Summary of
A
Meta-Analytic Test and Reformulation of Intergroup
Contact Theory
Thomas F. Pettigrew
and
Linda R. Tropp
����������� �Intergroup contact
theory maintains that a major means of reducing intergroup
prejudice is through contact between the groups under optimal conditions.� Expanding on the ideas of Williams (1947),
Gordon Allport (1954) formalized the theory, stating
that intergroup contact would lead to reduced intergroup prejudice if the contact situation embodies four
conditions: 1) equal status between the groups in the situation; 2) common
goals; 3) no competition between the groups; and 4) authority sanction for the
contact.
����������� Allport�s formulation of intergroup
contact theory has inspired extensive research over the past half century
(Pettigrew, 1998; Pettigrew & Tropp, 2000), and
it has received support across a variety of societies, situations, and
groups.� Contact studies have employed a
wide variety of research methods and procedures, including archival research,
field studies, laboratory experiments, and surveys. In addition to spanning
many disciplines, contact theory has been usefully applied to a host of
pressing social issues, ranging from the racial desegregation of schools
(Stephan & Rosenfield, 1978) and the resolution
of ethnopolitical conflicts (Chirot
& Seligman, 2001) to the educational mainstreaming of disabled children
(Harper & Wacker, 1985).
����������� Given
the diverse nature of research on intergroup contact,
a meta-analysis is needed to evaluate this work more thoroughly.� Yet, to our knowledge, no investigators have
conducted such an analysis on this vast and rich research literature.� This paper reports on such an effort,
utilizing 515 individual studies with 714 independent samples and 1,365
non-independent tests. Combined, 250,493 individuals from 38 nations
participated in the research.
Primary
Research Goals
����������� This analysis has been conducted
with several research goals in mind.
����������� Examining overall patterns of effects.� A first goal is to assess the overall effect
between�� intergroup
contact and prejudice, based on the population of empirical studies that
constitute the research literature.� Past
reviews of this extensive literature have been subjective, including only
selected subsets of cases.� Thus, a
meta-analysis is required to determine more conclusively the overall
relationship between intergroup contact and
prejudice.
����������� At the same
time, given the wide variety of contact studies, there is likely to be
substantial variability in outcomes.�
Therefore, we also move beyond the general analysis to consider the many
different types of variables that moderate the relationship between contact and
prejudice.� In so doing, we focus our
attention on variables that test for potential threats to our interpretations
of the obtained effects as well as variables that pertain to issues of broader
theoretical interest.
����������� Testing the effects of Allport�s
conditions.� We place special emphasis on testing the
utility of Allport�s proposed conditions, since they
have played a pivotal role in research and theory on intergroup
contact.� Specifically, we test whether Allport�s conditions contribute substantially to positive
contact outcomes, whether this facilitation occurs across a wide range of
circumstances, and whether such conditions are in fact necessary for positive
contact outcomes to occur.
����������� Exploring contact outcomes for different groups.� We also explore whether contact outcomes vary
across different groups and intergroup contexts.� Moreover, we examine contact outcomes for members
of different status groups who � depending on their group membership � may show
dramatically different responses to the same contact experience (Devine &
Vasquez, 1998; Robinson & Preston, 1976).
����������� We define intergroup contact as actual face-to-face interaction
between members of clearly defined groups.�
As such, we include only those studies that involve some degree of
direct interaction between members of discrete groups.� Additionally, we consider only those empirical studies in which: a) intergroup contact acts as a causal, independent variable
for intergroup prejudice; b) the prejudice dependent
variables are collected on individuals rather than simply as a total aggregate
outcome; and c) comparative data are available to evaluate any variability in
prejudice.� These comparisons make
use of pre-contact measures in within-group designs or control group data in
between-group designs.
An extensive five-year search of the psychological
(PsychLit, PsycInfo),
sociological (SocAbs, SocioFile),
political science (GOV), education (ERIC), dissertation (UMI
Dissertation Abstracts), and general research periodical (Current
Contents) databases yielded 515 studies, written between 1940 and the year
2000, that fit our inclusion criteria.��
Greater
intergroup contact is typically associated with less intergroup prejudice.� The mean estimate for the contact-prejudice
effect size, weighted for sample size, for both the 515 studies and 714 samples
is a Cohen�s d of -.47 (r
= -.23).� Additionally, 93% of the
studies in our analysis showed an inverse relationship between contact and
prejudice.� While the contact-prejudice
link could reflect a publishing bias that favors findings consistent with intergroup contact theory, two findings cast doubt on this
possibility.� First, it would require over
7,000 additional studies averaging no effect to erase the significance of the
negative contact-prejudice association at the 5% level of confidence.� Second, unpublished studies actually yield a
significantly larger mean effect than published work.
����������� The
more rigorous the research, the stronger the contact-prejudice relationship.� Experimental studies typically yield the
strongest effects, showing that contact can cause reductions in prejudice.� Studies with between-subject designs using
control groups that had no prior contact with the target group also reveal significantly
stronger effects than other studies.�
Moreover, when researchers measure the two key variables � contact and
prejudice � with experimental manipulations or reliable scales (a ≥ .70), they
attain significantly larger average effect sizes than when they employ measures
with low reliability.
����������� Intergroup contact
effects typically generalize beyond participants in the immediate contact situation
to the entire outgroup.� While
tests involving only the direct contact participants yield an average� d of -.42, those that measured
attitudes toward the entire outgroup provide a
comparable effect (d = -.41).�
The
effects of intergroup contact vary across different
types of outgroup targets and contact settings.� Reflecting contrasting types of contact as
well as diverse situations, mean effect sizes between contact and prejudice
vary widely across groups.� Contact
between ethnic and racial groups yielded robust effects, yet even larger
effects were found for contact between heterosexuals and gay men and lesbians.� Contact effects were also particularly strong
in laboratory, educational and recreational settings, while contact through tourism
and travel revealed weaker effects.
Intergroup contact effects differ among members of minority and
majority status groups.� ����While greater contact is generally
associated with less prejudice, these effects are markedly weaker for members
of minority status groups than for members of majority status groups,
suggesting that members of these groups may perceive and experience the same
contact situations in different ways.
����������� Global
indicators of Allport�s optimal conditions typically
lead to larger effect sizes between contact and prejudice.� Samples that experienced contact under
carefully structured situations designed to meet Allport�s
optimal conditions achieve a markedly higher mean effect size than other
samples.� Additionally, Allport�s conditions enhanced the positive effects of
contact across a wide range of research designs, and were especially effective
in contexts where participants had little or no choice in the contact.� Nonetheless, there remains considerable
variability in the degree to which Allport�s
conditions contribute to positive contact outcomes among members of different
status groups.� Thus, members of minority
and majority groups may subjectively perceive and experience intergroup contact in different ways, even when the contact
situation is objectively structured to maximize positive intergroup
outcomes.
����������� Together,
these findings suggest that while intergroup contact
typically promotes positive intergroup outcomes, we
must reconsider how intergroup contact is commonly
treated as a strategy for improving intergroup
relations. Most researchers have conceptualized conditions of intergroup contact as objective features of the contact
situation. But significant differences between the effects of contact for
members of minority and majority status groups indicates that these conditions
must be treated as elements that are perceived and experienced by people on
each side of the interaction, rather than being intrinsic to the contact
situation. Thus, as we encourage interactions across group boundaries, we must
focus greater attention on the subjective nature of intergroup
experiences, to gain a better understanding of those factors that may inhibit
the development of positive outcomes from intergroup
contact.
References
Allport, G. W. (1954). The nature of prejudice.
Chirot, D.,
& Seligman, M. E. P. (2001). Ethnopolitical
warfare: Causes, consequences, and
����������� possible solutions.
Devine, P. G., & Vasquez, K. A. (1998).�
The
rocky road to positive intergroup relations.� In
����������� J. L. Eberhardt & S. T. Fiske
(Eds.), Confronting racism: The problem and the response
����������� (pp.
234-262).�
Harper, D. C.,
& Wacker, D. P. (1985). Children�s attitudes toward disabled peers and
the
����������� effects of mainstreaming.�
Academic Psychology Bulletin, 7, 87-98.
Pettigrew, T. F. (1998). Intergroup contact theory.� Annual Review of Psychology, 49, 65-85.
Pettigrew, T. F., & Tropp, L. R. (2000).�
Does intergroup contact reduce prejudice?
Recent
����������� meta-analytic findings.�
In
����������� Social psychological perspectives (pp. 93-114).� Mahwah,
NJ: Erlbaum.
Robinson, J. W., &
����������� prejudice: A reexamination of the contact hypothesis.� Social Forces, 54, 911-924.
Stephan, W. G., & Rosenfield,
D. (1978).� Effects of desegregation on
race relations and self-esteem.�
����������� Journal of Educational Psychology, 70,
670-679.
Williams, R. M., Jr. (1947).�
The reduction of intergroup
tensions.
����������� Research Council.