve as a strong situational cue indicating the importance of information
related to team members’ functional backgrounds. Consistent with this, an effect for functional
background variety diversity and team performance was observed for cross-functional and
design teams (r = .16) but not for other types of teams. A team member may have been
assigned to a team because of his or her functional expertise (e.g., engineering, marketing),
thus signaling the team members area of expertise to other team members. Team members’
awareness of one anothers expertise is tied to the extent that team members share informa-
tion (Stewart & Stasser, 1995). It could be that team members are less aware of other team
members’ standings on other potentially task-related demographic variables, such as major
in college (i.e., educational background), if there are no specific situational cues drawing
their attention to the differences among team members. This lack of awareness of team
members’ standings on the demographic variable could reduce the sharing of unique infor-
mation, thus limiting the potential for variety and diversity on other demographic variables
to benefit team performance.
A third reason that effects were observed for functional background variety diversity and
team performance more so than other highly job-related variables may be the operational-
ization of diversity. Variety reflects differences in kinds of information, and it is consistent
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
732 Journal of Management / May 2011
with the informational diversity–cognitive resource perspective, which suggests that diversity
is beneficial to performance because diverse teams can draw from different pools of infor-
mation or resources. Although functional background diversity and educational background
diversity were consistently operationalized as an index reflecting the variety conceptualiza-
tion (e.g., Blau’s index), diversity on other variables was generally operationalized as an
index consistent with separation or disparity (e.g., standard deviation, coefficient of variation).
Harrison and Klein (2007) warned that a mismatch between theoretical conceptualization
and operationalization may lead to erroneous conclusions. The relationship between these
other variables considered to be highly job related and team performance may emerge when
an operationalization consistent with the variety conceptualization is used. In other words,
there was no relationship between educational level, organizational tenure, or team tenure diver-
sity and team performance; however, our meta-analysis revealed that these demographic
variables were rarely studied using the diversity conceptualization that can reflect breadth of
knowledge that can be applied to the task. Future research should conceptualize diversity as
variety when diversity on the variable should benefit team performance by increasing the pool
of knowledge that can be applied to the task. For example, research should examine if team
tenure has a positive relationship with team performance in terms of innovation when team
members have staggered entry into a team, consistent with a variety diversity conceptualiza-
tion. This may allow for the inclusion of fresh perspectives while minimizing the disruption to
team compilation (Kozlowski et al., 1999).
Our results also indicated that for some demographic variables, the team mean best pre-
dicted team performance (and perhaps reflected increased perspectives related to the task).
Specifically, there was a small effect between team mean organizational tenure and team
performance (r = .08), which was larger for team performance in terms of efficiency (r = .14).
The same trend was observed for mean team tenure and team performance (r = .09) and
efficiency (r = .11), although the confidence intervals included zero and the efficiency esti-
mate was based on few studies (k = 4). Taken together, our results suggest that there may be
cases when elevated levels of team demographic variables, rather than diversity, are more
reflective of task-relevant information. Future research is likely to benefit from more inte-
gration between team composition and team diversity research. We encourage researchers to
consider what team-level conceptualization of the variable (e.g., elevated levels, variety
diversity) best represents more task-relevant information and to operationalize the variable at
the team-level accordingly.
Our results bring some clarity to surface-level demographics historically tied to social
categorization (i.e., age, sex, race). Results of previous meta-analyses indicated either no
relationship between these diversity variables and team performance (e.g., Horwitz &
Hortwiz, 2007; Webber & Donahue, 2001) or small effects (Joshi & Roh, 2009) when con-
text was considered. Our results indicate that the relationship between diversity on surface-
level variables and team performance varied as a function of the particular variable. Specifically,
we found no support for the relationship between age diversity and team performance
regardless of the conceptualization of age diversity. It may be that generational differences
(Smola & Sutton, 2002) rather than diversity on chronological age have implications for
team performance. Our results suggest a small negative effect between race variety diversity
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Bell et al. / Getting Specific 733
and team performance (r = –.13) and sex variety diversity and team performance (r = –.09).
These negative effects were observed only in field settings.
The negative relationship between race diversity and team performance is interesting
because diversity was consistently operationalized as an index consistent with variety (e.g.,
Blau’s, Teachman), most likely because of the categorical nature of the variable. However,
race diversity is typically hypothesized to be related to team performance through processes
such as social categorization and intergroup bias, as well as the similarity–attraction hypoth-
esis (e.g., Williams & O’Reilly, 1998), which is consistent with a separation conceptualization
(Harrison & Klein, 2007). Maximum variety reflects a distribution wherein each member
within a unit comes from a unique category of the variable (e.g., race); that is, every member
is different from the others in terms of race. Maximum separation, however, reflects teams
that have polarized subgroups. Blau’s index was commonly used to operationalize race and
sex diversity. It is intended for and so reflects heterogeneity (or whether more categories are
represented), and it should be used when race and sex diversity is expected to be related to
team performance as more categories are represented. Researchers should consider how the
operationalization of race and sex diversity captures the mechanisms through which they
expect diversity on these variables to be related to team performance. Pearsall, Ellis, and
Evan’s (2008) recent lab research on gender faultlines captures the notion of sex separation.
Specifically, they compared homogeneous teams with teams equally split in terms of men
and women. For real-world organizational teams, researchers have the additional challenge of
considering the range of race and sex categories represented in their data set, and they likely
have less ability to manipulate the size and composition of the teams. Researchers should
consider whether moderate levels of variety or a different measure reflective of maximum
separation (e.g., faultlines) better captures the form of the differences suggested by their
theoretical arguments. Future research using data simulation and other methods is needed to
help understand these issues. The potential misapplication of Blau’s index to represent race
differences other than heterogeneity has been noted in the social sciences more broadly
(Rushton, 2008). Given the limited number of studies that examined race and sex separation
diversity, we question whether the majority of previous research used the most powerful
approach to examining the relationship between team performance and sex and race diversity.
There are at least two possible explanations why a negative effect was observed between race
diversity (which has the potential for multiple categories) and team performance, even though
race diversity was operationalized as an index consistent with variety. First, race diversity in field
settings may have been generally limited to more moderate levels of variety as the upper bound.
It is unlikely that most field studies included several teams wherein each member was from a
different race. If this was the case, then the minimum and moderate levels of variety diversity
may have mimicked the separation conceptualization. Future research using data simulations and
other methods should continue to explore the effects of sampling and range restriction (e.g.,
Allen, Stanley, Williams, & Ross, 2007) on the different conceptualizations, as well as the general
adequacy of the diversity indices used to represent specific diversity conceptualizations.
Second, race variety diversity may be predictive of team performance. For example, team
members from different backgrounds or representing a variety of races may be less likely to
develop a team identity, thus resulting in team members’ not sharing task-relevant information
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
734 Journal of Management / May 2011
with the team. A clearer understanding of how race and sex diversity influences team func-
tioning could be used to inform potential solutions for reducing the observed negative effect.
For example, if in-group/out-group biases, stereotypes, and prejudices are the cause for the
negative relationship between these types of diversities and team performance, a number of
solutions may be effective in reducing the negative influence such as diversity training and
increased awareness, a change in organizational culture, or the use of superordinate goals
(Thomas, 2005). If the lack of a shared identity is inhibiting information elaboration, tech-
niques such as persuading teams of the value of diversity may be more effective (Homan,
van Knippenberg, Van Kleef, & De Dreu, 2007).
Implications of Main Effect Findings
Our hypotheses were based on one path through which the variable would be related to
team performance (e.g., function background diversity in terms of variety would be posi-
tively related to team performance). We agree that a diversity variable may be related to team
performance through different mechanisms (e.g., information diversity-cognitive resource
perspective or social categorization) and conceptualizations in different situations (van
Knippenberg et al., 2004). For example, there may be situations in which differences on a
seemingly task-related variable results in two opposing camps or cliques in a team, causing
intergroup bias and poor team performance (e.g., functional differences in a team may be
negatively related to team performance in a politically charged organization with a negative
history between functional units). However, the results of our meta-analysis suggest that
there are some main effects (albeit small) between diversity on certain demographic vari-
ables, such as functional background, sex, and race, and team performance. In general, we
expect that surface-level differences are more likely to influence performance through pro-
cesses that elicit a psychological phenomenon (e.g., subgroup categorization) than serve as
a surrogate for a psychological attribute relevant to the task.
Second, we examined a subset of moderators of the demographic diversity and team
performance relationship (i.e., performance outcome, team type, study setting), but
agree that there are other potentially important moderators that deserve consideration in
future research (e.g., context; Joshi & Roh, 2009; time the team has been together;
Argote & McGrath, 1993). We hope that researchers will continue to examine the impor-
tance of moderators and mediators within the context of a specific demographic diver-
sity variable–team performance relationship. For example, we observed a consistent
positive effect between functional background variety diversity and team performance,
but theoretical guidance is limited regarding which functional differences are important.
Future research should examine the specific functional distinctions (e.g., marketing,
research and development) and contexts (e.g., industry, organizational structure) within
which these functional distinctions may be important. Future research could also inves-
tigate not only the circumstances within which functional background variety diversity
may be important for team performance (e.g., Cannella, Park, & Lee, 2008) but also
ways to capitalize on the benefits of functional background variety diversity (e.g.,
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Bell et al. / Getting Specific 735
Boone & Hendriks, 2009). Meta-analyses focused on a specific demographic variable
might also help researchers and practitioners understand how to capitalize on team
member differences (e.g., how can we best capitalize on functional differences to
improve the performance of design teams?).
Practical Implications
Our results have specific implications for practitioners in organizations that utilize work
teams. Staffing teams with members from different functional backgrounds (e.g., marketing,
engineering) may be beneficial, particularly in situations where diverse functional perspec-
tives are tied to the task such as in design teams or product development teams or when
creativity or innovation is of primary importance. Staffing TMTs or teams responsible for
innovation with individuals from a variety of educational specialties might lead to improved
performance. When efficiency (rather than innovation) is important, practitioners should
consider increasing the average organizational tenure of the team members while balancing
the need for performance with the need for developing new employees. However, diversity
in terms of race or sex appears to have negative consequences on team performance.
Therefore, staffing aimed specifically at increasing the racial or sex diversity of teams for
the purposes of increasing team performance is misguided.
Our findings offer guidance in terms of diversity training interventions that are intended
to reduce the negative consequences associated with team diversity. The effect of diversity is
not consistent across all variables. Therefore, the utility of diversity training will vary, depend-
ing on the relationship between the diversity variable and team performance. Specifically,
organizations may achieve greater utility by expending resources to bridge team members’
race and sex differences rather than their age differences. This is especially important in that
diversity training is likely to be the only intervention available to organizations that want to
reduce the negative effects of race or sex diversity, given the illegality of employment deci-
sions based on race and sex. Furthermore, this is an important consideration given the recent
movement toward a focus on generational differences in the workplace.
Limitations
A small number of correlations represented the relationships between demographic diver-
sity variables and team performance at some levels of the moderators. We do not intend for
our meta-analysis to be the final word on the study of demographic diversity and team per-
formance relationships but rather to organize and quantitatively summarize the current lit-
erature in a way that is consistent with prevailing frameworks, and to highlight areas in need
of future research. The meta-analytic effect size estimates were based on a small number of
studies (e.g., k < 5) and are in no way intended to represent the population; however, such
estimates do offer guidance into the possible trend and, more important, should serve as a call
to areas in need of further primary research or simulation.
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
736 Journal of Management / May 2011
Second, in addition to variables included in this meta-analysis, researchers have examined
other demographic diversity variables thought to be related to team performance, such as mari-
tal status (Harrison et al., 2002), industry experience (K. G. Smith et al., 1994), and nationality
(Kilduff, Angelmar, & Mehra, 2000). To date, a limited number of studies have investigated
these types of diversity. We limited our study to commonly studied demographic variables to
have meaningful meta-analytic estimates. Given that our results indicated variability in the
extent to which specific demographic diversity variables are related to team performance,
research should continue to explore these and other forms of diversity, carefully articulating
the theoretical connection among the specific variable, diversity conceptualization, and team
performance.
Finally, our focus was at the level of a specific attribute (e.g., educational background, race).
Researchers have suggested that diversity on multiple dimensions of diversity may create
strong divides (i.e., faultlines) among team members that disrupt team processes such as
information elaboration (Lau & Murnighan, 1998; Rico, Molleman, Sanchez-Manzanares, &
Van der Vegt, 2007). Given the recency of faultline research, we could not include measures
such as faultline strength for combinations of specific variables. We applaud these research
efforts and hope that our results can help to inform these researchers of the particular attri-
butes of diversity that may be meaningful in terms of affecting team performance.
Conclusion
As demographic diversity research has continued to accumulate, researchers have been
perplexed by the limited meta-analytic evidence of a demographic diversity–team performance
relationship despite its theoretical appeal. We revisited the team demographic diversity–team
performance relationship, moving beyond previous meta-analyses by examining the rela-
tionship for specific demographic variables (e.g., functional background, race) and concep-
tualizations of diversity (i.e., separation, variety, and disparity). We also investigated the
influence of study setting, team type, and performance criteria within those relationships. We
believe that our meta-analytic results provide a meaningful summary of the demographic
diversity and team performance literature to date by examining relationships more consistent
with prevailing theories and frameworks.
References
*
Articles contributing correlations to the meta-analysis are indicated with asterisks.
Allen, N. J., Stanley, D. J., Williams, H. M., & Ross, S. J. 2007. Assessing the impact of nonresponse on work group
diversity effects. Organizational Research Methods, 10: 262-286.
*
Alves, J. C. 2008. A multilevel analysis of the association among individual capabilities, team leadership behav-
iors, and performance in China (Doctoral dissertation). Retrieved from Dissertations and Theses: Full Text.
(Publication No. AAT 3315480)
*
Amason, A. C., Shrader, R. C., & Tompson, G. H. 2006. Newness and novelty: Related top management team
composition to new venture performance. Journal of Business Venturing, 21: 125-148.
*
Ancona, D. G., & Caldwell, D. F. 1992. Demography and design: Predictors of new product team performance.
Organization Science, 3: 321-341.
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Bell et al. / Getting Specific 737
Argote, L., & McGrath, J. E. 1993. Group processes in organizations: Continuity and change. In C. L. Cooper &
I. T. Robertson (Eds.), International review of industrial and organization psychology (Vol. 8, pp. 333-389).
New York: Wiley.
Arthur, W., Jr., Bennett, W., Jr., & Huffcutt, A. I. 2001. Conducting meta-analysis using SAS. Mahwah, NJ:
Erlbaum.
*
Balkundi, P., Kilduff, M., Barness, Z. I., & Micheal, J. H. 2007. Demographic antecedents and performance con-
sequences of structural holes in work teams. Journal of Organizational Behavior, 28: 241-260.
*
Bantel, K. A. 1993a. Strategic clarity in banking: Role of top management team demography. Psychological
Reports, 73: 1187-1201.
*
Bantel, K. A. 1993b. Top team, environment, and performance effects on strategic planning formality. Group and
Organizational Management, 18: 436-458.
*
Bantel, K. A., & Jackson, S. E. 1989. Top management and innovation in banking: Does the composition of the
top team make a difference? Strategic Management Journal, 10: 107-124.
*
Barsade, S. G., Ward, A. J., Turner, J. D. F., & Sonnenfeld, J. 2000. To your heart’s content: A model of affective
diversity in top management teams. Administrative Science Quarterly, 45: 802-836.
Beal, D. J., Cohen, R. R., Burke, M. J., & McLendon, C. L. 2003. Cohesion and performance in groups: A meta-
analytic clarification of the construct relationships. Journal of Applied Psychology, 88: 989-1004.
Bell, S. T. 2007. Deep-level composition variables as predictors of team performance: A meta-analysis. Journal of
Applied Psychology, 92: 595-615.
*
Boeker, W. 1997. Strategic change: The influence of managerial characteristics and organizational growth. Academy
of Management Journal, 40: 152-170.
*
Boone, C., & Hendriks, W. 2009. Top management team diversity and firm performance: Moderators of functional
background and locus of control diversity. Management Science, 55: 165-180.
*
Bow, E. D. 2001. Expectancy effects for group performance on diverse and non-diverse groups. Dissertation
Abstracts International, 62: 1625.
Bowers, C. A., Pharmer, J. A., & Salas, E. 2000. When member heterogeneity is needed in work teams: A meta-
analysis. Small Group Research, 31: 305-327.
Brewer, M. B. 1979. In-group bias in the minimal intergroup situation: A cognitive-motivational analysis. Psychological
Bulletin, 86: 307-324.
*
Bunderson, J. S. 2003. Team member functional background and involvement in management teams: Direct effects
and the moderating role of power centralization. Academy of Management Journal, 46: 458-474.
*
Bunderson, J. S., & Sutcliffe, K. M. 2002. Comparing alternative conceptualizations of functional diversity on
management teams: Process and performance effects. Academy of Management Journal, 45: 875-893.
*
Burpitt, W. J. 1995. Pro-active learning among project teams: The influence of leader behavior, functional diver-
sity, and cross-boundary communication (Doctoral dissertation). Retrieved from Dissertations & Theses: Full
Text. (Publication No. AAT 9605097)
Byrne, D. E. 1971. The attraction paradigm. New York: Academic Press.
*
Cady, S. H., & Valentine, J. 1999. Team innovation and perceptions of consideration: What difference does diver-
sity make? Small Group Research, 30: 730-750.
Cannella, A. A., Park, J. H., & Lee, H. U. 2008. Top management team functional background diversity and firm
performance: Examining the roles of team member colocation and environmental uncertainty. Academy of
Management Journal, 51: 768-784.
Carpenter, M. A., Geletkanycz, M. A., & Sanders, W. G. 2004. Upper echelons research revisited: Antecedents,
elements, and consequences of top management team composition. Journal of Management, 30: 749-778.
*
Cavarretta, F. 2008. Team diversity and performance variability (Doctoral dissertation). Retrieved from Dissertations &
Theses: Full Text. (Publication No. AAT 3329964)
*
Cavazotte, F. 2003. Diversity, efficacy beliefs and cooperation in small work groups: Exploring the role of identi-
fication and emotional competence. Dissertation Abstracts International, 63: 3630.
Chatman, J. A. 1991. Matching people and organizations: Selection and socialization in public accounting firms.
Administrative Science Quarterly, 36: 487-516.
Chattopadhyay, P., Glick, W. H., Miller, C., & Huber, G. P. 1999. Determinants of executive beliefs: Comparing func-
tional conditioning and social influence. Strategic Management Journal, 20: 763-789.
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
738 Journal of Management / May 2011
Cohen, J. 1992. A power primer. Psychological Bulletin, 112: 155-159.
Cox, T., & Blake, S. 1991. Managing cultural diversity: Implications for organizational competitiveness. Academy
of Management Executive, 5: 45-46.
*
Cummings J. N. 2004. Work groups, structural diversity, and knowledge sharing in a global organization.
Management Science, 50: 352-364.
*
Daily, B., Whatley, A., Ash, S. R., & Steiner, R. L. 1996. The effects of a group decision support system on cultur-
ally diverse and culturally homogeneous group decision making. Information & Management, 30: 281-289.
Dearborn, D. C., & Simon, H. A. 1958. Selective perception: A note on the departmental identification of execu-
tives. Sociometry, 21: 140-144.
Devine, D. J. 2002. A review and integration of classification systems relevant to teams in organizations. Group
Dynamics, 6: 291-310.
Devine, D. J., & Phillips, J. L. 2001. Do smarter teams do better. Small Group Research, 32: 507-532.
*
Drach-Zahavy, A., & Freund, A. 2007. Team effectiveness under stress: A structural contingency approach. Journal
of Organizational Behavior, 28: 423-450.
*
Drach-Zahavy, A., & Somech, A. 2001. Understanding team innovation: The role of team processes and structures.
Group Dynamics: Theory, Research, and Practice, 5: 111-123.
*
Drach-Zahavy, A., & Somech, A. 2002. Team heterogeneity and its relationship with team support and team effec-
tiveness. Journal of Educational Administration, 40: 44-66.
*
Earley, P. C., & Mosakowski, E. 2000. Creating hybrid team cultures: An empirical test of transnational team
functioning. Academy of Management Journal, 43: 26-49.
*
Ely, R. J. 2004. A field study of group diversity, participation in diversity education programs, and performance.
Journal of Organizational Behavior, 25: 755-780.
Fiske, S. T., & Neuberg, S. L. 1990. A continuum of impression formation, from category-based to individuating
processes: Influences of information and motivation on attention and interpretation. Advances in Experimental
Social Psychology, 23: 1-74.
Fiske, S. T., & Taylor, S. 2007. Social cognition: From brains to culture. New York: McGraw-Hill.
*
Foo, M. D., Wong, P. K., & Ong, A. 2005. Do others think you have a viable business idea? Team diversity and
judges’ evaluation of ideas in a business plan competition. Journal of Business Venturing, 20: 385-402.
*
Gellert, F. J., & Kuipers, B. S. 2008. Short- and long-term consequences of age in work teams. Career Development
International, 13: 132-149.
*
Giambatista, R. C. 2000. Reaping the fruits: Dimensions of diversity, communication technology, and team effec-
tiveness. Dissertation Abstracts International, 60: 3432.
*
Griest, D. L. 1996. The process and performance of diverse teams. Dissertation Abstracts International, 56: 4039.
Gully, S. M., Devine, D. J., & Whitney, D. J. 1995. A meta-analysis of cohesion and performance: Effects of level
of analyses and task interdependence. Small Group Research, 26: 497-520.
*
Hambrick, D. C., Cho, T. S., & Chen, M. J. 1996. The influence of top management team heterogeneity on firms
competitive moves. Administrative Science Quarterly, 41: 659-684.
Hambrick, D. C., & Mason, P. A. 1984. Upper echelons: The organization as a reflection of its top managers.
Academy of Management Review, 9: 193-206.
Harrison, D. A., & Klein, K. J. 2007. What’s the difference? Diversity constructs as separation, variety, or disparity
in organizations. Academy of Management Review, 32: 1199-1228.
*
Harrison, D. A., Price, K. H., Gavin, J. H., & Florey, A. T. 2002. Time, teams, and task performance: Changing effects
of surface- and deep-level diversity on group functioning. Academy of Management Journal, 45: 1029-1045.
*
Hayes, B. H. 2005. Brainstorming on diversity: The effects of racial diversity on group performance. Dissertation
Abstracts International, 65: 5459.
*
Henneke, D., & Luthje, C. 2007. Interdisciplinary heterogeneity as a catalyst for product innovativeness of entre-
preneurial teams. Creativity and Innovation Management, 16: 121-132.
*
Herrmann, P., & Datta, D. K. 2005. Relationships between top management team characteristics and international
diversification: An empirical investigation. British Journal of Management, 16: 69-78.
*
Herron, M. T. 1993. The effects of the ethnic and gender diversity of work teams on the perceptions of performance
outputs. Dissertation Abstracts International, 54: 2255.
Hewstone, M., Hantzi, A., & Johnston, L. 1991. Social categorization and person memory: The pervasiveness of race
as an organizing principle. European Journal of Social Psychology, 21: 517-528.
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Bell et al. / Getting Specific 739
Homan, A. C., van Knippenberg, D., Van Kleef, G. A., & De Dreu, C. K. W. 2007. Bridging faultlines by valuing
diversity: Diversity beliefs, information elaboration, and performance in diverse work groups. Journal of
Applied Psychology, 92: 1189-1199.
Horwitz, S. K., & Horwitz, I. B. 2007. The effects of team diversity on team outcomes: A meta-analytic review of
team demography. Journal of Management, 33: 987-1015.
Hunter, J. E., & Schmidt, F. L. 2004. Methods of meta-analysis: Correcting error and bias in research findings.
Newbury Park, CA: Sage.
*
Jackson, S. E., & Joshi, A. 2004. Diversity in a social context: A multi-attribute, multilevel analysis of team diver-
sity and sales performance. Journal of Organizational Behavior, 25: 675-702.
Jackson, S. E., Joshi, A., & Erhardt, N. L. 2003. Recent research on team and organizational diversity: SWOT
analysis and implications. Journal of Management, 29: 801-830.
Jackson, S. E., May, K. E., & Whitney, K. 1995. Understanding the dynamics of diversity in decision-making teams.
In R. A. Guzzo & E. Salas (Eds.) Team effectiveness and decision making in organizations (pp. 204-261). San
Francisco: Jossey-Bass.
*
Jehn, K. A. 1995. A multimethod examination of the benefits and detriments of organizational conflict. Administrative
Science Quarterly, 40: 256-282.
*
Jehn, K. A., & Bezrukova, K. 2004. A field study of group diversity, work group context, and performance. Journal
of Organizational Behavior, 25: 703-729.
*
Jordan, M. H. 2002. The relationship of individual difference and group process variables with self-managed team
performance: A field investigation. Dissertation Abstracts International, 62: 3469.
*
Joshi, A. A. 2002. How does context matter? Examining the process and performance outcomes of work team heteroge-
neity (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text. (Publication No. AAT 3055066)
Joshi, A. A., & Roh, H. 2009. The role of context in work team diversity research: A meta-analytic diversity.
Academy of Management Journal, 52: 599-627.
Katz, R. 1982. The effect of group longevity on project communication and performance. Administrative Science
Quarterly, 27: 81-104.
*
Kearney, E., & Gebert, D. 2009. Managing diversity and enhancing team outcomes: The promise of transforma-
tional leadership. Journal of Applied Psychology, 94: 77-89.
*
Kearney, E., Gebert, D., & Voelpel, S. C. 2009. When and how diversity benefits teams: The importance of the
team members’ need for cognition. Academy of Management, 52: 581-596.
*
Keller, R. T. 1994. Technology-information processing fit and the performance of R&D project groups: A test of
contingency theory. Academy of Management Journal, 37: 167-179.
*
Keller, R. T. 2001. Cross-functional project groups in research and new product development: Diversity, commu-
nications, job stress, and outcomes. Academy of Management Journal, 44: 547-555.
*
Kilduff, M., Angelmar, R., & Mehra, A. 2000. Top management-team diversity and firm performance: Examining
the role of cognitions. Organization Science, 11: 21-34.
*
Kirkman, B. L., Tesluk, P. E., & Rosen, B. 2004. The impact of demographic heterogeneity and team leader-team
member demographic fit on team empowerment and effectiveness. Group and Organization Management, 29:
334-368.
Kochan, T. Bezrukova, K., Ely, R., Jackson, S., Joshi, A. Jehn, K., et al. 2003. The effects of diversity on business
performance: Report of the diversity research network. Human Resource Management, 42: 3-21.
Kozlowski, S. W. J., Gully, S. M., Nason, E. R., & Smith, E. M. 1999. Developing adaptive teams: A theory of
compilation and performance across levels and time. In D. R. Ilgen & E. D. Pulakos (Eds.), The changing
nature of performance: Implications for staffing, motivation, and development (pp. 240-292). San Francisco:
Jossey-Bass.
Lau, D. C., & Murnighan, J. K. 1998. Demographic diversity and faultlines: The compositional dynamics of orga-
nizational groups. Academy of Management Review, 23: 325-340.
*
Lee, C., & Farh, J. 2004. Joint effects of group efficacy and gender diversity on group cohesion and performance.
Applied Psychology: An International Review, 53: 136-154.
Levine, J. M., & Moreland, R. L. 1990. Progress in small group research. Annual Review of Psychology, 41: 585-634.
*
L’Heureux, R. A. 1994. The effect if internationalization on the relationship between TMT characteristics and firm
performance (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text. (Publication No. AAT
9422466)
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
740 Journal of Management / May 2011
*
Li, J., & Hambrick, D. C. 2005. Factional groups: A new vantage on demographic faultlines, conflict, and disinte-
gration in work teams. Academy of Management Journal, 48: 794-813.
Locke, E. A., & Latham, G. P. 2002. Building a practically useful theory of goal setting and task motivation. A
35-year odyssey. American Psychologist, 57: 705-717.
*
Lovelace, K., Shapiro, D. L., & Weingart, L. R. 2001. Maximizing cross-functional new product teams’ innovative-
ness and constraint adherence: A conflict communications perspective. Academy of Management Journal, 44:
779-793.
Mannix, E., & Neale, M. A. 2005. The promise and reality of diverse teams in organizations. Psychological Science
in the Public Interest, 6: 31-55.
*
Marsteller, J. A. 2003. The relationship between non-racial diversity in team composition and performance and
creativity in a chronic illness care quality improvement intervention (Doctoral dissertation). Retrieved from
Dissertations & Theses: Full Text. (Publication No. AAT 3121597)
*
Mayo, M. 1999. Work team diversity: An examination of a process model using artificial intelligence and social
network methods. Dissertation Abstracts International, 59: 3543.
McGrath, J. E., Berdahl, J. L., & Arrow, H. 1995. Traits, expectations, culture, and clout: The dynamics of diversity
in work groups. In S. E. Jackson & M. N. Ruderman (Eds.), Diversity in work teams: Research paradigms for
a changing workplace (pp. 17-45). Washington, DC: American Psychological Association.
Messick, D., & Mackie, D. 1989. Intergroup relations. Annual Review of Psychology, 40: 45-81.
*
Michel, J. G., & Hambrick, D. C. 1992. Diversification posture and top management team characteristics. Academy
of Management Journal, 35: 9-37.
*
Miller, M. C. 1995. Relationships among cultural heterogeneity, group processes, and group outcomes: A model
and empirical study. Dissertation Abstracts International, 56: 271.
Milliken, F. J., Bartel, C. A., & Kurtzberg, T. R. 2003. Diversity and creativity in work groups. In P. B. Paulus &
B. A. Nijstad (Eds.), Group creativity through collaboration (pp. 32-62). New York: Oxford University Press.
Milliken, F. J., & Martins, L. L. 1996. Searching for common threads: Understanding the multiple effects of diver-
sity in organizations work groups. Academy of Management Review, 21: 402-433.
*
Mohammed, S., & Angell, L. C. 2004. Surface- and deep-level diversity in workgroups: Examining the moderating
effects of team orientation and team process on relationship conflict. Journal of Organizational Behavior, 25:
1015-1039.
*
Mosciski, P. D. 2004. The influence of top management team heterogeneity and internationalization of firm per-
formance (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text. (Publication No. AAT
3118030)
*
Myaskovsky, L., Unikel, E., & Dew, M. A. 2005. Effects of gender diversity on performance and interpersonal
behavior in small work groups. Sex Roles, 52: 645-657.
*
Nancarrow, L. 2001. The impact of work group diversity on organizational outcomes (MBA dissertation).
Retrieved from Dissertations & Theses: Full Text. (Publication No. AAT MQ68319)
*
Nandkeolyar, A. K. 2008. How do teams learn? Shared mental models and transactive memory systems as deter-
minants of team learning and effectiveness (Doctoral dissertation). Retrieved from Dissertations & Theses: Full
Text. (Publication No. AAT 3315997)
*
Nath, S. S. 1995. Top management team characteristics and firm R&D intensity (Unpublished doctoral disserta-
tion). University of Kentucky, Lexington.
*
O’Leary, B. J. 2005. The effects of racial diversity on group performance: Freeing reality from perception.
Dissertation Abstracts International, 66: 2332.
*
Olson, B. J., Parayitam, S., Twigg, N. W. 2006. Mediating role of strategic choice between top management team
diversity and firm performance: Upper echelons theory revisited. Journal of Business and Management, 12:
111-126.
*
Orlitzky, M., & Benjamin, J. D. 2003. The effects of sex composition on small-group performance in a business
school case competition. Academy of Management Learning and Education, 2: 128-138.
Orwin, R. G. 1983. A fail-safe n for effect size in meta-analysis. Journal of Educational Statistics, 8: 157-159.
*
Paletz, S. B. F., Peng, K., Erez, M., & Maslach, C. 2004. Ethnic composition and its differential impact on group
processes in diverse teams. Small Group Research, 35: 128-157.
*
Palmer, S. L. 2001. Executive group decision-making and demographic diversity: Exploring the roles of generating
issues and ideas and evaluating outcomes. Dissertation Abstracts International, 62: 1628.
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Bell et al. / Getting Specific 741
*
Pearsall, M. J., Ellis, A. P. J., & Evans, J. M. 2008. Unlocking the effects of gender faultlines on team creativity:
Is activation the key? Journal of Applied Psychology, 93: 225-234.
*
Pelled, L. H. 1993. Work group diversity and its consequences: The role of substantive and affective conflict.
Dissertation Abstracts International, 54: 2685.
Pelled, L. H. 1996. Demographic diversity, conflict, and work group outcomes: An intervening process theory.
Organization Science, 7: 615-631.
*
Pelled, L. H., Eisenhardt, K. M., & Xin, K. R. 1999. Exploring the black box: An analysis of work group diversity,
conflict, and performance. Administrative Science Quarterly, 44: 1-28.
*
Peters, L., & Karren, R. J. 2009. An examination of the roles of trust and functional diversity on virtual team
performance ratings. Group and Organization Management, 34: 479-504.
*
Phillips, K. W., Northcraft, G. B., & Neale, M. A. 2006. Surface-level diversity and decision-making in groups:
When does deep-level similarity help? Group Processes & Intergroup Relations, 9: 467-482.
*
Puck, J., Rygl, D., & Kittler, M. 2006. Cultural antecedents and performance consequences of open communication
and knowledge transfer in multicultural process-innovation teams. Journal of Organizational Transformation
and Social Change, 3: 223-241.
*
Raver, J. L., & Gelfand, M. J. 2005. Beyond the individual victim: Linking sexual harassment, team processes, and
team performance. Academy of Management Journal, 48: 387-400.
*
Reagans, R., & Zuckerman, E. W. 2001. Networks, diversity, and productivity: The social capital of corporate
R&D teams. Organization Science, 12: 502-517.
*
Richard, O. C., & Shelor, R. M. 2002. Linking top management team age heterogeneity to firm performance:
Juxtaposing two mid-range theories. International Journal of Human Resource Management, 13: 958-974.
Rico, R., Molleman, E., Sanchez-Manzanares, M., & Van der Vegt, G. S. 2007. The effects of diversity faultlines
and team task autonomy on decision quality and social integration. Journal of Management, 33: 111-132.
*
Ringseis, E. L. 2000. Lending a helping hand: The multilevel effects of diversity on performance. Dissertation
Abstracts International, 61: 1122.
*
Roberge, M. 2007. When and how does diversity increase group performance? A theoretical model followed by
an experimental study (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text. (Publication
No. AAT 3276680)
*
Rodriguez, R. A. 1998. Challenging demographic reductionism: A pilot study investigating diversity in group
composition. Small Group Research, 29: 744-759.
Rosenthal, R. 1979. The “file drawer problem” and tolerance for null results. Psychological Bulletin, 86: 638-641.
Rushton, M. 2008. A note on the use and misuse of the racial diversity index. Public Studies Journal, 36: 445-459.
Schippers, M. C., Hartog, D. N., Koopman, P. L., & Wienk, J. A. 2003. Diversity and team outcomes: The moder-
ating effects of outcome interdependence and group longevity and the mediating effect of reflexivity. Journal
of Organizational Behavior, 24: 779-802.
Schneider, B. 1987. The people make the place. Personnel Psychology, 40: 437-453.
Schneider, B., Goldstein, H. W., & Smith, D. B. 1995. The ASA framework: An update. Personnel Psychology, 48:
747-773.
*
Sessa, V. I. 1994. Conflict within small decision-making and problem-solving teams: A process model. Dissertation
Abstracts International, 55: 6730.
Sewell, W. H., & Shah, V. P. 1967. Socioeconomic status, intelligence, and the attainment of higher education.
Sociology of Higher Education, 40: 1-23.
*
Shin, S. J., & Zhou, J. 2007. When is educational specialization heterogeneity related to creativity in research
and development teams? Transformational leadership as a moderator. Journal of Applied Psychology, 92:
1709-1721.
*
Simons, T., Pelled, L. H., & Smith, K. A. 1999. Making use of difference: Diversity, debate, and decision compre-
hensiveness in top management teams. Academy of Management Journal, 42: 662-673.
*
Smith, A., Houghton, S. M., Hood, J. N., & Ryman, J. A. 2006. Power relationships among top managers: Does
top management team power distribution matter for organizational performance? Journal of Business Research,
59: 622-629.
*
Smith, K. G., Smith, K A., Olian, J. D., Sims, H. P, O’Bannon, D. P., & Scully, J. A. 1994. Top management team
demography and process: The role of social integration and communication. Administrative Science Quarterly,
39: 412-438.
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
742 Journal of Management / May 2011
Smola, K. W., & Sutton, C. D. 2002. Generational differences: Revisiting generational work values for the new
millennium. Journal of Organizational Behavior, 23: 363-382.
*
Somech, A. 2006. The effects of leadership style and team process on performance and innovation in functionally
heterogeneous teams. Journal of Management, 32: 132-157.
Stangor, C., Lynch, L., Duan, C., & Glas, B. 1992. Categorization of individuals on the basis of multiple social
features. Journal of Personality and Social Psychology, 62: 207-218.
*
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.
Stewart, D. D., & Stasser, G. (1995). Expert role assignment and information sampling during collective recall and
decision making. Journal of Social Psychology, 69: 619-628.
*
Stewart, M. M., & Johnson, O. L. 2009. Leader–member exchange as a moderator of the relationship between
work group diversity and team performance. Group and Organization Management, 34: 507-535.
Sturman, M. C. 2003. Searching for the inverted u-shaped relationship between time and performance: Meta-analyses
of the experience/performance, tenure/performance, and age/performance relationships. Journal of Management,
29: 609-640.
*
Szumal, J. L. 1995. An examination of the multiple-level effects of gender and race on group problem solving
(Unpublished doctoral thesis). University of Illinois at Chicago.
Tajfel, H. 1969. Cognitive aspects of prejudice. Journal of Social Issues, 25: 79-97.
Tajfel, H., & Turner, J. C. 1979. An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.),
The social psychology of intergroup relations (pp. 33-47). Monterey, CA: Brooks/Cole.
Thomas, K. M. 2005. Diversity dynamics in the workplace. Belmont, CA: Wadsworth.
Tsui, A. S., & Gutek, B. A. 1999. Demographic differences in organizations: Current research and future directions.
New York: Lexington Books/Macmillan.
Tsui, A. S., & O’Reilly, C. A. 1989. Beyond simple demographic effects: The importance of relational demography
in superior-subordinate dyads. Academy of Management Journal, 32: 402-423.
*
Tyran, K. L., & Gibson, C. B. 2008. Is what you see what you get? The relationship among surface-level and deep-
level heterogeneity characteristics, group efficacy, and team reputation. Group and Organizational Management,
33: 46-76.
*
Van der Vegt, G. S., & Bunderson, J. S. 2005. Learning and performance in multidisciplinary teams: The impor-
tance of collective team identification. Academy of Management Journal, 48: 532-547.
*
Van der Vegt, G. S., Bunderson, J. S., Kuipers, B. in press. Why turnover maters in self-managing work teams:
Learning, social integration, and task flexibility. Journal of Management.
van Ginkel, W. P., & van Knippenberg, D. A. 2008. Group information elaboration and group decision making: The
role of shared task representations. Organizational Behavior and Human Decision Processes, 105: 82-97.
van Knippenberg, D. A., De Dreu, C. K. W., & Homan, A. C. 2004. Work group diversity and group performance:
An integrative model and research agenda. Journal of Applied Psychology, 89: 1008-1022.
van Knippenberg, D. A., & Schippers, M. C. 2007. Work group diversity. Annual Review of Psychology, 58: 515-541.
Van Maanen, J., & Schein, E. H. 1979. Toward a theory of organizational socialization. In B. M. Staw & L. L. Cummings
(Eds.) Research in organizational behavior (Vol. 1, pp. 209-264). Greenwich, CT: JAI Press.
*
Wagner, W. G., Pfeffer, J., & O’Reilly, C. A. 1984. Organizational demography and turnover in top-management
groups. Administrative Science Quarterly, 29: 74-92.
*
Watson, W. E., Johnson, L., & Zgourides, G. D. 2002. The influence of ethnic diversity on leadership, group pro-
cess, and performance: An examination of learning teams. International Journal of Intercultural Relations, 26:
1-16.
Watson, W. E., Kumar, K., & Michaelsen, L. K. 1993. Cultural diversity’s impact on interaction process and per-
formance: Comparing homogeneous and diverse task groups. Academy of Management Journal, 36: 590-602.
Webber, S. S., & Donahue, L. M. 2001. Impact of highly and less job-related diversity on work group cohesion and
performance: A meta-analysis. Journal of Management, 27: 141-162.
*
Wegge, J., Roth, C., Neubach, B., Schmidt, K., & Kanfer, R. 2008. Age and gender diversity as determinants of
performance and health in a public organization: The role of task complexity and group size. Journal of Applied
Psychology, 93: 1301-1313.
at DEPAUL UNIV LIBRARIES on April 10, 2011jom.sagepub.comDownloaded from
Bell et al. / Getting Specific 743
West, M. A., & Farr, J. L. 1990. Innovation at work. In M. A. West, & J. L. Farr (Eds.), Innovation and creativity
at work: Psychological and organizational strategies (pp. 3-13). Chichester, UK: Wiley.