others' self-evaluations and interaction anticipation as

0 downloads 0 Views 958KB Size Report
3, 348-358. OTHERS' SELF-EVALUATIONS AND INTERACTION ANTICIPATION. AS DETERMINANTS OF SELF-PRESENTATION1. KENNETH J. GERGEN.
Journal ol Personality and Social Psychology 1965, Vol. 2, No. 3, 348-358

OTHERS' SELF-EVALUATIONS AND INTERACTION ANTICIPATION AS DETERMINANTS OF SELF-PRESENTATION1 KENNETH J. GERGEN

BARBARA WISHNOV 2

Harvard University

Duke University

Female Ss faced a partner who was either self-centered, self-derogating, or who evaluated herself in an average fashion. J of the Ss in each of these groups anticipated further interaction with the partner; i expected no further interaction. All Ss then made a number of self-ratings which were to be sent to their partner. As predicted, Ss rating themselves for the self-centered partner became more positive in their self-ratings, and the self-derogating partner caused Ss to emphasize more negative self-characteristics. Contrary to prediction, the average partner caused Ss to emphasize more negative features of self when future interaction was anticipated. Such modifications did not seem to depend on conscious mediation.

The information we present to others about ourselves is seldom selected at random. We constantly face the dilemma of choosing from a vast storehouse of self-knowledge the appropriate items for public display. One of the more crucial periods for such decision making is during the formative stage of a relationship. As Thibaut and Kelley (1959) have pointed out, the fate of any relationship depends to a great extent on the level of outcomes experienced by the respective members during the introductory stages. The present investigation was an attempt to explore the ways in which the attributes of one member of a dyad and the interaction context can modify the self presented by a second member during the initial stages of a relationship. One principal dimension along which selfreferent statements can vary is that of positiveness. The importance of this dimension is well reflected in the abundance of psychological studies using the variable of self-esteem. By far the major portion of the literature on self-esteem has found it convenient to assume relatively stable feelings of self-regard within any individual. Thus many of the studies in this area have been devoted to demonstrating the differential reactions to various situations of persons deemed to be either high or low in self-esteem. 1 This article is based on a thesis submitted by the second author in partial fulfillment for the BA degree with distinction at Duke University. 2 Now at Brandeis University.

On the other hand, the importance and to some extent the validity of a concept like self-esteem depends largely on the consistency of the overt behavior of the individual. If a person is completely capricious in his expressed feelings of self-regard, any single measure of self-regard would constitute a poor predictive device. And yet, we are all quite aware that persons do indeed vary in what they say about themselves from one situation to another. Many such variations have been discussed by Goffman (19S9); others have been more formally exposed in studies like those of Jourard and Lasakow (1958), and Katz (1942). There are two prominent questions which such studies pose: What are the systematic determinants of such variations? Under what circumstances do these overt variations reflect changes in the subjective state of the person? The present study was devoted primarily to shedding light on the former of these questions. In doing so, however, it was hoped that at least some evidence could be generated concerning the latter. There are a number of studies which have dealt with systematic changes in self-evaluations. Several investigations, for example, have explored the conditions under which success and failure affect a person's evaluations of self (cf. Stotland & Zander, 1958; Videbeck, 1960). From a different vantage point, Jones, Gergen, and Davis (1962) and Jones, Gergen, and Jones (1963) have demonstrated the effects of various interaction goals and status assignments on a person's

348

DETERMINANTS OF SELF-PRESENTATION public self-evaluation. In a further study (Gergen, 1965), the effects of social feedback on the positiveness of self-ratings were explored under systematically varied conditions. In this study it was possible to distinguish between those conditions giving rise to public versus private self-evaluations. However, it also seems apparent that a person's public self-evaluations are highly dependent on personal attributes of others in the same situation. For example, Davis (1962) demonstrated that the degree of interpersonal dominance displayed by one member of a dyad influenced the degree of dominance manifested by the other. In the present study the question was whether one can display the same social identity before another who is seen to be self-centered and egotistical, as before another who is self-effacing or self-demeaning? To shed light on this question subjects were exposed to a person who described herself in a very positive light, revealed both good and bad points, or who was very self-derogating. Each of these situations, it was felt, would pose for subjects a specific kind of problem in impression management (Goffman, 1959) and these problems would be resolved by differential self-evaluations. Before discussing the specific predictions a word must be said about the context of interaction. The personal attributes of another can seldom be considered in exclusion of the circumstances in which they are encountered. Although there are a number of context factors of relevance in the present case, one of the more intriguing aspects of a social relationship has to do with the participants' perceptions of how long the relationship will last. It has been noted, for example, that when interacting with a stranger a person is often more open and revealing than with a personal friend. Thibaut and Kelley (1959) have termed this the "stranger passant" phenomenon and have theorized that in a short-lived relationship the participants are not threatened by problems of dependency. The amount of anticipated interaction would thus seem to have some interesting implications for the present study. The solution to each of the problems mentioned above should be differentially altered

349

as a result of the subjects' perceptions of the duration of the relationship. We may now turn to a more complete description of the various problems and their probable resolutions. First of all, one major problem for a person finding himself faced with a highly positive other might be termed power restoration. Interacting with another who displays an impeccable personality can often be an intimidating experience. Such a person immediately usurps a position of seniority in a dyad and forces the other member into a usually undesirable, lowstatus position. Given little latitude of behavior, as subjects in the present experiment were, the optimum solution to such a problem would seem to be to counter in kind, that is, to become more positive about oneself. This anticipated self-enhancement should, however, be most dramatic when further interaction is anticipated. When the possibility of future intimidation and dependency is nonexistent, the necessity for redressing a power imbalance should not be as great. When faced with another who demonstrates that he has at least a number of worthy attributes and enough honesty to admit shortcomings, a different type of problem emerges. It would seem that such a person should cause one to seek out his friendship and the problem can thus be termed one of acquaintance seeking. One method of becoming better acquainted in the present experiment would be also to demonstrate both good and bad features of self, that is, "be oneself." However, as Thibaut and Kelley (1959) have suggested, when a long-term relationship is anticipated a person can ill afford to reveal weaknesses. It was thus predicted that subjects would be more positive about themselves when further interaction was anticipated than when no further interaction was expected. When another is encountered who is selfeffacing, still a third problem is generated. The self-derogating other immediately places himself in a position of dependency. The stronger partner in such a dyad is soon faced with the major share of the responsibility for the welfare of the dyad. The weaker member often acts as a parasite, thriving at the expense of the personal resources of the other. The person encounter-

350

KENNETH J. GEEGEN AND BARBARA WISHNOV

METHOD ing such another might be said to be faced with the problem of succorance avoidance. Subjects Before considering the solution to such a The subjects were 58 undergraduate females problem, it should also be noted that the enrolled in an introductory psychology course. During person who seems to have almost no good the early stages of the course, all students were qualities often elicits feelings of pity or required to fill out a self-evaluation measure to be described below. From this initial sample, sympathy. Such feelings greatly complicate persons who scored within approximately the upper the problem of succorance avoidance, for it is or lower 1$% of the sample were eliminated usually a very difficult matter to turn a from further consideration. Of the remaining group, cold shoulder on one who has admitted that 58 subsequently participated in the experiment. Selection of subjects and assignment to experiyou are everything and he is nothing. In mental conditions was carried out on a random such a situation the variation in interaction basis. One subject was excluded from the results anticipation should play a very important because of suspicion of the manipulations and role. If future dependency is expected, the three others were randomly eliminated in order to optimum solution should be to avoid the equalize the cell frequencies. self-derogating other at all costs. One seem- Self-Evaluation Measure ingly effective avoidance tactic would be The 30 items composing the positiveness of selfto emphasize as many differences between presentation measure were taken from a measure yourself and this person as possible. This developed by Dickoff (1961). Each item was in leads to the prediction that the self-derogat- the form of a short self-descriptive phrase, such ing other should cause subjects expecting as "more emotionally mature than average," "inof others," etc. Original items had been interaction to become more self-enhancing. considerate assigned scale values for their degree of positiveness However, when the relationship has no by a group of independent raters. Fifteen of the future, subjects should be free to express items were highly positive and 15 were considered the pity felt for the low self-esteem other. by the raters to represent negative attributes. The required a person to assign any numOne way of expressing such feelings might instructions ber of points from 0 to 10 to each item according be to say, "Don't feel so bad; when you to how representative he felt the item to be of get right down to it none of us are so good." himself. Either a large number of points assigned In the present study, such an expression to positive items or a small number to negative would be manifested in a decreased emphasis items would thus be indicative of high self-regard. on one's positive features. Experimental Procedure To summarize, the major predictions of The experimenter met with groups of 8-10 subthis study can be formally stated as follows: jects who were not generally acquainted with each 1. Under conditions of no anticipated other. He described the experiment as one designed interaction, the positiveness of self-presenta- to explore the way in which two people get to each other. After discussing some remotely tion will be a linear function of the perceived know related topics, such as, types of status relaegotism of the other. A highly positive other tionships among members of a military hierarchy, will induce self-rating changes in the posi- subjects were told that they would each be taken tive direction and a self-demeaning other will to a separate booth. Once in the booth each would communicate in writing to a partner, produce changes in the negative direction. ostensibly one of the other members of the group, 2. Under conditions of anticipated inter- whose identity would be unknown. Such anonymity action, the positiveness of self-presentation was said to eliminate the influence of previous will be a curvilinear function of the other's acquaintance and to allow a written record of all communications to be obtained. egotism. A highly positive and highly negative The subjects were further told that once in other will elicit positive self-presentation, and the booths, half of them would find a form (the self-evaluation measure and associated questions) the medium other will produce little change. on which they were to describe themselves. These 3. The positiveness of self-presentation descriptions were then to be taken by the experiunder conditions of anticipated interaction menter to a partner, a member of the other half will be displaced in a more positive direction of the group. The other half of the group (in actuality, the entire group) were to find magathan under conditions of no anticipated zines in their booths which they could read while waiting. After receiving the communication, the interaction.

351

DETERMINANTS OF SELF-PRESENTATION second partner would fill out an identical set of materials which were then to be returned to the first communicator. Half of the experimental groups were then told that the conditions of anonymity would be broken neither during nor after the communication exchange. Persons in these conditions, in other words, would never get to know the identity of the partner with whom they communicated (no-anticipation condition). The remaining experimental groups were told that they would meet the partner after they had finished the communication exchange. Further, the experimenter expressed his expectation that the various pairs would become much better acquainted with each other during the semester, and that later in the semester they would be asked to participate in additional experimentation together (anticipation condition). The experimenter then took each subject to a separate soundproof booth, where a new issue of a women's magazine was found. After approximately a 10-minute interval the experimenter brought the stimulus materials to the subjects. There were three different sets of stimulus materials, one of which was delivered to each subject on a random basis. Each set contained the self-evaluation measure and three handwritten paragraphs, ostensibly filled out by actual partners. Regarding the self-evaluation measure, the number of points assigned to items by the partners in the three relevant conditions was based on an assessment of the averages assigned to these items by the initial sample from which subjects were drawn. Whereas the average (medium-positive) partner assigned the same number of points to the items as the average student in the sample, the egotistical (high-positive) other was much more positive than the average and the humble (lowpositive) partner differed from the average by the same amount but in the opposite direction. Care was taken that the absolute or total number of points assigned to all items by the fictitious partner was the same for all conditions. In addition, it seemed that high or low numbers of points assigned to a particular item by the partner could introduce unwanted bias. Thus two sets of IS different numbers, one for positive and one for negative items, were constructed for each of the three sets of stimulus materials. Within each of these conditions the sets of numbers were then rotated through the items from subject to subject. Thus each subject received a different number of points on the various items but the total number of points assigned to positive and negative items within any condition remained the same. In order to maintain as much constancy as possible, each particular configuration of digits for each of these conditions appeared twice, once in the anticipation condition and once in the no-anticipation condition. As noted, in addition to the self-evaluation measure each subject received three paragraphs supposedly written by her partner. These paragraphs were in response to questions concerning

high-school or prep-school activities, relationships with peers, and future plans. The content of these paragraphs was also varied for the three partner types. In the high-positive condition, the partner said that she enjoyed high school, got along perfectly with others, and looked forward to the future. The partner in the mediumpositive condition, however, felt that she enjoyed high school most of the time, got along with others fairly well, and had at least some doubts about the future. The low-positive partner felt very depressed about her high-school days, her relations with others, and her future. In spite of the differences in feeling tone, the attempt was made to keep various other aspects of the paragraphs as similar as possible. For example, each partner had participated in the very same activities with equal success. In addition, the total number of written words was constant across conditions and all paragraphs were written in the same hand. After the subjects had an ample opportunity to digest these materials, the experimenter reappeared, bringing with him a fresh set of materials identical to those filled out by the supposed partner. He again reminded each subject that these were to be sent to her partner. After a 10-15 minute interval the experimenter again returned to the booth and picked up the subject's self-ratings and asked her to make a set of ratings of her partner, ratings which the partner was not to see. These ratings were designed to assess the subject's perceptions of the partner. After another 10-15 minute interval, the experimenter again returned and gave the subject an additional questionnaire to fill out. This questionnaire dealt with such matters as the subject's awareness of shifts in self-presentation and other aspects of her behavior in the situation. Upon completion of the questionnaire each subject was asked to return to the room in which the group had originally met. After all subjects had returned the experimenter attempted to assess suspicion and revealed the rationale behind the experiment. All subjects were asked to cooperate by not divulging the details of the experiment to their classmates.

RESULTS Validation of the Manipulations There was good independent evidence that the manipulation of the partner's characteristics was effective. One of the "rating of partner" measures which each subject filled out after communication with her partner consisted of a series of antonyms. Each antonym was separated from its opposite by a 10-point scale, and 4 of the 18 antonyms ("self-centered versus humble," "personally modest versus conceited," etc.) were designed to tap perceived egotism of the partner. Combining each subject's ratings on these

KENNETH J. GEEGEN AND BARBARA WISHNOV

352

four items, perceived egotism scores were derived. An analysis of variance revealed that subjects in the high-positive condition saw the partner as significantly (p < .001) more egotistical than subjects in the mediumpositive condition, and these latter subjects saw their partner as significantly (p < .001) more egotistical than those in the low-positive condition. The validation of interaction anticipation is more difficult to establish. Unfortunately the items on the postexperimental questionnaires could not be used to make a reasonable assessment of the effectiveness of this manipulation. However, in that the major results would be difficult to interpret without assuming that the instructions for the two conditions were perceived differently, it seems safe to assume that the subjects believed what they were told. Changes in the Positiveness of SelfPresentation As noted, positiveness of self-presentation was measured with a 30-item scale, of which IS items were stated in a positive manner and IS in a negative. Although it might be expected that changes in the number of points assigned to negatively phrased items would be a mirror image of changes found in the positive items, there seemed to be good reason for analyzing the two sets of data separately. For example, the most obvious way a person might indicate that he feels TABLE 1 ANALYSIS OP VARIANCE OP POSITIVENESS AS EXPEESSED ON POSITIVE ITEMS Source

df

MS

F

Between subjects High, medium, low (B) Anticipation versus no anticipation (C)

53 2

338.75 1,936.15

8.13**

BXC Error (b) Within subjects Before-after (A) AXB

AXC AXBXC Error (w) *t