ArticlePDF Available

Bridging Faultlines by Valuing Diversity: Diversity Beliefs, Information Elaboration, and Performance in Diverse Work Groups

Authors:

Abstract

Although there are numerous potential benefits to diversity in work groups, converging dimensions of diversity often prevent groups from exploiting this potential. In a study of heterogeneous decision-making groups, the authors examined whether the disruptive effects of diversity faultlines can be overcome by convincing groups of the value of diversity. Groups were persuaded either of the value of diversity or the value of similarity for group performance, and they were provided with either homogeneous or heterogeneous information. As expected, informationally diverse groups performed better when they held pro-diversity rather than pro-similarity beliefs, whereas the performance of informationally homogeneous groups was unaffected by diversity beliefs. This effect was mediated by group-level information elaboration. Implications for diversity management in organizations are discussed.
Bridging Faultlines by Valuing Diversity: Diversity Beliefs, Information
Elaboration, and Performance in Diverse Work Groups
Astrid C. Homan
University of Amsterdam
Daan van Knippenberg
Erasmus University Rotterdam
Gerben A. Van Kleef and Carsten K. W. De Dreu
University of Amsterdam
Although there are numerous potential benefits to diversity in work groups, converging dimensions of
diversity often prevent groups from exploiting this potential. In a study of heterogeneous decision-
making groups, the authors examined whether the disruptive effects of diversity faultlines can be
overcome by convincing groups of the value of diversity. Groups were persuaded either of the value of
diversity or the value of similarity for group performance, and they were provided with either homoge-
neous or heterogeneous information. As expected, informationally diverse groups performed better when
they held pro-diversity rather than pro-similarity beliefs, whereas the performance of informationally
homogeneous groups was unaffected by diversity beliefs. This effect was mediated by group-level
information elaboration. Implications for diversity management in organizations are discussed.
Keywords: diversity, faultlines, diversity beliefs, information elaboration, team performance
When important decisions have to be made, organizations often
turn to groups. Especially when group members differ with respect
to the information and expertise they bring to the table, groups may
outperform individuals in terms of the quality of the decisions they
reach (Argote, Gruenfeld, & Naquin, 2000; Ilgen, 1999; see also
Hinsz, Tindale, & Vollrath, 1997). Organizations therefore in-
creasingly rely on cross-functional work groups and project teams
in an attempt to stimulate innovation, solve problems, and make
decisions. Often, informational diversity within such teams comes
hand in hand with differences on other dimensions, such as demo-
graphic characteristics and deeply held values and beliefs (Harri-
son, Price, & Bell, 1998; Jehn, Northcraft, & Neale, 1999; Mil-
liken & Martins, 1996; Phillips, 2003; Williams & O’Reilly,
1998). When different dimensions of diversity converge (e.g.,
when all team members with technical expertise are male and
those with knowledge about marketing and sales are female),
so-called diversity faultlines emerge that may disrupt group pro-
cesses (Lau & Murnighan, 1998).
A number of studies have documented negative effects of di-
versity faultlines on group functioning (e.g., Homan, van Knip-
penberg, Van Kleef, & De Dreu, in press; Lau & Murnighan, 2005;
Phillips, Mannix, Neale, & Gruenfeld, 2004; Thatcher, Jehn, &
Zanutto, 2003). Accordingly, diversity faultlines are generally
believed to have a negative impact on group processes and per-
formance (for a review, see van Knippenberg & Schippers, 2007).
In this article, we challenge the widely shared assumption that
groups with diversity faultlines cannot benefit from their informa-
tional diversity (cf. Jehn et al., 1999) by focusing on the role of
group members’ beliefs about diversity. We argue and show ex-
perimentally that groups with diversity faultlines may effectively
use their informational diversity when group members believe in
the value of diversity.
Informational Diversity and Diversity Faultlines
Informational diversity is defined as “differences in knowledge
bases and perspectives that members bring to the group” (Jehn et
al., 1999, p. 743); it also has been referred to as functional or
knowledge diversity (Pelled, Eisenhardt, & Xin, 1999; Phillips et
al., 2004). Van Knippenberg, De Dreu, and Homan (2004) argued
that informational diversity can enhance group performance by
stimulating the elaboration of task-relevant information and per-
spectives (see also Cox, Lobel, & McLeod, 1991). Building on the
conceptualization of groups as information processors (Hinsz et
al., 1997), van Knippenberg, De Dreu, and Homan defined group
information elaboration as the exchange of information and per-
spectives, individual-level processing of the information and per-
spectives, feeding back the results of this individual-level process-
ing into the group, and discussion and integration of their
implications. This deeper and more extensive consideration of
task-relevant information may lead diverse groups to outperform
more homogeneous groups on tasks with clear information-
processing and decision-making requirements (e.g., Bowers,
Pharmer, & Salas, 2000; De Dreu, 2007; Jehn et al., 1999; Schol-
ten, van Knippenberg, Nijstad, & De Dreu, in press).
Astrid C. Homan, Gerben A. Van Kleef, and Carsten K. W. De Dreu,
Department of Psychology, University of Amsterdam, Amsterdam, the
Netherlands; Daan van Knippenberg, Rotterdam School of Management,
Erasmus University Rotterdam, Rotterdam, the Netherlands.
This research was partially funded by a Sustainability–Diversity (SUS-
DIV) grant from the European Union awarded to Carsten K. W. De Dreu.
We thank Katherine Klein for helpful comments on a draft of this article.
Correspondence concerning this article should be addressed to Astrid C.
Homan, who is now at the Institute for Psychological Research, Social and
Organizational Psychology, Leiden University, P.O. Box 9555, 2300 RB
Leiden, the Netherlands. E-mail: ahoman@fsw.leidenuniv.nl
Journal of Applied Psychology Copyright 2007 by the American Psychological Association
2007, Vol. 92, No. 5, 1189 –1199 0021-9010/07/$12.00 DOI: 10.1037/0021-9010.92.5.1189
1189
As an example of this process of elaboration in diverse groups,
consider cross-functional surgical teams consisting of surgeons,
radiologists, anaesthetists, and surgical nurses. High-quality per-
formance requires that team members use their own expertise to
inform the other team members about the different issues involved
in the specific operation (e.g., how long the operation will take,
what kind of surgical instruments are needed, where the fracture
is); carefully process the information, opinions, and perspectives
introduced by other team members to understand the implications
for their own area of medical expertise; feed these implications
back to the team; and integrate these implications to provide the
best possible care for the patient.
Although diverse perspectives within a team can lead to en-
hanced team functioning through information elaboration, this
effect may be reduced or even reversed when informational diver-
sity converges with other diversity dimensions such as gender,
personality differences, or attitudes and values. When different
dimensions of diversity converge, the covariation of differences
creates a diversity faultline that may elicit subgroup categoriza-
tion—an “us–them” distinction (Lau & Murnighan, 1998; van
Knippenberg, De Dreu, & Homan, 2004; cf. Turner, Hogg, Oakes,
Reicher, & Wetherell, 1987). Such subgroup categorizations can
disrupt group processes by rendering group members less trusting
of and motivated to cooperate with other group members and less
committed to the group, increasing interpersonal tensions and
conflict, and lowering communication (e.g., Earley & Mosa-
kowski, 2000; Lau & Murnighan, 2005; Li & Hambrick, 2005; for
a review, see van Knippenberg & Schippers, 2007). Thus, although
diversity may stimulate group performance through information
elaboration, it also may undermine group performance through
social categorization processes (Williams & O’Reilly, 1998); the
latter is more likely to occur when dimensions of diversity con-
verge to create diversity faultlines. However, and in contrast to this
commonly accepted view, we propose that the performance of
groups with diversity faultlines need not necessarily be impeded.
Rather, we argue that the performance of such groups depends on
group members’ beliefs about the value of diversity.
Diversity Beliefs
Several authors have noted that people may differ in their beliefs
about and attitudes toward diversity (Hostager & De Meuse, 2002;
Strauss, Connerley, & Ammermann, 2003; van Knippenberg &
Haslam, 2003), and that organizational climates and cultures may
differ in the extent to which they value diversity (Cox, 2003; Ely
& Thomas, 2001; Jackson & Associates, 1992; Kossek & Zonia,
1993; Mor Barak, Cherin, & Berkman, 1998). These studies have
advanced the theoretical notion that beliefs, attitudes, climates, or
cultures valuing diversity are needed to harvest the benefits of
diversity, and have focused on the measurement and the determi-
nants of diversity beliefs, attitudes, and climates. So far, however,
a quantitative test of the influence of diversity beliefs and related
constructs on group processes and performance in diverse teams
has not been conducted. The present study provided such a test.
Diversity beliefs can be defined as beliefs about the value of
diversity to work group functioning (van Knippenberg & Haslam,
2003). Contingent on such beliefs, diversity may affect the extent
to which one’s own work group is perceived as being a good
group, where good may refer to (expectations of) task performance
as well as to other aspects of group functioning. Diversity beliefs,
thus, may inform responses to work group diversity and lead
people to respond more favorably to work group diversity the more
they believe in the value of diversity for work group functioning
(van Knippenberg, Haslam, & Platow, 2004). In support of this
proposition, van Knippenberg, Haslam, and Platow (2004) showed
in a survey and a laboratory experiment that the relationship
between diversity and group members’ identification with their
work group was moderated by diversity beliefs. When individuals
believed that diversity was beneficial for the task at hand, diversity
was positively related to group identification, whereas diversity
tended to be negatively related to identification when individuals
believed in the value of similarity. In a similar vein, in a qualitative
study, Ely and Thomas (2001) observed that when an organiza-
tion’s perspective on diversity emphasized cultural diversity as a
valuable resource for the organization, members reported feeling
more valued and respected, reported a higher quality of intergroup
relations, and felt that they were more successful than when the
organization’s perspective was not focused on the potential value
of diversity.
Pro-diversity beliefs may thus remove an important barrier for
diverse groups to benefit from their informational diversity. Van
Knippenberg, De Dreu, and Homan (2004) proposed that inter-
group biases engendered by subgroup categorization disrupt the
elaboration of task-relevant information in diverse groups and,
thus, stand in the way of groups’ effective use of their informa-
tional resources. They further argued, however, that salient sub-
group categorizations (i.e., in a team with diversity faultlines) need
not necessarily elicit intergroup bias. Pro-diversity beliefs (as
compared with pro-similarity beliefs) may lead group members to
respond favorably to the group and its diverse membership. Pro-
diversity beliefs may thus increase the likelihood that groups
benefit from their diversity by inviting group members to actively
solicit new information and perspectives from fellow group mem-
bers and, thereby, stimulate performance.
The Present Study
Previous work has suggested that when diversity is seen as
valuable to group functioning, group members may respond more
positively to diversity. In the present study, we extended this line
of work by providing a quantitative test of the effects of diversity
beliefs on groups’ use of their informational diversity. Because the
potential positive effects of informational differences are more
likely to be impeded when dimensions of diversity combine to
form a faultline (Lau & Murnighan, 1998), we focused on groups
with a diversity faultline. It has been found that in teams with a
strong faultline, the negative effects of diversity in general out-
weigh the positive effects (e.g., Gibson & Vermeulen, 2003; Lau
& Murnighan, 1998; Thatcher et al., 2003). We thus tested the
impact of diversity beliefs in a context in which disruptive effects
might be expected in the absence of pro-diversity beliefs.
The positive effects of work group diversity on group perfor-
mance are likely to emerge primarily in groups performing rela-
tively complex tasks that require information processing, creativ-
ity, and collaborative decision making, where the exchange of
diverse task-related information and perspectives may stimulate
groups’ thorough consideration of the task at hand (Bowers et al.,
2000; Jehn et al., 1999; van Knippenberg, De Dreu, & Homan,
1190
HOMAN, VAN KNIPPENBERG, VAN KLEEF, AND DE DREU
2004). We, therefore, tested the interactive effects of diversity
beliefs and informational diversity in a group decision-making
context. We hypothesized that, even under faultline conditions,
groups may make good use of informational diversity when they
hold beliefs favoring group diversity rather than homogeneity.
Because especially informationally diverse groups need to elabo-
rate task-relevant information and perspectives to perform well
(i.e., they need to exchange and integrate diverse information and
perspectives that are already shared in informationally homoge-
neous groups), we expected performance to be affected more by
diversity beliefs in informationally diverse groups than in infor-
mationally homogeneous groups. On the basis of this reasoning,
we advanced the following three hypotheses:
Hypothesis 1: Diversity beliefs moderate the effect of infor-
mational diversity on group performance. Informationally
diverse groups perform better when group members believe
in the value of diversity rather than similarity, whereas the
performance of informationally homogeneous groups is less
affected by diversity beliefs.
Hypothesis 2: Diversity beliefs moderate the effect of infor-
mational diversity on elaboration of task-relevant informa-
tion. Informationally diverse groups engage in more elabora-
tion of task-relevant information when group members
believe in the value of diversity rather than similarity,
whereas elaboration in informationally homogeneous groups
is less affected by diversity beliefs.
Hypothesis 3: The effect of diversity beliefs on performance
in informationally diverse groups is mediated by elaboration
of task-relevant information.
These hypotheses were tested in an experimental study of four-
person groups that worked interactively on a complex decision-
making task in which they had to generate decision alternatives
and decide about the alternatives adopted. We adopted an exper-
imental approach for two reasons. First, testing our hypotheses in
an experimental setting allowed us to draw conclusions about
causality. Second, this approach allowed us to directly assess
relevant group processes through behavioral coding of audio and
video recordings of group interaction, rather than having to rely on
the retrospective self-report data that are more customary in field
research. This method thus yielded a more direct and objective
assessment of the group processes leading to group performance
(Weingart, 1997).
Method
Sample and Design
A total of 184 students (92 women and 92 men) of the Univer-
sity of Amsterdam participated in the experiment for course credit
or monetary compensation (10 euro, approximately $12 U.S.). The
mean age of the participants was 21 years. The participants were
randomly assigned to gender-diverse 4-person groups (always
consisting of 2 men and 2 women), and these groups were then
randomly assigned to one of the conditions of a 2 (informational
diversity: heterogeneity vs. homogeneity) 2 (diversity beliefs:
pro-similarity vs. pro-diversity) factorial design. A total of 46
groups participated in the experiment. One group could not be
videotaped because of technical problems.
The Task
The task required groups to generate and select ideas. The task
was inspired by a decision-making task developed by Johnson and
Johnson (1982). Groups were instructed to come up with as many
useful items as possible needed to survive in a desert on the basis
of information provided to them before the task. The only rules
were that (a) the items should be portable and (b) participants had
to explain why the selected items were important for surviving in
a desert. The groups thus had to work on a list of high-quality
options by generating options, discussing them, and determining
which options were good enough to put on the list. A pretest
showed that this task was not gender related.
1
Procedure
Creating a faultline. On arrival, participants were seated in a
room in same-gender pairs. At this time, they could not see the
other two group members. Participants individually read instruc-
tions stating that the study aimed to determine the effect of
personality on cooperation. Then, the participants were asked to
fill out a personality test. After filling out the questionnaire, their
answers were supposedly analyzed, and their personality type
determined. After about 10 min, the experimenter returned with
the results. The bogus feedback that male participants received
stated that they had an “M” personality type. Female participants
received bogus feedback stating that they had an “F” personality
type. They then received some superficial information about their
personality type. The description of the personality type consisted
of eight gender-specific traits (e.g., strong, adventurous for men;
emotional, considerate for women; Willemsen & Fischer, 1999).
Next, the participants read that there was a chance that they would
be working in a group with people with a different personality
type, and they were given three traits of people with a different
personality type (e.g., women read that people with an M person-
ality type were adventurous).
Finally, seating was used to make the faultline more salient.
After participants had read the information about the task, they
were seated in a new room in which the group would perform the
task. Same-gender group members were always seated next to each
other at a rectangular table, facing the opposite-gender members.
Such converging of diversity dimensions (i.e., gender, [bogus]
personality feedback, seating) results in high within-subgroup sim-
ilarity and high between-subgroups differences, which makes sub-
group categorization more likely (Gaertner, Mann, Murrell, &
1
A pretest with 23 students who did not participate in the main study
showed that there was no difference between men and women on task
performance, suggesting that the task was not gender related, F(1, 22)
0.29, ns,
2
.01. Also, we assessed whether the participants themselves
perceived performance on the task to be gender related (i.e., more “male”
or more “female”). A t test showed that perceptions (M 4.22, SD 0.90)
did not differ from the midpoint of the scale, t(22) 1.16, ns, indicating
that participants did not perceive the task to be gender related. Again, we
found no difference in responses between men and women, F(1, 22)
0.91, ns,
2
.10.
1191
DIVERSITY BELIEFS AND PERFORMANCE
Dovidio, 1989; Turner et al., 1987; van Knippenberg, De Dreu, &
Homan, 2004). We thus created a perfect faultline in the sense that
differences in gender, (bogus) personality feedback, and seating
arrangement were perfectly correlated (cf. Lau & Murnighan,
1998; Thatcher et al., 2003). Previous research has shown that this
specific convergence of diversity attributes elicits a strong faultline
and evokes disadvantageous group processes (Homan et al., in
press).
Manipulation of informational diversity. Then, participants re-
ceived the instructions for the decision-making task and some
information about surviving in the desert. This information was
used to manipulate informational diversity. On the basis of expert
information (Johnson & Johnson, 1982), 12 different categories of
information concerning surviving in a desert were distinguished
(e.g., make sure you do not dry out, create shade, batteries overheat
in the heat). In the information package, these 12 different cate-
gories were mentioned in such a way that thoughtful use of the
information could help in determining which items would be
useful for surviving in a desert. We used standard procedures to
manipulate the homogeneous versus heterogeneous information
within the groups (see, e.g., Gruenfeld, Mannix, Williams, &
Neale, 1996; Stasser & Titus, 1985). In the informationally homo-
geneous condition, all group members individually received the 12
different categories of information (i.e., all the available informa-
tion was shared among all the group members). In the informa-
tionally diverse conditions, 8 of these categories of information
were divided into two equally informative parts (Part A and Part
B).
2
Two group members received Part A, and two group members
received Part B. Same-gender group members always received the
same information (further enhancing the faultline). The 4 remain-
ing information items were given to all group members. The
individual group members in the informationally diverse condi-
tions thus received 8 information items. Four of these items were
given to all four group members, 4 information items were given
only to the female group members, and 4 information items were
given only to the male group members. At the group level, the
group thus received 12 items of information (i.e., 4 shared items
4 items for the male members 4 items for the female members).
In other words, groups in all conditions received the same set of
information; only the distribution of information across group
members differed between conditions.
Manipulation of diversity beliefs. After the personality test and
before the decision-making task, participants received some addi-
tional information about working in teams. Through this informa-
tion, we manipulated diversity beliefs. Groups in the pro-diversity
beliefs condition read that research had shown that gender-diverse
groups typically perform better on decision-making tasks and
experience more pleasant group processes than gender-
homogeneous groups. Conversely, groups in the pro-similarity
beliefs condition read that research had shown that gender-
homogeneous groups typically perform better on decision-making
tasks and experience more pleasant group processes than gender-
diverse groups. Following this information, the participants re-
ceived a short introduction to the task and some feedback about
their personality type. Finally, participants were given a question-
naire to check the manipulation of diversity beliefs. After reading
all the information, the participants were brought to another room
and were seated together. They were then given 30 min to work on
the decision-making task. Groups were videotaped during interac-
tion. On completion of the task, the experimenter administered a
questionnaire to check the manipulation of informational diversity.
Then, participants were debriefed and thanked.
Dependent Variables
Manipulation checks. We checked the manipulation of infor-
mational diversity with four items (e.g., “During the group task,
the group members regularly said things I did not know”). Prin-
cipal components analysis (PCA) revealed that these four items all
loaded on one factor (factor loadings between .80 and .84). Reli-
ability analysis showed that the four questions formed a reliable
scale (␣⫽.86; M 3.48, SD 1.19). We checked the manipu-
lation of diversity beliefs with four items (e.g., “Groups that are
diverse on gender usually perform better than groups that are
homogeneous on gender”). PCA revealed that these four items all
loaded on one factor (factor loadings between .82 and .84). The
four items formed a reliable scale (␣⫽.86; M 4.22, SD 1.42).
All responses were given on Likert-type scales ranging from 1
(totally disagree)to7(totally agree).
Performance. Performance was determined by calculating the
mean score per item for each group. In the original task, 12
categories of items were distinguished. On the basis of these
categories, we developed a coding scheme by which performance
could be calculated. Better items (i.e., items that were ranked
higher in the expert ranking reported by Johnson & Johnson, 1982)
received a higher score, with the highest possible score being 12
points. For example, when the group decided that it would bring a
magnetic compass, it received a score of 1 because the information
clearly indicated that groups should not walk. Two independent
raters coded the items generated by the groups, providing over-
lapping ratings of 21% of the groups to determine interrater
agreement. We assessed interrater reliability by computing intra-
class correlations (ICCs; Shrout & Fleiss, 1979). The average ICC
for the two raters was .91, which is considered excellent according
to the criteria developed by Cicchetti and Sparrow (1981). Finally,
because instructions emphasized quality rather than quantity, we
divided the score by the number of items generated to ensure that
groups that came up with a lot of low-quality items would not get
higher scores than groups that came up with fewer high-quality
2
In a pretest (n 22), we compared three different distributions of
information (only Part A vs. only Part B vs. Parts A and B) using the
original surviving in the desert exercise (Johnson & Johnson, 1982) and
found a significant effect on performance, F(2, 19) 6.73, p .01,
2
.42 (lower scores reflect better performance). Post hoc tests revealed that
students who received all the information (i.e., Parts A and B) did signif-
icantly better (M 25.00, SD 6.68) than did those who read only Part
A(M 35.50, SD 8.60) or Part B (M 39.67, SD 8.24), with the
latter two conditions not significantly differing from each other, F(1, 12)
0.83, ns,
2
.07. These findings suggest that (a) Parts A and B were
equally useful and informative when it came to listing items needed for
survival in the desert, and (b) people who were aware of the information
in both parts performed better than did those who knew only Part A or
Part B.
1192
HOMAN, VAN KNIPPENBERG, VAN KLEEF, AND DE DREU
items. Thus, performance scores reflect quality of performance,
not quantity.
3
Information elaboration. Group information elaboration in-
volves the degree to which information is shared, processed, and
integrated in group interaction (van Knippenberg, De Dreu, &
Homan, 2004; cf. Hinsz et al., 1997). Elaboration of information
was measured by coding the videotapes of 45 groups (1 group had
to be omitted from the analyses because of technical problems).
The difference between the heterogeneous and homogeneous in-
formation conditions lies in the distribution of 8 of the 12 infor-
mation items that were divided in Part A (4 items) and Part B (4
items). In the heterogeneous information condition, these subsets
were given to the two female and to the two male group members,
respectively. Differences in information elaboration between con-
ditions should, therefore, be evident primarily for these 8 pieces of
information. Accordingly, our measure of information elaboration
focused on these 8 information items.
The coding scheme was constructed as follows. The higher the
score, the more an information item was elaborated on. A score of
0 was given when an information item was not mentioned at all
during the discussion. A score of 1 was given when information
was mentioned, but none of the other members reacted to it (i.e.,
if the information was only exchanged). A score of 2 was given
when one of the members mentioned an item of information and at
least one of the other members reacted to it (e.g., by saying
something like “OK” or by nodding), but after this the group still
failed to ask questions about it or integrate it with the other
information. A score of 3 was given when a piece of information
was mentioned by one of the group members, and one or more
other members clearly responded by asking a question about it
(e.g., “Why is it important to give light signals?”). A score of 4
was given when the mentioning of an information item resulted in
a conclusion about whether something was important or not (e.g.,
“Ah, a mirror must be important, you can use the light of the sun
to signal with that”). Finally, a score of 5 was given when the
information item was combined with another piece of information
by one of the other group members (e.g., “Wait a minute, we need
protection from the sun as well, right? Why don’t we take an
aluminum tent with us? That will create shade and will reflect light
as well”). We used the highest level of information elaboration for
each information item (from 0 to 5); the total elaboration was then
determined by computing the sum of information elaboration for
the eight information categories. It is important to stress that
information elaboration was coded in the same way in all condi-
tions (i.e., regardless of whether information items were shared or
unshared). Thus, groups in all conditions could obtain scores
between 0 and 5 for all eight information items. The maximum
number of points that could be obtained thus was eight items 5
points for 40 points. Two independent raters, blind to the experi-
mental conditions and hypotheses, coded the videotapes. They
provided double ratings of 20% of the videotapes to check inter-
rater reliability. The average ICC for the two raters was .96, which
is considered excellent (Cicchetti & Sparrow, 1981).
Results
Treatment of the Data
We used analysis of variance to test Hypotheses 1 and 2 and
regression analysis to test Hypothesis 3. Manipulation checks were
measured at the individual level, but because the individuals were
working in four-person groups, their answers are probably not
independent (Kashy & Kenny, 2000). Therefore, we aggregated
individuals’ answers to the group level. To control whether this
aggregation was appropriate, we computed ICC(1), ICC(2), and
r
wg
(e.g., Bliese, 2000). First, to test whether the groups could be
reliably differentiated on the manipulation checks, we estimated
ICC(2) values. Following Glick’s (1985) recommendations,
ICC(2) values were acceptable (.79 for the informational diversity
check; .68 for the diversity beliefs check). To further support
aggregation to the group level, we calculated ICC(1) and r
wg
values. For both the manipulation check of informational diversity,
ICC(1) .49, F(45, 138) 4.81, p .01,
2
.61, and r
wg
.71, and manipulation check of diversity beliefs, ICC(1) .35,
F(45, 138) 3.14, p .01,
2
.51, and r
wg
.70, the obtained
values justified aggregation to the group level (George, 1990).
Manipulation Checks
Informational diversity. Groups in the homogeneous informa-
tion condition (M 2.63, SD .67) indicated that the information
that the group members received was less diverse than did groups
in the heterogeneous information condition (M 4.33, SD .98),
F(1, 42) 64.40, p .01,
2
.59. The manipulation check for
informational diversity was not influenced by diversity beliefs,
F(1, 42) 0.55, ns,
2
.01, nor by the interaction between
informational diversity and diversity beliefs, F(1, 42) 2.26, ns,
2
.05.
Diversity beliefs. Groups with pro-diversity beliefs (M 4.85,
SD .84) indicated that diverse teams would perform and coop-
erate better than groups with pro-similarity beliefs (M 3.64,
SD .74), F(1, 42) 25.87, p .01,
2
.38. The manipulation
check for diversity beliefs was not influenced by informational
diversity, F(1, 42) 0.02, ns,
2
.00, or the interaction between
informational diversity and diversity beliefs, F(1, 42) 0.40, ns,
2
.01.
Performance
There was no main effect of informational diversity on perfor-
mance: Groups with diverse information (M 6.67, SD 0.98)
showed similar levels of performance as groups with homoge-
neous information (M 6.62, SD 0.97), F(1, 42) 0.28, ns,
2
.00. We did find a significant main effect of diversity beliefs,
indicating that groups with pro-diversity beliefs performed better
(M 6.95, SD 0.96) than did groups with pro-similarity beliefs
(M 6.36, SD 0.89), F(1, 42) 5.27, p .05,
2
.11.
Moreover, in support of Hypothesis 1, the interaction between
diversity beliefs and informational diversity was significant, F(1,
42) 4.25, p .05,
2
.09. Means and standard deviations
pertaining to this interaction are shown in Table 1. Simple effects
analysis showed that groups with diverse information and pro-
diversity beliefs outperformed groups with diverse information
and pro-similarity beliefs, F(1, 42) 4.27, p .05,
2
.18.
Groups with homogeneous information were not influenced by
3
Analysis of variance revealed no effects of the experimental manipu
-
lations on the number of items generated, F(1, 42) 0.41, ns,
2
.02.
1193
DIVERSITY BELIEFS AND PERFORMANCE
diversity beliefs, F(1, 42) 0.03, ns,
2
.00. This interaction is
depicted in Figure 1.
Information Elaboration
We obtained a significant main effect of informational diversity
on information elaboration, showing that groups with diverse
information (M 28.29, SD 6.39) elaborated more information
than did groups with homogeneous information (M 24.13, SD
3.91), F(1, 41) 7.13, p .05,
2
.14. We also found a
significant main effect of diversity beliefs, revealing that groups
with pro-diversity beliefs (M 27.59, SD 6.21) elaborated more
information than did groups with pro-similarity beliefs (M
24.61, SD 4.54), F(1, 41) 4.39, p .05,
2
.08. More
important, these main effects were qualified by a significant Di-
versity Beliefs Informational Diversity interaction, F(1, 41)
6.61, p .05,
2
.33 (see Table 1 for means and standard
deviations). As predicted in Hypothesis 2, groups with diverse
information and pro-diversity beliefs elaborated more information
than did groups with diverse information and pro-similarity be-
liefs, F(1, 41) 10.96, p .01,
2
.21. Groups with homoge
-
neous information were not influenced by diversity beliefs, F(1,
41) 0.22, ns,
2
.01. This interaction is shown in Figure 2
.
4
Mediation Analysis
Hypothesis 3 predicted that diversity beliefs would moderate the
effect of informational diversity on performance through their
impact on information elaboration in informationally diverse
groups. To test this proposed pattern of mediation, we followed
procedures suggested by Baron and Kenny (1986) and extended by
Hull, Tedlie, and Lehn (1992). According to Baron and Kenny
(1986), four requirements should be met to establish mediation.
First, there should be a significant effect of the independent vari-
able(s) on the dependent variable. Second, there should be an
effect of the independent variable(s) on the mediator. Third, the
mediator should predict the dependent variable. Finally, the effect
of the independent variable(s) should be reduced to nonsignifi-
cance when controlling for the mediator.
More recently, Hull et al. (1992; also see Muller, Judd, &
Yzerbyt, 2005; Yzerbyt, Muller, & Judd, 2004) proposed an im-
portant extension of these procedures. They suggested that in
mediation analysis, it also may be relevant to control for the
possibility that the proposed mediator is not linearly related to the
dependent variable but instead is more strongly related to the
dependent variable under certain conditions than under others.
When this is the case, entering the mediator as a linear covariate
violates the statistical assumption of homogeneity of regression
slopes (i.e., the assumption that the slopes of the regression lines
are the same in each group). Inclusion of the “covariate interac-
tion” (i.e., the interaction between an independent variable and the
4
Analysis of variance of elaboration of the four items that were shared
prior to group interaction in all conditions revealed no main effect of
informational diversity, F(1, 41) 3.82, ns,
2
.09; no main effect of
diversity beliefs, F(1, 41) 0.06, ns,
2
.001; and no interaction, F(1,
41) 1.10, ns,
2
.01. Consistent with the results for information
elaboration based on the subset of information that was unshared in the
heterogeneous information conditions, analysis of variance involving the
full set of information revealed (a) no effect of the manipulation of
diversity beliefs (pro-diversity beliefs: M 41.95, SD 9.52, vs. pro-
similarity beliefs: M 39.22, SD 6.79), F(1, 41) 1.48, ns,
2
.04;
(b) a main effect of the manipulation of informational diversity (informa-
tionally heterogeneous: M 44.29, SD 8.05, vs. informationally homo-
geneous: M 37.29, SD 7.08), F(1, 41) 10.34, p .01,
2
.20; and
(c) an interaction between informational diversity and diversity beliefs,
F(1, 41) 5.04, p .05,
2
.12. Simple effects analysis showed that
groups with diverse information and pro-diversity beliefs (M 47.91,
SD 6.84) elaborated more information than did groups with diverse
information and pro-similarity beliefs (M 40.30, SD 7.63), F(1, 41)
5.89, p .05,
2
.23. Groups with homogeneous information were not
influenced by diversity beliefs (pro-diversity beliefs: M 36.00, SD
8.01, vs. pro-similarity beliefs: M 38.39, SD 6.21), F(1, 41) 0.66,
ns,
2
.03. Note that in this analysis the mean levels of information
elaboration are higher than the means reported in the article because we
calculated information elaboration for all 12 items.
5.4
5.6
5.8
6
6.2
6.4
6.6
6.8
7
7.2
7.4
Homogeneous
information
Heterogeneous
information
Performance
Pro-similarity belief
Pro-diversity belief
Figure 1. Group performance as a function of informational diversity and
diversity beliefs.
Table 1
Effects of Diversity Beliefs and Informational Diversity on Performance and Information Elaboration
Measure
Informationally heterogeneous Informationally homogeneous
Pro-diversity belief Pro-similarity belief Pro-diversity belief Pro-similarity belief
MSDMSDMSDMSD
Performance 7.20
a
0.82 6.04
b
0.74 6.70
ab
1.07 6.64
ab
0.94
Information elaboration 31.55
a
5.82 24.70
b
5.08 23.64
b
3.56 24.54
b
4.29
Note. Means within a row with a different subscript differ at p .05. Performance represents the mean number of points obtained per item and ranges
from 0 ( poor)to12(excellent). Information elaboration ranges from 0 (no elaboration)to40(high elaboration).
1194
HOMAN, VAN KNIPPENBERG, VAN KLEEF, AND DE DREU
proposed mediator; Hull et al., 1992) then yields a more appropri-
ate test of mediation (i.e., a test that does not assume homogeneity
of regression slopes across conditions; cf. Stevens, 1996) than an
analysis that includes only the “main effect” of the proposed
mediator (Hull et al., 1992; Muller et al., 2005; Yzerbyt et al.,
2004). Our theoretical analysis points to the possibility that infor-
mation elaboration is more positively related to performance under
conditions of informational diversity (i.e., where groups need to
exchange and integrate information to reach optimal decisions)
than under conditions of informational homogeneity (i.e., where
groups in principal can rely more on the pooling of prediscussion
preferences, and performance, therefore, may be less contingent on
information elaboration). We, therefore, controlled for this possi-
bility by including the covariate interaction between informational
diversity and elaboration in our mediational analysis.
In sum, to test our mediation model, we followed the four steps
described by Baron and Kenny (1986) but also included the
covariate interaction between information elaboration and infor-
mational diversity in the analysis, as suggested by Hull et al.
(1992). We have already established that informational diversity
and diversity beliefs interact to affect performance (Step 1; see
analysis under Performance), and that pro-diversity beliefs in-
spired greater elaboration in informationally diverse groups than
pro-similarity beliefs (Step 2; see analysis under Information Elab-
oration). Next, we aimed to establish that the proposed mediator,
information elaboration, predicted performance (Step 3). To do so,
following Hull et al., we regressed performance on information
elaboration (centered) as well as on informational diversity
(dummy coded) and the interaction between information elabora-
tion and informational diversity. The statistics pertaining to this
analysis are summarized in Table 2. This analysis revealed no
main effects of information elaboration and informational diver-
sity, but it did reveal a significant interaction. Simple slopes
analysis showed that informationally diverse groups performed
better when they elaborated more information, ␤⫽.47, t(41)
2.60, p .05, whereas performance of informationally homoge-
neous groups was not affected by information elaboration, ␤⫽
–.50, t(41) –1.82, ns. This analysis thus confirms that more
information elaboration may be associated with better perfor-
mance, but that this is only the case in informationally diverse
groups. These results point to the need to include the covariate
interaction between information elaboration and informational di-
versity in the final step of the mediation analysis (Hull et al., 1992;
Muller et al., 2005; Stevens, 1996; Yzerbyt et al., 2004).
For the final step of the mediation analysis, we regressed per-
formance on diversity beliefs and informational diversity (both
dummy coded) and their interaction, as well as on information
elaboration and on the interaction between information elaboration
and informational diversity. This analysis yielded a significant
Information Elaboration Informational Diversity interaction,
and the originally significant interaction between diversity beliefs
and informational diversity was reduced to nonsignificance (see
Table 2). In line with Hypothesis 3, this pattern of results indicates
that the effect of diversity beliefs on performance in information-
ally diverse groups is mediated by information elaboration.
Discussion
Diverse information and perspectives in work groups can po-
tentially boost group performance, but diverse groups are often
unable to benefit from their diversity. Addressing this issue, we
proposed that groups are more likely to effectively use their
informational resources when group members believe in the value
of diversity. We put this proposition to the test under conditions in
which groups were characterized by diversity faultlines (Lau &
Murnighan, 1998)—a situation widely assumed to stand in the way
of groups’ effective use of information (see van Knippenberg &
Schippers, 2007). In support of our proposition, diversity beliefs
moderated the relationship between informational diversity and
performance, such that informationally diverse (but not informa-
15
17
19
21
23
25
27
29
31
33
Homogeneous
information
Heterogeneous
information
Information elaboration
Pro-similarity belief
Pro-diversity belief
Figure 2. Information elaboration as a function of informational diversity
and diversity beliefs.
Table 2
Summary of Hierarchical Regression Results of Mediational Analysis
Predictor BSE tR
2
Performance Step 3 .20
*
Informational diversity 0.06 0.29 .30 0.19
Information elaboration 0.09 0.05 .50 1.82
Informational Diversity Information Elaboration 0.17 0.06 .77 2.93
*
Performance Step 4 .27
*
Diversity beliefs 0.46 0.39 .24 1.18
Informational diversity 0.02 0.37 .01 0.05
Diversity Beliefs Informational Diversity 0.91 0.65 .39 1.40
Information elaboration 0.09 0.05 .49 1.83
Informational Diversity Information Elaboration 0.13 0.07 .58 2.02
*
Note. The degree of freedom (df) for the t tests listed for Performance Step 3 is 41; the df for the t tests listed for Performance Step 4 is 39.
*
p .05.
1195
DIVERSITY BELIEFS AND PERFORMANCE
tionally homogeneous) groups performed better when they held
pro-diversity beliefs rather than pro-similarity beliefs.
Although a number of scholars have argued that diversity beliefs
and related constructs play an important role in teams (e.g., Ely &
Thomas, 2001; Kossek & Zonia, 1993; van Knippenberg &
Haslam, 2003), a quantitative test of the moderating effect of
diversity beliefs on the relation between diversity and group pro-
cesses and performance was lacking. The present study thus pro-
vides an important next step in research on diversity beliefs,
attitudes, perspectives, and climates. More generally, the present
findings may be viewed as a contribution to attempts to identify
the contingencies of the effects of work group diversity. Diversity
research has not been overly successful in mapping the effects of
work group diversity, and several authors have attributed this to
the main effects approach that has characterized a lot of diversity
research (Pelled et al., 1999; van Knippenberg, De Dreu, & Ho-
man, 2004; van Knippenberg & Schippers, 2007). The present
focus on diversity beliefs as moderator of the effects of work group
diversity, thus, also may be seen as testifying to the value of a
focus on the contingencies of the effects of diversity.
In addition, we were able to establish that the effect of diversity
beliefs in informationally diverse groups was mediated by group
elaboration of task-relevant information. Van Knippenberg, De
Dreu, and Homan (2004) proposed that information elaboration is
the core process underlying the positive effects of diversity on
group performance; thus, the present finding that informationally
diverse groups were dependent on elaboration to perform well may
be interpreted as important first evidence for their theoretical
analysis. Moreover, the fact that pro-diversity beliefs engendered
elaboration in informationally diverse groups corroborates our
proposition that pro-diversity beliefs invite group members to
actively capitalize on their group’s diversity.
Clearly, the difference between the heterogeneous information
and homogeneous information condition lies in the subset of
information that was unshared prior to group interaction in the
heterogeneous information condition but given to all group mem-
bers in the homogeneous information condition. Differences in
information elaboration between conditions should, therefore, ma-
terialize primarily for this subset of information, and, accordingly,
our measure of elaboration focused on this subset of information.
Arguably, however, we should test whether differences in infor-
mation elaboration are limited to this subset rather than assume
this to be the case. We, therefore, also analyzed elaboration of the
four pieces of information that were given to all group members in
all conditions and elaboration of the total set of information.
Corroborating our analysis, results (reported in Footnote 4)
showed that elaboration of the subset of information that was given
to all group members in both conditions was unaffected by the
experimental manipulations, whereas findings for elaboration of
the total set of information yielded similar conclusions as the
analysis that concentrated on the information items that were not
shared in the heterogeneous condition.
It is important to emphasize that we should not conclude too
much from the main effect of informational diversity on informa-
tion elaboration (i.e., overall, groups with heterogeneous informa-
tion elaborated more). First, by the very nature of the task, there is
a greater need to engage in information elaboration in informa-
tionally heterogeneous groups than in informationally homoge-
neous groups. Individual members of informationally diverse
teams do not possess all the relevant information prior to group
interaction; therefore, they need to exchange information with the
other group members and elaborate on this information to get a
thorough understanding of the task. Members of informationally
homogeneous groups, in contrast, have a lesser need to exchange
information because all the members of the group possess all the
relevant information before group interaction. This is an inherent
difference between informationally homogeneous and informa-
tionally heterogeneous groups. Second, the pattern of means for
the interaction shows that the main effect of informational diver-
sity on information elaboration is fully qualified by the interaction
with diversity beliefs: It can be attributed solely to groups with
heterogeneous information and pro-diversity beliefs. Finally, the
finding that performance is more contingent on diversity beliefs in
the heterogeneous information conditions than in the homoge-
neous information conditions indicates that it is the interaction
between informational diversity and diversity beliefs, rather than
the main effect of informational diversity, that predicts perfor-
mance.
Because faultlines were a constant in our study, we should be
careful not to conclude that the present findings pertain to the
effects of faultlines. Rather, our results concern group processes
and performance under faultline conditions. Given what we know
about the effects of faultlines (e.g., Lau & Murnighan, 2005; Li &
Hambrick, 2005), the current focus on performance under faultline
conditions provides a test of the effects of diversity beliefs in
conditions under which diversity often has been found to disrupt
group processes. If pro-diversity beliefs can stimulate groups to
use their informational diversity under faultline conditions, we
might expect that pro-diversity beliefs also are able to boost
diverse groups’ performance under conditions that are less condu-
cive to the disruptive effects of work group diversity.
In a related vein, the fact that diversity beliefs affected group
performance under faultline conditions corroborates another point
raised by van Knippenberg, De Dreu, and Homan (2004). They
argued that salient subgroup categorizations in diverse groups as
such are not problematic. The problem is intergroup bias that may,
but need not, be engendered by subgroup categorization. Faultlines
are generally assumed to render subgroup categorization salient,
and the present findings, thus, might be interpreted as showing that
salient subgroups need not be detrimental to group performance,
and that diversity beliefs moderate the extent to which salient
subgroups elicit detrimental group processes.
Although experiments are not conducted in a quest for external
validity (Dipboye, 1990; Mook, 1983), reports of experimental
research tend to elicit questions of external validity among their
readership. Obviously, then, confidence in the conclusions ad-
vanced here could be bolstered when the current results are repli-
cated in a study of teams in actual organizations, and this would
indeed seem an important avenue for future research. In this
respect, it is noteworthy that a previous study of work group
diversity found similar effects of diversity beliefs on identification
in both a laboratory experiment and a field study (van Knippen-
berg, Haslam, & Platow, 2004). Likewise, Ely and Thomas’ (2001)
qualitative analysis of the role of diversity perspectives in organi-
zations suggests that the effects of diversity beliefs observed in the
present study also occur in the field. As always, however, the proof
of the pudding is in the eating, and it would be valuable if future
1196
HOMAN, VAN KNIPPENBERG, VAN KLEEF, AND DE DREU
research would focus on the effects of diversity beliefs on team
performance in the field.
An interesting possibility in this respect is that the influence of
diversity beliefs on diverse teams’ performance might actually be
greater in the field. In organizations, informationally diverse teams
typically have a larger pool of information and perspectives than
informationally homogeneous teams. Accordingly, when circum-
stances are conducive to the elaboration of task-relevant informa-
tion, informationally diverse teams should be able to outperform
informationally homogeneous teams (e.g., Jehn et al., 1999). This
means that pro-diversity beliefs might lead informationally diverse
teams to outperform informationally homogeneous teams because
the former are likely to have more information at their disposal
than the latter. This possibility could not be addressed in the
present study because providing informationally diverse groups
with more information than informationally homogeneous groups
would have confounded informational diversity with the amount of
information available to the group. Although this confounding is
likely to occur in organizations, it was undesirable for our purposes
because the aim of the present study was to show that diversity
beliefs (and not the amount of information available) moderate the
diversity–performance relationship. We expect that, in organiza-
tions, informationally diverse groups may outperform information-
ally homogeneous groups partly because they will often have more
information, and we expect this to happen especially when they
hold pro-diversity beliefs.
From an applied perspective, an important implication of the
present findings is that the effective management of a diverse
workforce should involve the management of diversity beliefs. For
example, managers may foster pro-diversity beliefs by communi-
cating their belief in the value of diversity and by explaining how
task performance can benefit from diversity of information and
perspectives. We should realize, however, that the present findings
were obtained in ad hoc groups of students with presumably no
strong beliefs about the implications of gender diversity for per-
formance in decision-making tasks. Even though earlier studies on
diversity beliefs and cultures in organizations have suggested that
diversity beliefs are malleable (Ely & Thomas, 2001; van Knip-
penberg, Haslam, & Platow, 2004), we should expect constraints
with regard to the extent and the ease with which organizational
members can be convinced of the value of particular dimensions of
diversity. For instance, diversity is more valuable for more com-
plex, knowledge-intensive tasks than for more simple, routine
tasks (van Knippenberg & Schippers, 2007). Arguably, then, it
should be more feasible for managers to explain the value of
diversity in more complex tasks than in more routine tasks. Ste-
reotypic beliefs may also stand in the way of pro-diversity beliefs
(i.e., it might be hard to convince a person with racist beliefs of the
value of ethnic diversity). Accordingly, although the present study
suggests that fostering pro-diversity beliefs may be an important
aspect of the successful management of diversity, more work
clearly needs to be done to develop our understanding of the
possibilities to foster pro-diversity beliefs in organizations.
In a related way, current results also have implications for
diversity training programs. Diversity training could, in principle,
exert an important influence on diversity beliefs. However, most
diversity training programs seem to be limited to making people
aware of their stereotypes about other groups and changing peo-
ple’s feelings and ideas about those groups (e.g., Karp & Sam-
mour, 2000; Kossek & Lobel, 1996; Rynes & Rosen, 1995). The
current findings suggest that it is also important to manage peo-
ple’s feelings about diversity itself (rather than about different
others) and to make them aware of the potential value of being a
member of a diverse team. It would, therefore, seem worthwhile to
extend diversity training programs beyond this focus on stereo-
types to include a focus on beliefs about and attitudes toward
diversity itself (cf. van Knippenberg, Haslam, & Platow, 2004).
Given the potential benefits of a workforce endorsing pro-
diversity beliefs, an important direction for future research would
seem to be to develop theory about the origins of diversity beliefs.
At least three partly related antecedents of diversity beliefs are
suggested by previous research. First, van Knippenberg, Haslam,
and Platow (2004) identified task requirements as a source of
diversity beliefs, showing that individuals working on a task that
required diverse perspectives developed more positive attitudes
toward diversity than did individuals who worked on a task that
required homogeneous perspectives (cf. the notion that diversity is
more beneficial on more knowledge-intensive tasks; e.g., van
Knippenberg, Haslam, & Platow, 2004). Second, prior experience
would seem to be a source of diversity beliefs. When people have
positive experiences with working in a diverse group, it is likely
that those experiences will shape their beliefs about diversity in the
future. Finally, Flynn (2005) and Strauss et al. (2003) noted that
individual difference variables may affect beliefs about different
ethnic groups and diversity in general. One variable that might be
of interest in this context is openness to experience (e.g., Flynn,
2005). In sum, then, it would seem appropriate to explore individ-
ual differences as well as situational influences both internal and
external to the work group as determinants of diversity beliefs.
Considering these and other potential precursors of diversity be-
liefs may help to lay the foundations for successful diversity
management in organizations.
References
Argote, L., Gruenfeld, D., & Naquin, C. (2000). Group learning in orga-
nizations. In M. E. Turner (Ed.), Groups at work: Theory and practice
(pp. 369–211). Mahwah, NJ: Erlbaum.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable
distinction in social psychological research: Conceptual, strategic, and
statistical considerations. Journal of Personality and Social Psychology,
51, 1173–1182.
Bliese, P. D. (2000). Within-group agreement, non-independence, and
reliability: Implications for data aggregation and analysis. In K. J. Klein
& S. W. J. Kozlowski (Eds.), Multilevel theory, research, and methods
in organizations: Foundations, extensions, and new directions (pp. 349
381). San Francisco: Jossey-Bass.
Bowers, C. A., Pharmer, J. A., & Salas, E. (2000). When member homo-
geneity is needed in work teams: A meta-analysis. Small Group Re-
search, 31, 305–327.
Cicchetti, D. V., & Sparrow, S. S. (1981). Developing criteria for estab-
lishing the interrater reliability of specific items in a given inventory:
Applications to assessment of adaptive behavior. American Journal of
Mental Deficiency, 86, 127–137.
Cox, T. H. (2003). Cultural diversity in organizations: Theory, research
and practice. San Francisco: Berrett-Koehler.
Cox, T. H., Lobel, S. A., & McLeod, P. L. (1991). Effects of ethnic group
cultural differences on cooperative and competitive behavior on a group
task. Academy of Management Journal, 34, 827–847.
De Dreu, C. K. W. (2007). Cooperative outcome interdependence, task
1197
DIVERSITY BELIEFS AND PERFORMANCE
reflexivity, and team effectiveness: A motivated information processing
perspective. Journal of Applied Psychology, 93, 628 638.
Dipboye, R. L. (1990). Laboratory vs. field research in industrial and
organizational psychology. International Review of Industrial and Or-
ganizational Psychology, 5, 1–34.
Earley, P. C., & Mosakowski, E. (2000). Creating hybrid team cultures: An
empirical test of transnational team functioning. Academy of Manage-
ment Journal, 43, 26 49.
Ely, R. J., & Thomas, D. A. (2001). Cultural diversity at work: The effects
of diversity perspectives on work group processes and outcomes. Ad-
ministrative Science Quarterly, 46, 229 –273.
Flynn, F. J. (2005). Having an open mind: The impact of openness to
experience on interracial attitudes and impression formation. Journal of
Personality and Social Psychology, 88, 816 826.
Gaertner, S. L., Mann, J. A., Murrell, A. J., & Dovidio, J. F. (1989).
Reduction of intergroup bias: The benefits of recategorization. Journal
of Personality and Social Psychology, 57, 239–249.
George, J. M. (1990). Personality, affect, and behavior in groups. Journal
of Applied Psychology, 75, 107–116.
Gibson, C., & Vermeulen, F. (2003). A healthy divide: Subgroups as a
stimulus for team learning behavior. Administrative Science Quarterly,
48, 202–239.
Glick, W. H. (1985). Conceptualizing and measuring organizational and
psychological climate: Pitfalls in multilevel research. Academy of Man-
agement Journal, 10, 601– 616.
Gruenfeld, D. H., Mannix, E. A., Williams, K. Y., & Neale, M. A. (1996).
Group composition and decision making: How member familiarity and
information distribution affect process and performance. Organizational
Behavior and Human Decision Processes, 67, 1–15.
Harrison, D. A., Price, K. H., & Bell, M. P. (1998). Beyond relational
demography: Time and the effects of surface- and deep-level diversity
on work group cohesion. Academy of Management Journal, 41, 96–107.
Hinsz, V. B., Tindale, R. S., & Vollrath, D. A. (1997). The emerging
conceptualization of groups as information processors. Psychological
Bulletin, 121, 43– 64.
Homan, A. C., van Knippenberg, D., Van Kleef, G. A., & De Dreu,
C. K. W. (in press). Interacting dimensions of diversity: Cross-
categorization the functioning of diverse work groups. Group Dy-
namics: Theory, Research, and Practice.
Hostager, T. J., & De Meuse, K. P. (2002). Assessing the complexity of
diversity perceptions: Breadth, depth, and balance. Journal of Business
and Psychology, 17, 189 –206.
Hull, J. G., Tedlie, J. C., & Lehn, D. A. (1992). Moderator variables in
personality research: The problem of controlling for plausible alterna-
tives. Personality and Social Psychology Bulletin, 18, 115–117.
Ilgen, D. R. (1999). Teams embedded in organizations: Some implications.
American Psychologist, 54, 129 –139.
Jackson, S. E., & Associates (1992). Diversity in the workplace. New
York: Guilford Press.
Jehn, K. A., Northcraft, G. B., & Neale, M. A. (1999). Why differences
make a difference: A field study of diversity, conflict, and performance
in workgroups. Administrative Science Quarterly, 44, 741–763.
Johnson, D. W., & Johnson, F. P. (1982). Joining together: Group theory
and group skills (2nd ed.). Englewood Cliffs, NJ: Prentice Hall.
Karp, H. B., & Sammour, H. Y. (2000). Workforce diversity: Choices in
training programs and dealing with resistance to diversity. College
Student Journal, 34, 451– 459.
Kashy, D. A., & Kenny, D. A. (2000). The analysis of data from dyads and
groups. In H. T. Reis & C. M. Judd (Eds.), Handbook of research
methods in social and personality psychology (pp. 451– 477). Cam-
bridge, England: Cambridge University Press.
Kossek, E. E., & Lobel, S. A. (1996). Managing diversity: Human resource
strategies for transforming the workplace. Oxford, England: Blackwell.
Kossek, E. E., & Zonia, S. C. (1993). Assessing diversity climate: A field
study of reactions to employer efforts to promote diversity. Journal of
Organizational Behavior, 14, 61– 81.
Lau, D., & Murnighan, J. K. (1998). Demographic diversity and faultlines:
The compositional dynamics of organizational groups. Academy of Man-
agement Review, 23, 325–340.
Lau, D., & Murnighan, J. K. (2005). Interactions with groups and sub-
groups: The effects of demographic faultlines. Academy of Management
Journal, 48, 645– 659.
Li, J., & Hambrick, D. C. (2005). Factional groups: A new vantage on
demographic faultlines, conflict, and disintegration in work teams. Acad-
emy of Management Journal, 48, 794 813.
Milliken, F. J., & Martins, L. L. (1996). Searching for common threads:
Understanding the multiple effects of diversity in organizational groups.
Academy of Management Review, 21, 402– 433.
Mook, D. G. (1983). In defense of external invalidity. American Psychol-
ogist, 5, 379 –387.
Mor Barak, M. E., Cherin, D. A., & Berkman, S. (1998). Organizational
and personal dimensions of diversity climate: Ethnic and gender differ-
ences in employee perceptions. Journal of Applied Behavioral Sciences,
31, 82–104.
Muller, D., Judd, C. M., & Yzerbyt, V. Y. (2005). When moderation is
mediated and mediation is moderated. Journal of Personality and Social
Psychology, 89, 852– 863.
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.
Phillips, K. W. (2003). The effects of categorically based expectations on
minority influence: The importance of congruence. Personality and
Social Psychology Bulletin, 29, 3–13.
Phillips, K. W., Mannix, E. A., Neale, M. A., & Gruenfeld, D. H. (2004).
Diverse groups and information sharing: The effects of congruent ties.
Journal of Experimental Social Psychology, 40, 497–510.
Rynes, S. L., & Rosen, B. (1995). A field survey of factors affecting the
adoption and perceived success of diversity training. Personnel Psychol-
ogy, 48, 247–270.
Scholten, L., van Knippenberg, D., Nijstad, B. A., & De Dreu, C. K. W. (in
press). Motivated information processing and group decision making:
Effects of process accountability on information sharing and decision
quality. Journal of Experimental Social Psychology.
Shrout, P. E., & Fleiss, J. L. (1979). Intraclass correlations: Uses in
assessing rater reliability. Psychological Bulletin, 86, 420 428.
Stasser, G., & Titus, W. (1985). Pooling of unshared information in group
decision making: Biased information sampling during discussion. Jour-
nal of Personality and Social Psychology, 48, 1467–1478.
Stevens, J. (1996). Applied multivariate statistics for the social sciences
(3rd ed.). Mahwah, NJ: Erlbaum.
Strauss, J. P., Connerley, M. L., & Ammermann, P. A. (2003). The “Threat
Hypothesis,” personality, and attitudes toward diversity. Journal of
Applied Behavioral Science, 39, 32–52.
Thatcher, S. M. B., Jehn, K. A., & Zanutto, E. (2003). Cracks in diversity
research: The effects of diversity faultlines on conflict and performance.
Group Decision and Negotiation, 12, 217–241.
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S.
(1987). Rediscovering the social group: A self-categorization theory.
Oxford, England: Blackwell.
van Knippenberg, D., De Dreu, C. K. W., & Homan, A. C. (2004). Work
group diversity and group performance: An integrative model and re-
search agenda. Journal of Applied Psychology, 89, 1008 –1022.
van Knippenberg, D., & Haslam, S. A. (2003). Realizing the diversity
dividend: Exploring the subtle interplay between identity, ideology, and
reality. In S. A. Haslam, D. van Knippenberg, M. Platow, & N. Ellemers
(Eds.), Social identity at work: Developing theory for organizational
practice (pp. 61–77). New York: Psychology Press.
1198
HOMAN, VAN KNIPPENBERG, VAN KLEEF, AND DE DREU
van Knippenberg, D., Haslam, S. A., & Platow, M. J. (2004, April). Unity
through diversity: Value-in-diversity beliefs as moderator of the rela-
tionship between work group diversity and group identification. Paper
presented at the Society for Industrial and Organizational Psychology
19th Annual Conference, Chicago.
van Knippenberg, D., & Schippers, M. C. (2007). Work group diversity.
Annual Review of Psychology, 58, 515–541.
Weingart, L. R. (1997). How did they do that? The ways and means of
studying group processes. Research in Organizational Behavior, 19,
189 –239.
Willemsen, T. M., & Fischer, A. H. (1999). Assessing multiple facets of
gender identity: The Gender Identity Questionnaire. Psychological Re-
ports, 84, 561–562.
Williams, K. Y., & O’Reilly, C. A. (1998). Demography and diversity in
organizations. Research in Organizational Behavior, 20, 77–140.
Yzerbyt, V. Y., Muller, D., & Judd, C. M. (2004). Adjusting researchers
approach to adjustment: On the use of covariates when testing interac-
tions. Journal of Experimental Psychology, 40, 424 431.
Received August 1, 2005
Revision received October 8, 2006
Accepted January 12, 2007
New Editors Appointed, 2009 –2014
The Publications and Communications Board of the American Psychological Association an-
nounces the appointment of six new editors for 6-year terms beginning in 2009. As of January 1,
2008, manuscripts should be directed as follows:
Journal of Applied Psychology (http://www.apa.org/journals/apl), Steve W. J. Kozlowski,
PhD, Department of Psychology, Michigan State University, East Lansing, MI 48824.
Journal of Educational Psychology (http://www.apa.org/journals/edu), Arthur C. Graesser,
PhD, Department of Psychology, University of Memphis, 202 Psychology Building, Memphis,
TN 38152.
Journal of Personality and Social Psychology: Interpersonal Relations and Group Processes
(http://www.apa.org/journals/psp), Jeffry A. Simpson, PhD, Department of Psychology,
University of Minnesota, 75 East River Road, N394 Elliott Hall, Minneapolis, MN 55455.
Psychology of Addictive Behaviors (http://www.apa.org/journals/adb), Stephen A. Maisto,
PhD, Department of Psychology, Syracuse University, Syracuse, NY 13244.
Behavioral Neuroscience (http://www.apa.org/journals/bne), Mark S. Blumberg, PhD, De-
partment of Psychology, University of Iowa, E11 Seashore Hall, Iowa City, IA 52242.
Psychological Bulletin (http://www.apa.org/journals/bul), Stephen P. Hinshaw, PhD, Depart-
ment of Psychology, University of California, Tolman Hall #1650, Berkeley, CA 94720.
(Manuscripts will not be directed to Dr. Hinshaw until July 1, 2008, as Harris Cooper will
continue as editor until June 30, 2008.)
Electronic manuscript submission: As of January 1, 2008, manuscripts should be submitted
electronically via the journal’s Manuscript Submission Portal (see the website listed above with
each journal title).
Manuscript submission patterns make the precise date of completion of the 2008 volumes
uncertain. Current editors, Sheldon Zedeck, PhD, Karen R. Harris, EdD, John F. Dovidio, PhD,
Howard J. Shaffer, PhD, and John F. Disterhoft, PhD, will receive and consider manuscripts through
December 31, 2007. Harris Cooper, PhD, will continue to receive manuscripts until June 30, 2008.
Should 2008 volumes be completed before that date, manuscripts will be redirected to the new
editors for consideration in 2009 volumes.
1199
DIVERSITY BELIEFS AND PERFORMANCE
... Materials and procedure. To test our hypothesis and guarantee the internal validity of the research findings, we administered a translated version of the desert survival test of Homan et al. (2007) using back-translation and conducted a situational simulation in the laboratory. The desert survival test requires decision making through coordination and interaction between team members. ...
... We used the individual task to assign different sets of task-related information and to help the participants to remember these information. There are 12 crucial pieces of information in Homan et al. (2007)'s experiment to facilitate the participants' decision making process. Of these 12 pieces of information, 4 were shared within the team. ...
... In the original experiment,Homan et al. (2007) attempted to activate different levels of team faultlines based on these three attributes. However, they did not determine which of these attributes indeed causes a team to fracture; the results only revealed that the activated faultline did, in fact, divide the teams into different subgroups. ...
Article
Full-text available
In this study, we closely examined the transformation from task conflict to relationship conflict and the conditions under which this transformational process is more likely to occur. Based on social identity approach, we suggest that when differing opinions originate from outgroup rather than ingroup members, team members will tend to misattribute the motivations of the conflicting behaviors, causing task conflict to evolve into relationship conflict. We conducted 2 studies to test our hypothesis. In Study 1, 60 4-person teams participated in a simulated task. In Study 2, we used 45 operational teams to further confirm our hypothesis and validate the generalization of the results. The results of both studies support our hypothesis that under a high level of subgroup perception, task conflict is more likely to transform into relationship conflict, which also demonstrates the significance of information source is just as important as, or even more important than the conflicting ideas themselves. The theoretical implications of this study and the new insights that it offers are noted.
... The information elaboration account endogenizes this variable by arguing that it is caused by sociodemographic diversity. Note that the only substantive deviation of Figure 2 from Figure 1 is the arrow from sociodemographic diversity to open and effective communication, the latter of which is now replaced with the term "information elaboration" for consistency with the terminology used in the relevant empirical and theoretical literature (Homan et al. 2007;van Kooji-de Bode et al. 2008;Steel et al. 2021). ...
... Here, I included information elaboration as an individual-level variable since it can be conceived and possibly measured at the individual level. The handful of the relevant empirical (experimental) literature measures information elaboration only at the group level(Homan et al. 2007; van Kooji-de Bode 2008), but this is most likely because of the practical difficulty/cost of measuring it for each participant and not because it is a concept that is only defined at the group level. ...
Preprint
Full-text available
Feminist epistemology aims to propose epistemic reasons for increasing the representation of women or socially subordinated people in science. This is typically done-albeit often only implicitly-by positing a causal mechanism through which the representation of sociodemographic minorities exerts a positive effect on scientific advancement. Two types of causal theories can be identified: The 'epistemic diversity thesis' presents a causal path from sociodemographic diversity to scientific progress mediated by epistemic diversity. The 'thesis of epistemic advantage' proposes a causal path from social subjugation to capacity for inquiry. The latter theory is defined with substantial ambiguity in the existing literature, and I present an explicit causal reformulation that disambiguates it. The epistemic diversity thesis focuses on the effect of group composition on collective epistemic performance and is largely silent about what kind of characteristics lead to individual epistemic excellence. On the other hand, the thesis of epistemic advantage seeks to identify sociodemographic background conditions that make certain epistemic agents strictly better knowers or inquirers than others and pays little attention to the synergistic effects of diverse group composition. Such a difference in the causal structure reflects the diverging political characteristics of the two theories.
... Teams must be able to take advantage of the diversity of thought resulting from functional diversity. Positive attitudes toward diversity have been found to facilitate creative performance in diverse teams (Homan et al., 2007;Nakui et al., 2011). Hoever et al. (2012) found that perspective taking helped facilitate the positive effects of diversity in teams. ...
Article
Synopsis The Research Problem This study examined whether and how top management team (TMT) fault lines formed by diversity and division affect stock price crash risk (hereafter, crash risk). Motivation Prior studies have documented how a single characteristic or the demographic diversity of managers affects real economic activities. However, TMT fault lines, formed by the structure of executives’ demographic diversity, remain untapped in crash risk research. Given the potential monopoly of the TMT on information, it is important to determine whether TMT fault lines influence the distribution of bad news. This study therefore explored how TMT fault lines reduce crash risk through the private leakage of bad news by subgroups and the moderating role of analysts in this process. The Test Hypotheses We hypothesized that firms with a strong TMT fault line have low stock price crash risk. Moreover, subgroups divided by TMT fault lines have a greater chance of leaking bad news when firms are followed by more analysts, thereby reducing the stock price crash risk to a greater extent. Target Population We examined 19,228 annual observations of Chinese publicly listed firms from 2008 to 2021. Adopted Methodology We applied ordinary least squares (OLS) regression and moderation analysis to our study. Analyses We evaluated the microgovernance factors affecting crash risk from a novel perspective — internal TMT conflict. First, we performed regression analysis on a sample of Chinese publicly listed firms to determine whether TMT fault lines reduce crash risk and the effect of analysts on this relationship. Second, we performed cross-sectional variation tests for the main regression. Finally, we investigated the effect of a CEO-CFO coalition on crash risk, given that the unequal status of subgroups may reduce the monitoring capacity of any subgroup in a weaker position. Findings We found that TMT task-related fault lines are negatively associated with crash risk and positively moderated by analysts. Meanwhile, the TMT biodemographic fault lines do not have a significant influence on crash risk. We also demonstrated that the monitoring effect of TMT task-related fault lines is stronger when external governance is weaker. Finally, additional analysis implies that the monitoring effect of a CEO-CFO coalition fault line remains effective, but this effect is weaker.
Article
Full-text available
Purpose This paper investigates the interaction effects of gender diversity and diversity beliefs on group-level personality fit (“collective personality fit”) and group organizational citizenship behavior (GOCB). It seeks to provide a sufficient understanding of the under-researched area, such as how group composition impacts group behavioral outcomes. Design/methodology/approach Data were collected from 453 employees and their leaders in 63 teams of Korean private sector firms using a moderated-mediation model. Findings This study found that high diversity beliefs can weaken the negative effects of gender diversity on collective personality fit and further enhance GOCB. The results confirm the moderated-mediation effect of diversity beliefs in the relationship between gender diversity and GOCB. Practical implications Management should realize that the negative effect of workforce diversity on GOCB can be reduced by boosting collective personality fit in the team. In this process, enhancing diversity beliefs may relieve the adverse effects on GOCB caused by workgroup gender differences. Originality/value This study develops a group-level model proposing that the interaction effects of gender diversity and high diversity beliefs enable a high level of collective personality fit, enhancing GOCB. It attempts to investigate the effects of gender diversity at the group level under boundary conditions.
Article
Full-text available
Deviants are pivotal to sparking social change but their influence is often hindered by group dynamics that serve to maintain the status quo. This paper examines the influence of a group's value in diversity in deviant's ability to spark social change, with a unique focus on the experience and anticipation of group dynamics that enable minority influence. Hypotheses were tested in three studies (NTotal = 674), which varied in their use of ad-hoc conversation groups or existing friend groups, and whether deviants were newcomers, or existing group members. We demonstrated social influence of a vegan deviant increased to the extent that participants perceived their group to value diversity. Furthermore, group value in diversity related to experienced and anticipated group dynamics that enabled minority influence: decreased conformity pressure, increased attentive listening, and, importantly, an increased search for agreement with the deviant. We discuss the importance of studying group dynamics for understanding what valuing diversity entails.
Article
Full-text available
The purpose of this study is to investigate in-group and out-group dynamics within multinational corporations in India and their influence on organizational performance. The paper uses concurrent triangulation. Primary data were gathered via a purposive sample of 500 employees across various demographic backgrounds, ensuring diversity. Respondents participated in an online survey and a few in-depth interviews, administered in-person or online, capturing their experiences of group dynamics. Survey data were analyzed using R, SPSS, and SAS, while thematic analysis was applied to in-depth interview transcripts using NVivo software. The study also employed descriptive statistics, ANOVA, MANOVA for the survey and thematic analysis for interviews, revealing the depth and complexity of in-group favoritism. In-group members have reported significantly higher levels of communication, trust, cooperation, conflict management, and innovation, indicating the impact of group dynamics on organizational performance. The outcome of the multivariate analysis of variance (MANOVA) demonstrates a substantial impact of in-group/out-group categorization on the amalgamated dependent variables (evidenced by a Wilks' Lambda value of 0.68, F (5,94) = 4.17, p < .001). This suggests a noteworthy influence of in-group/out-group affinities on overall team efficacy. The insights offer concrete strategies to enhance inclusivity amid diversity within multinational corporations, thereby mitigating group bias and promoting organizational performance. This study not only expands the understanding of group dynamics in Indian multinational corporations but also offers valuable guidance for managerial practices.
Article
Full-text available
In this article, we attempt to distinguish between the properties of moderator and mediator variables at a number of levels. First, we seek to make theorists and researchers aware of the importance of not using the terms moderator and mediator interchangeably by carefully elaborating, both conceptually and strategically, the many ways in which moderators and mediators differ. We then go beyond this largely pedagogical function and delineate the conceptual and strategic implications of making use of such distinctions with regard to a wide range of phenomena, including control and stress, attitudes, and personality traits. We also provide a specific compendium of analytic procedures appropriate for making the most effective use of the moderator and mediator distinction, both separately and in terms of a broader causal system that includes both moderators and mediators. (46 ref) (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Article
Full-text available
In this article we address issues of diversity within organizational groups by discussing and summarizing previous approaches and by introducing a new variable-faultlines-which depends on the alignment of individual member characteristics. By analyzing a group's faultlines, we focus attention on the underlying patterns of group member characteristics, which can be an important determinant of subgroup conflict, particularly when the group's task is related to one of its faultlines. We discuss the dynamics of faultlines from the early to later stages of a group's development and show how they may be strongest and most likely when diversity of individual member characteristics is moderate.
Article
We examined the impact of surface-level (demographic) and deep-level (attitudinal) diversity on group social integration. As hypothesized, the length of time group members worked together weakened the effects of surface-level diversity and strengthened the effects of deep-level diversity as group members had the opportunity to engage in meaningful interactions.
Article
Organizational and psychological climate research has been plagued by cross-level inference problems. This paper advocates treating the organization as the unit of theory for organizational climate while preserving the individual as the unit of theory for psychological climate. It examines multilevel conceptual problems in climate research and discusses strategies for improving the validity and assessing the reliability of measurement. Additional multilevel research on climate and other areas of organizational science, particularly organizational culture, is encouraged.
Article
Focusing on hybrid team culture within transnational teams as a facilitator of group interaction, we hypothesized a curvilinear relationship between team heterogeneity on nationality and effective performance. Through a qualitative field study, we developed a mediation model of the effects of transnational team dynamics. Two confirmatory laboratory studies followed. The hypothesized curvilinear relationship was confirmed, with homogeneous and highly heterogeneous teams outperforming moderately heterogeneous ones in the long run. Drawing from conceptual work on status hierarchies, group "faultlines," and group membership, we discuss implications for team structures in transnational settings.