PSY 301

Annu. Rev. Psychol. 1998. 49:65–85 Copyright © 1998 by Annual Reviews Inc. All rights reserved INTERGROUP CONTACT THEORY Thomas F. Pettigrew Department of Psychology, University of California, Santa Cruz, California 95064; e-mail: [email protected] KEY WORDS: affective ties, ingroup reappraisal, group categorization ABSTRACT Allport specified four conditions for optimal intergroup contact: equal group status within the situation, common goals, intergroup cooperation and authority support. Varied research supports the hypothesis, but four prob- lems remain. 1. A selection bias limits cross-sectional studies, since preju- diced people avoid intergroup contact. Yet research finds that the positive ef- fects of cross-group friendship are larger than those of the bias. 2. Writers overburden the hypothesis with facilitating, but not essential, conditions. 3.

The hypothesis fails to address process. The chapter proposes four processes:

learning about the outgroup, changed behavior, affective ties, and ingroup reappraisal. 4. The hypothesis does not specify how the effects generalize to other situations, the outgroup or uninvolved outgroups. Acting sequentially, three strategies enhance generalization—decategorization, salient categori- zation, and recategorization. Finally, both individual differences and societal norms shape intergroup contact effects. The chapter outlines a longitudinal intergroup contact theory. It distinguishes between essential and facilitating factors, and emphasizes different outcomes for different stages of contact. CONTENTS INTRODUCTION....................... ................................... 66 Allport’s Intergroup Contact Hypothesis..................................... 66 INITIAL EMPIRICAL EVIDENCE......... ................................... 67 RECENT EMPIRICAL EVIDENCE........ ................................... 68 PROBLEMS............................ ................................... 69 The Causal Sequence Problem.............................................. 69 Independent Variable Specification Problem..................................69 Unspecified Processes of Change Problem.................................... 70 The Generalization of Effects Problem.......................................70 FOUR PROCESSES OF CHANGE THROUGH INTERGROUP CONTACT........... 70 THREE TYPES OF GENERALIZATION.... ................................... 73 0066-4308/98/0201-0065$08.0065 A REFORMULATION OF INTERGROUP CONTACT THEORY................... 75 Individual Differences Shape Contact Effects..................................77 Societies Shape Contact Effects.............................................78 GENERAL EVALUATIONS OF CONTACT EFFECTS........................... 79 SUMMARY............................ ................................... 80 INTRODUCTION Social scientists began to theorize about intergroup contact after World War II (Watson 1947, Williams 1947). Allport’s (1954) hypothesis proved the most in- fluential by specifying the critical situational conditions for intergroup contact to reduce prejudice. His hypothesis has received extensive attention both for its rare theoretical status and policy importance (Pettigrew 1971). Oddly, for a disci- pline that focuses on face-to-face interaction, social psychology rarely decom- poses situations into their basic components. Allport’s attempt is a prominent exception. And it has proven useful in applied settings, such as in the distinc- tion between racial desegregation and integration in schools (Pettigrew 1975).

Allport’s Intergroup Contact Hypothesis Allport (1954) held that positive effects of intergroup contact occur only in situations marked by four key conditions: equal group status within the situa- tion; common goals; intergroup cooperation; and the support of authorities, law, or custom.

EQUAL STATUS Allport stressed equal group statuswithinthe situation. Most research supports this contention, although “equal status” is difficult to define and has been used in different ways (Cagle 1973, Riordan 1978). It is impor- tant that both groups expect and perceive equal status in the situation (Cohen & Lotan 1995, Cohen 1982, Riordan & Ruggiero 1980, Robinson & Preston 1976). Some writers emphasize equal group statuscoming intothe situation (Brewer & Kramer 1985). Thus, Jackman & Crane (1986) show negative ef- fects from contact with outgroup members of lower status. Yet Patchen (1982), in research on racially mixed high schools, found this to be less important than equal status within the situation. The meta-analytic results of Mullen et al (1992) clarify these disparities. They noted that ingroup bias increased with relative status in laboratory groups but decreased in field research with real groups. COMMON GOALS Prejudice reduction through contact requires an active, goal-oriented effort. Athletic teams furnish a prime example (Chu & Griffey 1985, Miracle 1981, Patchen 1982). In striving to win, interracial teams need each other to achieve their goal. Goal attainment, such as a winning season, furthers this process. 66 PETTIGREW INTERGROUP COOPERATION Attainment of common goals must be an inter- dependent effort without intergroup competition (Bettencourt et al 1992).

Sherif (1966) demonstrated this principle vividly in his Robbers’ Cave field study. Intergroup cooperation in schools provides the strongest evidence (Brewer & Miller 1984, Desforges et al 1991, Johnson et al 1984, Schofield 1989, Slavin 1983, Slavin & Madden 1979). Drawing on this thinking, Aron- son’s jigsaw classroom technique structures classrooms so that students strive cooperatively for common goals (Aronson & Patnoe 1997). This technique has led to positive results for a variety of children: Australians (Walker & Crogan 1997), Germans (Eppler & Huber 1990), Japanese (Araragi 1983), and Mexi- can Americans (Aronson & Gonzalez 1988). SUPPORT OF AUTHORITIES, LAW, OR CUSTOM The final condition concerns the contact’s auspices. With explicit social sanction, intergroup contact is more readily accepted and has more positive effects. Authority support establishes norms of acceptance. Field research underscores its importance in military (Landis et al 1984), business (Morrison & Herlihy 1992), and religious (Parker 1968) institutions. INITIAL EMPIRICAL EVIDENCE Allport (1954) derived his hypothesis from early field research. An Alabama study revealed negative effects when all four conditions were violated (Sims & Patrick 1936). White college students from the North increased on average in antiblack prejudice with each year spent in the South.

Other studies investigated optimal conditions. After desegregation of the Merchant Marine in 1948, interdependency developed on ships and in the mari- time union. The more voyages the white seamen took with blacks under these conditions, the more positive their racial attitudes became (Brophy 1946).

Similarly, white police in Philadelphia who had worked with black colleagues had fewer objections to black police joining their districts, teaming with a black partner, and taking orders from qualified black officers (Kephart 1957).

Studies of public housing provided robust evidence. Deutsch & Collins (1951) compared racially desegregated housing projects in New York City with similar but segregated projects in Newark. Sharp differences emerged.

Desegregated white housewives held their black neighbors in higher esteem and favored interracial housing more (75% to 25%). When asked to name black faults, they listed such personal issues as feelings of inferiority. Segre- gated white women voiced stereotypes such as “rowdy” and “dangerous.” Later public housing research extended these findings (Wilner et al 1955).

Favorable racial attitudes developed among only one third of the white tenants who just had casual greetings with their black neighbors. But half who entered INTERGROUP CONTACT THEORY 67 into conversations and three fourths who had multiple interactions developed positive racial views. Social norms are crucial. In the desegregated projects, whites expected approval from their neighbors for their friendly interracial be- havior. In the segregated projects, they feared social ostracism from other whites for such behavior. “Contact and perceived social climate tend to rein- force each other when their influence operates in the same direction, and to cancel each other out when their influence works in the opposite direction” (Wilner et al 1955, p. 106).

RECENT EMPIRICAL EVIDENCE Allport’s formulation continues to receive support across a variety of situa- tions, groups, and societies. Some research, conducted in situations that do not provide key conditions, uncovers negative effects. Poorly arranged entry of black workers into London’s public transportation, for example, led to hostility by white workers (Brooks 1975). Interracial housing research sometimes reveals few effects on racial attitudes (Bradburn et al 1971, Meer & Freedman 1966, Zuel & Humphrey 1971). These findings support Allport, although some writers mistakenly view their results as falsifying his hypothesis. The erroneous notion that the hypothesis holds that intergroup contact “will of itself produce better re- lations between…groups” still persists (McGarty & de la Haye 1997, p. 155).

Most studies report positive contact effects, even in situations lacking key conditions. Although it concentrates on school and housing situations (Ford 1986), the contact literature ranges from Chinese students in the United States (Chang 1973) and interracial workers in South Africa (Bornman & Mynhardt 1991) to German and Turkish school children (Wagner et al 1989) and Austra- lians (McKay & Pitman 1993) and Americans (Riordan 1987) getting to know Southeast Asian immigrants. It involves attitudes toward a wide range of tar- gets beyond ethnic groups—the elderly (Caspi 1984, Drew 1988), homosexu- als (Eskilson 1995, Herek & Capitanio 1996), the mentally ill (Desforges et al 1991), disabled persons (Anderson 1995), victims of AIDS (Werth & Lord 1992), and even computer programmers (McGinnis 1990). In addition, diverse research methods yield supporting results—field (Meer & Freedman 1966, Ohm 1988), archival (Fine 1979), survey (Jackman & Crane 1986; Pettigrew 1997a,b; Robinson 1980; Sigelman & Welch 1993), and laboratory (Cook 1978, 1984; Desforges et al 1991).

Research with African Americans is also supportive. Works (1961) found that both black wives and husbands in a desegregated housing project felt more positively about their white neighbors than those in a segregated project. With longitudinal data, Smith (1994) found contact meeting Allport’s conditions decreased prejudice among both black and white neighbors, though there were 68 PETTIGREW group differences in contact effects. In a national probability sample, interra- cial friendships proved to be a strong predictor of blacks’ racial attitudes (Elli- son & Powers 1994, Powers & Ellison 1995).

These varied investigations broaden the application of the hypothesis. They also raise the question: Why does intergroup contact usually have positive ef- fects even when the situation does not attain all of Allport’s conditions? In ad- dition, they typically do not address the basic problems with the original con- tact hypothesis—problems to which I now turn.

PROBLEMS The Causal Sequence Problem Selection bias limits the interpretation of many cross-sectional studies of con- tact. Instead of optimal contact reducing prejudice, the opposite causal se- quence could be operating. Prejudiced people may avoid contact with out- groups. Three methods overcome this limitation. 1. Find an intergroup situa- tion that severely limits choice to participate. In the Merchant Marine, police, and public-housing research, little choice was available to participants. 2. Sta- tistical methods borrowed from econometrics allow researchers to compare the reciprocal paths (optimal contact lowers prejudice and prejudice decreases contact) with cross-sectional data. Powers & Ellison (1995) use endogenous switching regression models. Later I review the use of a nonrecursive struc- tural equation model. 3. Although not without problems (Clogg 1986), longi- tudinal designs are best (Pettigrew 1996). Yet such research is rare in inter- group research; Sherif’s (1966) study is a famous exception. Indeed, the major findings of the Robbers’ Cave study would not have emerged without its longi- tudinal design (Pettigrew 1991a). The initial intergroup contact situations had little effect. Only after repeated treatments did the positive results cumulate.

Independent Variable Specification Problem Allport’s hypothesis risks being an open-ended laundry list of conditions— ever expandable and thus eluding falsification (Pettigrew 1986, Stephan 1987). Researchers keep advancing new situational factors for optimal con- tact. From Germany, Wagner & Machleit (1986) concluded that positive ef- fects require a common language, voluntary contact, and a prosperous econ- omy. From Israel, Ben-Ari & Amir (1986) held that the group’s initial views of one another cannot be too negative. From the United States, Cook (1978) in- sisted that stereotype disconfirmation is crucial.

This growing list of limiting conditions threatens to remove all interest from the hypothesis. Too many factors would exclude most intergroup situa- tions. The hypothesis would rarely predict positive results from contact, al- INTERGROUP CONTACT THEORY 69 though research typically finds positive results. The problem is that writers of- ten confusefacilitatingwithessentialconditions. Many factors suggested for optimal contact may not be essential but relate to the underlying mediating processes. This point leads to the third problem.

Unspecified Processes of Change Problem The original hypothesis says nothing about theprocessesby which contact changes attitudes and behavior. It predicts onlywhencontact will lead to posi- tive change, nothowandwhythe change occurs. A broader theory of inter- group contact requires an explicit specification of the processes involved, which I provide below.

The Generalization of Effects Problem Nor does the hypothesis specify how the effects generalize beyond the imme- diate situation. Such generalization is pivotal if intergroup contact is to have broad and lasting consequences. There are three distinct types of generaliza- tion:Situational—do the changes generalize across situations?Individual to group—do the changes generalize from the specific outgroup members with whom there is contact to the outgroup?To uninvolved outgroups—do the changes toward the outgroup generalize to other outgroups not involved in the contact? (For analyses of the generalization of different effects, see Brewer & Miller 1988 and Hewstone 1996.) Thus, a broader theory of intergroup contact also requires explicit predictions of how the contact effects will generalize, an- other task I undertake below.

FOUR PROCESSES OF CHANGE THROUGH INTERGROUP CONTACT Recent work suggests that four interrelated processes operate through contact and mediate attitude change: learning about the outgroup, changing behavior, generating affective ties, and ingroup reappraisal.

LEARNING ABOUT THE OUTGROUP Initial theory held this process to be the major way that intergroup contact has effects. When new learning corrects negative views of the outgroup, contact should reduce prejudice. Support for such a benign process is available, though plausible rival explanations remain.

Consider Jeffries & Ransford’s (1969) findings on middle-class white reac- tions to the Watts race riot in Los Angeles. Those who had prior interracial contact were significantly less fearful of blacks, less punitive, and less likely to view the riot as caused by outside agitators.

Yet cognitive research has uncovered a host of mechanisms that limit learn- ing material that counters our attitudes and stereotypes. Writing from this per- 70 PETTIGREW spective, Rothbart & John (1985) conclude that disconfirming evidence alters stereotypes only if (a) the outgroup’s behavior is starkly inconsistent with their stereotype and strongly associated with their label, (b) occurs often and in many situations, and (c) the outgroup members are seen as typical. These re- strictions eliminate most intergroup contact situations.

Nonetheless, new information about an outgroup can improve attitudes.

Stephan & Stephan (1984) found that contact allowed Anglo students to learn more about Chicano culture that in turn led to more positive attitudes toward Chicano classmates. “Ignorance,” they assert, “promotes prejudice…” (Stephan & Stephan 1984, p. 238). Other studies with the cultural assimilator technique of Triandis (1994) provide further evidence that learning about an outgroup can improve intergroup attitudes and stereotypes (Gardiner 1972, Weldon et al 1975).

Still, the dominant consensus of cognitive analyses denies the likelihood of positive effects from most contact situations. Yet the research literature sug- gests that positive effects are more common than either the contact hypothesis or cognitive analyses predict. Why the contradiction? The basic reason is that learning about the outgroup is only one of several processes involved. Cogni- tive analyses are not so much wrong as they are incomplete. Other processes are also involved.

CHANGING BEHAVIOR Optimal intergroup contact acts as a benign form of be- havior modification. Behavior change is often the precursor of attitude change.

New situations require conforming to new expectations. If these expectations include acceptance of outgroup members, this behavior has the potential to produce attitude change. We can resolve our dissonance between old preju- dices and new behavior by revising our attitudes (Aronson & Patnoe 1997).

This behavioral process also benefits from repeated contact, preferably in var- ied settings (Jackman & Crane 1986). Repetition makes intergroup encounters comfortable and “right.” Repetition itself leads to liking (Zajonc 1968). Ap- propriate rewards for the new behavior enhances the positive effects further. GENERATING AFFECTIVE TIES Emotion is critical in intergroup contact. Anxi- ety is common in initial encounters between groups, and it can spark negative reactions (Islam & Hewstone 1993; Stephan 1992; Stephan & Stephan 1985, 1989, 1992; Wilder 1993a,b). Such anxious, negative encounters can occur even without intergroup prejudice (Devine et al 1996). Continued contact gen- erally reduces anxiety, though bad experiences can increase it.

Positive emotions aroused by optimal contact also can mediate intergroup contact effects. Empathy plays a role here. Reich & Purbhoo (1975) found that school contact improved cross-group role-taking ability among both majority and minority Canadian students. And empathy for a stigmatized outgroup INTERGROUP CONTACT THEORY 71 member—a young woman with AIDS, a homeless man or a convicted mur- derer—can improve attitudes toward the whole outgroup (Batson et al 1997).

Positive emotions aroused by intergroup friendship also can be pivotal. The Oliners (1988) found that non-Jews who risked their lives to save Jews during World War II reported more close friendships as children with other groups.

Similarly, Rippl (1995) found friendship to be decisive in shaping contact ef- fects between West and East Germans. These findings support earlier claims by Amir (1976) concerning the importance of intimacy in intergroup contact.

The most extensive data on intergroup friendship derive from 1988 surveys in Western Europe (Pettigrew 1997a,b; Pettigrew & Meertens 1995). (For sur- vey details, see Reif & Melish 1991; for a different, though consistent, analysis of these data, see Hamberger & Hewstone 1997). Over 3800 majority group re- spondents in seven probability samples of France, Great Britain, the Nether- lands, and West Germany were asked their attitudes toward major minority groups in their country and whether they had friends of another nationality, race, culture, religion, or social class. In all samples, Europeans with outgroup friends scored significantly lower on five prejudice measures even after con- trolling for seven variables. The largest effect occurred for a two-item measure of affective prejudice. Those with intergroup friends significantly more often reported having felt sympathy and admiration for the outgroup. Few studies in the contact literature have used affective dependent variables. When they have, similar results emerge (Wright et al 1997).

Figure 1 compares the paths between intergroup friends and affective preju- dice. Note that living in an intergroup neighborhood makes it more likely that a European will have an outgroup friend (+.356). There is no direct relationship, however, between mixed neighborhoods and affective prejudice. This allows a test of the paths between friendship and affective prejudice (Bollen 1989, Heist 1975). As in other research (Herek & Capitanio 1996), the prejudiced avoid intergroup contact (-.137). But the path from friendship to reduced af- fective prejudice is significantly stronger (-.210), a finding consistent with that of Powers & Ellison (1995). In short, like prejudice, contact involves both cognition and affect.

INGROUP REAPPRAISAL Optimal intergroup contact provides insight about ingroups as well as outgroups. Ingroup norms and customs turn out not to be the only ways to manage the social world. This new perspective can reshape your view of your ingroup and lead to a less provincial view of outgroups in general (“deprovincialization”). In the European surveys, outgroup friendship related to significantly less “pride” in nationality even after education, age, and political conservatism are controlled (Pettigrew 1997c). (A comparable 72 PETTIGREW analysis to Figure 1 reveals similar results with a significant path from out- group friends to less national pride.) Part of this process involves having less contact with the ingroup as a result of more contact with the outgroup. Wilder & Thompson (1980) covaried con- tact with the ingroup and outgroup in a laboratory study. While it had no im- pact on ingroup ratings, less ingroup contact led to less bias toward the out- group. This finding is consistent with meta-analytic results that show ingroup bias is positively related with ingroup salience (Mullen et al 1992).

THREE TYPES OF GENERALIZATION GENERALIZATION ACROSS SITUATIONS World War II research showed major improvements in white attitudes toward black soldiers after combat together (Stouffer et al 1949). Yet whites continued to favor racially separate post ex- changes (military stores). Other early studies also demonstrated limited gener- alization across situations (Biesanz & Smith 1951, Deutsch & Collins 1951, Harding & Hogrefe 1952, Minard 1952, Reitzes 1953, Saenger & Gilbert 1950).

Recall that the first cooperative encounters of the Robbers’ Cave study did not generalize either. Only the cumulative effect of repeated optimal situations INTERGROUP CONTACT THEORY 73 Figure 1Friends—prejudice model altered the attitudes of the rival groups. Similarly, only after the US Army of- fered many types of optimal interracial situations could it conclude that its pro- gram of racial desegregation “works” (US Department of Defense 1955, Mos- cos & Butler 1996).

GENERALIZATION FROM THE OUTGROUP INDIVIDUAL TO THE OUTGROUP Brown & Turner (1981) questioned the generalization of effects betweenin- terpersonalandintergroupphenomena. They doubted whether positive ef- fects from getting to know an outgroup member (interpersonal) could affect at- titudes about the outgroup (intergroup).

Following this reasoning, Hewstone & Brown (1986) theorized that contact effects generalize to the outgroup only when group membership is salient. When group saliency is low, the situation is interpersonal and no intergroup effects should result. Only when the interactants view one another as group representa- tives does the contact become an intergroup event. Research that supports this salient categorization strategyshows stereotype change generalizes best to the intergroup level when the individuals involved are typical group members (Johnston & Hewstone 1992, Vivian et al 1995, Weber & Crocker 1983, Wilder 1984). As typical members, their group memberships were more salient.

This poses a problem. Typical members of real groups are different in many ways, but people with similar interests and status seek each other out—the similarity principle (Byrne 1971). People from different groups who have con- tact, then, are more likely to share similar interests and values. And outgroup members with similar interests to the ingroup often will not typify their group or make group membership salient. Hence, Brewer & Miller (1984) advocate a decategorization strategy.The opposite of salient categorization, it holds that intergroup contact is most effective when group saliency is low.

So, given real differences between groups, those most likely to have inter- group contact are atypical of their groups. Yet contact effects generalize best when the participants are typical group members. Thus, people most likely to engage in intergroup contact are the least likely to evoke changes that general- ize to their groups.

How, then, can effective contact take place? First, the similarity principle is not the whole story. Durkheim (1960) pointed out that similarity (mechanical solidarity) is only one form of social bond. Society could not exist without bonds across reciprocal roles—parents and children, clerks and customers. So differences (organic solidarity) also are important for social bonds. These dif- ferences, widespread in modern society, guarantee that contact takes place be- tween dissimilar people. Further, under some conditions, optimal contact leads to positive changes that generalize even when atypical members are involved (Hamburger 1994, Werth & Lord 1992).

74 PETTIGREW Second, the time sequence is crucial. Conflicting as they seem, both strate- gies are possible if they occur sequentially. Diminished saliency of group cate- gories can be important when intergroup contact is initiated. Once established, salient group categorization is required for the effects to generalize to the inter- group level (Van Oudenhoven et al 1996).

In time, the Common Ingroup Identity Model emphasizes thatrecategori- zationbecomes possible (Anastasio et al 1997; Gaertner et al 1993, 1994). Af- ter extended contact, people can begin to think of themselves in a larger group perspective. Recategorization adopts an inclusive category that highlights similarities among the interactants and obscures the “we” and “they” boundary (Perdue et al 1990). Note, however, that recategorization is the final state many interacting groups never reach. The progression through these three stages of categorization is not automatic, and recategorization into a single group often will not be attained.

Wright and his colleagues propose a further form of generalization involv- ing friendship. Using both questionnaire and experimental methods, they show that even knowledge of an ingroup member’s friendship with an outgroup member relates to more positive attitudes toward the outgroup (Wright et al 1997). Such an “extended contact effect” does not require intergroup friend- ship for the perceiver. GENERALIZATION FROM THE IMMEDIATE OUTGROUP TO OTHER OUTGROUPS This higher-order form of generalization presupposes the other forms and is seldom studied because many regard it as highly unlikely (but see Reich & Purbhoo 1975, Weigert 1976). Nonetheless, the European surveys show that such generalization is possible. Having an ingroup friend related to greater ac- ceptance of minorities of many types (Pettigrew 1997a,b). The 3800 respon- dents rated how favorable they were toward eight outgroups, many of whom were not in their country. In all samples, significant relationships emerged even after seven variables were controlled. For each outgroup, those with in- tergroup friends were significantly more positive in their views. And a test like that of Figure 1 showed that the path from friendship to reduced prejudice is significantly stronger than the prejudice to less friendship path. These findings challenge Rose’s (1981) assertion that intimate relationships cannot generalize to different persons and groups.

A REFORMULATION OF INTERGROUP CONTACT THEORY These considerations provide direction for a reformulation of Allport’s hy- pothesis. At least four processes, not one, are involved, and these processes may well overlap and interact in complex ways. Intergroup friendship is potent INTERGROUP CONTACT THEORY 75 because it potentially invokes all four mediating processes. This suggests that constructive contact relates more closely to long-term close relationships than to initial acquaintanceship—a dramatic shift for the intergroup contact re- search literature.

Optimal intergroup contact requires time for cross-group friendships to de- velop. Past work has focused chiefly on short-term intergroup contact—the very condition that Sherif’s (1966) Robbers’ Cave field experiment found mini- mally effective. Once we adopt a long-term perspective that allows cross-group friendship to develop and the full decategorization, salient categorization, and recategorization sequence to unfold, we can expect striking results. Such a re- vised perspective explains why extended intergroup contact often has more positive results than either the contact hypothesis or cognitive analyses predict.

The power of cross-group friendship to reduce prejudice and generalize to other outgroups demands a fifth condition for the contact hypothesis:The con- tact situation must provide the participants with the opportunity to become friends.Such opportunity implies close interaction that would make self- disclosure and other friendship-developing mechanisms possible. It also im- plies the potential for extensive and repeated contact in a variety of social con- texts. Allport (1954) alluded to this point when he favored intimate to trivial contact; Cook (1962) called it “acquaintance potential.” These European re- sults suggest that “friendship potential” is an essential, not merely facilitating, condition for positive intergroup contact effects that generalize. Further, they suggest that Allport’s conditions are important in part because they provide the setting that encourages intergroup friendship.

Instead of a list of conditions, then, I propose a longitudinal model as sche- matically outlined in Figure 2. Note first that this version of intergroup contact theory involves the meso-level of analysis. Yet it is placed within the micro- and macrolevel contexts of (B) the participants’ experiences and characteris- tics as well as the larger societal setting of the situation. The basic features of this reformulated version consist of (A) the essential and facilitating situational factors and (C, D, E) the time dimension. Each of these features involves de- tails not shown in Figure 2.

(A) From the previous discussion, the theory posits Allport’s four conditions and friendship potential as essential situational factors for positive intergroup outcomes—less negative stereotyping, prejudice, and discrimination. An array of additional factors, such as equivalent group status outside the situation, act as facilitating factors for such effects. Complicating the picture further, such factors might prove important at different stages of the intergroup contact.

(C, D, E) in Figure 2 designate only three stages. As diagrammed, this time dimension is obviously oversimplified. The stages will overlap, and at any point the groups can break off contact. This heuristic, however, following the 76 PETTIGREW three research traditions on group categorization, allows the theory to predict different outcomes at different stages.

We need research on several points. While the recategorization of the unified group, if attained, appears to yield the maximum in prejudice reduction, we have little data on how this effect generalizes. Further, we must understand how the four contact processes of change are activated and become important at each stage. The thrust of the time dimension is to underline the need for longitudinal research on intergroup contact. Though rare in this literature, this reformulated theory holds that such research designs are necessary for further progress.

Conceptualizing intergroup contact as a meso-level, situational phenome- non links it to both the microindividual and macrosocietal levels of analysis (Pettigrew 1991b, 1996). Research has shown how important these cross-level links can be.

Individual Differences Shape Contact Effects Prior attitudes and experiences influence whether people seek or avoid inter- group contact, and what the effects of the contact will be. Figure 1 showed that prejudiced Europeans are less likely to have outgroup friends. Other character- istics also deter contact. Value differences shaped differential readiness for in- INTERGROUP CONTACT THEORY 77 Figure 2Reformulated contact theory. tergroup contact among both Israeli Arabs and Jews (Sagiv & Schwartz 1995).

Cook (1984) found large individual differences in the effects of even optimal contact; 40% of his experimental subjects, compared with 12% of controls, evinced sweeping attitude changes. Yet other experimental subjects averaged little change.

High intergroup anxiety and threat also can impede both contact and its positive effects (Islam & Hewstone 1993; Stephan 1992; Stephan & Stephan 1985, 1989, 1992, 1996; Wilder 1993a,b; Wilder & Shapiro 1989). Such emo- tions often derive from no prior experience with the outgroup. Black and white high school students who had the most favorable earlier interracial experience were more positive toward the other race (Patchen 1982). Thus, intergroup contact and its effects are cumulative—we live what we learn. Braddock (1989) found that black graduates of segregated high schools were signifi- cantly less likely later to work with whites. Since jobs with white co-workers have better pay and promotions, this result helps to explain why black gradu- ates of interracial schools do better in later life than those from segregated schools (Braddock 1989, Braddock et al 1984, Braddock & McPartland 1987).

Societies Shape Contact Effects Situations are embedded in social institutions and societies. Thus, institu- tional and societal norms structure the form and effects of contact situations (Kinloch 1981, 1991). Consider intergroup strife in Northern Ireland and Que- bec. These societal contexts severely limit all forms of intergroup contact.

Moreover, they render the contact that does occur less than optimal. Implicit in Allport’s equal-status condition is equivalent group power in the situation.

This is difficult to achieve when a struggle over power fuels the larger inter- group conflict.

The meager equal-status contact between groups that takes place in such societies is typically subversive in character. In Northern Ireland, neighbor- ing Catholic and Protestant farmers cooperate in their agricultural pursuits but remain apart in other activities (Harris 1972, Kirk 1993). Even conversations are circumscribed. In both Quebec and Northern Ireland, intergroup interac- tion focuses on local issues and avoids divisive group concerns (Taylor et al 1986, Trew 1986). It is at best constrained discussion, not the easy banter of friends.

Russell (1961) showed how societal norms of discrimination poison inter- group contact. She tested a racially mixed neighborhood at the height of South Africa’sApartheidpolicy of intense racial segregation. This rare area had 50% whites, 20% Coloureds, and 30% Indians. Even here, modest improvements emerged in white attitudes toward their neighbors of color. Yet the larger so- cial context constrained these effects. The improved attitudes did not general- 78 PETTIGREW ize to Coloureds and Indians as groups, and whites were defensive about their interracial contact. Some avoided it, and the exchange that took place was not reciprocal. Whites received neighborly aid and entered nonwhite homes far more than the reverse. Many whites rationalized their interracial behavior with the exploitative nature of the relationship. All were aware that the then stern South African norms punished equal-status interracial contact. Such norms erode true neighborliness.

Alternatively, when a society embraces intergroup harmony, equal-status contact between groups is no longer subversive. Normative support makes at- tainment of other optimal conditions far easier.

GENERAL EVALUATIONS OF CONTACT EFFECTS Consider this reformulation in light of two efforts to test the chief contentions of contact theory, one with surveys and the other experimental.

In meta-analytic fashion, the European surveys tested the link between in- tergroup friendship and five different measures of prejudice across seven sam- ples (Pettigrew 1997a). Weighting for sample size, the average correlations ranged between-.14 and-.32 without controls. With controls for seven vari- ables and weighting for sample size, the average correlations ranged between -.10 and-.25. These results are consistent with the preliminary findings of a meta-analysis of the full contact literature now under way (Pettigrew et al 1998).

Cook (1984) provided the most extensive laboratory test of the intergroup contact hypothesis. He set up an ideal interracial situation featuring equal status, a stereotype-disconfirming minority member, an interdependent task, task success, high friendship potential, and equalitarian social norms. Highly prejudiced white subjects worked with two partners who were confederates, one black and one white. They operated an imaginary railroad system over 40 sessions, sharing successes and failures but eventually winning bonus money.

To enhance generalization, the black confederate told of a personal experience with racial discrimination after rapport had developed. Later, the white con- federate disapproved of the discriminatory practice.

Cook recorded major changes. Subjects reported highly favorable opinions of their black partner, and these views generalized. In another context, the sub- jects evinced on average more positive racial views toward African Americans than controls. Observe the time sequence. First, the prejudiced subjects found the stereotype-disconfirming black as similar and grew to like her in the opti- mal situation. Next the black confederate made race salient by telling of past discrimination. Then the white confederate emphasized the link between their friend’s bad treatment and racial discrimination. INTERGROUP CONTACT THEORY 79 SUMMARY Allport’s hypothesis specified four conditions for optimal contact. The situa- tion must allow equal group status within the situation, common goals, inter- group cooperation, and authority support. Recent research adds another: The contact situation must have “friendship potential.” Varied research supports the hypothesis—from field and archival studies to national surveys and laboratory experiments. However, cross-sectional studies suffer from a selection bias. Prejudiced people avoid intergroup contact, so the causal link between contact and prejudice is two-way. Yet those studies that have tested both paths find that the positive effects of cross-group friendship are larger than those of the bias.

Three additional problems limit the contact hypothesis. 1. Writers have overburdened the hypothesis with too many facilitating, but not essential, con- ditions. 2. The hypothesis does not address process. It predicts only when posi- tive contact effects will occur, not how and why. The chapter details four inter- related processes underlying contact effects: learning about the outgroup, changing behavior, generating affective ties, and ingroup reappraisal. Inter- group friendship has strong positive effects, because it potentially entails all four processes.

3. The hypothesis does not specify how the effects generalize to other situa- tions, the entire outgroup or uninvolved outgroups. Many effects do not gener- alize beyond the immediate contact situation and participants. There are three strategies to enhance generalization: decategorization, salient group categori- zation, and recategorization. Thinking of these strategies acting sequentially removes the apparent contradiction between them. Since similarity attracts, initial stages of intergroup contact benefit from not making group membership salient. Later, as anxiety and threat subside, group membership must become salient to maximize the generalization of positive effects beyond the immedi- ate situation. Then recategorization becomes possible if the participants adopt an all-encompassing group identification.

Both individual differences and societal norms shape intergroup contact ef- fects. The deeply prejudiced both avoid intergroup contact and resist positive effects from it. Societies suffering intergroup conflict both restrict and under- cut intergroup contact. From these considerations, the chapter advances a lon- gitudinal reformulation of the intergroup contact hypothesis. Within the con- texts of the participants’ characteristics and the situation’s societal setting, the chapter outlines a meso-level theory with two key features. It distinguishes be- tween essential and facilitating situational factors. And it emphasizes the time dimension with different outcomes predicted for different stages of intergroup contact.

80 PETTIGREW ACKNOWLEDGMENTS I wish to thank my research colleagues over the past decade, James Jackson of the University of Michigan and Roel W. Meertens of the University of Amster- dam, for their invaluable collaboration with the European survey research. I also greatly appreciate the helpful comments on earlier drafts of this chapter by John Darley, Susan Fiske, Miles Hewstone, Marylee Taylor, and Linda Tropp. Visit theAnnual Reviews home pageat http://www.AnnualReviews.org. INTERGROUP CONTACT THEORY 81 Allport GW. 1954.The Nature of Prejudice.

Reading, MA: Addison-Wesley. 537 pp.

Amir Y. 1976. The role of intergroup contact in change of prejudice and race relations.

InTowards the Elimination of Racism,ed.

PA Katz, pp. 245–80. New York: Perga- mon. 444 pp.

Anastasio P, Bachman B, Gaertner S, Dovidio J. 1997. Categorization, recategorization and common ingroup identity. See Spears et al 1997, pp. 236–56 Anderson LS. 1995.Outdoor adventure rec- reation and social integration: a social- psychological perspective.PhD thesis.

Univ. Minn., Minneapolis, MN. 271 pp.

Araragi C. 1983. The effect of the jigsaw learning method on children’s academic performance and learning attitude. Jpn. J.

Educ. Psychol.31:102–12 Aronson E, Gonzalez A. 1988. Desegregation, jigsaw, and the Mexican-American expe- rience. See Katz & Taylor 1988, pp.

301–14 Aronson E, Patnoe S. 1997.The Jigsaw Class- room.New York: Longman. 150 pp. 2nd ed.

Batson CD, Polycarpou MP, Harmon-Jones E, Imhoff HJ, Mitchener EC, et al. 1997. Em- pathy and attitudes: can feeling for a mem- ber of a stigmatized group improve feel- ings toward the group?J. Pers. Soc. Psy- chol.72:105–18 Ben-Ari R, Amir Y. 1986. Contact between Arab and Jewish youth in Israel: reality and potential. See Hewstone & Brown 1986, pp. 45–58 Bettencourt BA, Brewer MB, Rogers-Croak MR, Miller N. 1992. Cooperation and the reduction of intergroup bias: the role of re-ward structure and social orientation.J.

Exp. Soc. Psychol.28:301–19 Biesanz J, Smith LM. 1951. Race relations of Panama and the Canal Zone.Am. J. Sociol.

57:7–14 Bollen KA. 1989.Structural Equations with Latent Variables.New York: Wiley. 514 pp.

Bornman E, Mynhardt JC. 1991. Social identi- fication and intergroup contact in South Africa with specific reference to the work situation.Genet. Soc. Gen. Psychol.

Monogr.117:437–62 Bradburn N, Sudman S, Gockel GL. 1971.

Side by Side: Integrated Neighborhoods in America.Chicago: Quadrangle Books.

209 pp.

Braddock JH Jr. 1989. Social psychological processes that perpetuate racial segrega- tion: the relationship between school and employment desegregation.J. Black Stud.

19:267–89 Braddock JH Jr, Crain RL, McPartland JM.

1984. A long-term view of racial desegre- gation: some recent studies of graduates as adults.Phi Delta Kappan66:259–64 Braddock JH Jr, McPartland JM. 1987. How minorities continue to be excluded from equal employment opportunities: research on labor market and institutional barriers.

J. Soc. Issues43:5–39 Brewer MB, Kramer RM. 1985. The psychol- ogy of intergroup attitudes and behavior.

Annu. Rev. Psychol.36:219–43 Brewer MB, Miller N. 1984. Beyond the con- tact hypothesis: theoretical perspectives on desegregation. See Miller & Brewer 1984, pp. 281–302 Brewer MB, Miller N. 1988. Contact and co- Literature Cited 82 PETTIGREW operation: when do they work? See Katz & Taylor 1988, pp. 315–26 Brooks D. 1975.Race and Labour in London Transport.London: Oxford Univ. Press.

389 pp.

Brophy IN. 1946. The luxury of anti-Negro prejudice.Public Opin. Q.9:456–66 Brown RJ, Turner JC. 1981. Interpersonal and intergroup behavior. InIntergroup Behavior,ed. J Turner, H Giles, pp.

33–65. Chicago, IL: Univ. Chicago Press.

277 pp.

Byrne D. 1971.The Attraction Paradigm.

New York: Academic. 474 pp.

Cagle LT. 1973. Interracial housing: a reas- sessment of the equal-status contact hy- pothesis.Sociol. Soc. Res.57:342–55 Caspi A. 1984. Contact hypothesis and inter- age attitudes: a field study of cross-age contactSoc. Psychol. Q.47:74–80 Chang H. 1973. Attitudes of Chinese students in the United States.Sociol. Soc. Res.58:

66–77 Chu D, Griffey D. 1985. The contact theory of racial integration: the case of sport.Sociol.

Sport J.2:323–33 Clogg CC. 1986. Invoked by RATE.Am. J.

Sociol.92:696–706 Cohen EG. 1982. Expectation states and inter- racial interaction in school settings.Annu.

Rev. Sociol.8:209–35 Cohen EG, Lotan RA. 1995. Producing equal- status interaction in the heterogeneous classroom.Am. Educ. Res. J.32:99–120 Cook SW. 1962. The systematic analysis of socially significant events: a strategy for social research.J. Soc. Issues18:66–84 Cook SW. 1978. Interpersonal and attitudinal outcomes in cooperating interracial groups.J. Res. Dev. Educ.12:97–113 Cook SW. 1984. Cooperative interaction in multiethnic contexts. See Miller & Brewer 1984, pp. 155–85 Desforges DM, Lord CG, Ramsey SL, Mason JA, Van Leeuwen MD, et al. 1991. Effects of structured cooperative contact on changing negative attitudes toward stig- matized social groups.J. Pers. Soc. Psy- chol.60:531–44 Deutsch M, Collins M. 1951.Interracial Housing: A Psychological Evaluation of a Social Experiment.Minneapolis: Univ.

Minn. Press. 173 pp.

Devine PG, Evett SR, Vasquez-Suson KA.

1996. Exploring the interpersonal dynam- ics of intergroup contact. InHandbook of Motivation and Cognition: The Interper- sonal Context,ed. RM Sorrentino, ET Hig- gins, 3:423–64. New York: Guilford. 646 pp.Drew B. 1988.Intergenerational contact in the workplace: an anthropological study of relationships in the secondary labor market.PhD thesis. Rutgers Univ., New Brunswick, NJ. 285 pp.

Durkheim E. 1960. (1893).The Division of La- bor. Glencoe, IL: Free Press. 439 pp.

Ellison CG, Powers DA. 1994. The contact hy- pothesis and racial attitudes among Black Americans.Soc. Sci. Q.75:385–400 Epler R, Huber GL. 1990. Wissenserwerb im Team: Empirische Untersuchung von Ef- fekten des Gruppen-Puzzles.Psychol. Er- zieh. Unterr.37:172–78 Eskilson A. 1995.Trends in homophobia and gender attitudes: 1987–1993.Presented at Annu. Meet. Am. Sociol. Assoc., 90th, Washington, DC Fine GA. 1979. The Pinkston settlement: an historical and social psychological investi- gation of the contact hypothesis.Phylon 40:229–42 Ford WS. 1986. Favorable intergroup contact may not reduce prejudice: inconclusive journal evidence 1960–1984.Sociol. Soc.

Res.70:256–58 Gaertner SL, Dovidio JF, Anastasio PA, Bach- man BA, Rust MC. 1993. The common in- group identity model: recategorization and the reduction of intergroup bias.Eur. Rev.

Soc. Psychol.4:1–26 Gaertner SL, Rust MC, Dovidio JF, Bachman BA, Anastasio PA. 1994. The contact hy- pothesis: the role of a common ingroup identity on reducing intergroup bias.Small Group Res.25:224–49 Gardiner GS. 1972. Complexity training and prejudice reduction.J. Appl. Soc. Psychol.

2:326–42 Hamberger J, Hewstone M. 1997. Inter-ethnic contact as a predictor of prejudice: tests of a model in four West European nations.Br.

J. Soc. Psychol.In press Hamburger Y. 1994. The contact hypothesis reconsidered: effects of the atypical out- group member on the outgroup stereotype.

Basic Appl. Soc. Psychol.15:339–58 Harding J, Hogrefe R. 1952. Attitudes of white department store employees toward Negro co-workers.J. Soc. Issues8:18–28 Harris R. 1972.Prejudice and Tolerance in Ulster.Manchester, UK: Manchester Univ. Press. 234 pp.

Heist DR. 1975.Causal Analysis.New York:

Wiley. 301 pp.

Herek GM, Capitanio JP. 1996. “Some of my best friends”: intergroup contact, conceal- able stigma, and heterosexuals’ attitudes toward gay men and lesbians.Pers. Soc.

Psychol. Bull.22:412–24 INTERGROUP CONTACT THEORY 83 Hewstone M. 1996. Contact and categoriza- tion: social-psychological interventions to change intergroup relations. InFounda- tions of Stereotypes and Stereotyping,ed.

CN Macrae, C Stagnor, M Hewstone, pp.

323–68. New York: Guilford. 462 pp.

Hewstone M, Brown R, eds. 1986.Contact and Conflict in Intergroup Encounters.

Oxford: Blackwell. 231 pp.

Islam MR, Hewstone M. 1993. Dimensions of contact as predictors of intergroup anxiety, perceived out-group variability, and out- group attitude: an integrative model.Pers.

Soc. Psychol. Bull.19:700–10 Jackman MR, Crane M. 1986. “Some of my best friends are black…”: interracial friendship and whites’ racial attitudes.

Public Opin. Q.50:459–86 Jeffries V, Ransford HE. 1969. Interracial so- cial contact and middle-class white reac- tion to the Watts riot.Soc. Probl.16:

312–24 Johnson DW, Johnson RT, Maruyama G.

1984. Goal interdependence and inter- personal-personal attraction in heteroge- neous classrooms: a meta-analysis. See Miller & Brewer 1984, pp. 187–212 Johnston L, Hewstone M. 1992. Cognitive models of stereotype change: III. Subtyp- ing and the perceived typicality of discon- firming group members.J. Exp. Soc. Psy- chol.28:360–86 Katz PA, Taylor DA, eds. 1988.Eliminating Racism: Profiles in Controversy.New York: Plenum. 380 pp.

Kephart WM. 1957.Racial Factors and Ur- ban Law Enforcement.Philadelphia: Univ.

Pa. Press. 209 pp.

Kinloch GC. 1981. Comparative race and eth- nic relations.Int. J. Comp. Sociol.22:

257–71 Kinloch GC. 1991. Inequality, repression, dis- crimination and violence: a comparative study.Int. J. Cont. Sociol.28:85–98 Kirk T. 1993.The polarisation of Protestants and Roman Catholics in rural Northern Ireland: a case study of Glenravel Ward, County Antrim 1956 to 1988.PhD thesis.

Queen’s Univ., Belfast, North. Ireland.

541 pp.

Landis D, Hope RO, Day HR. 1984. Training for desegregation in the military. See Miller & Brewer 1984, pp. 257–78 McGarty C, de la Haye AM. 1997. Stereotype formation: beyond illusory correlation.

See Spears et al 1997, pp. 144–70 McGinnis SP. 1990.Descriptive and evalua- tive components of stereotypes of computer programmers and their determinants.PhD thesis. New York: City Univ. NY. 248 pp.McKay S, Pitman J. 1993. Determinants of Anglo-Australian stereotypes of the Viet- namese in Australia.Aust. J. Psychol.45:

17–23 Meer B, Freedman E. 1966. The impact of Ne- gro neighbors on white house owners.Soc.

Forces45:11–19 Miller N, Brewer MB, eds. 1984.Groups in Contact: The Psychology of Desegrega- tion.Orlando, FL: Academic. 316 pp.

Minard RD. 1952. Race relations in the Poca- hontas coal field.J. Soc. Issues8:29–44 Miracle AW. 1981. Factors affecting interra- cial cooperation: a case study of a high school football team.Hum. Organ.40:

150–54 Morrison EW, Herlihy JM. 1992. Becoming the best place to work: Managing diversity at American Express Travel related serv- ices. InDiversity in the Workplace,ed. SE Jackson, pp. 203–26. New York: Guilford.

356 pp.

Moscos CC, Butler JS. 1996.All That We Can Be.New York: Basic Books. 198 pp.

Mullen B, Brown R, Smith C. 1992. Ingroup bias as a function of salience, relevance, and status: an integration.Eur. J. Soc. Psy- chol.22:103–22 Ohm RM. 1988.Constructing and recon- structing social distance attitudes.PhD thesis. Ariz. State Univ., Tempe. 316 pp.

Oliner SP, Oliner PM. 1988.The Altruistic Personality: Rescuers of Jews in Nazi Europe.New York: Free Press. 419 pp.

Parker JH. 1968. The interaction of Negroes and whites in an integrated church setting.

Soc. Forces46:359–66 Patchen M. 1982.Black-White Contact in Schools: Its Social and Academic Effects.

West Lafayette, IN: Purdue Univ. Press.

387 pp.

Perdue CW, Dovidio JF, Gurtman MB, Tyler RB. 1990. Us and them: social categoriza- tion and the process of intergroup bias.J.

Pers. Soc. Psychol.59:475–86 Pettigrew TF. 1971.Racially Separate or To- gether?New York: McGraw-Hill. 371 pp.

Pettigrew TF. 1975. The racial integration of the schools. InRacial Discrimination in the United States,ed. TF Pettigrew, pp.

224–39. New York: Harper & Row. 429 pp.

Pettigrew TF. 1986. The contact hypothesis revisited. See Hewstone & Brown 1986, pp. 169–95 Pettigrew TF. 1991a. The importance of cu- mulative effects: a neglected emphasis of Sherif’s work. InSocial Judgment and In- tergroup Relations: Essays in Honor of Muzafer Sherif,ed. D Granberg, G Sarup, 84 PETTIGREW pp. 89–103. New York: Springer-Verlag.

280 pp.

Pettigrew TF. 1991b. Toward unity and bold theory: Popperian suggestions for two per- sistent problems of social psychology. In The Future of Social Psychology,ed. CW Stephan, W Stephan, TF Pettigrew, pp.

13–27. New York: Springer-Verlag. 121 pp.

Pettigrew TF. 1996.How to Think Like a So- cial Scientist.New York: Harper-Collins.

198 pp.

Pettigrew TF. 1997a. Generalized intergroup contact effects on prejudice.Pers. Soc.

Psychol. Bull.23:173–85 Pettigrew TF. 1997b. The affective compo- nent of prejudice: empirical support for the new view. InRacial Attitudes in the 1990s:

Continuity and Change,ed. SA Tuch, JK Martin, pp. 76–90. Westport, CT: Praeger.

In press Pettigrew TF. 1997c. Ingroup reappraisal: an- other intergroup contact process that re- duces prejudice. Univ. Calif., Santa Cruz.

Unpublished manuscr.

Pettigrew TF, Meertens RW. 1995. Subtle and blatant prejudice in western Europe.Eur.

J. Soc. Psychol.25:57–75 Pettigrew TF, Wright S, Tropp L. 1998. Inter- group contact and prejudice: a meta- analytic test of Allport’s hypothesis. Dep.

Psychol., Univ. Calif., Santa Cruz. Unpub- lished manuscr.

Powers DA, Ellison CG. 1995. Interracial con- tact and black racial attitudes: the contact hypothesis and selectivity bias.Soc.

Forces74:205–26 Reich C, Purbhoo M. 1975. The effect of cross-cultural contact.Can. J. Behav. Sci.

7:313–27 Reif K, Melich A. 1991.Euro-Barometer 30:

Immigrants and Out-Groups in Western Europe, 1988.Ann Arbor, MI: Inter-Univ.

Consort. Polit. Soc. Res. 780 pp.

Reitzes DC. 1953. The role of organizational structures: union versus neighborhood in a tension situation.J. Soc. Issues9(1):37–44 Riordan C. 1978. Equal-status interracial con- tact: a review and revision of the concept.

Int. J. Intercult. Relat.2:161–85 Riordan C. 1987. Intergroup contact in small cities.Int. J. Intercult. Relat.11:143–54 Riordan C, Ruggiero J. 1980. Producing equal-status interracial interaction: a repli- cation.Soc. Psychol. Q.43:131–36 Rippl S. 1995. Vorurteile und personliche Be- ziehungen zwischen Ost- und West- deuschen.Z. Soziol.24:273–83 Robinson JL Jr. 1980. Physical distance andracial attitudes: a further examination of the contact hypothesis.Phylon41:325–32 Robinson JW Jr, Preston JD. 1976. Equal- status contact and modification of racial prejudice.Soc. Forces54:911–24 Rose TL. 1981. Cognitive and dyadic pro- cesses in intergroup contact. InCognitive Processes in Stereotyping and Intergroup Behavior,ed. DL Hamilton, pp. 145–81.

Hillsdale, NJ: Erlbaum. 366 pp.

Rothbart M, John OP. 1985. Social categoriza- tion and behavioral episodes: a cognitive analysis of the effects of intergroup con- tact.J. Soc. Issues41:81–104 Russell MJ. 1961.Study of a South African Inter-Racial Neighbourhood.Durban, S.

Afr: Univ. Natal, Durban. 167 pp.

Saenger G, Gilbert E. 1950. Customer reac- tions to the integration of Negro sales personnel.Int. J. Opin. Attent. Res.4:

57–76 Sagiv L, Schwartz SH. 1995. Value priorities and readiness for out-group social contact.

J. Pers. Soc. Psychol.69:437–48 Schofield JW. 1989.Black and White in School: Trust, Tension, or Tolerance?

New York: Teachers College Press. 255 pp.

Sherif M. 1966.In Common Predicament.

Boston: Houghton Mifflin. 192 pp.

Sigelman L, Welch S. 1993. The contact hy- pothesis revisited: black-white interaction and positive racial attitudes.Soc. Forces 71:781–95 Sims VM, Patrick JR. 1936. Attitude toward the Negro of northern and southern college students.J. Soc. Psychol.7:192–204 Slavin RE. 1983.Cooperative Learning.New York: Longman. 147 pp.

Slavin RE, Madden NA. 1979. Social prac- tices that improve race relations.Am.

Educ. Res. J.16:169–80 Smith CB. 1994. Back and to the future: the in- tergroup contact hypothesis revisited.So- ciol. Inq.64:438–55 Spears R, Oakes PJ, Ellemers N, Haslam SA, eds. 1997.The Social Psychology of Stereotyping and Group Life.Oxford, UK:

Blackwell. 422 pp.

Stephan CW. 1992. Intergroup anxiety and in- tergroup interaction.Cult. Divers. Sch.2:

145–58 Stephan CW, Stephan WG. 1992. Reducing intercultural anxiety through intercultural contact.Int. J. Intercult. Relat.16:89–106 Stephan WG. 1987. The contact hypothesis in intergroup relations.Rev. Pers. Soc. Psy- chol.9:13–40 Stephan WG, Stephan CW. 1984. The role of INTERGROUP CONTACT THEORY 85 ignorance in intergroup relations. See Miller & Brewer 1984, pp. 229–56 Stephan WG, Stephan CW. 1985. Intergroup anxiety.J. Soc. Issues41:157–75 Stephan WG, Stephan CW. 1989. Antecedents of intergroup anxiety in Asian-Americans and Hispanic-Americans.Int. J. Intercult.

Relat.13:203–19 Stephan WG, Stephan CW. 1996.An inte- grated threat theory of prejudice.Pre- sented at Int. Psychol. Congr., 26th, Mont- real Stouffer SA, Suchman EA, DeVinney LC, Star SA, Williams RM Jr. 1949.The American Soldier: Adjustment During Army Life.Princeton, NJ: Princeton Univ.

Press. 756 pp.

Taylor DM, Dube L, Bellerose J. 1986. Inter- group contact in Quebec. See Hewstone & Brown 1986, pp. 107–18 Trew K. 1986. Catholic-Protestant contact in Northern Ireland. See Hewstone & Brown 1986, pp. 93–106 Triandis HC. 1994.Culture and Social Behav- ior.New York: McGraw-Hill. 330 pp.

US Department of Defense. 1955.A Progress Report on Integration in the Armed Serv- ices, 1954.Washington, DC: US Gov.

Print. Off. 165 pp.

Van Oudenhoven JP, Groenewoud JT, Hew- stone M. 1996. Cooperation, ethnic sali- ence and generalisation of interethnic atti- tudes.Eur. J. Soc. Psychol.26:649–61 Vivian J, Brown R, Hewstone M. 1995.

Changing attitudes through intergroup contact: the effects of group membership salience. Univ. Kent & Cardiff, Wales.

Unpublished manuscr.

Wagner U, Hewstone M, Machleit U. 1989.

Contact and prejudice between Germans and Turks.Hum. Relat.42:561–74 Wagner U, Machleit U. 1986. “Gestarbeiter” in the Federal Republic of Germany: con- tact between Germans and migrant popula- tions. See Hewstone & Brown 1986, pp.

59–78 Walker I, Crogan M. 1997.Academic per- formance, prejudice, and the jigsaw class- room: new pieces to the puzzle.Presented at Annu. Meet. Soc. Aust. Soc. Psychol., 3rd, Wollongong Watson G. 1947.Action for Unity.New York:

Harper. 165 pp.

Weber R, Crocker J. 1983. Cognitive pro- cesses in the revision of stereotypic be- liefs.J. Pers. Soc. Psychol.45:961–77 Weigert KM. 1976. Intergroup contact and at- titudes about third-group: a survey ofBlack soldiers’ perceptions.Int. J. Group Tens.6:110–24 Weldon DE, Carlston DE, Rissman AK, Slo- bodin L, Triandis HC. 1975. A laboratory test of effects of culture assimilator train- ing.J. Pers. Soc. Psychol.32:300–10 Werth JL, Lord CG. 1992. Previous concep- tions of the typical group member and the contact hypothesis.Basic Appl. Soc. Psy- chol.13:351–69 Wilder DA. 1984. Intergroup contact: the typi- cal member and the exception to the rule.

J. Exp. Soc. Psychol.20:177–94 Wilder DA. 1993a. The role of anxiety in fa- cilitating stereotypic judgments of out- group behavior. InAffect, Cognition, and Stereotyping: Interactive Processes in Group Perception,ed. DM Mackie, DL Hamilton, pp. 87–109. San Diego: Aca- demic. 389 pp.

Wilder DA. 1993b. Freezing intergroup evaluations: anxiety fosters resistance to counterstereotypic information. InGroup Motivation: Social Psychological Per- spectives,ed. MA Hogg, D Abrams, pp.

68–86. London: Harvester Wheatsheaf.

240 pp.

Wilder DA, Shapiro PN. 1989. Role of competition-induced anxiety in limiting beneficial impact of positive behavior by an outgroup member.J. Pers. Soc. Psy- chol.56:60–69 Wilder DA, Thompson JE. 1980. Intergroup contact with independent manipulations of in-group and out-group interaction.J.

Pers. Soc. Psychol.38:589–603 Williams RM Jr. 1947.The Reduction of Inter- group Tensions.New York: Soc. Sci. Res.

Counc. 153 pp.

Wilner DM, Walkley R, Cook SW. 1955.Hu- man Relations in Interracial Housing: A Study of the Contact Hypothesis.Minnea- polis: Univ. Minn. Press. 167 pp.

Works E. 1961. The prejudice-interaction hy- pothesis from the point of view of the Ne- gro minority group.Am. J. Sociol.67:

47–52 Wright SC, Aron A, McLaughlin-Volpe T, Ropp SA. 1997. The extended contact ef- fect: knowledge of cross-group friendships and prejudice. Univ. Calif., Santa Cruz.

Unpublished manuscr.

Zajonc RB. 1968. Attitudinal effects of mere exposure.J. Pers. Soc. Psychol.9(Suppl.

2):1–27 Zuel CR, Humphrey CR. 1971. The integra- tion of black residents in suburban neigh- borhoods.Soc. Probl.18:462–74