Frequency of contact and sterotyping.

Six samples of Ss *ere tested with instrument which utilized a semantic differential format. The "concepts" v -2 "Americans in general tend to be" and "Greeks in general tend to be. The scales were characteristics obtained from unstructured interviews of Americans and Greeks woi 'ing together in jobs requiring face-to-face social relations. The six samples varied in the degree of contact. Maximum contact groups consisted of Americans and Greeks working together; medium contact groups consisted of Americans living in Athens taking a university course, and Greek university students studying in Illinois; minimum contact groups consisted of American students in Illinois and Greek students at the University of Athens. The autostereotypes and heterostereotypes of the six samples were investigated. It was found that the autostereotype of the American samples having contact with Greeks is more favorable than the autostereotype of the Americans having no contact; the autostereotypes of the three Greek samples show no differences. The heterostereotype of Americans concerning Greeks Is less favorable for the maximum than the minimum contact groups; the heterostereotype of Greeks concerning Americans is more favorable for the maximum than for the minimum contact groups,. A theoretical integration of these results is offered. Frequency of Contact and Stereotypinp Harry C. Trlancis Vasso Vassiliou University oJf Illinois Athenian Institute of Anthropos Stereotypes are boiiöfs about the characteristics of groups cf people. The present study explores the effects of intercultural contact on the (a) favorableness of stereotypes and (b) within sample agreement concerning the stereotypes. Two kinds of stereotypes are examined: autosterectypes (members of culture A thinking about members of culture A) and heterostereotypes (members of culture A thinking about members of culture B). Thus, favorableness is examined both with respect to autostereotypes and to heterostereotypes. Within sample agreement is consideretil as a measure of "clarity" of the stereotype, for when most members of a group agree, this irdicates they have no doubts about the particular stereotype. Such agreement, however, can occur simply when adequate communication exists within a given group, Thut, agreement in itself is not necessarily an index of the "validity" of the stereotype. On the ether hand, when a "clear" autostereotype agrees with a "clear" heterostereotype, and there is no substantial amount of communication between the two groups, this condition may be thought of as providing presumtive evidence that the particular stereotype has validity. The validation of such stereotypes, however, requires experiments in which Ss froc. the various cultures behave as predicted from the stereotypes. Such experiments were beyond the purpose of the present study. Some of the data for this study were collected while Triandis was a Ford Foundation Faculty Fellow in 1964-65. Some of the data were collected under contract NR 177-472, Nonr 1834(36) with the Advanced Research Projects Agency and the Office of Naval Research to study "Communication, Cooperation and Negotiation in Culturally Heterogeneous Groups (Fred E. Fiedler, Lawrence k!. Stolurow, and Harry C. Triandis, Principal Investigators). E. E. Davis, Uriel Foa, C. E. Osgood, and D, Summers made valuable comments on an earlier version of this paper. A great deal of research has been completed about stereotypes since the early Katz and Braly (1933) study. This research suggests tha; stereotypes are learned primarily through interaction with family, friends, teachers, etc., so that they constitute "cognitive norms" for thinking about a group of people, but they may also develop through interaction with the persons being stereotyped. It appears reasonable to expect certr.in stereotypes to develop primarily as a result of Interaction with members of one's in-group^and other stereotypes to develop as a result of interaction with the persons being btereotyped. For example, if an American hoars about "irtatn atrocities committed by members of group X, he might attribut* the characteristic "cruel" to them without any interaction with this group. But, he may have no stereotype concerning the characteristic tendency "to pry into personal affairs" and might not acquire such a stereotype unless he has personal contact with members of group X. The basic hypotheses of the study were based on the view that the greater the degree oi contact between two groups the greater will be the emount of Information that one group will have about the other, so that the heterostereotypes should be more differentiated or "clearer." At the same time, since each group would have a chance to compare its own selfimage with that of another cv.tural group, greater contact should result in more differentiation in ehe autostereotypes as well as in the heterostereotypes. On the matter of favorability of the stereotypes, the present view wouiu suggest that the more group A succeeds in reaching goals that are admired by group B, the more will the heterostereotype of B concerning A be favorable. Conversely, the less group A reaches group goals admired by group B, the leas favorable will be the heterostereotype of B concerning A, Finally, if group A reaches a desirable goal, which group B does not reach,