اثر تماس با قبول واقعیت بین فرهنگی: یک مطالعه میدانی

Abstract A field study was conducted to assess the effect of an intervention designed to promote contact between international and Australian students in an Australian university residential hall on their subsequent intercultural contact with, and acceptance of, unfamiliar outgroup members. In addition, the research was designed to assess the extent to which three variables (i.e. cultural stereotypes, cultural knowledge, and cultural openness) mediated the effect of intercultural contact. Seventy-six Australian and international students in one residential hall (vs 71 Australian and international students in a control residential hall) experienced an intervention designed to promote intercultural contact over a 7 month period. The results revealed considerable support for the intercultural contact hypothesis — the pattern of residential hall contact tended to impact directly upon the dependent measures. However, the results also indicated that the intervention impacted differentially upon the responses of the Australian and international students and that the most plausible explanation for this effect related to the students’ intercultural knowledge and openness.

1. Introduction The most influential social psychological approach to facilitating positive intercultural relations has been the “contact hypothesis” (Schofield, 1995). According to Allport (1954), positive contact between members of different groups should improve intercultural relations and, in particular, should reduce negative out-group stereotyping. Following Allport, many researchers have sought to identify the aspects of the contact situation which are most important in determining whether positive intergroup relations develop. The aspects which now appear to be most influential include equal status contact (e.g. Cohen, 1972, Hewstone and Brown, 1986 and Watson, 1950), opportunity to get to know outgroup members and disconfirm negative stereotypes (e.g. Amir and Ben-Ari, 1985 and Desforges et al., 1991), co-operative vs competitive task interaction (e.g. Bettencourt et al., 1992 and Johnson and Johnson, 1984), explicit support of relevant authority figures (Schofield, 1995), and situations with equalitarian social norms (e.g. Hewstone & Brown, 1986). Given the support for the contact hypothesis, it has provided the foundation for social policies designed to promote intergroup integration in schools, housing, and sport and recreation (Wagner, Hewstone & Machleitt, 1989). At the same time, however, it needs to be recognized that the research support for the hypothesis is not unambiguous (e.g. Barnard and Benn, 1987, Desforges et al., 1991 and Jackman and Crane, 1986). A number of studies have failed to support the hypothesis, even when the preceding conditions have been met or, alternatively, the findings have supported reverse contact effects; that is, greater rather than less prejudice following contact (e.g. Amir, 1976, Butler and Wilson, 1978, Mitchell, 1968, O’Driscoll et al., 1983, Ray, 1980, Ray, 1983 and Schaefer, 1975). Furthermore, much of the supportive research has limited external validity because researchers have experimentally manipulated the group membership of subjects instead of using members of existing (e.g. racial, gender) social groups (e.g. Bettencourt et al., 1992, Gaertner et al., 1990, Gaertner et al., 1989 and Miller et al., 1985). In addition, the actual contact has often lacked realism and has usually been very brief (e.g. Barnard and Benn, 1987, Bettencourt et al., 1992, Damico and Sparks, 1986, Desforges et al., 1991, Gaertner et al., 1990, Johnson and Johnson, 1984, Katz and Zalk, 1978, Masson and Verkuyten, 1993 and Werth and Lord, 1992). And, aside from research which has demonstrated the effectiveness of intergroup contact in schools, particularly when it involves co-operative learning groups (e.g. Johnson et al., 1984, Miller and Davidson-Podgorny, 1987 and Slavin, 1983), most of the studies which have assessed the contact hypothesis in situations of more everyday intergroup contact (e.g. communities, the workplace) have been ex post facto in nature with measurements being taken of ongoing contact such that inferences of causality are difficult to draw with any confidence (Hewstone & Brown, 1986). Finally, despite all the extant research on the contact hypothesis, a major limitation is that there is actually little evidence that intergroup contact promotes positive attitudes which generalize beyond those group members in the contact setting to unacquainted out-group members (see Amir, 1976, Brown, 1995, Cook, 1978 and Hewstone and Brown, 1986). Indeed, even those studies which have reported a generalization effect have typically been based on an artificial, short-term laboratory interaction (e.g. Bettencourt et al., 1992 and Desforges et al., 1991). While the preceding discussion highlights the equivocal nature of the intercultural contact literature, it also serves to emphasize the difficulty of undertaking an unambiguous test of the hypothesis which has external validity. In particular, the review highlights the need for research that assesses the contact hypothesis in an everyday real-world environment in which, at the same time, the degree of contact between the two groups is at least controlled, if not experimentally manipulated, yet lasts over a considerable period of time. Most importantly, the effect of the contact needs to be assessed in an environment outside that in which the contact occurs, and with unfamiliar members of the target group. The first aim of the present research was to provide such a controlled test of the contact hypothesis in an everyday environment which emphasized the five favorable conditions of contact noted earlier (e.g. Amir, 1969, Cook, 1978 and Hewstone and Brown, 1986). Importantly, however, the present study was designed to assess the efficacy of an explicit intervention which was designed to promote intercultural contact, beyond that which might otherwise be prompted by the five favorable contact conditions. The rationale for developing and assessing this intervention arose from previous research indicating that mere contact, even under the favorable contact conditions noted above, has frequently been found to be insufficient to promote positive intercultural attitudes, especially towards those outside the contact setting (e.g. Cook, 1978 and Nesdale and Todd, 1997). As several writers have noted, one implication of these findings is that the contact situation may need to be augmented by the presence of additional factors in order for generalizable, positive intergroup attitudes to be realized ( Hewstone and Brown, 1986 and Nesdale and Todd, 1997). While several such factors have been assessed, including superordinate goals (e.g. Gaertner et al., 1993, Gaertner et al., 1996 and Sherif, 1966), multiple group membership and cross-cutting social categories (e.g. Brewer, 1996, Deschamps and Doise, 1978 and Marcus-Newhall et al., 1993) and expectation states (e.g. Cohen and Roper, 1972 and Riordan and Ruggiero, 1980), the present intervention took its direction from Hewstone & Brown’s (1986) model of intergroup contact. Briefly, compared with interpersonal models of contact which argue for de-emphasizing group or category divisions in order to facilitate personalized contact between group members (e.g. Brewer and Miller, 1984 and Gaertner et al., 1993), Hewstone and Brown’s (1986)intergroup model emphasizes the retention of group divisions so that, as much as possible, contact occurs between members as representatives of their group, rather than as individuals. According to the latter model, the generalization of the favorable outcomes of intergroup contact to outgroup members is made more likely because the change in attitudes relate to all outgroup members, not just those who were involved in the contact situation. Consistent with the broad thrust of Hewstone and Brown’s (1986) model, the intervention program assessed in the present study focused on explicitly promoting intercultural contact in a range of everyday contexts, and between individuals whose group membership was kept salient in the situation. That is, the contact between group members was not simply left to emerge, and group membership was not treated as if it did not exist (i.e. the interpersonal approach). While the first aim of the present research was therefore to assess the efficacy of this intervention in promoting intercultural contact, the second and related aim was to examine the extent to which the effects of the contact generalized beyond the individuals in, and the setting of, the contact situation. As noted earlier, this issue is important for there is actually little extant evidence that positive intergroup contact actually generalizes to unacquainted outgroup members (see Brown, 1995). The third aim of the study was to shed some light on the mechanism(s) by which successful intercultural contact exerts its impact. Among several possibilities, the present research focused on three variables which, separately or in combination, might plausibly mediate the contact effect: cultural stereotypes, cultural knowledge, and cultural openness. Specifically, given that an intervention maintained the salience of the group divisions, one possible effect of positive contact is that the negativity of outgroup stereotypes might be reduced such that new and unfamiliar outgroup members are perceived more positively (e.g. Allport, 1954, Amir and Ben-Ari, 1985, Johnston and Hewstone, 1992, Rothbart, 1996 and Weber and Crocker, 1983). In contrast, a second possibility draws upon attitude change research indicating that familiarity influences evaluation (e.g. Fazio and Zanna, 1981 and Zajonc, 1968). That is, another possible effect of positive contact is that the members of one group might learn more about the cultural background of members of another group with the result that they might feel more positive towards them. A third and related possibility is that positive contact reduces ingroup members’ tendencies towards ethnocentrism (i.e. an evaluative preference for all aspects of their own group relative to other groups) and, instead, enhances their tendencies to be open to, and interested in, the similarities and differences between their own group and other groups (e.g. Drapela, 1975). If positive contact promotes intercultural openness, then increased generalization to new and unfamiliar outgroup members might be anticipated. Although this expectation has not been explicitly tested, consistent with it are the results of studies which have reported increased intercultural openness following programmes designed to facilitate intercultural interaction (e.g. Mio, 1989 and Zatzick et al., 1994). In sum, the aims of the present research were threefold. First, to assess the efficacy of an intervention designed to promote intercultural contact in an everyday environment. Second, to assess the extent to which the resulting contact effect generalized to unacquainted outgroup members. Third, to assess the extent to which the positive contact effect was mediated by, separately or in combination, three plausible variables (i.e. cultural stereotypes, cultural knowledge, cultural openness). The research was carried out in a number of halls of residence at an Australian university. As with many other Australian and overseas universities, international fee-paying students from primarily Asian countries (mainly Malaysia, Hong Kong, Indonesia, Singapore) are recruited by the university in question and comprise up to 12% of its total student enrolments. Many of these students (more than 70% of whom are ethnic Chinese) live in the university’s six halls of residence, along with local Australian students (the overwhelming majority of whom are from an Anglo-Australian cultural background), at least for the first year of their university study, after which many tend to move into private accommodation. The halls of residence have considerable similarities in philosophy, size, facilities and services, social and recreational activities, tutorial programs, and accommodation costs, as well as the positive attitudes of the halls’ administrators towards the recruitment of international students. The residential halls actually combine together in their recruitment activities and are similarly concerned with facilitating good relationships between the two student groups. As a result, it is perhaps not surprising that international students tend to distribute fairly evenly across the residential halls and typically comprise a similar proportion (i.e. 40–60%) of the total residents in each hall. Halls of residence have a particular relevance to the assessment of the contact hypothesis. To some extent, they can be seen as mini-societies in which people eat, live, work and socialize, but in a comparatively small geographical area. Each year sees the arrival of the new cohorts of the two student groups, each typically comprised of more females than males, the two cohorts having had no previous contact with each other. The subsequent extent of contact between the two groups may range from little to great, with some amount of contact being simply unavoidable. For example, there are a number of unavoidable points of incidental contact (e.g. in corridors and recreation facilities, in shared bathrooms and kitchens, and in dining halls) as well as contact arising in more formal settings (e.g. residential hall tutorial programs). Students may choose, however, to constrain their level of contact to this incidental and minimal level or to develop contacts which may become deep and long-lasting relationships. The students are typically resident for at least one academic year thus enabling assessment of the contact hypothesis over a lengthy period of time; indeed, for a length of time over which the assessment of the contact hypothesis has rarely been attempted. Finally, the crucially important generalizability issue can be addressed by examining the extent to which residential hall contact is reflected in the level of intercultural contact in the wider university environment and in the students’ level of intercultural acceptance. Of greatest significance, however, is the fact that the residential hall environment meets the five criteria for favorable intergroup contact noted earlier. For example, while international students have a minority status as a cultural group, their status as students within the residential colleges is explicitly equivalent to that of Australian students in terms of such factors as: entrance criteria, obligations and responsibilities, social standing, fees, and involvement in student administration, and most come from a middle class background and are similarly well-off financially (see Allport, 1954, Kramer, 1950 and Pettigrew, 1971). Further, residential halls provide a daily 24 h opportunity for members of the two groups to get to know each other and disconfirm negative stereotypes. Indeed, the halls emphasize collegiality and equalitarian social norms and encourage co-operation and collaboration in both academic and recreational activities. Thus, residential halls provide an excellent opportunity to examine the interaction (or otherwise) of two broad groups of people who are culturally, linguistically and physically dissimilar, with the members of one group being part of the cultural majority. This is of special significance because, as noted above, much of the existing research on the contact hypothesis has focussed on artificially created groups who have shared a common culture (i.e. American college students) rather than two or more groups possessing different, cultural heritages. Considerably less support for the contact hypothesis has been revealed when the group members belong to culturally dissimilar groups, especially when the members of one group belong to the cultural majority (e.g. Amir, 1976, Mitchell, 1968, O’Driscoll et al., 1983, Ray, 1980 and Ray, 1983); that is, the situation typically faced by temporary or permanent immigrants. In the present study, one of the residential halls was chosen to be the recipient of the contact intervention, while three other residential halls were used as a comparison/control group. Given the similarities between the residential halls which were noted earlier, it was anticipated that there would not be any initial differences in attitudes between the students each attracted, nor that there would be any differences between the responses of the students in the three halls comprising the comparison/control group at the end of the study. And, as is detailed below, both expectations were confirmed. Given that the intervention program was successfully implemented, it was expected that: 1. there would be greater intercultural contact between the Australian and international students in the intervention residential hall (i.e. Hall “I”) vs the comparison residential halls (i.e. Hall “C”); 2. there would be greater intercultural contact in the wider university campus between the Australian and international students who were resident in Hall I vs Hall C; 3. there would be greater intercultural acceptance (operationalized as “intercultural friendship”) on the university campus between the Australian and international students who were resident in Hall I vs Hall C. No directional hypotheses were specified in relation to the three possible mediating variables (i.e. cultural stereotypes, knowledge, openness), given the largely exploratory nature of this aspect of the research. At the same time, however, given that each was identified as a plausible mediator of the contact effect, it was anticipated that one or more of these variables would reveal an intervention effect. That is, Hall I vs Hall C international and Australian students might be anticipated to reveal more cultural openness and/or cultural knowledge and/or less negativity in their outgroup stereotypes. It should be noted that since the latter is not a straightforward measure, the strategy adopted was to compare the positivity or negativity of own group stereotype with the stereotype attributed by an outgroup, in order to derive a measure of greater or lesser negativity following the intervention. Given that individuals typically see themselves more positively than they are perceived by members of an outgroup (Brewer and Silver, 1978 and Doise et al., 1972), the implication of this strategy was that, if stereotypes were positively affected by the intervention, the difference in ratings of members of one student group by the international vs Australian students should be less for those who experienced the contact program vs those who did not.

Results 3.1. Effectiveness of intercultural contact intervention The first issue addressed was whether the intervention to promote contact between international and Australian students within Hall I actually instigated more contact than occurred in Hall C. Analysis of subjects’ responses to the question on the amount of intercultural contact in their residential hall revealed a significant student group main effect, F(1,133)=7.85, P<0.01, which was qualified by a student group×hall interaction, F(1,133)=4.97, P<0.05. As indicated in Table 1, Australian students in Hall I vs Hall C indicated that they had significantly more intercultural contact, whereas there was no difference in the amount of intercultural contact revealed by international students in Hall I vs Hall C. Table 1. Summary of cell means in significant student group×hall interaction effects on the dependent measuresa Dependent measures Australian Ss International Ss Hall I Hall C Hall I Hall C Hall contact 6.30∗ 4.84† 4.76† 4.71† University contact 20.67∗ 14.94† 18.37∗† 19.71∗ Intercultural acceptance 11.44∗ 8.23† 12.39∗ 11.61∗ Cultural stereotypes Australian – – – – Singaporean – – – – Malaysian 45.60∗† 41.65† 47.32∗† 50.06∗‡ Cultural knowledge 5.15∗ 3.38† 4.82∗ 5.44∗ Cultural openness 57.17∗ 51.76† 56.50∗ 56.16∗ a Cell means with asterisks, daggers and double daggers are significantly different, Duncan’s Multiple Range Test, P<0.05. Table options The students’ responses on the question administered at both the commencement and termination of the project (i.e. extent to which they would be interested in taking part in activities designed to promote contact between Australian and international students) was analysed in a 2 (student group)×2 (hall)×2 (gender)×2 (time: beginning/end) ANOVA. This analysis revealed main effects for hall, F(1/356)=25.08, P<0.001; student group, F(1/356)=38.32, P <0.001; gender, F(1/356)=14.47, P<0.001; and time, F(1/356)=22.25, P<0.001. In addition, each main effect was qualified by a two-factor interaction, and there was one three-factor interaction. A significant residential hall×time interaction, F(1/356)=11.94, P<0.001 indicated that whereas there was no difference between Hall I (M=5.60) and Hall C (M=5.35) students at the outset of the project, Hall I students (M=5.43) were significantly more interested than Hall C students (M=4.15) in promoting contact, by the end of the project. A significant student group×gender interaction, F(1/356)=18.98, P<0.001, indicated that Australian females (M=5.15) vs Australian males (M=4.30) were significantly more interested in promoting contact, whereas there was no difference in the responses of international males (M=5.60) and females (M=5.67). A significant student group×gender×residential hall interaction, F(1/356)=4.48, P<0.05, indicated that male international vs Australian students in Hall I (Ms=5.70 and 4.86, respectively) and Hall C (Ms=5.49 and 3.67, respectively) as well as female international vs Australian students in Hall C (Ms=5.44 and 4.91, respectively) expressed greater interest in promoting contact, whereas there was little difference between female international (M=5.98) and Australian students in Hall I (M=5.70) in promoting intercultural contact. 3.2. Generalization of residential hall contact 3.2.1. University intercultural contact To examine the extent to which residential hall contact generalized to university contact, the subjects were asked to reveal the extent of their intercultural contact on five questions relating to the university in general, as well as in lectures and tutorials, assignments, leisure activities, and travel. Analysis of the students’ summed responses to these items revealed a significant student group×hall interaction, F(1/133)=8.05, P<0.006. Comparison of the cell means in the interaction (see Table 1) indicated that Hall I vs Hall C Australian students had more intercultural contact, while there was no difference between Hall I and Hall C international students. 3.2.2. Intercultural acceptance The extent to which intercultural contact facilitated intercultural acceptance was assessed via subjects’ responses to three questions on intercultural friendship (i.e. friendship importance, friendship attempted, and formed friendships). Analysis of the students’ summed responses on these questions revealed significant main effects for hall, F(1/133)=28.94, P<0.001, and student group, F(1/133)=34.98, P<0.001, which were qualified by a student group×hall interaction, F(1/133)=16.21, P<0.001. As indicated in Table 1, Australian students in Hall I vs Hall C considered intercultural friendship to be more important, while there was no difference in the views of international students in Hall I and Hall C. 3.3. Variables mediating the intercultural contact effect 3.3.1. Cultural stereotypes 3.3.1.1. Australians Analysis of the international and Australian students’ summed ratings of Australians on the 15 bi-polar trait scales revealed no significant effects. 3.3.1.2. Singaporeans Analysis of the summed ratings of Singaporeans by the Australian and Singaporean students yielded only a significant main effect for student group, F(1/48)=7.19, P<0.01; Singaporean students (M=46.54) rated themselves significantly more positively than did Australian students (M=41.38). 3.3.1.3. Malaysians The summed ratings of Malaysians by Australian and Malaysian students revealed a significant main effect for student group, F(1/59)=10.28, P<0.01, which was qualified by two significant interactions. A significant student group×hall interaction, F(1/59)=4.17, P<0.05, revealed (see Table 1) no difference between the ratings by the Malaysian and Australian students in Hall I, whereas Malaysian students in Hall C rated themselves significantly more positively than did the Australian students in Hall C. A significant student group×gender interaction, F(1/59)=5.45, P<0.03, indicated no difference between the ratings by the Malaysian (M=46.37) and Australian (M=44.21) female students, whereas Malaysian male students (M=51.06) rated themselves significantly more positively than did Australian male students (M=41.88). 3.3.2. Cultural knowledge Cultural knowledge was assessed via a question which asked the students whether, compared with when they commenced their university studies, they felt they knew the same, or more, about the background, culture, and interests of the other group. Analysis of these data indicated a significant effect for student group, F(1/133)=11.35, P<0.001 which was qualified by a significant student group×hall interaction, F(1/133)=24.84, P<0.001. The interaction cell means (see Table 1) revealed that Australian students in Hall I had more intercultural knowledge than the same students in Hall C, whereas there was no difference between the international students in Halls I and C. 3.3.3. Cultural openness Analysis of the subjects’ summed responses on the 11 questions comprising the cultural openness scale revealed a significant residential hall main effect, F(1/133)=6.24, P<0.02, which was qualified by a significant student group×hall interaction, F(1/133)=6.62, P<0.02. Comparison of the cell means (see Table 1) indicated that Hall I Australian students displayed more openness than Hall C Australian students, whereas there was no difference in openness between Hall I and Hall C international students.