Article Assignment

Introduction to Business Research

Spring 2022

Article Critique Assignment

On a typed word document, type your name, date, section, and the article name.

After reading the article assigned, answer the following questions using complete sentences.

1. What is the overview of this article?

2. What is the research problem?

3. What are the research questions?

4. What is the significance of the research?

5. What are the propositions, hypotheses of the article?

6. What research method are the authors using? Do you think this research method is best fit for the study? Why or why not?

7. What are the article findings?

8. What are the strengths of the article? List at least 2.

9. What is the weakness of the article? List at least 2.

10. What implications can be made to the article?

See discussions, stats, and author profiles for this publication at:

Why Differences Make a Difference: A Field Study of Diversity, Conflict and

Performance in Workgroups

Article  in  Administrative Science Quarterly · December 1999

DOI: 10.2307/2667054




3 authors, including:

Some of the authors of this publication are also working on these related projects:

Prenups in Start-ups: How to Write and Frame Contracts Among Founders to Ensure Start-up Success View project

Group Faultlines in Strategic Alliances View project

Karen Jehn

University of Melbourne



Margaret A. Neale

Stanford University



All content following this page was uploaded by Margaret A. Neale on 27 May 2014.

The user has requested enhancement of the downloaded file.

Why Differences Make a Difference: A Field Study of Diversity, Conflict, and
Performance in Workgroups

Karen A. Jehn; Gregory B. Northcraft; Margaret A. Neale

Administrative Science Quarterly, Vol. 44, No. 4. (Dec., 1999), pp. 741-763.

Stable URL:

Administrative Science Quarterly is currently published by Johnson Graduate School of Management, Cornell University.

Your use of the JSTOR archive indicates your acceptance of JSTOR’s Terms and Conditions of Use, available at JSTOR’s Terms and Conditions of Use provides, in part, that unless you have obtained
prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in
the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic
journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers,
and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take
advantage of advances in technology. For more information regarding JSTOR, please contact [email protected]
Thu Nov 1 14:40:32 2007

Why Differences Make
a Difference: A Field
Study of Diversity,
Conflict, and
in Workgroups

Karen A. J e h n
University of Pennsylvania
G r e g o r y B . Northcraft
University of Illinois,
M a r g a r e t A. Neale
Stanford University

01999 by Cornell University.
0001-839219914404-0741/$1 .OO

W e appreciate the research support pro-
vided by t h e Reginald H. Jones Center
for Management Policy, Strategy, and
Organization. W e would also like t o thank
Sandeep Murthy, Emma Vas and Sara
Yang for their assistance w i t h data entry
and analysis and the Wharton Meso
Working Paper Series (especially Sherry
Thatcher, Anne Cummings, Keith
Weigelt, and Steffanie Wilk), Associate
Editors Dan Brass and Keith Murnighan,
and three anonymous AS0 reviewers for
their helpful comments on earlier drafts.

A multimethod field study o f 92 workgroups explored the
influence of three types of workgroup diversity (social
category diversity, value diversity, and informational di-
versity) and t w o moderators (task type and task interde-
pendence) o n workgroup outcomes. Informational diver-
sity positively influenced group performance, mediated
b y task conflict. Value and social category diversity, task
complexity, and task interdependence all moderated this
effect. Social category diversity positively influenced
group member morale. Value diversity decreased satis-
faction, intent t o remain, and commitment t o the group;
relationship conflict mediated the effects of value diver-
sity. We discuss the implications of these results for
group leaders, managers, and organizations wishing t o
create and manage a diverse workforce successfully.’

In response to changing economic conditions, organizations
recently have embraced new structural forms designed to
reduce costs while simultaneously maximizing flexibility and
responsiveness to customer demands (e.g., Boyett and
Conn, 1991 ; Byrne, 1993; Donnellon, 1996). The resulting
flatter, more decentralized organizational forms tend to be
built around groups and depend on rich synchronous com-
munication provided by teams and task forces to a much
greater extent than more traditional hierarchical and central-
ized organizations (Nohria, 1991 ). In addition, groups have
become important vehicles for identifying high-quality solu-
tions to emerging organizational problems (Dumaine, 1991 ).

While groups have become central to organizations, they
present their own intrinsic problems of coordination, motiva-
tion, and conflict management (Gladstein, 1984; Jehn, 1995).
In large part, the use of groups as fundamental building
blocks of organizational structure and strategy seems to be
premised on the assumption that groups can gather together
the diversity of information, backgrounds, and values neces-
sary to make things happen (Jackson, 1992), to produce ef-
fective organizational action. If groups are to provide forums
for sharing information across functional and cultural bound-
aries (Lipnack and Stamps, 1993), however, the diverse
views and backgrounds members bring with them to the
group must be successfully managed. Moreover, the work-
force is becoming increasingly diverse on a number of di-
mensions (e.g., age, gender, ethnicity). Although differences
among members of workgroups are the norm, Byrne’s
(1 971) similarity-attraction theory suggests that people prefer
similarity in their interactions. Likewise, theories of selection
(Chatman, 1991) and socialization (Van Maanen and Schein,
1979) promote similarity in values and demographics as the
basis for maintaining effective work environments. Recently,
however, diversity theorists (Jackson, 1992; Williams and
O’Reilly, 1998), group researchers (Lipnack and Stamps,
1993; Gruenfeld, 1995; Gruenfeld et al., 1996), and creativity
theorists (Amabile, 1994; Oldham and Cummings, 1998)
have been singing the praises of diversity in workgroups.
But empirical research on the effects of diversity has pro-
duced mixed results.

In some studies, diverse groups have been shown to outper-
form homogenous groups (Hoffman and Maier, 1961; Hoff-

741/Administrative Science Quarterly, 44 (1999): 741-763

man, 1978; Nemeth, 1986; Jackson, 1992). In contrast, other
studies have demonstrated that homogenous groups avoid
the process loss associated with poor communication pat-
terns and excessive conflict that often plague diverse groups
(Steiner, 1972; O’Reilly and Flatt, 1989; Ancona and Cald-
well, 1992). These inconsistent results should not be all that
surprising. No theory suggests that a workgroup’s diversity
on outward personal characteristics such as race and gender
should have benefits except to the extent that diversity cre-
ates other diversity in the workgroup, such as diversity of
information or perspective. For instance, social category di-
versity may not always reflect other types of diversity (Tsui
and O’Reilly, 1989)-age does not necessarily reflect values
or even work experience. Even when workgroups do pos-
sess that “other” diversity (e.g., information or perspective),
performance benefits should be expected only to the extent
that workgroup members successfully manage the difficul-
ties of interacting effectively with dissimilar others (e.g., Tsui
and O’Reilly, 1989).

In light of these concerns, it is also not surprising that Wil-
liams and O’Reilly’s (1998) review of forty years of diversity
research concluded that there are no consistent main effects
of diversity on organizational performance. They proposed
that a more complex framework and a more complex con-
ceptualization of the nature of diversity are needed to study
the impact of diversity. Specifically, they called for the incor-
poration of contextual aspects (e.g., task and organizational
characteristics), types of diversity (informational and demo-
graphic), and intervening variables (e.g., communication and
conflict). Our study addresses these concerns by examining
the effects of three specific types of diversity (informational
diversity, social category diversity, and value diversity), a key
intervening process (conflict), and t w o contextual moderators
of these effects (task interdependence and task type) on
workgroup outcomes. W e thus provide a more detailed
model of the process by which various types of workgroup
diversity affect performance than past theorizing. For ex-
ample, differences in gender may not affect member satis-
faction if all members express similar values, and informa-
tion diversity may have little effect on performance when
tasks are highly routine.

Our research builds on prior research investigating various
aspects of contextual and intervening variables to articulate a
more comprehensive understanding of the relationship be-
tween diversity and performance. Pelled (1 996a, 1996b1, for
example, suggested that a workgroup’s social category di-
versity (group differences in social category membership)
enhances its performance and that task conflict-disagree-
ment about task issues-mediates the effects of social cat-
egory diversity. In contrast to task conflict, however, relation-
ship conflicts, which are often caused by social category
diversity, can negatively influence group outcomes (Jehn,
1995). Thus, while personality conflicts may interfere with
task performance, conflict about the best way to perform
the task may lead to insights that increase task performance.
We investigate three types of conflict to determine how dif-
ferent types of diversity influence workgroup outcomes.

742JASQ. December 1999



Diversity and Conflict

Researchers have devoted considerable attention to how
workgroups can generate knowledge and insights beyond
the reach of their individual members (e.g., Murray, 1983;
Doise and Mugny, 1984; Perret-Clermont, Perret, and Bell,
1991; Garton, 1992). This research on emergent knowledge
in groups suggests that social interaction among diverse per-
spectives can lead t o the emergence of new insights
through conceptual restructuring within the groups (e.g., Le-
vine and Resnick, 1993). The creation of knowledge and the
discovery of insight by groups appears to depend on the
presence of diverse viewpoints and perspectives about the
task (Damon, 1991 ; Levine and Resnick, 1993; Nonaka and
Takeuchi, 1995). We explore three categories of diversity
discussed in past research on groups: informational diversity,
social category diversity, and value diversity. These three
types of diversity are not always distinct in practice. For ex-
ample, two individuals from different races (social category
diversity) may (though not necessarily) have experienced dif-
ferent educational cultures (informational diversity) and may
consequently espouse different values (value diversity). Each
of these different kinds of diversity implies different chal-
lenges and opportunities for workgroups, and consequently,
each should differentially influence workgroup outcomes.

lnformational diversity. lnformational diversity refers to dif-
ferences in knowledge bases and perspectives that mem-
bers bring to the group. Such differences are likely to arise
as a function of differences among group members in edu-
cation, experience, and expertise. These differences in edu-
cational background, training, and work experience increase
the likelihood that diverse perspectives and opinions exist in
a workgroup (Stasser, 1992). Recent research has demon-
strated that differences in educational background lead to an
increase in task-related debates in work teams (Jehn, Chad-
wick, and Thatcher, 1997). Task-related debates can be
about either the content or the process of the task. Task
content is about what to do (e.g., a new marketing cam-
paign), in contrast to task process, which is about how to do
it (e.g., delegation of responsibilities). Following Jehn (1995,
19971, w e refer to disagreements about task content as task
conflict and disagreements about task process as process
conflict. We expect that informational diversity will increase
the potential for task conflict:

Hypothesis l a (Hla): lnformational diversity will increase task con-
flict in workgroups.

Workgroups often fail to realize the potential benefits of in-
formational diversity and task conflict for t w o reasons. First,
when groups form naturally in organizations, the most com-
mon bases for group formation are similarity (e.g., New-
comb, 1960; Ancona and Caldwell, 1992), proximity (e.g.,
Festinger, Schachter, and Back, 1950), and familiarity (e.g.,
Tenbrunsel et al., 1994; Mannix, Goins, and Carroll, 1996).
These natural group formation processes typically overselect
members from the same social networks. Because the
knowledge, experiences, and perspectives of group mem-
bers from the same social networks may be more redundant

743/ASQ, December 1999

than diversified (Granovetter, 1973), naturally formed groups
are likely to lack diversity, undermining their potential for
learning, insight, and problem-solving effectiveness (Jackson,

Organizations often counter the tendency of groups to form
based on shared social networks (i.e., similarity, proximity,
familiarity) by creating cross-functional teams, or teams with
members of different functional training, to enhance the in-
formational diversity available in the group (Northcraft et al.,
1995). Even when group membership is specifically man-
aged to enhance informational diversity, however, the poten-
tial of this diversity often is not realized (Steiner, 1972; Hack-
man, 1990). Dougherty (1992), for example, found that
cross-functional new product teams had difficulty getting
their products to market, and Ancona and Caldwell (1992)
found managers’ ratings of innovativeness to be lower when
teams were functionally diverse than when they were homo-
geneous. Similarly, O’Reilly and Flatt (1989) found that top
management teams with homogeneous patterns of organiza-
tional tenure were more creative than teams whose tenure
patterns were more diverse.

Groups with diverse members often prove ineffective at
capitalizing on the potential benefits of their informational
diversity (Stasser and Titus, 1985, 1987). Managers have ex-
pressed frustration with the time and resource demands of
functionally diverse teams, while team members have be-
moaned the difficulty of motivating their members t o work
together effectively (Dumaine, 1994). Even in groups demon-
strating performance benefits from membership diversity,
group members report finding the experience frustrating and
dissatisfying (e.g,. Baron, 1990; Amason and Schweiger,
1 994).

The second reason groups often fail to realize the benefit of
informational diversity is that what makes a group informa-
tionally diverse may also prevent the group from realizing the
benefits of its informational diversity. Disagreements in
workgroups could be disagreements about task content (task
conflict), but they could also be disagreements about how to
do the task or how to delegate resources, reflecting process
conflict (Jehn, 1997). For example, a group member with an
engineering background will probably want to proceed differ-
ently (in terms of how t o identify potential courses of action
and choose among them) than a group member with a mar-
keting or accounting background. Therefore, process con-
flict-disagreements about delegation of duties and re-
sources-are often distinct from task content conflicts-
potentially productive disagreements about the task or
problem at hand, such as the interpretation of market analy-
sis. Recent research has demonstrated that groups with
members of diverse educational majors experience more
difficulty defining how to proceed than groups in which
members have similar educational backgrounds (Jehn, Chad-
wick, and Thatcher, 1997). This gives rise to a second hy-
Hypothesis Ib (Hlb): Informational diversity will increase process
conflict in workgroups.
Social category diversity. While informational diversity is
clearly an important resource for organizations, social cat-

744/ASQ, December 1999


egory diversity is most often what people are referring t o
when talking about diversity (McGrath, Berdahl, and Arrow,
1996). Social category diversity refers to explicit differences
among group members in social category membership, such
as race, gender, and ethnicity (Jackson, 1992; Pelled, 1996a).
Explicit social category membership characteristics provide a
particularly salient basis by which individuals can categorize
themselves and others. Social category diversity is likely to
influence group interactions by virtue of social identity ef-
fects (e.g., Tajfel and Turner, 1986).

According to social identity theory, group members establish
positive social identity and confirm affiliation by showing fa-
voritism to members of their own social category (e.g., Billig
and Tajfel, 1973), an effect, via discrimination and self-segre-
gation, that disrupts group interaction. Social category mem-
bership provides naturally occurring lines along which con-
flicts can be drawn; categorizing individuals into different
groups can provoke hostility or animosity within the work-
group. This intragroup hostility can surface as relationship
conflict-conflict over workgroup members’ personal prefer-
ences or disagreements about interpersonal interactions,
typically about nonwork issues such as gossip, social events,
or religious preferences (Jehn, 1995, 1997). This leads t o
another hypothesis:
Hypothesis 2 (H2): Social category diversity will increase relation-
ship conflict in workgroups.

Value diversity. Value diversity occurs when members of a
workgroup differ in terms of what they think the group’s real
task, goal, target, or mission should be. In many cases,
these differences can lead to task conflict-disagreements
about task content such as disagreements about appropriate
advertisements (Jehn, 1994). They also could lead to pro-
cess conflicts-disagreements about delegation and re-
source allocation. For instance, group members who value
effectiveness (e.g., quality) are likely to have disagreements
about duty and resource allocation with group members who
value efficiency (e.g., units produced). In addition, similarity
in group members’ goals and values enhances interpersonal
relations within the group (Hackman, 1990). This similarity of
values will likely decrease relationship conflict among mem-
bers (Jehn, 1994). This leads to a third hypothesis:
Hypothesis 3 (H3): Value diversity will increase task conflict, pro-
cess conflict, and relationship conflict in workgroups.

Diversity and Performance

Research addressing the determinants of group performance
in organizations suggests that success often hinges on the
ability of the workgroup to embrace, experience, and man-
age (rather than avoid) disagreements that arise (Tjosvold,
1991; Gruenfeld et al., 1996). Considerable evidence points
to the detrimental effects of unmanaged conflicts (e.g.,
Pruitt and Rubin, 1986; Bettenhausen, 1991 ; Jehn, 1997).
Schwenk and Valacich (1 994) found that evaluating and cri-
tiquing-engaging conflicts about the task-yielded better
decisions in workgroups than when members avoided con-
flicts or smoothed over their disagreements. Similarly, Put-
nam (1 994) showed that explicit task disagreements helped
group members better identify issues, and Baron (1 991)

745/ASQ, December 1999

showed that disagreements within groups encouraged group
members to develop new ideas and approaches.

Mischel and Northcraft (1 997) noted that a workgroup’s suc-
cess depends not only on its ability to do the task but also
on the group’s ability to manage its own interactions effec-
tively, including communicating, cooperating, and coordinat-
ing its collective efforts. Similarly, Nonaka and Takeuchi
(1 995), in their discussion of the organizational conditions
that facilitate group performance in knowledge-creating com-
panies, suggested that informational diversity can offer little
benefit to a workgroup whose members cannot work to-
gether effectively to capitalize on it. They suggested that to-
tal diversity among workgroup members is not desirable;
rather, some similarity in perspective among group members
is necessary to ensure enough common ground to facilitate
successful group interaction. Given the aforementioned
negative effects of value and social category diversity (i.e.,
increased relationship conflict), similarity is likely to be most
effective in the areas of value and social category diversity.
In effect, low value diversity and low social category diver-
sity create conditions for a workgroup to take advantage of
its informational diversity, which should be reflected in work-
group performance:

Hypothesis 4 (H4): The effects of informational diversity on work-
group performance will be moderated by value diversity and social
category diversity within the group; informational diversity is more
likely t o increase workgroup performance when value diversity and
social category diversity in the group are low than when they are

Performance is not the only outcome of interest to organiza-
tional workgroups. Also at stake are the morale and commit-
ment of the workers, which have long-term implications for
group performance as well as for costs associated with ab-
senteeism and turnover. Individuals do not enjoy being im-
mersed in interpersonal conflict (Walton and Dutton, 1969;
Peterson, 1983; Ross, 1989), and such conflict makes indi-
viduals less likely to remain (Pervin and Rubin, 1967; Em-
mons, Diener, and Larsen, 1986; Chatman, 1991 ). Signifi-
cantly, it is not necessarily differences resulting from
informational diversity in how to solve the problem or make
the decision that creates the ill-will and bad feelings leading
to physical or psychological withdrawal; rather, it typically
comes from the relationship conflict often caused by social
category diversity and value diversity:

Hypothesis 5 (H5): High value diversity and social category diver-
sity will decrease worker morale.

Moderators of Diversity Effects

The effects of workgroup diversity on workgroup perfor-
mance are likely to be affected by structural aspects of the
task (e.g., Brehmer, 1976; Van de Ven and Ferry, 1980). Evi-
dence suggests that when a task is simple and well under-
stood, group members can rely on standard operating proce-
dures. Under these circumstances, debates about task
strategy are unnecessary and likely t o prove disruptive and
counterproductive (Barnard, 1938; Gladstein, 1984; Jehn,
1995). This is consistent with Jehn’s (1997) finding that pro-
cess conflict interferes with effective performance of simple,

746/ASQ, December 1999


routine tasks. When a task is complex and not well under-
stood, however, discussing and debating competing per-
spectives and approaches is essential for group members t o
identify appropriate task strategies and t o increase the accu-
racy of members’ assessments of the situation (e.g., Fiol,
1994; Amason and Schweiger, 1994; Putnam, 1994; Jehn,
1995). Such complex tasks require problem solving, have a
high degree of uncertainty, and have f e w set procedures
(Van de Ven, Delbecq, and Koenig, 1976), while routine tasks
have a low level of variability, are repetitive (Hall, 19721, and
are generally familiar and done the same way each time
(Thompson, 1967). The constructive discussions and debates
needed t o accomplish complex tasks depend on the avail-
ability of informational diversity:

Hypothesis 6 (H6): Informational diversity is more likely to increase
workgroup performance when tasks are complex rather than rou-

Prior research also suggests that task interdependence can
influence diversity effects in workgroups. Task interdepen-
dence is the extent t o which group members rely on one
another t o complete their jobs (Van de Ven, Delbecq, and
Koenig, 1976). When tasks are interdependent, the demand
for smooth interaction among group members (communica-
tion, cooperation, and coordination of effort) is heightened
(Thibaut and Kelley, 1959; Salancik and Pfeffer, 1977; Saave-
dra et al., 1993). The disruptive effect of value diversity and
social category diversity will be exacerbated when tasks are

Hypothesis 7 (H7): The moderating effects of value diversity and
social category diversity on the relationship between informational
diversity and workgroup performance will be stronger when tasks
are interdependent rather than independent.

The inhibiting effect of value and social diversity on the posi-
tive relationship between informational diversity and perfor-
mance (H4) will be increased when members must interact
closely t o perform a task. Similarly, because task interdepen-
dence heightens the disruptive roles of value diversity and
social category diversity on group interaction, task interde-
pendence also should strengthen the negative effects of
value diversity and social category diversity on worker mo-

Hypothesis 8 (H8): Value diversity and social category diversity will
be more likely t o decrease morale when tasks are interdependent
than when they are independent.

Mediators of Diversity Effects

Finally, because w e have hypothesized that informational,
value, and social category diversity give rise to conflict in
workgroups and that conflict in turn has been linked t o work-
group performance (e.g., Jehn, 1995), w e also hypothesize
that the effects of workgroup diversity will be mediated by
the types of conflict in the workgroup they give rise to,
based on the previous discussions. Relationship and process
conflict have been negatively linked to performance and mo-
rale, while task conflict has been shown t o have positive ef-
fects on performance (Jehn, 1995, 1997; Amason, 1996).
Therefore, w e propose the following hypotheses:

747/ASQ, December 1999

Hypothesis 9a (H9a): Task conflict will mediate the effects of infor-
mational diversity on workgroup performance.

Hypothesis 9b (H9b): Process conflict will mediate the effects of
informational diversity on workgroup performance.

Hypothesis 9c (H9c): Process conflict will mediate the effects of
value diversity on worker morale.

Hypothesis 9d (H9d): Relationship conflict will mediate the effects
of value diversity and social category diversity on worker morale.

The hypotheses were tested in a field study of organizational


Research Site and Sample

The sample consisted of 545 employees in one of the top
three firms in the household goods moving industry. The
sample (as reported in Jehn, 1995) was taken from the inter-
national headquarters for this firm, which houses all func-
tional areas: divisions include marketing and sales, account-
ing, information systems, domestic and international
operations, etc. The featured diversity constructs and mea-
sures are unique t o this study.

This firm had formally designated work units (teams). A work
unit is defined in the organization as a group in which all per-
sonnel report directly to the same supervisor and interact t o
complete unit tasks. W e verified the organization’s delinea-
tion of work units by examining departmental reports and
organizational charts, which indicated that members were
batched together t o perform tasks and were seen by others
as a group. The organization’s delineation of work units was
quite accurate and corresponded with the supervisors’ and
employees’ view of who their fellow group members were.

Work units completed all functions within the organization,
from sorting and delivering mail to making corporate strat-
egy. The work units included sales units selling services t o
corporations moving their employees to other domestic and
international locations, data entry and coding units that pro-
cess this information, and groups that oversee the govern-
mental regulations on state and national cross-border transit.
This organization provides a fitting arena in which t o test our
hypotheses, since it has well-delineated work units that vary
on a wide range of demographic variables and our other vari-
ables of interest (e.g., conflict, task type, interdependence)
yet were relatively similar in size (x = 6.21, s.d. = .47).

Survey Procedure

W e distributed a survey to all employees in the firm. Al-
though the survey was voluntary, the chief executive officer
requested that all employees participate in the confidential
study, supervisors and employees were told in advance that
w e would be there t o administer the survey, and employees
were given company time t o complete it. The response rate
of the survey (89 percent, 485 employees) was quite high
and included 92 complete work units. Later, w e followed up
with employees w h o were absent or off-site (e.g., sales
teams) when w e administered the survey. The high re-
sponse rate allowed us t o include in the analysis only units

748/ASQ, December 1999


with a 100-percent response rate. Thirteen units that did not
achieve full response were dropped from the study.

The survey consisted of 85 self-report, Likert-style questions,
randomly ordered. We used personnel records to verify the
demographic information collected by the survey and, at the
same time, collected archival data, such as performance ap-
praisals and departmental output reports. Sixty supervisors,
managers, and vice presidents received and returned a
packet of materials to evaluate their work unit(s). Information
collected in this packet included organizational charts, group
and individual effectiveness ratings, and departmental output


Diversity. Perceived value diversity among group members
was measured by six 5-point Likert scales anchored by 1 =
“Strongly disagree” and 5 = “Strongly agree.” Members
were asked if the values of all group members were similar,
if the work unit as a whole had similar work values, if the
work unit as a whole had similar goals, whether members
had strongly held beliefs about what was important within
the work unit, whether members had similar goals, and if -all
members agreed on what was important to the group. The
coefficient alpha for this scale was .85. Items were reverse-
coded so that higher scores reflected higher diversity.

Following past research (e.g., McGrath, Berdahl, and Arrow,
1996; Jehn, Chadwick, and Thatcher, 1997), informational
diversity measures assessed heterogeneity of education (i.e.,
major), functional area in the firm (e.g., marketing, mailroom,
operations), and position in the firm (i.e., hourly employee or
management). Social category diversity measures assessed
heterogeneity of sex and age. The firm’s executives declined
to provide data on the ethnic background or nationality of the

As is typical in the treatment of categorical variables, w e
used the entropy-based index (Teachman, 1980; Ancona and
Caldwell, 1992) to form an aggregate measure of the infor-
mational and social category diversity within workgroups:

Diversity = E – PJln Pi),

where Pi represents the proportion of the work unit that has
each diversity characteristic. If a demographic characteristic
is not represented in the team, the value assigned is zero.
Thus, the diversity index represents the sum of the products
of each characteristic’s proportion in the work unit’s makeup
and the natural log of its proportion. The higher the diversity
index, the greater the distribution of characteristics within
the work unit. If the work unit is composed of six individu-
als, one female and five male, their diversity index is ,4506;
if all six members are female, the diversity index is 0.00; and
if three members are female and three are male, the diver-
sity index is ,6931. Likewise, a group with three engineers
and three accountants would have a diversity index of ,6931,
and if all members are engineers, the diversity index is 0.00.

lntragroup conflict. We used the items of the intragroup
conflict scale developed by Jehn (1995) to measure the
amount and type of perceived relationship and task conflict

749/ASQ, December 1999

in the work units. The 12 items on the presence of conflict
were rated on a 5-point Likert scale anchored by 1 = “None”
and 5 = “A lot.” Four items measured relationship conflict
(“How much friction is there among members in your work
unit?” “How much are personality conflicts evident in your
work unit?” “How much tension is there among members
of your work unit?” and “How much emotional conflict is
there among members in your work unit?”). Examples of the
five items measuring task conflict include the following:
“How frequently are there conflicts about ideas in your work
unit?” and “How often do people in your work unit disagree
about opinions?” The coefficient alphas for relationship and
task conflict were .90 and .88, respectively.

Three items measuring process conflict were taken from
Shah and Jehn (1993): “How often do members of your
work unit disagree about who should do what?” “How fre-
quently do members of your work unit disagree about the
way to complete a group task?” and “How much conflict is
there about delegation of tasks within your work unit?” The
coefficient alpha for process conflict was .78.

High correlations among the conflict variables led us to con-
duct a number of analyses to examine the discriminant va-
lidities of the conflict variables, using Howell’s (1 987) ap-
proach. While it is important to discriminate between these
measures in our analyses, it is not unreasonable to expect
that the different types of conflict may overlap. For example,
conflicts originating in personal relationships have been
shown t o spill over into disagreements about how to do the
task (Jehn, 1997). The test of discriminant validity computes
the upper limit for the confidence interval of the observed
correlations and assesses whether this limit is smaller than
the maximum possible correlation between the scores as
computed from their reliability coefficients. All of the conflict
construct pairs meet the discriminant validity test at p <
.0013. In addition, in conducting a factor analysis with ob-
lique rotation, w e found results similar to Shah and Jehn
(1 9931, Amason (1 996), and others (Amason and Sapienza,
1997; Janssen, Van De Vliert, and Veenstra, 1998) who used
the intragroup conflict scale (Jehn, 1995) and found that rela-
tionship, task, and process conflict items load separately
(see Simons and Peterson, 1999, for a review of these stud-
ies and the intercorrelations between the types of conflict).

Task moderator variables. To measure task interdepen-
dence, w e used Van de Ven, Delbecq, and Koenig’s (1976)
workflow interdependence scale, which provides diagrams
depicting the workflow within a unit to measure interdepen-
dence. Group members indicated on a 5-point Likert scale
the degree to which the level of interdependence in their
work unit was similar to the diagram. The average standard
deviation among members within units was quite low (s.d. =
.34), indicating that members viewed their level of task inter-
dependence similarly. We also included Likert-style ques-
tions on task interdependence: “Within my work unit, people
have one-person jobs: that is, people can complete most of
the jobs on their own, with no help from others” (reverse
coded); “Often, all the work unit members meet together to
discuss how each task, case, or claim should be performed
or treated in order to do the work in this unit.” The Cron-

7501ASQ. December 1999


bach alpha of these items (including the diagram Likerts)
was .78. In addition, we verified the reported interdepen-
dence with respondents’ supervisors and via observation.

Task type was measured using an adaptation and combina-
tion of Perrow’s (1 970) index of routinization and Van de
Ven, Delbecq, and Koenig’s (1 976) dimension of task variety.
Examples of items from the 12-item, agree-disagree 5-point
scale are ” I encounter a lot of variety in my normal working
day” (reverse coded), “The methods I follow in my work are
about the same for dealing with all types of work, regardless
of the activity,” “My job is very routine,” and ” I feel like I
am doing the same thing over and over again,” and are simi-
lar to Jehn’s (1995) routineness adaptation of the same
scales. The coefficient alpha for this scale was .94, with high
scores reflecting routineness. We verified the scores for the
reported team task type scores with supervisors’ reports and
observation of the task units at work.

Worker morale. We used three different measures of
worker morale: satisfaction, intent to remain, and commit-
ment. Individual satisfaction with the group was measured
by a 5-point Likert question (“How satisfied are you working
in this work unit?”) anchored by 1 = “Not at all” and 5 =
“Very” and the Kunin Faces Scale (1955). Members re-
sponded t o the Kunin Faces Scale by circling the face that
indicated how happy they were working in their groups. The
coefficient alpha for the t w o satisfaction items was .85.
Members also reported on their intent to remain in the
group by responding to Kraut’s (1975) measure of tenure
intentions: “How long do you expect t o stay in this work
unit?” “If you have your own way, will you be working in
this same work unit three years from now?” and “Do you
want to change work units?” The Cronbach alpha of the
three-item scale was .96.

We rated the commitment of group members by the degree
to which members agreed or disagreed on a 5-point Likert
scale with the following items: ” I talk up this work unit to
my friends as a great group to work in,” ” I am very commit-
ted to my work unit,” “I am proud to tell others that I am
part of this work unit,” and ” I feel a sense of ownership for
this work unit rather than being just an employee.” This ad-
aptation of O’Reilly and Chatman’s (1986) commitment ques-
tionnaire had a coefficient alpha of .85.

Workgroup performance. Perceived group performance
was measured as members’ responses to the following
questions on a 5-point Likert scale: “How well do you think
your work unit performs?” and “How effective is your work
unit?” The coefficient alpha was .93. Actual group perfor-
mance was assessed by departmental records (computer-
ized production records and error reports) provided and stan-
dardized by the firm, and efficiency was assessed by
supervisors’ ratings of the groups. This firm has developed
well-established outcome measures that are comparable
across work units and that are updated biannually. Its Quality
Assurance department is specifically designed to assess the
productivity of work units. For example, to assess the perfor-
mance of one top management team, Quality Assurance de-
signed a 360-degree feedback system that included ratings

751/ASQ, December 1999

of members’ performance by one another, users of their
work (e.g., subordinates), and their vice presidents. We put
this outcome into a standardized form to compare with other
work units. To measure the performance of more routine
task groups, such as one data entry group, the Quality As-
surance team measured the number of data fields entered in
a specified time period and deducted for data errors, along
with measuring other unit tasks. Once again, w e put the out-
come into a standardized form for comparison with other
units. This firm is considered a leader by others in the indus-
try for the performance measures developed by its Quality
Assurance department.

Workgroup efficiency was assessed by supervisors’ ratings
of t w o items measured on 7-point Likert scales, “How effec-
tive is this group at getting things done quickly?” and “How
efficient is this work unit?” (1 = “Not at all Effective” to 7 =
“Very Effective”). The Cronbach alpha for this two-item mea-
sure was .88.


Table 1 provides the means, standard deviations, and corre-
lations for all variables in the model. Our three types of di-
versity are all statistically independent of each other.

Diversity and Conflict

Table 2 provides the regression analyses that tested H I
through H3. Supporting H I a, informational diversity was
positively related to task conflict in workgroups. H I b, pre-
dicting that informational diversity would increase process
conflict, was not supported. Support also was found for HZ:
social category diversity increased relationship conflict in
workgroups. As predicted by H3, value diversity was posi-
tively and significantly related to all three types of conflict.
Social category diversity and value diversity explained 21.9
percent of the variance in relationship conflict within the
groups. Informational and value diversity explained 13.9 per-
cent of the variance in task conflict; value diversity alone ex-
plained 10.3 percent of the variance in process conflict
within workgroups.

Impact of Diversity o n Performance and Worker Morale

We conducted regression analyses to test our hypotheses
predicting the effects of workgroup diversity on worker mo-
rale and performance (H4 through H8). As shown in tables 3
and 4, below, the hypothesized relationships explain be-
tween 6.6 percent (workgroup efficiency) and 37.8 percent
(commitment to workgroup) of workgroup performance and
worker morale. Utilizing a procedure for cross-level analysis
(Rousseau, 1985), w e averaged individual responses on each
of the independent and moderator variables for each work
unit to create a group-level measure for the analysis of
group-level dependent variables only (i.e., workgroup perfor-
mance). We identified workgroups from a listing of who re-
ports t o whom, which was verified by the unit members.
The average intragroup interrater agreement for each vari-
able aggregated for the group performance equations was
between .75 and .87. In addition, w e calculated the c2,
which indicates whether any t w o people in the same group

752/ASQ, December 1999


Table 1

Means, Standard Deviations, and Intercorrelations*

Variable 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3

1. Value diversity
2. Informational diversity
3. Social category diversity
4. Relationship conflict
5 . Process conflict
6 . Task conflict
7. Interdependence
8. Task type
9. Commitment

10. Satisfaction
11. Intent to remain
12. Perceived performance
13. Actual group performance
14. Group efficiency


* Correlations above .09 are significant at the .05 level.

Table 2

Regression Analyses Predicting Conflict ( N = 518)

Relationship Task Process
conflict conflict conflict

Informational diversity ( H I )
Social category d~versity (H2)
Value d~vers~ty (H3)
Adjusted R2

are more similar than t w o people who are members of dif-
ferent groups (Florin et al., 1990). Our results, averaging .54,
exceeded Georgopoulos’s (1986) minimum criteria of .20,
indicating that it was appropriate to aggregate the variables
into group-level variables for the analysis of workgroup per-

Workgroup performance. Table 3 presents the hierarchical
regression analyses conducted to test the hypotheses about
informational diversity and workgroup performance. Step 1
of the hierarchical regression includes the main effects of
informational diversity, value diversity, social category diver-
sity, and task type; step 2 includes the three hypothesized
interactions (informational diversity x value diversity; informa-
tional diversity x social category diversity; and informational
diversity x task type).

Informational diversity was positively related to actual work-
group performance. In support of H4, value diversity moder-
ated the effect of informational diversity on actual perfor-
mance and efficiency; informational diversity was more
beneficial when there were low levels of value diversity than
when there were high levels. In further support of H4, infor-
mational diversity was more positively related to efficiency
when social category diversity was low. In support of H6,
the interaction between informational diversity and task type

753/ASQ, December 1999

Hierarchical Regression Analyses Predicting Workgroup Performance

Step 1: Main effects

Informational diversity (ID)

Social category diversity (SC)

Value diversity (V)

Task type (T)


Step 2: Interactions
H4 ID x V
H4 ID x SC
H6 ID x T

Change in R2
F change
Adjusted R2

‘p i.05; ” p i.01; “‘p i,001.

Perceived Actual group Group
performance performance efficiency

( N = 508) ( N = 87) ( N = 90)

was significant for all three measures of workgroup perfor-
mance-perceived, actual, and efficiency; informational diver-
sity was more likely to increase performance and efficiency
when tasks were complex.

Hypothesis 7 predicted that the moderating effects of value
diversity and social category diversity on the relationship be-
tween informational diversity and workgroup performance
would be strongest when tasks are highly interdependent.
Given the binormal distribution of interdependent and inde-
pendent task groups, to get a clear picture of the three-way
interactions, w e dichotomized the groups into those that
were highly interdependent ( N = 57) and those that were
highly independent ( N = 35). In partial support of H7, the in-
teraction between informational diversity and value diversity
was more strongly related to performance when groups
were interdependent ( B = -.35, p < .01) than when mem-
bers were independent ( B = .09, n.s.1, but interdependence
did not similarly moderate the effects of the social category
diversity and informational diversity interaction on workgroup

Worker morale. H5 predicted that high value diversity and
social category diversity would decrease the morale of work-
group members. As shown in table 4, more value diversity in
the workgroup decreased satisfaction, intent to remain, and
commitment of group members. In contrast, a higher level
of social category diversity increased satisfaction, intent to
remain, and commitment, opposite to what w e had hypoth-
esized. Hypothesis 8 predicted that value diversity and social
category diversity would be more likely to affect morale
when task interdependence is high. As shown in table 4,
members in interdependent groups were more satisfied and
felt more committed when high levels of social category di-
versity were present; however, these interactions did not
significantly add to the variance explained by the main ef-

754/ASQ, December 1999


Table 4

Hierarchical Regression Analyses Predicting Worker Morale

Satisfaction to remain Commitment

( N = 491) ( N = 412) ( N = 488)

Step 1: Main effects
Informational diversity (ID)
H5 Social category diversity (SC)
H5 Value diversity (V)
Interdependence ( I )

Step 2: Interactions
H 8 V x I
H8 SC x I

Change in R2
F change
Adjusted R2

‘p < .05; ‘ p < .01; “‘p < ,001.

fects. The interaction between value diversity and task inter-
dependence was not significant.
Mediators of Diversity Effects
Hypothesis 9a predicted that task conflict would mediate the
effects of informational diversity on workgroup performance.
Using the procedure suggested by Baron and Kenny (1 986),
w e found that the significant effect of informational diversity
on actual group performance (5 = .30, p < .001) became
nonsignificant (5 = .07, n.s.1 when task conflict was con-
trolled for. Thus, the mediating role of task conflict between
informational diversity and actual group performance was
confirmed. The results did not confirm hypothesis 9b, that
process conflict would mediate the effects of informational
diversity on workgroup performance. Results did confirm hy-
pothesis 9c, which predicted that process conflict would me-
diate the effects of value diversity on worker morale. Value
diversity was significantly related t o the following dependent
variables: satisfaction ( B = – . I 1, p < .05), intent t o remain ( B
= -.19, p < .001), commitment (5 = -.19, p < ,011, per-
ceived performance ( B = -.lo, p < .05), actual group perfor-
mance (5 = -.I2, p < ,051, and group efficiency (5 = -.I7, p
< .01). The effect of value diversity became nonsignificant
when process conflict was included in the regression analy-
ses on satisfaction (5 = -.04) , intent t o remain (5 = -.05),
commitment (5 = -.05), perceptual performance (5 = .04),
and actual group performance (5 = .03), meaning that value
diversity accounts for the variation in these outcome vari-
ables through process conflict. Process conflict thus has a
mediating role in the relationship of value diversity t o satis-
faction, intent t o remain, commitment, perceptual perfor-
mance, and actual group performance.
H9d predicted that relationship conflict would mediate the
effects of value diversity and social category diversity on
worker morale. Relationship conflict (mediator) was re-
gressed on value diversity and social category diversity (inde-
pendent variables) and found t o be significant ( 5 s = . I 5, .08,

755/ASQ, December 1999

p < .01, .05, respectively). Second, value diversity was sig-
nificantly related to the following dependent variables: satis-
faction (5 = – . I 1, p < .05), intent t o remain (5 = -.19, p <
.001), commitment (5 = – . I 9, p < .01), perceptual perfor-
mance (5 = -.lo, p < .05), actual group performance (5 =
-.I2, p < .05), and group efficiency (5 = -.I7, p < ,011. So-
cial category diversity was significantly related to the follow-
ing dependent variables: satisfaction (5 = .14, p < ,011, in-
tent t o remain (5 = . I 2, p < ,011, commitment (5 = .16, p <
.001), and perceptual performance (5 = .16, p < .001). Value
diversity’s effect became nonsignificant when relationship
conflict was included in the regression analyses for satisfac-
tion (5 = .01), intent to remain (5 = -.02), perceptual perfor-
mance (5 = .04), and actual group performance (5 = .02),
meaning that value diversity accounts for the variation in
these outcome variables through relationship conflict. Thus,
for satisfaction and intent to remain, relationship conflict me-
diates between value diversity and worker morale.
Social category diversity’s effect also became nonsignificant
when relationship conflict was included in the regression
analyses for satisfaction (5= -.01), intent to remain (5 =
.05), commitment (5 = .06), and perceptual performance (5
= .01). Thus, the mediating role of relationship conflict be-
tween social category diversity and worker morale was con-
firmed for satisfaction, intent to remain, and commitment.

The purpose of this study was to explore the differential im-
pact of three group compositional factors (social category
diversity, value diversity, and informational diversity) and t w o
moderating variables (task type and task interdependence)
on workgroup performance. With few exceptions (see
Gruenfeld et al., 1996; Jehn, Chadwick, and Thatcher, 1997),
past research has lumped social category diversity and infor-
mational and value diversity under the general heading of
diversity in attempting to understand the impact of diversity
on workgroup performance. In addition, even when distinc-
tions have been made about types of diversity, previous re-
search has typically been limited to studying one type of di-
versity but not others (e.g., O’Reilly and Flatt, 1989; Ancona
and Caldwell, 1992). As a result, it is not surprising that a
review of this literature produces different results across
studies that purport to study the same thing-diversity and
its impact on performance.
The present study was successful in distinguishing among
three types of diversity and their impact on workgroup per-
formance. While previous research has demonstrated the
influence of conflict on workgroup outcomes (Jehn, 1995,
1997; Amason, 1996), the study described here takes the
additional step of exploring how different types of diversity
evoke conflict. The results show that different forms of di-
versity exacerbate different forms of conflict (within different
task configurations), which in turn affects perceived perfor-
mance, actual performance, satisfaction, intent to remain,
and commitment. How these different types of diversity ulti-
mately influence performance, both perceived and actual, is
no simple story.
Before w e review that story, however, w e wish to acknowl-
edge the limitations of this study. Because w e did not have

7561ASQ. December 1999


access to the ethnic diversity of the participants in this
study, our social category diversity results are based on the
measurement of only t w o factors: age and gender. In addi-
tion, the study is cross-sectional, so that no causal infer-
ences can be drawn. Finally, some of the variable measures
are self-reported, and w e cannot rule out the possibility of
response-response bias in some of our analyses. To mini-
mize this possiblity, however, w e also included measures
from archival data and multiple sources. For example, while
different types of conflict, value diversity, the moderators,
and many of the affective dependent variables (e.g., mea-
sures of morale, intent t o remain, etc.) were self-reported,
the t w o other diversity measures (social category and infor-
mational diversity), actual group performance, and workgroup
efficiency were based on archival data or supervisory ratings.
In addition, besides conducting multiple tests t o assess the
discriminant validity of the three types of conflict, future re-
search should examine the transformation of one type of
conflict into another. For instance, arguments about who is
capable of doing what can often lead t o relationship con-
flicts, and vice versa.

While most of our hypotheses received support in the pre-
dicted direction, w e did have an unexpected finding. The
finding that social category diversity resulted in increased
relationship conflict, even though group members reported
increased morale, runs counter to both conventional wisdom
and past research. For example, Jehn (1995, 19971, among
others (Pruitt and Rubin, 1986; Bettenhausen, 1991 ; Sch-
wenk and Valacich, 1994), illustrated how relationship con-
flict is associated with a general reduction in worker morale.
One explanation for this finding may be the particular vari-
ables that compose the social category diversity variable in
this study. It may be that at least for age, diversity on this
factor reflects lower levels of intragroup competition, as
workers are more likely t o be competing with similar (in age)
others for various valued organizational resources. But this
inconsistency, coupled with the cross-sectional nature of the
data, suggests at least one plausible, alternative explanation.
It may be that high performance leads t o high morale and
low task conflict rather than, in our interpretation, that low
task conflict leads t o high morale and high performance.

W e explored this alternative explanation by positing that a
third variable, group performance, may be affecting the rela-
tionship between social category diversity and worker mo-
rale such that it would overwhelm the negative effects of
relationship conflict on morale. In fact, follow-up analyses
demonstrated that performance mediated the impact of so-
cial category diversity on morale. While, as noted above, so-
cial category diversity was significantly related t o satisfac-
tion, intent t o remain, perceived performance, and
commitment, these effects became nonsignificant when
group performance was included in the regression analyses.
Thus, the mediating role of performance between social cat-
egory diversity and worker morale was confirmed for satis-
faction, intent to remain, perceived performance, and com-
mitment. Diverse groups performed better and perhaps,
therefore, were more pleased with the group in which they

757/ASQ, December 1999

were working, independent of its level of social category di-

From this study w e can identify the types of diversity that
are associated with various types of performance. For a
team t o be effective, members should have high information
diversity and low value diversity. For a team t o be efficient,
members should have low value diversity. For a team to
have high morale (higher satisfaction, intent to remain, and
commitment) or t o perceive itself as effective, it should be
composed of participants with low value diversity. What
these consistent findings suggest is the value, for most
measures of group performance, of low value diversity
among members. Moreover, it may also be that value diver-
sity, which is often not immediately discernible, becomes
more important as a predictor of group performance over
time, while age and gender diversity, characteristics that are
readily apparent, become less relevant over time. The impor-
tance of low value diversity on workgroup performance over
time is also supported by results of a recent field study of
research and development teams (Owens and Neale, 1999).

The most arresting aspect of this study may be the window
it provides into our understanding of the importance of value
diversity t o both workgroup performance and worker morale.
Thus, it seems that certain types of similarity are dramati-
cally more important than others, despite the assumption
that people generally strive for similarity among those with
whom they interact (Byrne, 1971 ). It is the diversity associ-
ated with values, and not social category, that causes the
biggest problems in and has the greatest potential for en-
hancing both workgroup performance and morale.

This study suggests, like Williams and O’Reilly (19981, that
the impact of diversity goes well beyond simple main ef-
fects. Task interdependence and task type moderate the re-
lationships between diversity and various measures of per-
formance. Informational diversity is more likely t o lead t o
improved performance when tasks are nonroutine. Again,
social category diversity unexpectedly led t o greater satisfac-
tion and commitment when task interdependence was high
than when it was low. It is more difficult here t o explain
away this finding by deferring to the performance-morale
path. It may actually be that social category diversity results
in higher morale in interdependent tasks. Being able t o work
together successfully, even when the group is diverse with
respect to age and gender composition, may result in
greater morale because the group has overcome a serious
challenge to its effectiveness. Further, these groups may
have discovered that the social category differences were
not good signals of value diversity. This interpretation re-
ceived support from earlier research. For example, several
studies in the 1960s (Byrne and Wong, 1962; Stein, Hardyck,
and Smith, 1965) found that whites preferred blacks with
attitudes similar t o their own over whites with opposing atti-
tudes, but this effect of value similarity on racial attitudes
apparently has been ignored in recent years, as researchers
have used similarity in attributes such as race or gender as
surrogates for value similarity. Further, it appears from the
work of Owens and Neale (1999) that groups are aware of
some of the impact of different types of diversity on perfor-

758/ASQ, December 1999


mance and can select team members for their contributions
along multiple diversity dimensions that enhance group per-
formance. Taken as a whole, these results provide direction
for creating and managing diverse teams to enhance perfor-
mance. Our results suggest that diversity itself is not
enough t o ensure innovation; the nature of the team’s diver-
sity is critical. For group members t o be willing to engage in
the difficult and conflictful processes that may lead t o inno-
vative performance, it seems that group members must
have similar values.

Our results also shed light on the difficulty of studying social
category diversity. One problem associated with attempting
t o make predictions about the effects of social category di-
versity on workgroup performance is that social category di-
versity may represent informational diversity, value diversity,
both, or neither. Since social category diversity is not neces-
sarily associated with either informational or value diversity,
it poses prediction problems for researchers and signaling
problems for group members. What does being the only
woman in an otherwise all-male group mean about the
unique perspectives that an individual brings t o the group? If
the task of the group is t o define a strategic direction for the
organization, and all group members have backgrounds in
finance, it is not likely that the gender of one member will
make a significant difference in the information that an indi-
vidual brings t o the group. If the group must select product
features for a new model of automobile, however, the expe-
rience of being a woman may bring a different orientation t o
the discussion, even if that woman is an engineer, just like
everyone else in the group. Finally, are all categorical vari-
ables equally influential? Are there ebbs and flows of influ-
ence of these categorical variables as teams age and
evolve? While clearly important questions, w e must await
future research for the answers.

Unlike demographic characteristics, the characteristics of
value and informational diversity are not easily discernible
from a quick physical inspection of fellow group members
the way that social category characteristics often are. Thus,
because of their ease of observation, demographic character-
istics, in particular, are more likely to be incorporated into
the heuristic information processing of group members as
they develop mechanisms t o manage group processes and
complete assigned tasks. Just as past researchers may have
relied on social category diversity as a surrogate for informa-
tional and value diversity, social category similarity may lead
group members to overlook important sources of informa-
tional and value diversity or t o assume similarity where it
does not exist.

Our findings-specifically, distinguishing among different
types of diversity and their differential effects-may help
reconcile some of the inconsistencies in past research. If the
type of diversity measured is informational diversity, group
performance may be enhanced by diversity. If the type of
diversity measured is social category diversity, the most
positive effects will likely be on worker morale (satisfaction,
intent t o remain, commitment, and perceived performance).
In contrast, groups that have greater diversity as measured
in terms of values may suffer significant performance decre-

759/ASQ, December 1999

ments (being less effective and efficient as well as having
poorer perceived performance) and diminished worker mo-
rale (decreased satisfaction, commitment, and intent to re-
main in the group). While the story told in previous research,
even with its contradictory findings and inconsistent empiri-
cal support, may have been easier to tell-heterogeneity
leads to better workgroup performance and homogeneity
leads to easier workgroup process-the more complex rep-
resentation of these relationships as provided by this paper
should enhance our understanding of the ways t o create,
intervene in, and manage high-performance groups and


Amabile, Teresa M.
1994 “The atmosphere of pure

work: Creativity in research
and development.” In William
R. Shadish and Steve Fuller et
al. (eds.), The Social Psychol-
ogy of Science: 31 6-328.
New York: Guilford Press.

Amason, Allen C.
1996 “Distinguishing the effects of

functional and dysfunctional
conflict on strategic decision
making: Resolving a paradox
for top management teams.”
Academy of Management
Journal, 39: 123-1 48.

Amason, Allen C., and Harry J.
1997 “The effects of top manage-

ment team size and interac-
tion norms on cognitive and
affective conflict.” Journal of
Management, 23: 495-516.

Amason, Allen C., and David M .
1994 “Resolving the paradox of

conflict, strategic decision
making and organizational per-
formance.” International Jour-
nal of Conflict Management,
5: 239-253.

Ancona, Deborah, and David
1992 “Demography and design:

Predictors of new product
team performance.” Organiza-
tion Science, 3: 321-341.

Barnard, Chester I.
1938 The Functions of the Execu-

tive. Cambridge, MA: Harvard
University Press.

Baron, R. A.
1990 “Countering the effects of

destructive criticism: The rela-
tive efficacy of four intewen-
tions.” Journal of Applied Psy-
chology, 75: 235-245.

1991 “Positive effects of conflict: A
cognitive perspective.” Em-
ployees Responsibilities and
Rights Journal, 4: 25-36.


Baron, R. M., and D. A. Kenny
1986 “The moderator-mediator vari-

able distinction in social psy-
chological research: Concep-
tual, strategic, and statistical
considerations.” Journal of
Personality and Social Psy-
chology, 51 : 1 173-1 182.

Bettenhausen, Kenneth L.
1991 “Five years of group research:

What w e have learned and
what needs to be addressed.”
Journal of Management, 17:

Billig, Michael, and Henri Tajfel
1973 “Social categorization and

similarity in intergroup behav-
ior.” European Journal of So-
cial Psychology, 3 (1 ): 27-52.

Boyett, J. H., and H. P. Conn
1991 Workplace 2000: The Revolu-

tion Reshaping American Busi-
ness. New York: Dutton.

Brehmer, Berndt
1976 “Social judgement theory and

the analysis of interpersonal
conflict.” Psychological Bulle-
tin, 83 (6): 985-1 003.

Byrne, D.
1971 The Attraction Paradigm. New

York: Academic Press.

Byrne, D., and T. J. Wong
1962 “Racial prejudice, interper-

sonal attraction, and assumed
dissimilarity of attitudes.”
Journal of Abnormal Psychol-
ogy, 65: 246-253.

Byrne, John A.
1993 horizontal corporation,^^

Business Week, December
20: 76-81.

Chatman, Jennifer A.
1991 hi^^ people and organi-

zations: Selection and social-
ization in public accounting
firms.” Administrative Science
Quarterly, 36: 45S484.

Damon, William
1991 “Problems of direction in so-

cially shared cognition.” In
Lauren B. Resnick, John M .
Levine, and S. D. Teasley
(eds.), Perspectives on So-
cially-Shared Cognition: 384-
397. Washington, DC: Ameri-
can Psychological Association.

Doise, W., and G. Mugny
1984 The Social Development of

the Intellect. Oxford: Perga-
mon Press.

Donnellon, Anne
1996 Team Talk: The Power of Lan-

guage in Team Dynamics.
Cambridge, MA: Haward Busi-
ness School Press.

Dougherty, Deborah
1992 “Interpretive barriers to suc-

cessful product innovation in
large firms.” Organization Sci-
ence, 3: 179-202.

Dumaine, Brian
1991 “The bureaucracy busters.”

Fortune, June 17: 36-50.
1994 “The trouble with teams.”

Fortune, September 5: 52-59.

Emmons, Robert A., Ed Diener,
and Randy J. Larsen
1986 “Choice and avoidance of ev-

eryday situations among men
and women in managerial,
professional, and technical
positions.” Journal of Applied
Psychology, 75: 539-546.

Festinger, Leon, S. Schacter, and
K. Back
1950 Social Pressure in Informal

Groups. New York: Harper
and Brothers.

Fiol, C. Marlene
1994 “Consensus, diversity, and

learning in organizations.” Or-
ganization Science, 5: 403-

760/ASQ, December 1999


Florin, Paul, Gary A. Giamartino,
David A. Kenny, and Abraham
1990 “Levels of analysis and ef-

fects: Clarifying group influ-
ence and climate by separat-
ing individual and group
effects.” Journal of Applied
Social Psychology, 20: 881-

Garton, A. F.
1992 Social Interaction and the De-

velopment of Language and
Cognition. Hillsdale, NJ:

Georgopoulos, Basil Spyros
1986 Organizational Structure, Prob-

lem Solving, and Effective-
ness. San Francisco: Jossey-

Gladstein, Deborah L.
1984 “A model of task group effec-

tiveness.” Administrative Sci-
ence Quarterly, 29: 499-517.

Granovetter, Mark
1973 Getting a Job: A Study of

Contacts and Careers. Cam-
bridge, MA: Harvard Univer-
sity Press.

Gruenfeld, Deborah H.
1995 “Status, ideology, and integra-

tive complexity on the U.S.
Supreme Court: Rethinking
the politics of political decision
making.” Journal of Personal-
ity and Social Psychology, 68:
5-2 0.

Gruenfeld, Deborah H., Elizabeth
A. Mannix, Katherine Y. Williams,
and Margaret A. Neale
1996 “Group composition and deci-

sion making: How member
familiarity and information dis-
tribution affect process and
performance.” Organizational
Behavior and Human Decision
Processes, 67: 1-1 5.

Hackman, J. Richard
1990 Groups that Work (and Those

that Don’t). San Francisco:

Hall, R. H.
1972 Organizations, Structure, and

Process. Englewood Cliffs,
NJ: Prentice-Hall.

Hoffman, L. R.
1978 “The group problem-solving

process.” In L. Berkowitz
(ed.), Group Processes: 101-
1 14. New York: Academic

Hoffman, L. R., and N. R. F. Maier
1961 “Quality and acceptance of

problem solutions by mem-
bers of homogeneous and
heterogeneous groups.” Jour-
nal of Abnormal and Social
Psychology, 62: 401-407.

Howell, R. D.
1987 “Covariance structure model-

ing and measurement issues:
A note on interrelations
among a channel entitity’s
power sources.” Journal of
Marketing Research, 24: 11 9-

Jackson, Susan
1992 “Team composition in organi-

zations.” In S. Worchel, W.
Wood, and J. Simpson (eds.),
Group Process and Productiv-
ity: 1-12. London: Sage.

Janssen, Onne, Evert Van De
Vliert, and Christian Veenstra
1998 “How task and person conflict

shape the role of positive in-
terdependence in manage-
ment teams.” Journal of Man-
agement, 25: 1 17-142.

Jehn, Karen A.
1994 “Enhancing effectiveness: An

investigation of advantages
and disadvantages of value-
based intragroup conflict.”
International Journal of Con-
flict Management, 5: 223-238.

1995 “A multimethod examination
of the benefits and detriments
of intragroup conflict.” Admin-
istrative Science Quarterly, 40:

1997 “A qualitative analysis of con-
flict types and dimensions in
organizational groups.” Admin-
istrative Science Quarterly, 42:

Jehn, Karen A,, Clint Chadwick,
and Sherry Thatcher
1997 “To agree or not to agree:

Diversity, conflict, and group
outcomes.” International Jour-
nal of Conflict Management,
8: 287-306.

Kraut, Allen I.
1975 “Predicting turnover of em-

ployees for measured job atti-
tudes.” Organizational Behav-
ior and Human Decision
Processes, 13: 233-243.

Kunin, Theodore
1955 “The construction of a new

type of attitude measure.”
Personnel Psychology, 8: 65-

Levine, John M., and L. B. Resnick
1993 “Social foundations of cogni-

tions.” Annual Review of Psy-
chology, 41 : 585-61 2. Palo
Alto, CA: Annual Reviews.

Lipnack, J., and J. Stamps
1993 The Teamnet Factor: Bringing

the Power of Boundary Cross-
ing in the Hearts of Your Busi-
ness. Essex Junction, VT:
Oliver Wright.

Mannix, Elizabeth A., Sheila
Goins, and S. Carroll
1996 “Starting at the beginning:

Linking team performance to
process and performance.”
Working paper, School of
Business, Columbia Univer-

McGrath, Joseph E., Jennifer L.
Berdahl, and Holly Arrow
1996 “No one has it but all groups

do: Diversity as a collective,
complex, and dynamic prop-
erty of groups.” In Susan E.
Jackson and Marian N. Ruder-
man (eds.), Diversity in Work
Teams: Research Paradigms
for a Changing World: 42-66.
Washington, DC: APA Publica-

Mischel, Leann Joan, and Gregory
B. Northcraft
1997 ” 1 think w e can, I think we

can . . . : The role of self-effi-
cacy beliefs in group and
team effectiveness.” In Barry
Markovsky and Michael J.
Lovaglia (eds.), Advances in
Group Processes, 14: 177-
197. Greenwich, CT: JAI

Murray, F. B.
1983 “Learning and development

through social interaction and
conflict: A challenge to social
learning theory.” In L. Liben
(ed.), Piaget and the Founda-
tion of Knowledge: 231-247.
Hillsdale, NJ: Erlbaum.

Nemeth, Charlan
1986 “Differential contributions of

majority and minority influ-
ence.” Psychological Review,
93: 23-32.

Newcomb, T. M.
1960 The Acquaintance Process.

New York: Holt, Rinehart, and

Nohria, Nitin
1991 “Garcia-Pont, Carlos Global

strategic linkages and industry
structure.” Strategic Manage-
ment Journal, 12: 105-1 24.

Nonaka, Ikujiro, and H. Takeuchi
1995 The Knowledge-Creating Com-

pany. New York: Oxford Uni-
versity Press.

Northcraft, Gregory B., Jeffrey T.
Polzer, Margaret A. Neale, and
Roderick Kramer
1995 “Productivity in cross-func-

tional teams: Diversity, social
identity, and performance.” In
Susan E. Jackson and Marian
N. Ruderman (eds.), Diversity
in Work Teams: Research
Paradigms for a Changing
World: 69-96. Washington,
DC: APA Publications.

761 /ASQ, December 1999

Oldham, Greg, and Anne
1998 “Creativity in the organiza-

tional context.” Productivity,
39 (2): 187-1 94.

O’Reilly, Charles A.,and Jennifer
1986 “Organizational commitment

and psychological attachment:
The effects of compliance,
identification, and internaliza-
tion of prosocial behavior.”
Journal of Applied Psychology,
71: 492-499.

O’Reilly, Charles A,, and Sylvia
1989 “Executive team demography:

Organizational innovation and
firm performance.” Working
paper, School of Business,
University of California at

Owens, David A,, and Margaret A.
1999 “The dubious benefit of group

heterogeneity in highly uncer-
tain situations: Too much of a
good thing?” Working paper,
School of Business, Vanderbilt

Pelled, Lisa
1996a “Demographic diversity, con-

flict, and work group out-
comes: An intervening pro-
cess theory.” Organization
Science, 17: 61 5-631.

1996b “Relational demography and
perceptions of group conflict
and performance: A field in-
vestigation.” International
Journal of Conflict Resolu-
tion, 7: 230-246.

Perret-Clermont, A,, Jean-Francois
Perret, and Nancy Bell
1991 “The social construction of

meaning and cognitive activity
in elementary school chil-
dren.” In Lauren B. Resnick,
John M. Levine, and S. D.
Teasley (eds.), Perspectives
on Socially-Shared Cognition:
41-62. Washington, DC:
American Psychological Asso-

Perrow, Charles
1970 Organizational Analysis: A So-

ciological View. Belmont, CA:

Pervin, Lawrence A,, and Donald
B. Rubin
1967 “Student dissatisfaction with

college and the college drop-
out: A transactional ap-
proach.” Journal of Social Psy-
chology, 72: 285-295.

Peterson, Donald R.
1983 “Conflict.” In Harold H. Kelley

et al. (eds.), Close Relation-
ships: 360-396. New York:
W.H. Freeman.

Pruitt, Dean G., and Jeffrey Z.
1986 Social Conflict: Escalation,

Stalemate, and Settlement.
New York: Random House.

Putnam, Linda
1994 “Productive conflict: Negotia-

tion as implicit coordination.”
International Journal of Con-
flict Management, 5: 285-299.

Ross, Raymond
1989 “Conflict.” In R. Ross and J.

Ross (eds.), Small Groups in
Organizational Settings: 139-
178. Englewood Cliffs, NJ:

Rousseau, Denise
1985 “Issues of level in organiza-

tional research.” In L. L.
Cummings and Barry M. Staw
(eds.), Research in Organiza-
tional Behavior, 7: 1-37.
Greenwich, CT: JAI Press.

Saavedra, Richard, P. Christopher
Earley, Linn Van Dyne, and
Cynthia Lee
1993 “Complex interdependence in

task-performing groups.” Jour-
nal of Applied Psychology, 71 :

Salancik, Gerald R., and Jeffrey
1977 “A social information process-

ing approach to job attitudes
and task design.” Administra-
tive Science Quarterly, 23:

Schwenk, Charles, and Joseph S.
1994 “Effects of devil’s advocacy

and dialectical inquiry on indi-
viduals versus groups.” Orga-
nizational Behavior and Hu-
man Decision Processes, 59:

Shah, P. P., and Karen A. Jehn
1993 “Do friends peform better

than acquaintances: The inter-
action of friendship, conflict,
and task.” Group Decision and
Negotiation, 2: 149-1 66.

Simons, T., and R. Peterson
1999 “Task conflict and relationship

conflict in top management
teams: The pivotal role of in-
tragroup trust.” Journal of Ap-
plied Psychology (in press).

762/ASQ, December 1999

Stasser, Garold
1992 “Information salience and the

discovery of hidden profiles by
decision-making groups: A
‘thought experiment.”‘ Organi-
zational Behavior and Human
Decision Processes, 52: 156-

Stasser, Garold, and William Titus
1985 “Pooling of unshared informa-

tion in group decision making:
Biased information sampling
during discussion.” Journal of
Personality and Social Psy-
chology, 48: 1467-1478.

1987 “Effects of information load
and percentage of shared in-
formation on the dissemina-
tion of unshared information
during group discussion.”
Journal of Personality and So-
cial Psychology, 53: 81-93.

Stein, D. D., J. E. Hardyck, and
M . B. Smith
1965 “Race and belief: An open

and shut case.” Journal of
Personality and Social Psy-
chology, 1: 21-289.

Steiner, Ivan D.
1972 Group Process and Productiv-

ity. San Diego: Academic

Tajfel, Henri, and J. C. Turner
1986 “The social identity theory of

intergroup behavior.” In S.
Worchel and W. G. Austin
(eds.), Psychology of Inter-
group Relations: 33-47. Chi-
cago: Nelson-Hall.

Teachman, J. D.
1980 “Analysis of population diver-

sity.” Sociological Methods
and Research, 8: 341-362.

Tenbrunsel, Ann E., Kim A.
Wade-Benzoni, Joseph Moag, and
Max H. Bazerman
1994 “The costs of friendship: The

effects of strong ties on
matching efficiency, individual
effectiveness and barriers to
learning.” Unpublished manu-
script, Kellogg Graduate
School of Management
(DRRC), Northwestern Univer-

Thibaut, John W., and Hal H.
1959 The Social Psychology of

Groups. New York: Wiley

Thompson, James D.
1967 Organizations in Action. New

York: McGraw-Hill.

Tjosvold, Dean
1991 The Conflict Positive Organiza-

tion. Reading, MA: Addison-


Tsui, Anne S., and Charles A.
1989 “Beyond simple demographic

effects: The importance of
relational demography in supe-
rior-subordinate dyads.” Acad-
emy of Management Journal,
32: 402-423.

Van de Ven, Andrew H., Andre
Delbecq, and R. Koenig
1976 “Determinants of coordination

modes within organizations.”
American Sociological Review,
41: 322-338.

Van de Ven, Andrew H., and
Diane Ferry
1980 Measuring and Assessing Or-

ganizations. New York: Wiley.

Van Maanen, John, and Edgar
1979 “Toward a theory of organiza-

tional socialization.” In L. L.
Cummings and B. M. Staw
(eds.), Research in Organiza-
tional Behavior, 1 : 209-264.
Greenwich, CT: JAI Press.

Walton, Richard E., and John M.
1969 “The management of interde-

partment conflict: A model
and review.” Administrative
Science Quarterly, 14: 73-84.

Williams, Katherine Y., and
Charles A. O’Reilly
1998 “Demography and diversity in

organizations.” In Barry M .
Staw and Robert M. Sutton
(eds.), Research in Organiza-
tional Behavior, 20: 77-140.
Stamford, CT: JAI Press.

763/ASQ, December 1999

You have printed the following article:

Why Differences Make a Difference: A Field Study of Diversity, Conflict, and Performance
in Workgroups
Karen A. Jehn; Gregory B. Northcraft; Margaret A. Neale
Administrative Science Quarterly, Vol. 44, No. 4. (Dec., 1999), pp. 741-763.
Stable URL:

This article references the following linked citations. If you are trying to access articles from an
off-campus location, you may be required to first logon via your library web site to access JSTOR. Please
visit your library’s website or contact a librarian to learn about options for remote access to JSTOR.


Distinguishing the Effects of Functional and Dysfunctional Conflict on Strategic Decision
Making: Resolving a Paradox for Top Management Teams
Allen C. Amason
The Academy of Management Journal, Vol. 39, No. 1. (Feb., 1996), pp. 123-148.
Stable URL:

Demography and Design: Predictors of New Product Team Performance
Deborah Gladstein Ancona; David F. Caldwell
Organization Science, Vol. 3, No. 3, Focused Issue: Management of Technology. (Aug., 1992), pp.
Stable URL:

Matching People and Organizations: Selection and Socialization in Public Accounting Firms
Jennifer A. Chatman
Administrative Science Quarterly, Vol. 36, No. 3. (Sep., 1991), pp. 459-484.
Stable URL:

– Page 1 of 3 –

Interpretive Barriers to Successful Product Innovation in Large Firms
Deborah Dougherty
Organization Science, Vol. 3, No. 2. (May, 1992), pp. 179-202.
Stable URL:

Consensus, Diversity, and Learning in Organizations
C. Marlene Fiol
Organization Science, Vol. 5, No. 3. (Aug., 1994), pp. 403-420.
Stable URL:

Groups in Context: A Model of Task Group Effectiveness
Deborah L. Gladstein
Administrative Science Quarterly, Vol. 29, No. 4. (Dec., 1984), pp. 499-517.
Stable URL:

Covariance Structure Modeling and Measurement Issues: A Note on “Interrelations among a
Channel Entity’s Power Sources”
Roy D. Howell
Journal of Marketing Research, Vol. 24, No. 1. (Feb., 1987), pp. 119-126.
Stable URL:

A Multimethod Examination of the Benefits and Detriments of Intragroup Conflict
Karen A. Jehn
Administrative Science Quarterly, Vol. 40, No. 2. (Jun., 1995), pp. 256-282.
Stable URL:

A Qualitative Analysis of Conflict Types and Dimensions in Organizational Groups
Karen A. Jehn
Administrative Science Quarterly, Vol. 42, No. 3. (Sep., 1997), pp. 530-557.
Stable URL:

– Page 2 of 3 –

Global Strategic Linkages and Industry Structure
Nitin Nohria; Carlos Garcia-Pont
Strategic Management Journal, Vol. 12, Special Issue: Global Strategy. (Summer, 1991), pp.
Stable URL:

Demographic Diversity, Conflict, and Work Group Outcomes: An Intervening Process Theory
Lisa Hope Pelled
Organization Science, Vol. 7, No. 6. (Nov. – Dec., 1996), pp. 615-631.
Stable URL:

A Social Information Processing Approach to Job Attitudes and Task Design
Gerald R. Salancik; Jeffrey Pfeffer
Administrative Science Quarterly, Vol. 23, No. 2. (Jun., 1978), pp. 224-253.
Stable URL:

Beyond Simple Demographic Effects: The Importance of Relational Demography in
Superior-Subordinate Dyads
Anne S. Tsui; Charles A. O’Reilly III
The Academy of Management Journal, Vol. 32, No. 2. (Jun., 1989), pp. 402-423.
Stable URL:

Determinants of Coordination Modes within Organizations
Andrew H. Van De Ven; Andre L. Delbecq; Richard Koenig, Jr.
American Sociological Review, Vol. 41, No. 2. (Apr., 1976), pp. 322-338.
Stable URL:

The Management of Interdepartmental Conflict: A Model and Review
Richard E. Walton; John M. Dutton
Administrative Science Quarterly, Vol. 14, No. 1. (Mar., 1969), pp. 73-84.
Stable URL:

– Page 3 of 3 –

View publication statsView publication stats

Needs help with similar assignment?

We are available 24x7 to deliver the best services and assignment ready within 3-4 hours? Order a custom-written, plagiarism-free paper

Get Answer Over WhatsApp Order Paper Now