Social distance as categorization of intergroup interaction[1]

Carolyn W. Sherif[2}
Pennsylvania State University

Black students categorized 50 situations describing black-white interaction according to advisable social distance, using any number of categories and frequency distribution of situations they deemed appropriate (own categories procedure). Personal involvement was indicated by number of categories and frequency of noncommitment As predicted, black fraternity members revealed less involvement than independents, and males more involvement than females, There were significant interaction effects between these variables and college context (small, largely black versus large, overwhelmingly white institutions). The interpersonal versus institutional setting of situations was less important in black students' categorizations of social distance than expressed attitudes and actions of white participants and reactions by black peers.

Social distance is one of a few concepts in social psychology with a cumulative research history approaching the half-century mark. As proposed by Robert E. Park and measured by Emory Bogardus, social distance became a research tool for assessing the relative intimacy or rejection between members of diverse groups both for purposes of longitudinal study (e.g., Bogardus, 1947, 1967; Hyman & Sheatsley, 1964), interregional and interpersonal comparison (e.g., Hartley, 1946), and cross-cultural examination (e.g., Triandis & Triandis, 1965).

The fruitfulness of the social distance concept in research depends upon the researcher's acumen in selecting situations that reflect the prevailing order of social intimacy and distance. The importance of such knowledge, intuitive or empirical, is reflected in two of the more salient criticisms of the concept and existing assessment techniques, which may be summarized as follows:

1. Situations comprising a social distance test are ordered as a cumulative scale, in that accepting members of a group into a situation with a particular rank on the scale is assumed to indicate acceptance in all more distant situations. Yet, cross-cultural research indicates differences from one culture to another in the relative intimacy of the same situations. For example, interaction with a family friend is judged more intimate than with a friendship group of peers in Greece but not in the United States (Triandis & Triandis. 1965). Similarly, unexpected reversals in the order of situations within the same society are found during periods of rapid change. For example, Rollins (1970) found that members of six nationality groups in Rhode Island accepted members of certain other groups as friends or in-laws while rejecting members of the same groups as landlords.

2. The expression of social distance through conventional techniques need not relate closely to reactions to actual changes in contacts between the groups in question (e.g., Hill, 1953; Killian, 1962; Williams, 1964). Significant changes in actual contacts between black and white have occurred, for

( 328) example, in communities and institutions whose members expressed social distance sufficiently great to preclude such contacts. However, the salience of maintaining social distance has seldom been systematically studied. While the continuing struggles over desegregation of schools and housing indicate that intergroup attitudes are not irrelevant in interaction between groups, such findings suggest that in times of rapid change the criteria for social distance and its salience are proper objects of investigation.

The present research was undertaken with the aims of, on the one hand, introducing the problem of the salience of intergroup interaction to the person and, on the other, inquiring into the definitions of intimacy-distance by black college students. The situations studied were contacts between black and white students that frequently arise on a college campus.

The problems of attitudinal reactions in a test situation and of the efficacy of legitimized changes in contact situations are complex (cf, Sherif & Sherif, 1969, Ch. 21). A survey of available research on social distance indicated that the degree of the person's involvement in intergroup contact was often neglected. Under a variety of labels (e.g., salience, centrality, meaning, importance, relevance), the phenomenal referents of the concept ego-involvement include much recent research on attitude and related behaviors (e.g., Brock & Becker, 1965; Kiesler, Collins, & Miller, 1969; Rhine & Severance, 1970; Walster & Festinger, 1962; Whittaker, 1967). Despite conceptual confusion, the phenomenal referents of the concept in the present context are relatively clear: The more an ongoing situation involves those attitudes that define the person relative to his reference persons and sets, including their relationships with other groups, the greater the motivationalemotional arousal (cf. Bieri, 1967; Sherif & Cantril, 1947). The outcome is more characteristic patterns of behavior that are, in some important aspect, indicative of the person's attitudes.

The present research studied the degree of ego-involvement of black students in interracial interaction as a function of their previous and current white contacts. In the United States, the black student's decision whether or not to interact with whites in a college setting arouses considerable ego-involvement. The general hypothesis was that the lower the probability of black-white interaction and the stronger the ties with black reference groups in the "conventional" social scheme (fraternal organizations), the lower the black student's ego-involvement in interaction situations with whites. In addition, we predicted that the kinds of situations viewed by black students as acceptable or objectionable would relate more closely to the certainty of acceptance as a black person than to criteria defining the situation as formal and institutionalized, on one hand, or as informal, on the other. Specific hypotheses were generated from previous research on social judgment and attitudes, hence require an explanation of the method of investigation, whose validation was a primary concern.

Data were collected during the academic years of 1966 and 1967, before unified black movements emerged on the campuses studied. Doubtless, the categorizations of social distance have since changed, owing primarily to the active growth of efforts toward black student solidarity and education. There is little reason to expect, however, that the effects of the major variable have diminished. On the contrary, the black social movement may have heightened personal involvement in issues of black-white interaction.


Specific Hypotheses

The twin objectives of exploring black students' ordering of interaction situations in terms of social distance and of assessing relative ego-involvement suggested the method of investigation, namely the own categories procedure (Sherif, Sherif, & Nebergall, 1965). The method had the following advantages for our purposes:

1. It is unobtrusive and nonconfrontive in that the person's task is to make objective judgments.

In this case, the task was to judge the advisability that a black student decide to interact with white individuals in a series of situations, each described briefly. The person's attitude was not directly questioned. The task was to classify a set of descriptions into any number of categories the person regarded as suitable to differentiate among the situations according to how advisable it would be for the black student to interact with the whites described in the situations. The judge was free to distribute the various situations into the categories thus generated with any frequency he (she) wished.


2. On the basis of previous research, exact prediction., could he made about the personís relative ego-involvement, as indicated by the number of categories generated and by the distribution of items into those categories. Previous research indicated that persons highly involved in a bipolar issue (here., whether to associate or not) use fewer categories than those less involved, even when matched for education and ethnic background (Reich & Sherif. 1963) and regardless of whether they are for or against the attitude object (La Fave & Sherif, 1968; Vaughan, 1961). Further, the same person uses fewer categories when classifying a highly involving stimulus set than when a neutral series of items is presented (Glixman, 1965; Sherif, 1961).

After completing the sorting task, the person is asked to label the categories on the basis of criteria relevant to the stimulus material. However, instructions do not require that every category be labeled, leaving it up to the person to decide which to label. Those categories that are not labeled, when the person is not forced to label all categories, constitute the latitude of noncommitment. Previous research indicated that the size of the noncommitment latitude (frequency of items in unlabeled categories) is inversely related to the degree of involvement, that is, the greater the latitude of noncommitment, the less the person's involvement (Sherif et al., 1965).

Specifically, then, the prediction was that the greater the probability that the person frequently encountered interaction situations with whites and, concommitantly, the greater the lack of black reference persons and groups, the greater would be that person's involvement as revealed by the use of fewer categories and a smaller latitude of noncommitment.

The variables used in the research as indicative of differential probabilities of white encounters were as follows:

1. Institutional setting: Samples from a large state university, where the proportion of black to white students was about 1/200, were compared with students at a small, predominantly black college in the same state, where the ratio was at least 3/1

2. Sex role: Male students, being more mobile than females, were expected to encounter whites, particularly in heterosexual situations, more frequently than females.

3. Group membership in black social fraternities: The independent student, lacking close ties with a social center of fellow blacks, was more likely to encounter white interaction than the fraternity (sorority) member, whose social life was firmly rooted with other black students.

4. Relative frequency of past interactions with whites in educational, community, and social settings: The black student with a previous history in predominantly black settings was expected to be less involved than the student previously exposed to predominantly white settings. A check list of past interaction situations and a record of transfer from a predominantly white to a predominantly black college were indicators of this background variable.


Stimulus situations. A set of 50 descriptions of black-white interaction situations was constructed from a pool of episodes reported by black students. Each description, in 35-60 words, described a black student of the same sex as the student who was faced with a decision about whether to enter or continue interaction with one or more whites. The focal person was designated as Negro, rather than black, in keeping with the terminology used by the students themselves at the time. Descriptions for males and females differed only in altering the sex of the focal person and other participants so that own sex and heterosexual relationships were comparable. Names of organizations that were clearly sex-linked (e.g., Association of Women Students) were also varied. For each description, the subject's task was to assess the advisability that the focal person decide to enter interaction or interact further with the white persons described. Some situations included the information that black peers disapproved of the black-white interaction. Others involved only the focal person's choice.

The following descriptions represent the extremes of the set of situations, the first typically being viewed as advisable and the fourth as altogether inadvisable. The second and third descriptions were categorized with considerable variability, although the frequencies of judgments that further white interaction would be inadvisable were large. For females, the sex of the focal person was changed.

1. A Negro male student has been selected for an Honors program sponsored by his University. The faculty member who directs the program invites the students selected to come to his house for a celebration of their success. The student knows that if he attends the party, he will be the only Negro present.

2. A Negro fraternity member is active in several campus organizations where there are few other Negroes. His fraternity elects him as president. If lie serves as president, he will not have time to continue his campus activities.

3. A Negro senior is being considered for a teaching job in a small Northern town that has no Negro residents. The superintendent calls to request an interview in a nearby city, explaining that it might "cause talk" for the student to come to the town before he was hired, but that he sincerely wants to hire a Negro teacher. The student wonders whether to go for the interview.

4. A Negro student has a white roommate who is friendly and congenial. Just before Dad's Day, when their fathers are coming to see their room, he suggests that they take turns showing the room to their fathers, since his father doesn't know that his roommate is Negro. The Dad's Day schedule allows only half an hour for room visits.


A total of 315 students performed the own categories task, 150 males and 165 females. Eighty-eight were black students at an overwhelmingly white, large state university. The 50 female and 38 male

( 330) students represented all available in the university dormitories (including the membership of two sororities housed in the dormitories) and most members of two black fraternities living in their own houses. The fraternal organizations were paid small sums as an organization for securing the cooperation of members. All other students were volunteers solicited in the dormitories at regular floor meetings. The remaining black student population at the time can only be estimated. owing to lack of records. but probably numbered no more than 50-60 students living in rooms or apartments. Sixty of the students were fraternity members, and 28 were independents.

At the small, predominantly black college in the same state. 147 students (72 males, 75 females) were secured by six students in a social science course, each assigned to administer the task to 25 students of their own sex. The selected sample consisted of 44 fraternity members and 103 independents about equally male and female in each classification and included a total of 32 students who had transferred to the college from predominantly white institutions.

Eighty white students (40 male. 40 female) from an introductory psychology class at the large state university performed the task to provide a comparison with their black counterparts. Their task was the same, namely to assess the advisability of a black student interacting with whites.

Own Categories Task

Nine administrators collected data. They were thoroughly instructed on the procedures but not acquainted with the theoretical background or predictions of this study. In every case except the white sample, the administrator and subjects were the same sex. Black administrators worked with black subjects with two exceptions: Data from female black students at the large university were obtained by a white dormitory coordinator who had unusually good relations with them. One of the six administrators at the small college was white, since he volunteered with the others as a social science student. Afterward, he reported that the experience was one of the most frustrating he had ever had. After securing 9 subjects, he found that the remaining 16 were difficult to enlist, to the point that he had to "coax and persuade." Since the data analysis did not differ significantly when his sample was eliminated, it is retained for the present report. None of the black administrators encountered such resistance.

The task was performed individually or together with others so located that observation of each other's performance was not feasible. The instructions emphasized the need for realistic, objective judgments on the advisability of interacting further with white persons in the situations described. End categories, to be used if the student found them appropriate, were labeled "Altogether advisable" and "Not at all advisable." The number of categories to be used was not specified.

Upon completing the task (usually within 40 to 55 minutes), each student numbered the categories in order. Only then was any affective reaction requested, namely to label any categories "in which you would feel quite comfortable" or "definitely uncomfortable," with the option to label any other categories if desired. The unlabeled categories were defined as the latitude of noncommitment.

Finally a questionnaire was administered concerning participation in a variety of previous and current school and community activities, the racial composition of other participants being rated in four categories ("all Negro" to "mostly white").


Judgments on the advisability of interacting with whites in the 50 situations were analyzed in terms of the number of categories used and the frequencies of situations placed into categories with different labels. In addition, item analyses were performed by calculating the mean number of the category into which each item was placed by students using 1, 2, 3.... n categories.

The number of categories used ranged from 2 to 13. Fifteen of the 16 persons using 2 categories were black. One white female used 13 categories. Only 3 of the 19 students who used 7 or more categories were male.

Table I presents results on the number of categories used by the various samples and the frequency of items in categories that were not labeled as either comfortable or uncomfortable (latitude of noncommitment). The rows in the table are ordered on the basis of the probability of black-white encounters for the various subsamples. The expected order of the samples for number of categories used and latitude of noncommitment at the large state university was independent students, followed by members of fraternal organizations, then white students. At the small college, independent students were again expected to be more involved in the situations, hence to use fewer categories and to exhibit less noncommitment, followed by fraternity members. In both institutions, the predicted order was that males would use fewer categories and exhibit less noncommitment than females.

The findings in Table 1 indicate that the number of categories used to categorize the situations and the relative sizes of the latitude of noncommitment are related to frequency of black-white interactions. The order of the mean percentage of the situations that were unlabeled (noncommitment) follows the predicted order by frequency of black-white

( 331) interactions with but one exception, the black independent females at the small college, who exhibited less noncommitment than their male counterparts. These female black students also used fewer categories than their male counterparts in the same institution.

As a sole indicator of involvement, the number of categories proved to be less discriminating than the two indicators together. A two-way analysis of variance comparing mean number of categories used by the different sex and color samples at the large institution yielded no main effect for color; however, overall sex differences were statistically significant (F = 13.89, p < .01) as was the Sex X Color interaction (F = 7.75, p < .01). In effect, then, the differences in number of categories used by students at the large institution are accountable by the tendency for males to use significantly fewer categories than females, the sex difference being sufficiently larger in the black than in the white samples to produce the significant Sex X Color interaction and to reduce the probability of a significant main effect by color.

A three-way analysis of variance on the mean number of categories used by black students according to their institution, sex, and membership or nonmembership in social fraternities indicated that kind of institution per se was not significantly associated with number of categories. Fraternity members, whether male or female. used significantly more categories than independents (F = 65.32, p < .001). Males used fewer categories than females (F = 5.38, p < .05). The role of the institutional context was revealed through a significant Institution X Sex interaction (F = 5.47, p < .05). As indicated by the means in Table 1, this significant interaction reflects the fact that sex differences were larger at the large university than at the predominantly black college and, further, that independent females at the small college used fewer categories than male independents, contrary to prediction.

Predicted order of samples No. categories Latitude of noncommitment
(mean % of items)
Large University
B independent of M 2.80 0.00
B independent of F 3.50 1.30
B fraternity M 4.67 3.44
B fraternity F 5.57 4.56
W-M 4.53 12.97
W-F 4.75 20.49
Small College
B independent of M 4.05 21.31
B independent of F 3.72 11.87
B fraternity M 4.53 26.00
B fraternity F 4.86 23.43
Note:  B = black, W = white, M = male, F = female

The questionnaire responses on prior black-white associations were analyzed in terms of whether or not the composition had been predominantly black or white, taking into account differences among the students in absolute frequencies of reported participations of any kind. However, analysis of variance of the number of categories used according to the students' reports of their prior black-white interaction revealed no significant differences.

On the other hand, black students at the small college who had transferred there from predominantly white institutions did differ significantly from nontransfers at the same college, depending primarily on whether or not they joined a fraternity. In a three-way analysis of variance (transfer status, fraternity membership, sex), the only significant main effect was fraternity membership versus independent status (F = 18.05, p < .01). However, the two variable interactions and the triple interaction were all significant. The interaction effects can be regarded as merely suggestive. owing to small ns in certain classifications. The trend was for transfers who joined fraternities to use even more categories than nontransfers. These trends were strongest for males. They suggest the importance of the total context of prior and current experiences.

Analysis of variance for related measures was performed to compare the frequencies of items in categories given various labels by the students ("advisable-comfortable,"' "less than comfortable," "inadvisable-uncomfortable") or not labeled (noncommitment). The significant differences related entirely to the latitude of noncommitment (see Table 1). As predicted, the noncommitment latitude

(332) was significantly greater for white than for black students (p < .01), for female than for male students (p < .025), and for black students at the small, predominantly black college than for students at the large university (p < .01).

As data in Table 1 suggest, the number of categories generated and the frequency of items in unlabeled categories (noncommitment) were related aspects of the categorization process. By examining the distribution of situations into categories and the mean category placement of items by students using the same number of categories, it was evident that students who used a large number of categories typically had a large latitude of noncommitment. This relationship reflected a tendency for less involved students to differentiate very finely among situations intermediate between those labeled almost unanimously as "advisable" and those labeled "inadvisable" by almost all students. Further, the less involved students used more categories in dealing with "inadvisable" situations. The highly involved student, on the other hand, lumped "inadvisable" situations into fewer categories, clearly separated the intermediate situations as either advisable or inadvisable to some degree, thus was noncommittal on few situations and handled the set of situations with fewer categories.

Analysis of the category placement of each situation was made separately for students using each number of categories (2-13). The measure was the mean category number in which a situation was placed by students using the same number of categories. Separate analyses of variance of situation placement by black students revealed no significant differences attributable to sex or institutional context for students using 2 categories. Significant variations according to sex and Sex X Institution were found for those using 3, 4, or 5 categories. Variations in category placement by students using 6 or more categories were attributable to individual differences, primarily among the 38 females of the total number (51) who differentiated so finely among the 50 situations.

The significant variations among black students in their categorizations of specific situations included several trends worth noting for future research.

1. In every analysis according to number of categories used, certain situations were categorized differently by males and females. Specifically, the female students consistently categorized heterosexual relationships with whites in informal relationships as less advisable than male students (for example, playing cards in the student building, dancing with a white at a dance). On the other hand, the males categorized heterosexual relationships with whites as less advisable when they were more formal and public in nature (for example, involving school situations such as study arrangements and employment).

2. In same-sex situations, the categorizations of both male and female black students included both formal and informal contacts among those considered advisable and among those judged inadvisable. The major contingency determining whether a more informal and "intimate" situation was judged advisable was not the presence of whites alone, but common attitudes exhibited by the whites about the black situation, civil rights issues, and their public commitment to nondiscriminatory interaction.

3. Both male and female black students reacted negatively to situations in which their decision to achieve prominence in a white organization implied neglect of the other blacks or their overt disapproval. The single exception to their general judgment of "not advisable" was a situation in which a white roommate asks a black student to delay dinner for a short time so that they could eat together, while other black students are urging the person to forget the roommate and go to dinner with them at an earlier time. On the whole, women students judged the delay to wait for the white roommate as advisable, while male students typically placed this item in their median category or beyond (judging the delay as less advisable).

The white students typically placed informal, interpersonal relationships in intermediate, noncommittal categories whereas, as noted, black students differentiated among such situations. White students also were more likely to be noncommittal concerning employment opportunities where the black

(333) applicant would be the only black employee. White students unanimously placed the one situation involving possible interracial marriage as highly inadvisable. For black students, on the other hand, this situation was categorized with greater variability.


The procedures used in this research to study personal categorization of intergroup interaction in terms of social distance appear to hold promise for further research. First, particularly in times of rapid change, it provides a basis for determining the order of sets of situations as to their intimacy-comfort and relative distance-discomfort as perceived by the persons who are actual or potential participants in such situations. As the present research found, this order may not correspond to that constructed on the basis of sheerly formal definitions of the situations, such as whether the situation is informal or institutional in context. Further, both the order and the categorizations of situations varied significantly according to sex and to the insulating effects of the person's reference sets and groups.

Second, through the related indicators of number of categories used and extent of noncommitment, the own categories method is predictive of the degree of personal involvement in the interaction situations. a variable that has not been amenable to assessment through traditional methods for studying social distance. There is research evidence that persons with large latitudes of noncommitment are more susceptible to change than highly involved individuals, who tend to accept or reject categorically (Tittler, 1967). In view of the findings in this research, it would seem no accident that the spring following the data collection, it was chiefly black males and independent students at the large state university who organized the first meetings of a campus-wide black movement with outside speakers, closed to whites. Active student movement on the smaller college campus, predominantly black, began still a year later and was directed primarily against its own black administration.

Third, the social distance situations used in this research included contingencies pertaining to the attitudes and treatments of the outgroup members, as well as their social status. Variations that have been used with conventional social distance techniques (e.g., Triandis, 1967) presented combinations of somewhat abstract characterizations (e.g., a Negro female physician, 50 years old) in comparison with the greater naturalness of interaction situations from which intentions and reactions of the stimulus person can readily be inferred.

Fourth, the method is unobtrusive and nonconfrontive. Especially with today's research-wise students and public, these are advantages both for black researchers wishing to explore the attitudes of blacks and white researchers studying white social distance judgments. It would be of considerable interest to study further the social distance norms of whites in the United States today, including such contingencies of interaction as those described in the situations used here. More than one white liberal has been heard to protest, perhaps too much, that he is perfectly willing to interact with blacks but objects to black reluctance or treatment. Conversely, the black movement has placed some black students with broad latitudes of noncommitment about white interaction into conflict situations of varying intensity. In one case known to the writer, a black student with a white roommate was leading a double life. While rooming with the white student, he never appeared in public with him with considerable inner conflict as a result. These and other problems of intergroup interaction could be studied more properly with such an indirect, unobtrusive method than through direct questioning, in which the presentation of self as a group member becomes highly salient.

The findings of this research indicate once more that one's attitudes are related to one's reference groups and to the actual interaction situations encountered. Thus, while males typically exhibit more involvement than females and independents more involvement in intergroup interaction than members of closely knit social groups, the effects of sex role and group membership depend to an important extent upon the nature of the interaction setting. Lacking frequent white

( 334) interactions on campus, the students at the predominantly black college were uncertain about their reactions to large numbers of situations. Likewise, the white students at the large state university found the roleplaying experience of judging for a black student both difficult and bewildering, as revealed in their much larger latitudes of noncommitment. Further, there were significant divergences between the situations about which whites and blacks were noncommittal. White students concentrated their noncommitment on informal associations, many of which black students distributed across several categories of advisability or inadvisability. At both institutions, the black social fraternity at the time served an insulating function, such that fraternity members were less concerned with intergroup interaction than their independent counterparts.

While not a time-saving efficiency technique for large-scale surveys of attitudes. the own categories method holds promise for truly social-psychological study of individual attitudes in that it yields information on the structure of the person's beliefs, their evaluation, and the degree of personal involvement in the attitude objects, In addition to social distance, it would seem eminently suited for the study of even more intimate personal attitudes including, for example, sex roles and relationships.


BIER , J. Attitudes and arousal: Affect and cognition in personality functioning, In C. W. Sherif & M. Sherif (Eds.), Attitude, ego-involvement and change. New York: Wiley, 1967.

BOGARDUS. E. Changes in racial distances. International Journal of Opinion and Attitude Research, 1947, 1 (1), 55-62.

BOGARDUS, E. A forty-year racial distance study. Los Angeles: University of Southern California Press, 1967.

BROCK, T. C., & BICKER, L. A. Ineffectiveness of "overhead" counter-propaganda. Journal of Personality and Social Psychology, 1965, 2, 654-660.

GLIXMAN, A. R. Categorizing behavior as a function of meaning domain. Journal of Personality and Social Psychology, 1965, 2, 370-377.

HARTLEY, E. L. Problems of prejudice. New York: King's Crown Press. 1946.

HILL, M. Some problems of social distance in intergroup relations. In M. Sherif & M. O. Wilson (Eds.), Group relations at the crossroads. New York: Harper & Row, 1953.

HYMAN, H. H., & SHEATSLEY, P. B. Attitudes toward desegregation. Scientific American, 1964, 211(1), 16-23.

KIESLER, C. A,, COLLINS, B. E., & MILLER, N. Attitude change: A critical analysis of theoretical approaches. New York: Wiley, 1969.

KILLIAN, L. M. Leadership in the desegregation crisis An institutional analysis. In M. Sherif (Ed.), Intergroup relations and leadership. New York: Wiley, 1962,

LA FAVE, L., & SHERIF, M. Reference scale and placement of items with the own categories technique. Journal of Social Psychology, 1968, 76, 75-82.

REICH, J., & SHERIF, M. Ego-involvement as a factor in attitude assessment by the own categories technique. Norman, Okla.: Author, 1963. Reported in C. W. Sherif, M. Sherif, & R. Nebergall, Attitude and attitude change. Philadelphia: Saunders, 1965.

RHINE, R. J., & SEVERANCE, L. J. Ego-involvement, discrepancy, source credibility and attitude change. Journal of Personality and Social Psychology, 1970, 16, 175-190.

ROLLINS, J. H. Reference identification of youth of different ethnicity. Unpublished doctoral dissertation, University of Oklahoma, 1970.

SHERIF, C. W. Established reference scales and series effects in social judgment. Dissertation Abstracts, 1961, 22, No. 6.

SHERIF. C. W., SHERIF, M., & NEBERGALL, R. Attitude and attitude change. Philadelphia: Saunders, 1965.

SHERIF, M., & CANTRIL. H. The psychology of ego-involvements. New York: Wiley, 1947.

SHERIF, M., & SHERIF, C. W. Social psychology. New York: Harper & Row, 1969.

TITTLER, B. I. The relationship between attitude change and ego-involvement and its relevance to sex differences in attitude change. Unpublished master's thesis, Pennsylvania State University, 1967.

TRIADIS, H. C. Towards an analysis of the components of interpersonal attitudes. In C. W. Sherif & M. Sherif (Eds.), Attitudes, ego-involvement and change. New York: Wiley, 1967.

TRIANDIS, H. C., & TRIANDIS, L. M. Some studies of social distance. In I. Steiner & M. Fishbein (Eds.), Current studies in social psychology. New York: Holt, Rinehart & Winston, 1965.

VAUGHAN, K. R. A disguised instrument for the assessment of intergroup attitudes. Unpublished master's thesis, Texas College of Arts and Industries, Kingsville, 1961.

WALSTER, E., & FESTINGER, L. The effectiveness of "overhead" communications. Journal of Abnormal and Social Psychology, 1962, 65, 395-402.

WHITTAKER, J. O. Resolution of the communication-discrepancy issue in attitude change. In C. W. Sherif & M. Sherif (Eds.), Attitude, ego-involvement and change. New York: Wiley, 1967.

WILLIAMS, R. M. Strangers next door. Englewood Cliffs, N.J.: Prentice-Hall, 1964.

(Received September 14, 1971)


1 The research was conducted with the support of a grant from the Rockefeller Foundation to Muzafer Sherif. In designing the research materials, the help of Edith Gray was invaluable through her preliminary drafts of black-white situations based on her close experience with students and in collecting data from women at Pennsylvania State University. The unqualified cooperation of Estella Scott Johnson of Cheyney State College made data collection possible there. Along with her students (Geneva Caleb, Carl E. Geisler, Jr., Sheila Hickman, George Nicholas, Gene Sharp, and Paul L. Stephenson), Estella Johnson insured the best possible conditions for the research. At Pennsylvania State University, Ken Roy collected data from black male students. Others who assisted in data collection, tabulation, and analysis were Merrilea Kelly, Gian Sarup, Don Granberg, Ron Botto, Michael Saks, George Dunlap, Sue Sherif, and Joan Sherif.

2 Requests for reprints should be sent to the author, Department of Psychology, Pennsylvania State University, University Park, Pennsylvania 16802.



Valid HTML 4.01 Strict Valid CSS2