Orientations toward relationships: Choosing a situation in

19
JOURNAL OF EXPERIMENTAL SOCIAL PSYCHOLOGY 21, 544-562 (1985) Orientations toward Relationships: Choosing a Situation in Which to Begin a Relationship PETER GLICK Lawrence University Received April 8, 1985 An investigation was conducted to examine how individuals choose situations in which to initiate relationships. It was hypothesized that individuals’ self- monitoring orientation and the characteristics of their relationship partners would interact to determine situation choices. Accordingly, low- and high-self-monitoring men were asked to rate their preferences for interacting in romantic and nonromantic situations with particular types of partners. The characteristics of the partners were varied along two dimensions: physical attractiveness and personality de- sirability. Results demonstrated that low-, relative to high-, self-monitoring in- dividuals’ preferences for interacting in romantic situations were more influenced by the personality characteristics of potential partners. In contrast, high-, relative to low-, self-monitoring individuals’ preferences for interacting in romantic situations were more influenced by the physical attractiveness of potential partners. Neither of these effects occurred when preferences for nonromantic situations were assessed. Implications of these differences for the initiation, maintenance, and dissolution of relationships were discussed. c’ 1985 Academic press. 1~. Systematic study of close personal relationships is increasing due to growing recognition of the importance of relationships in people’s lives (e.g., Berscheid & Peplau, 1983). Central to this effort is the formulation of conceptual frameworks for the scientific study of close relationships (Hinde, 1979; Kelley et al., 1983). In particular, Kelley et al. have argued that the nature of the interaction within a relationship is influenced by two types of causal conditions-personal and environmental-and their interaction. While investigators to date have focused almost exclusively This research was undertaken as a doctoral dissertation at the University of Minnesota. I gratefully acknowledge the guidance of my mentors-Mark Snyder and Ellen Berscheid- in the conduct of this research and writing of this manuscript. I also acknowledge the assistance of Nancy Welch in the preparation and editing of this manuscript, Maryann Westland for her assistance in conducting the experiment, and Alana Matwychuk, David Hamilton, and two anonymous reviewers for their helpful comments on earlier versions of this manuscript. Requests for reprints should be sent to Peter Glick. Department of Psychology, Lawrence University, Appleton, WI 54912. 544 0022-1031/85 $3.00 Copyright 0 1985 by Academic Press, Inc. AU rights of reproduction in any form reserved.

Transcript of Orientations toward relationships: Choosing a situation in

JOURNAL OF EXPERIMENTAL SOCIAL PSYCHOLOGY 21, 544-562 (1985)

Orientations toward Relationships: Choosing a Situation in Which to Begin a Relationship

PETER GLICK

Lawrence University

Received April 8, 1985

An investigation was conducted to examine how individuals choose situations in which to initiate relationships. It was hypothesized that individuals’ self- monitoring orientation and the characteristics of their relationship partners would interact to determine situation choices. Accordingly, low- and high-self-monitoring men were asked to rate their preferences for interacting in romantic and nonromantic situations with particular types of partners. The characteristics of the partners were varied along two dimensions: physical attractiveness and personality de- sirability. Results demonstrated that low-, relative to high-, self-monitoring in- dividuals’ preferences for interacting in romantic situations were more influenced by the personality characteristics of potential partners. In contrast, high-, relative to low-, self-monitoring individuals’ preferences for interacting in romantic situations were more influenced by the physical attractiveness of potential partners. Neither of these effects occurred when preferences for nonromantic situations were assessed. Implications of these differences for the initiation, maintenance, and dissolution of relationships were discussed. c’ 1985 Academic press. 1~.

Systematic study of close personal relationships is increasing due to growing recognition of the importance of relationships in people’s lives (e.g., Berscheid & Peplau, 1983). Central to this effort is the formulation of conceptual frameworks for the scientific study of close relationships (Hinde, 1979; Kelley et al., 1983). In particular, Kelley et al. have argued that the nature of the interaction within a relationship is influenced by two types of causal conditions-personal and environmental-and their interaction. While investigators to date have focused almost exclusively

This research was undertaken as a doctoral dissertation at the University of Minnesota. I gratefully acknowledge the guidance of my mentors-Mark Snyder and Ellen Berscheid- in the conduct of this research and writing of this manuscript. I also acknowledge the assistance of Nancy Welch in the preparation and editing of this manuscript, Maryann Westland for her assistance in conducting the experiment, and Alana Matwychuk, David Hamilton, and two anonymous reviewers for their helpful comments on earlier versions of this manuscript. Requests for reprints should be sent to Peter Glick. Department of Psychology, Lawrence University, Appleton, WI 54912.

544 0022-1031/85 $3.00 Copyright 0 1985 by Academic Press, Inc. AU rights of reproduction in any form reserved.

ORIENTATION TOWARD RELATIONSHIPS 545

on the influence of personal causal conditions and have generally ignored contextual factors and their interaction with personal factors, Berscheid (1985) has noted that as attraction researchers move to the study of relationship development, maintenance, and dissolution “it will be es- pecially disastrous to continue to do so.” (p. 467).

The context in which a relationship is initiated is particularly important because “the determinants and outcomes of an initial social encounter may cast their influence over the entire relationship-influencing not just whether it dies aborning, but if it lives, both its nature and its course” (Berscheid, 1985, p. 460). Through the processes of behavioral confirmation (see Darley & Fazio, 1980; Snyder, 1984, for recent reviews), the conditions surrounding the initial encounter are likely to continue to affect a rela- tionship. Therefore, in order to fully understand why a relationship is initiated and maintained, one must know the context in which it began- and this context may, more often than not, have been chosen by the individual who initiated the relationship.

Given that individuals often choose the situations in which they initiate a relationship, what factors govern situation choice? It has been suggested by Snyder (1981, 1983) and Snyder & Ickes (1985) that individuals may often choose to enter situations that facilitate the enactment of their own personalities and goals. For example, if the individual’s goal is the de- velopment of a romantic relationship with a particular partner, he or she may attempt to begin that relationship over dinner at a candlelit restaurant. This situation may reasonably be expected to be more likely to promote the development of a romantic relationship than a situation such as playing tennis. By choosing to begin the relationship in a romantic situation, the individual may attempt to use the influence of the situation on his or her own behavior and on the behavior of his or her partner to promote the development of a romantic relationship when such a relationship is desired with a particular partner. If, however, the individual does not wish to develop a romantic relationship with a specific person, such situations may be avoided.

Thus, the choice of a situation in which to begin a relationship may reflect the individual’s motivation to develop a certain type of relationship with a particular partner. Recent research into the initiation of dating relationships in which individuals were allowed to choose relationship partners from a set of alternatives (Snyder, Berscheid, & Glick, 1985) suggested that systematic differences may exist in the manner in which individuals choose relationship partners. One such difference was found to be related to individuals’ self-monitoring propensities (Snyder, 1974). Low-self-monitoring individuals, who are known to be generally concerned with acting in ways that reflect their own underlying attitudes, traits, values, and dispositions (Snyder, 1979), were found to be particularly concerned with the personality characteristics of potential dating partners.

546 PETER GLICK

In contrast, high-self-monitoring individuals, who are known to be par- ticularly concerned with managing the images of self they project to others and who tend to modify their behavior based upon interpersonal and situational appropriateness (Snyder, 1979) were found to be particularly concerned with the physical attractiveness of potential partners, at least in the dating context investigated. In fact, when forced to sacrifice one dimension for the other, over 80% of low-self-monitoring individuals preferred to date a partner who had a desirable personality even though that partner was physically unattractive while almost 70% of high-self- monitoring individuals chose a partner who was physically attractive even though that person possessed undesirable personality traits.

Given that low- and high-self-monitoring individuals are motivated to develop romantic relationships with different types of partners, these individuals should also differ in their preferences for interacting in romantic situations with particular types of partners. Specifically, low-self-monitoring individuals, who preferred to interact with personally desirable partners in a dating context, may attempt to initiate relationships with this type of partner in romantic situations and avoid romantic situations if they must interact with partners who do not possess desirable personality characteristics. Similarly, high-self-monitoring individuals may attempt to begin relationships with their preferred type of partner (physically attractive) in romantic situations and avoid such situations if they must interact with a physically unattractive partner. This behavior, if it does occur, would demonstrate that high- and low-self-monitoring individuals strategically employ situational forces in order to facilitate the development of romantic relationships with the type of relationship partner each finds most rewarding in such relationships.

However, because of the nature of information about a person’s internal attributes in comparison to the nature of physical attractiveness information, low-self-monitoring individuals may tend to be more reluctant to initiate relationships in romantic situations with their preferred type of partner (desirable personality) in comparison to high-self-monitoring individuals with their preferred type of partner (physically attractive). Information about another person’s internal attributes is generally less vivid and less complete than the information provided by the physical appearance of that individual. A picture is said to be worth a thousand words; visual information, such as the physical appearance of another person, is quickly and easily perceived and is generally more compelling than abstract information (Nisbett & Ross, 1980). Information about another person’s interior attributes is by its nature more abstract than visual information since such information must always be inferred from behavior, self- report, or second-hand observations. Furthermore, visual information about a person’s appearance is generally more complete than information gained about someone’s internal attributes-we quickly perceive what

ORIENTATION TOWARD RELATIONSHIPS 547

another person looks like on the outside, it takes much longer to get to know the many aspects of a person’s internal attributes.

Because the personality of a potential romantic partner is particularly important for low-self-monitoring individuals, they may be reluctant to steer a relationship in a romantic direction until they are convinced that a partner possesses a desirable personality. Thus, even when they have reason to believe that a partner may be personally desirable on some dimensions, this information may not be sufficiently compelling or complete enough for low-self-monitoring individuals. These individuals may first want to assess other important dimensions of the person’s personality through interaction in situations without romantic overtones before deciding whether or not to pursue a romantic relationship. In contrast, since the information that is most important for high-self-monitoring individuals is readily apparent and complete when it is visually represented, these individuals may be quite willing to steer a relationship with a physically attractive partner immediately into romantic situations regardless of the other’s personality characteristics.

The present study was conducted to examine the effects of the indi- vidual’s self-monitoring orientation and the characteristics of the partner on the individual’s choice of a situation in which to begin a relationship. It was hypothesized that when determining preferences for romantic situations the personality of the partner would be relatively more important to low-, than to high-, self-monitoring individuals and the physical at- tractiveness of the partner would be relatively more important to high-, than to low-, self-monitoring individuals. It was also hypothesized that when anticipating interaction with their preferred type of dating partners, low-self-monitoring individuals would be relatively less willing than high- self-monitoring individuals to initiate their relationships in romantic sit- uations. Lastly, it has been suggested that the differing orientations of high- and low-self-monitoring individuals toward relationships may not be engaged when considering interaction in nonromantic situations because such situations do not provide the same opportunities for the development of romantic relationships (see Snyder et al., 1985). Accordingly, it was hypothesized that the interactions between self-monitoring orientation and the characteristics of the partner that are expected to affect romantic situation preferences should not affect preferences for the nonromantic situations specified in this study.

In traditional laboratory experiments on the formation of dating re- lationships, experimenters have generally determined the context in which the acquaintance process takes place. Unlike the traditional laboratory paradigm in which the experimenter controls and manipulates the nature of the situation, this study employed a paradigm which allows the researcher to study the active role individuals have in determining the situations in which they live their lives (cf. Snyder, 1981). In the present study this

548 PETER GLICK

was accomplished by allowing individuals to choose the situation in which a proposed interaction with a particular kind of partner was to take place, These individuals, known to be relatively high or low in self-monitoring and not committed to a steady dating partner, were placed in an exper- imental situation with the following features: (a) participants were led to believe that they would interact with one of four potential partners who represented all possible combinations of two levels of physical at- tractiveness (high and low) and two levels of personality desirability (high and low), and (b) participants rated, for each potential partner, how much they would prefer to interact in a variety of romantic and nonromantic situations if they were to interact with that particular partner.

Participants

METHOD

Fifty male University of Minnesota undergraduates participated in individual sessions for extra credit in their introductory psychology course. Scores on the Self-Monitoring Scale (Snyder, 1974), collected previously in a large questionnaire study, were available for all participants. Twenty-five low- (scores < 11) and 25 high- (scores 3 15) self-monitoring individuals, none of whom had a steady dating partner, participated.’

Procedure Participants were contacted by phone and asked if they wished to participate in a “social

interaction” study in which participants would get acquainted with another person. When each participant arrived, a male experimenter (blind to the participant’s self-monitoring classification) explained that in order to study “social interaction processes in everyday, real-life settings,” he would be arranging an interaction between the participant and a female undergraduate to take place at a later date, after which the participant would fill out some questionnaires about the experience. The experimenter explained further that he wanted to give the participant as much choice as possible in the selection of the situation in which the interaction would take place. The participant was told that whichever 1 of 12 possible situations he was assigned, all arrangements and expenses would be taken care of by the experimenter. In addition, the experimenter said he realized that the type of situation the participant might prefer could depend a lot on what his partner was like, and, for this reason, before the participant rated his situational preferences, he would be allowed to examine information about his potential partners. The experimenter then retrieved four tile folders which, he explained, contained information about four women who were currently available as interaction partners, one of whom, depending on scheduling considerations, would be assigned as the participant’s partner. Since (supposedly) which of the four potential partners would be assigned to the participant could not be determined until everybody’s schedules were examined, the participant was asked to look at each file one at a time and then rate his preferences for the 12 situations if the interaction was going to take place with that particular partner described in the file. This procedure was repeated for each of the potential partners (presented in a randomly assigned order), so that preference for each of the 12 situations was rated four times (once for each potential partner).

Potential partners. Each file contained one page of information about a fictitious female

’ These cutoffs represent approximately kO.5 standard deviations from the mean score on the Self-Monitoring Scale.

ORIENTATION TOWARD RELATIONSHIPS 549

undergraduate. This information included a small (4.76 x 3.49~cm) yearbook-type photograph, ID number, class, first name of the person, and a summary of results from the Sociability, Emotionality and Other-Directedness scales of the (fictitious) “Minnesota Personality In- ventory.” All four of the potential partners were similar in race (all, like the participants, were white) and class in school (all sophomores). What was varied between the four potential partners was their physical attractiveness and the desirability of their personality characteristics such that between the four alternatives all possbile combinations of two levels of physical attractiveness (low and high) and two levels of personality desirability (low and high) were represented.

More specifically, the two physically attractive women (one described as having a desirable personality and the other an undesirable personality) averaged a physical attractiveness rating of 5.56 on a scale of 1 (physically unattractive) to 7 (physically attractive) as rated by eight independent pretest judges. In contrast, the physically unattractive women (one of whom possessed desirable, and the other undesirable, personality traits) averaged a rating of 2.56, t(7) = 9.46. p < .OOOl. Within each level of physical attractiveness the women were rated alike (each differed by less than one-third of a point, both t’s ns).

The personality scores in each folder were explained by a separate “Score Interpretation Sheet.” The scores of two of the women (one physically attractive, the other unattractive) were constructed so that they fell into the same categories on each of the three “subscales” of the personality inventory. Thus, for both of these women, participants read identical descriptions on the interpretation sheet. These women were described under the heading of “Sociability” as having a good ability to interact with others and as being open, relaxed, and honest with people. Under “Emotionality” they were said to have stable, easygoing dispositions. Finally, in terms of “Other-Directedness” these women were said to be concerned with getting along with others and willing to listen to other people. In contrast, the two remaining women (one physically attractive and one unattractive) each scored so that they fell into categories which described undesirable personality traits. Both of these women were described under the heading of “Sociability” as tending to adopt a reserved attitude toward strangers. Furthermore, their “Emotionality” score indicated that they tended to be moody and to react emotionally to events. Finally, the “Other-Directedness” score revealed that these women tended to be more concerned with themselves than with other people.

Situations. After the participant examined a file, he rated each of the 12 situations to indicate how much he would prefer to interact in the situation if he was going to interact with the person whose file he had just examined. The ratings were done on a 7-point scale which ranged from -3 (would very much prefer not to interact in this situation) to +3 (would very much prefer to interact in this situation). The 12 situations represented natural contexts within which interactions might normally take place. Six situations were intended to have little or no romantic overtones. They were (a) play a two-person sport such as racquetball, (b) have coffee at a croissant shop, (c) talk in a student lounge, (d) attend a spectator sports event at the University, (e) have lunch at a university cafeteria, and (f) attend a lunchtime lecture. The other six situations were intended to be romantic situations. They were (a) go dancing at a local bar, (b) go to a movie in the evening, (c) attend a concert, (d), (e), and (f) all involved dinner at a nice restaurant. As rated by 14 independent pretest judges on a 7-point scale ranging from -3 (not at all romantic) to +3 (highly romantic), the former 6 situations were judged to be nonromantic (M = -0.48), while the latter 6 situations were judged to be highly romantic (M = 1.93), t(13) = 4.95, p < .ool.

After the participant indicated his situational preferences for each potential partner, the experimenter informed him that more than one of the women might have a compatible schedule and, if this occurred, the experimenter would try to assign the participant a partner he preferred. The participant indicated his preferences for each of the partners (irrespective of the situation) on a 7-point scale ranging from -3 (would very much prefer

550 PETER GLICK

not to interact with this person) to +3 (would very much prefer to interact with this person). Next, the participant rated the physical attractiveness and personality desirability of the four potential partners. These ratings were both made on a 7-point scale ranging from - 3 (the most negative assessment) to + 3 (the most positive assessment). After these ratings were completed the experimenter probed for suspicion. All participants expressed the belief that they would be interacting with one of the women. Participants were then informed about the purpose of the experiment and told that no interaction would actually take place.

RESULTS

Manipulation Checks

To ensure that the potential partners were perceived as representing two levels of physical attractiveness (low and high) and two levels of personality desirability (low and high), participants’ ratings of the stimulus persons were analyzed. As expected, the alternatives who were intended to be physically attractive were rated as being much more attractive (M = 2.10) than the two who were intended to be unattractive (M = - 0.26), t(49) = 9.90, p < .OOOl. Moreover, within each level of attractiveness the alternatives were perceived to be about equally attractive (both t’s < 1.5, ns). In addition, the two alternatives who were intended to be perceived as having desirable personalities were rated as having better personalities (M = 2.01) than the two women who were intended to be perceived as having undesirable personalities (M = 0.14), t(49) = 8.61, p < .OOOl. Thus, both the physical attractiveness and the personality manipulations were successful.

Situation Preferences

Situation preferences were analyzed using a design that employed all factors simultaneously; a 2 (self-monitoring) X 2 (type of situation) x 2 (physical attractiveness of partner) x 2 (personality of partner) analysis of variance in which self-monitoring (high and low) was treated as a between-subjects factor and all remaining factors-type of situation (ro- mantic and nonromantic), physical attractiveness of partner (attractive and unattractive), and personality of partner (desirable and undesirable)- were treated as repeated measures. Means for this full analysis are reported in Table 1.

While Table 1 has been included to give readers a clear picture of the complete analysis, for the sake of clear exposition the theoretically relevant results of this analysis are presented individually as answers to a series of questions that follow from the previously stated hypotheses. In the course of this presentation all statistically significant results are discussed.

Preferences for Romantic and Nonromantic Situations

Was there any difference between preferences for the romantic and nonromantic situations? The analysis of situation preferences revealed

ORIENTATION TOWARD RELATIONSHIPS 551

TABLE I MEAN SITUATION PREFERENCES

Type of situation

Type of partner Romantic Nonromantic

Low self-monitors Physically attractive

Personally desirable 1.14 Personally undesirable 0.67

Physically unattractive Personally desirable 0.75 Personally undesirable 0.19

High self-monitors Physically attractive

Personally desirable 1.49 Personally undesirable 1.15

Physically unattractive Personally desirable 0.68 Personally undesirable 0.47

Note. N = 25 for each self-monitoring category. Range = -3 to +3.

0.48 0.33

0.46 0.15

0.43 0.20

0.52 0.29

a main effect for type of situation such that the romantic situations (A4 = 0.82) were preferred significantly more than the nonromantic situations (M = 0.36), F(1. 48) = 215.00, p < .OOOl. This preference for romantic situations may, at least in part, have been due to the procedure employed in the investigation, which effectively removed most of the costs (both psychological and monetary) of selecting a romantic situation. Not only did the experimenter promise to cover all expenses involved in the in- teraction, thus making the generally more expensive romantic situations (e.g., dinner at a restaurant, attending a concert) more attractive than the less expensive nonromantic situations (e.g., having lunch at a cafeteria, playing a two-person sport), but since the experimenter arranged and legitimized the interaction, any fears of rejection by potential partners should have been eliminated.

Effects of Partner’s Characteristics on Sifuafion Preferences

Were certain types of situations preferred when participants anticipated interaction with certain kinds of partners? The situation preference analysis revealed main effects for both physical attractiveness of the partner, F( I, 48) = 38.30, p < .OOOl, and for personality of the partner, F(1, 48) = 45.40, p < .OOOl, such that romantic situations were preferred when participants anticipated interaction with physically attractive and with personally desirable partners. These effects, however, must be interpreted with respect to two interactions. First, the analysis revealed a physical

552 PETER GLICK

attractiveness of partner x type of situation interaction, F(1, 48) = 31.57, p < .OOOl. Means for this interaction are presented in Table 2.

Individual contrasts revealed that this interaction occurred because romantic situations were preferred more when participants anticipated being with physically attractive, in contrast to physically unattractive, partners, F(1, 48) = 58.86, p < .OOOl, while there was no effect of physical attractiveness on nonromantic situation preferences, F( 1, 48) = 1.19, ns. Thus. the physical attractiveness main effect was carried by the influence of the partner’s appearance on the romantic situation preferences.

The second interaction revealed by the analysis was a personality of partner x type of situation interaction, F(1, 48) = 7.25, p < .Ol. Means for this interaction are reported in Table 3. Specific contrasts revealed that participants rated both the romantic situations, F(1, 48) = 41.95, p < .OOOl, and the nonromantic situations, F(I) 48) = 10.74, p < .Ol, more highly when they anticipated being with a partner who had a desirable, as opposed to an undersirable, personality. However, the fact that the overall interaction was significant indicates that the influence of the per- sonality of the partner was significantly stronger on romantic, than on nonromantic, situation preferences.

Effect of Self-Monitoring on Situation Preferences

Did high- or low-self-monitoring individuals differentially prefer one type of situation over the other? The analysis revealed no main effect for self-monitoring, F(1, 48) = 1.53, ns. The self-monitoring x type of situation interaction was also nonsignificant, F(1, 48) = 1.01, ns. Thus, there were no differences between high- and low-self-monitoring individuals in their general preferences for the different types of situations.

Effect of Self-Monitoring and Physical Attractiveness of Partner on Situation Preferences

It was predicted that high-, compared to low-, self-monitoring individuals would be relatively more influenced by the physical attractiveness of potential partners when rating their preferences for romantic situations,

TABLE 2 EFFECT OF PHYSICAL ATTRACTIVENESS OF PARTNER ON SITUATION PREFERENCES

Type of situation

Type of partner Romantic Nonromantic

Physically attractive 1.14 0.35 Physically unattractive 0.52 0.36

No&. N = 50. Range = -3 to +3.

ORIENTATION TOWARD RELATIONSHIPS 553

TABLE 3 EFFECT OF PERSONALITY OF PARTNER ON SITUATION PREFERENCES

Type of situation

Type of partner Romantic Nonromantic

Desirable personality 1.02 0.47 Undesirable personality 0.64 0.24

N&e. N = 50. Range = - 3 to +3.

but that this interaction would not occur when rating their preferences for nonromantic situations. If this was the case, then there should have been a significant three-way interaction between individuals’ self-monitoring orientation, the physical attractiveness of the partner, and the type of situation.2 As predicted, a significant self-monitoring x physical attrac- tiveness of partner x type of situation interaction was found, F(1, 48) = 7.08, p = .Ol. To test whether this interaction was of the precise nature predicted, the situation preferences of high- and low-self-monitoring in- dividuals were analyzed separately. Means for these analyses are reported in Table 4.

The analyses revealed a significant type of situation x physical at- tractiveness of partner interaction for the situation preferences of high- self-monitoring individuals, F(1, 24) = 27.35, p < .OOOl, as well as for low-self-monitoring individuals, F(1, 24) = 7.52, p < .05. A further breakdown of the interaction among the preferences of high-self-monitoring individuals revealed that romantic situations were preferred more strongly when high-self-monitors anticipated interaction with physically attractive, as opposed to physically unattractive, partners, F(1, 24) = 25.59, p < .OOOl, but the physical attractiveness of the partner had no effect on high-self-monitoring individuals’ preferences for nonromantic situations,

’ The reader may feel that the two predicted three-way interactions found here indicate that there is a four-way interaction between self-monitoring, the situation being rated, the personality of the partner, and the physical attractiveness of the partner. This, however, is not the case. A four-way interaction would mean that the three-way interaction between (for instance) self-monitoring, the situation being rated, and the attractiveness of the partner was modified by the fourth variable (the personality of the partner). There was no reason to suspect that this would occur. In other words, the greater preference that high-self- monitoring individuals were expected to show for romantic situations when anticipating interaction with a physically attractive partner should have occurred whether DY not that partner possessed a desirable personality. Similar reasoning holds true for the three-way interaction between self-monitoring, the situation being rated, and the personality of the partner. Therefore, a four-way interaction was not expected and, in fact, did not occur, F( 1, 48) = 0.24, ns. Furthermore, none of the three-way or two-way interactions involving both personality of the partner and physical attractiveness of the partner as factors were significant, all F’s < 1.10, ns.

554 PETER GLICK

TABLE 4 EFFECT OF SELF-MONITORING AND PHYSICAL ATTRACTIVENESS OF PARTNER ON SITUATION

PREFERENCES

Type of situation

Type of partner Romantic Nonromantic

Low self-monitors Physically attractive 0.92 0.40 Physically unattractive 0.46 0.31

High self-monitors Physically attractive 1.34 0.31 Physically unattractive 0.57 0.41

Note. N = 25 for each self-monitoring category. Range = -3 to +3.

F(1, 24) = 0.76, ns. Similarly, a breakdown of the type of situation x attractiveness of partner interaction among low-self-monitoring individuals revealed that they, too, preferred romantic situations more when antic- ipating interaction with a physically attractive, as opposed to an unat- tractive, partner, F(1. 24) = 14.44, p < .OOl, and that the physical attractiveness of the partner had no effect on low-self-monitors’ preferences for nonromantic situations, F( 1, 24) = 1.19, ns.

Thus, the two-way interactions between type of situation and attrac- tiveness of partner were composed of similar simple effects for both high- and low-self-monitoring individuals. However, the existence of a significant three-way self-monitoring x type of situation x attractiveness of partner interaction indicates that the type of situation x physical attractiveness of partner interaction was significantly stronger for high-, as opposed to low-, self-monitoring individuals.

Furthermore, a planned comparison of high- and low-self-monitoring individuals’ increased desire to interact in romantic situations when an- ticipating interaction with physically attractive, as opposed to unattractive, partners demonstrated that the romantic situation preferences of high-, in comparison to low-, self-monitoring individuals were affected significantly more by the physical attractiveness of the partner. Difference scores were obtained by subtracting romantic situation preferences when an- ticipating interaction with physically unattractive partners from preferences for such situations when anticipating interaction with physically attractive partners. The mean difference score of high-self-monitoring individuals was significantly higher than that of low-self-monitoring individuals, t(48) = 1.81, p < .05. Difference scores constructed in the same manner for the nonromantic situation preferences of high- and low-self-monitoring in- dividuals yielded no significant difference, t(48) = - 1.38, ns.

ORIENTATION TOWARD RELATIONSHIPS 555

Effect of Self-Monitoring and Personality of Partner on Situation Preferences

It was predicted that low-, in comparison to high-, self-monitoring individuals would be relatively more influenced by the personality of potential partners when rating their preferences for interacting in romantic situations but that this effect would not occur when individuals rated their preferences for interacting in nonromantic situations. As predicted, this three-way self-monitoring x personality of partner x type of situation interaction was significant, F(1) 48) = 3.97, p = .05. Once again, separate analyses of situation preferences were performed for high- and low-self- monitoring individuals in order to determine whether the interaction consisted of the predicted effects. Means for these analyses are presented in Table 5. These analyses revealed that, as predicted, the three-way interaction consisted of a type of situation x personality of partner interaction among low-self-monitors, F(1, 24) = 10.15, p < .Ol, and an absence of this interaction among the situation preferences of high-self- monitors, F(1, 24) = 0.52, ns.

Individual contrasts revealed that low-self-monitoring individuals pre- ferred romantic situations significantly more when anticipating interaction with a partner who possessed a desirable, as opposed to a partner who possessed an undesirable, personality, RI, 24) = 42.12, p < .OOOl. This effect was also observed among low-self-monitoring individuals’ preferences for nonromantic situations, F( 1, 24) = 10.37, p < .Ol. However, the significant type of situation X personality of partner interaction among low-self-monitoring individuals’ situation preferences indicates that the effect of the personality of the partner was significantly greater on the romantic, as compared to the nonromantic, situation preferences of low- self-monitors.

While high-self-monitoring individuals similarly preferred nonromantic situations more when anticipating interaction with personally desirable,

TABLE 5 EFFECT OF SELF-MONITORING AND PERSONALITY OF PARTNER ON SITUATION PREFERENCES

Type of situation

Type of partner Romantic Nonromantic

Low self-monitors Desirable personality 0.94 0.47 Undesirable personality 0.44 0.24

High self-monitors Desirable personality 1.09 0.47 Undesirable personality 0.83 0.24

Note. N = 25 for each self-monitoring category. Range = -3 to +3.

556 PETER GLICK

as opposed to undesirable, partners, F(I, 24) = 11.22, p < .Ol, this effect did not significantly increase when high-self-monitoring individuals indicated their preferences for interacting in romantic situations, F(1, 24) = 15.52, p < .OOl. That these two contrasts were not significantly different from each other is indicated by the fact that the type of situation x personality of partner interaction was nonsignificant for high-self- monitoring individuals.

Furthermore, a planned comparison of difference scores revealed that the increased desire to interact in romantic situations when anticipating interaction with personally desirable, as opposed to personally undesirable, partners was greater for low-, in comparison to high-, self-monitoring individuals, t(48) = 1.79, p < .05. A similar contrast for the nonromantic situation preferences revealed no significant differences between low- and high-self-monitoring individuals, t(48) = -0.26, ns. Thus, the per- sonality of the partner had a significantly greater effect on the romantic situation preferences of low-, in comparison to high-, self-monitoring individuals.

Effect of Self-Monitoring on Romantic Situation Preferences When Anticipating Interaction with Preferred Partners

Another prediction regarding preferences for romantic situations was that low-self-monitoring individuals would be relatively reluctant to initiate relationships in romantic situations, even when they had reason to believe that a partner possessed some desirable personality traits, in comparison to high-self-monitoring individuals who anticipated interaction with phys- ically attractive partners. In other words, it was hypothesized that when anticipating interaction with their most preferred types of partners (Snyder et al., 1985),3 low-self-monitoring individuals would be less willing than

3 Snyder, Berscheid, and Glick (1985) demonstrated that high-, relative to low-, self- monitoring individuals’ choices of dating partners are based more on physical attractiveness of the partner and that low-, relative to high-, self-monitoring individuals tend to base their partner choices on personality considerations. In the present investigation participants rated their preferences for interacting with each of the four types of potential partners. Of these four, one possessed both an attractive appearance and a desirable personality and one possessed neither of these attributes. Since, in these two cases. choices based either on appearances or personality would have been the same. one would expect that both high- and low-self-monitoring individuals would have preferred the former partner and wished to avoid the latter. In fact, both high- (M = 2.16) and low- (M = 2.52) self- monitoring individuals preferred the physically attractive partner with the desirable personality over all other alternatives (pairwise comparisons with all other alternatives were all significant for both high- and low-self-monitoring individuals, all r’s(49) > 2.14, p < .05). Similarly, neither high- (M = -0.70) nor low- (M = -0.16) self-monitoring individuals desired to interact with the physically unattractive partner who possessed an undesirable personality (pairwise comparisons with all other alternatives were all significant for both high- and low-self-monitoring individuals, all l’s(49) < - 2.93, p < .Ol). Among preferences for the remaining two partners, however, self-monitoring differences should have emerged if the

ORIENTATION TOWARD RELATIONSHIPS 557

high-self-monitoring individuals to interact in romantic situations. While low-self-monitoring individuals preferred (irrespective of the situation) to be with partners with desirable personalities (M = 1.80) as much as high-self-monitoring individuals preferred partners who were physically attractive (M = 1.84), t(48) = 0.17, ns, a planned comparison revealed that when anticipating interaction with their preferred partners, high-self- monitoring participants (M = 1.34) preferred romantic situations signif- icantly more than did low-self-monitoring participants (M = 0.94), r(48) = 2.13, p < .05.

DISCUSSION

The present investigation examined high- and low-self-monitoring in- dividuals’ preferences for two types of situations-romantic and non- romantic-in which to begin personal relationships with particular types of partners. Clearly, the situations employed in this investigation differ on more than just one dimension. The romantic situations were judged to be typical contexts for romantic relationships. As such, they differ from the nonromantic situations in many ways. For instance, the romantic situations would probably involve longer interactions, require greater monetary expense, provide less structured interaction, take place at a different time of day, and offer more opportunities to pursue future interaction. In fact, it seems likely that many of these differences contribute to the reasons why these situations are typical contexts for romantic relationships-they may generally be chosen as preferred situations in which to develop romantic relationships because of the opportunities they provide. Accordingly, it was hypothesized that factors that lead individuals to pursue romantic relationships with others would also lead them to prefer to interact in romantic situations in order to take advantage of the particular opportunities that these situations provide.

Several findings supported this prediction. First, high-self-monitoring individuals, who are known to place greater emphasis on having a romantic partner who is physically attractive (Snyder et al., 1985), were significantly

results of Snyder, Berscheid, and Click (1985) are correct. Specifically, high-self-monitoring individuals should have been relatively more influenced by information about physical appearances in comparison to low-self-monitoring individuals. In contrast, low-self-monitoring individuals should have been relatively more influenced by personality information than high-self-monitoring individuals. A 2 (self-monitoring) x 2 (type of partner) analysis of variance in which type of partner (physically attractive/undesirable personality and desirable personality/physically unattractive) was treated as a repeated measures factor revealed that this was indeed the case. The analysis revealed a highly significant self-monitoring x type of partner interaction, F(l,48) = 6.60, p < .02, with means in the expected directions- supporting the prediction that high-self-monitoring individuals were more influenced by physical attractiveness and low-self-monitoring individuals by personality information. These results serve as a replication of the findings of Snyder, Berscheid, and Glick (1985).

558 PETER GLICK

more sensitive than low-self-monitoring individuals to the physical at- tractiveness of their partner when rating their preferences for interacting in romantic situations. Similarly, low-self-monitoring individuals, who are known to place greater emphasis on having a romantic partner who is personally desirable (Snyder et al., 1983, were significantly more senstive than high-self-monitoring individuals to the personality char- acteristics of their partner when rating their preferences for interacting in romantic situations. Neither of these effects occured when participants rated their preferences for interacting in nonromantic situations.4

These results are informative not only about what kinds of partners different types of individuals find rewarding for romantic relationships, but also about how these individuals actively structure their interactions with desirable partners in order to create romantic relationships. By choosing to enter romantic situations, the individual may deliberately use the power of the situation over his or her own behavior and the behavior of a relationship partner to further the individual’s own goals for the relationship.5 Furthermore, this experiment demonstrates that

4 For researchers in the interpersonal relationships area it is important to take note of the lack of influence of the individual’s orientation toward relationships and the characteristics of the partner on preferences for interacting in nonromantic situations. When anticipating interaction in nonromantic situations, it appears that the differing orientations of high- and low-self-monitoring individuals toward romantic relationships simply were not engaged. Furthermore, a factor known to be important in romantic relationships-the physical at- tractiveness of the partner (Berscheid & Walster, 1979)--had no effect on preferences for the nonromantic situations employed in this study. This implies that the considerations individuals typically. have when contemplating a romantic relationship are likely to be operative only when the relationship may take place in a romantic context. Thus, these results not only provide evidence for the argument that relationships must be studied in natural contexts (cf. Berscheid, 1985), but also suggests that the context of a proposed interaction should, in addition to being naturalistic, be typical of the type of relationship being studied. The nonromantic situations in the present study, which were naturalistic, still did not engage romantic considerations. This is also consistent with the findings of Snyder, Berscheid, and Glick (1985).

’ There is an alternative explanation of the results which would suggest that high-self- monitoring individuals were only responding to immediate concerns about how they might be seen by others in the chosen situation. Specifically, since high-self-monitoring individuals are more concerned with the impressions other people form of them and because being associated romantically with a physically attractive partner can lead to more favorable impressions (Sigall & Landy. 1973). the high-self-monitoring participants may only have been concerned about being seen in romanric situations with attractive partners rather than any long-term implications their situation choice might have for the formation of a relationship. However, while this is certainly a viable alternative explanation for the behavior of the high-self-monitoring participants, it does not explain why the low-self-monitoring participants preferred to interact in romantic situations with personally desirable partners. Therefore, at least half of the participants did seem to consider the implications that interaction in romantic situations might have for the formation of ongoing relationships, and it is parsimonious to conclude that such considerations were, at least in part, responsible for the behavior of high-self-monitoring individuals as well. In fact, both explanations may

ORIENTATION TOWARD RELATIONSHIPS 559

these goals are the product of an interaction between the individual’s personality (in this case, their self-monitoring orientation) and the char- acteristics of the potential relationship partner (in this case, the partner’s physical and personality attributes).

Implications for the Study of Personality and Social Behavior

Recent theoretical discussions have focused on the active role people may play in choosing the situations in which they live their lives and the implications that this activity has for understanding the role of per- sonality in social behavior (Snyder, 1981, 1983; Snyder & Ickes, 1985). In particular, Snyder and Ickes (1985) have argued that people may choose situations that facilitate the expression of their own personalities and goals. Because people may typically avoid situations that might pressure them to behave in ways that are incongruent with their personality traits, the traditional methods of social psychology in which the nature of the situation is manipulated by the experimenter rather than chosen by the subject may underestimate the role of personality in determining social behavior in everyday life.

While a number of investigations (e.g., Snyder & Gangestad, 1982) have demonstrated that people do choose to enter situations that encourage the expression of their personality traits, little attention has been paid to the way in which the individual’s personality and the characteristics attributed to a relationship partner interact to determine situation choices (for an exception see Cantor, Mackie. & Lord, 1983-1984). The results of the present experiment, like the results achieved by Cantor et al. (1983-1984), demonstrate that the individual’s choice of situation cannot be separated from considerations of the partner with whom the individual expects to interact in the situation. Individuals not only choose situations and choose partners, but match situations and partners. Thus, while the study of how people choose situations is to be encouraged, it should be undertaken with the recognition that people’s situation choices are likely to be based not only upon considerations of how the situation will affect their own behavior, but also the impact the situation is likely to have on their partners’ behavior and, therefore, on the nature of their rela- tionships with those partners. Thus, the appropriate unit of study may not be the individual’s choice of situations, but rather the choice of particular partner-situation combinations.

be correct. High-self-monitoring individuals’ preference for forming romantic relationships with physically attractive partners may be mediated by the social rewards (i.e., favorable impressions) that such an association creates. The high-self-monitoring participants in this investigation stood to gain more of this type of reward over both the short-term and the long-term by choosing to be in romantic situations with physically attractive partners.

560 PETER GLICK

Implications for Relationship Development

This investigation, coupled with the previous research of Snyder et al., (1985), followed the process of relationship initiation through two important steps: the choice of a partner and the choice of a situation in which to begin the relationship. At each step the paths chosen by low- and high-self-monitoring individuals have been seen to diverge-setting their relationships on courses which may become increasingly different.

Consider first the case of low-self-monitoring individuals who tend to seek out partners possessing particular internal attributes. Even though low-self-monitoring participants in the present study had access to highly diagnostic personality information, they were still relatively reluctant to initiate relationships in romantic situations with their most preferred type of partner (personally desirable) when compared to high-self-monitoring individuals who anticipated interaction with their preferred type of partner (physically attractive). Under normal circumstances, information about internal attributes must be gained through the time-consuming process of interpersonal interaction. Therefore, given the importance they attach to such information, low-self-monitoring individuals’ tendency to be less willing to begin relationships in romantic settings is likely to be even stronger than was observed in this experiment. Therefore, low-self-mon- itoring individuals may tend to interact many times with a potential partner to discover what that person is “really like” before attempting to steer the relationship in a romantic direction. In contrast, because what high-self-monitoring individuals value most in a dating partner (physical attractiveness) can be so quickly perceived, they appear to feel free to immediately define relationships with physically attractive partners as romantic ones by choosing to begin those relationships in romantic situations.

How might this difference in the manner in which high- and low-self- monitoring individuals initiate relationships affect the subsequent de- velopment of those relationships? The greater amount of time and effort that low-, in contrast to high-, self-monitoring individuals may devote to the development of their romantic relationships may lead to enhanced commitment by increasing the amount that low-self-monitoring individuals have invested in their relationships (Kelley, 1983; Rosenblatt, 1977; Rusbult, 1980). Also, because the more time spent pursuing a single relationship, the less time a person has to pursue other relationships, low-, in comparison to high-, self-monitoring individuals may have fewer alternative partners available once they define a relationship as a romantic one (Kelley, 1983). Both of these factors may make the relationships of low-self-monitoring individuals less vulnerable to dissolution than the relationships of high- self-monitoring individuals (Berscheid & Campbell, 1981; Levinger, 1976; Thibaut & Kelley, 1959).

Evidence for these conjectures does, in fact, exist. Snyder and Simpson (1984) found that among people who were going steady with a single

ORIENTATION TOWARD RELATIONSHIPS 561

dating partner, low-self-monitoring individuals had dated their partners for almost twice as long, on average, as high-self-monitoring individuals. Furthermore, high-, in comparison to low-, self-monitoring individuals reported a greater willingness to develop dating relationships with al- ternative dating partners. These results are consistent- with the notion that the manner in which they initiate relationships leads low-self-monitoring individuals to develop relationships that are characterized by greater commitment and that last longer than the relationships of high-self-mon- itoring individuals.

Thus, the results of the present experiment are entirely consistent with a larger pattern of results that employed both laboratory experiments (Snyder et al., 1985) and more naturalistic studies (Snyder & Simpson, 1984). Taken together, these investigations demonstrate that differences observed at the later stages of relationships may be traced to the continuing influences of differences in the manner in which individuals initiate re- lationships which, in turn, follow from individuals’ characteristic ori- entations toward relationships. Through their choice of partners and their choice of situations in which to begin relationships, high-, and low-self- monitoring individuals start, from the very beginning, to follow paths that become increasingly divergent, but which follow logically from dif- ferences in the manner in which these individuals begin their relationships. By examining the processes involved in the initiation of relationships it not only has been possible to predict how their courses may diverge, but also to demonstrate how, through their choices of situations in which to interact, individuals strive to use the influence of situations over their own and their partner’s behavior to actively direct the courses of their relationships in ways that further the enactment of their own personalities and goals.

REFERENCES

Berscheid, E. (1985). Interpersonal attraction. In G. Lindzey & E. Aronson (Eds.), Handbook of social psychology (3rd ed.). Reading, MA; Addison-Wesley.

Berscheid, E., & Campbell, B. (1981). The changing longevity of heterosexual close re- lationships: A commentary and forecast. In M. Lerner (Ed.). The justice motive in times of scarcity and change. New York: Plenum.

Berscheid. E., & Peplau, L. A. (1983). The emerging science of relationships. In H. H. Kelley, E. Berscheid, A. Christensen, J. H. Harvey, T. L. Huston, G. Levinger, E. McClintock, L. A. Peplau, & D. R. Peterson, Close relationships. San Francisco: Freeman.

Berscheid. E., & Walster, E. (1978). Interpersonal attraction (2nd ed.). Reading, MA: Addison-Wesley.

Cantor, N.. Mackie, D.. & Lord, C. G. (1983-1984). Choosing partners and activities: The social perceiver decides to mix it up. Social Cognition. 2(3), 256-272.

Darley. J. M., & Fazio. R. H. (1980). Expectancy confirmation processes arising in the social interaction sequence. American Psychologist, 35, 867-881.

Hinde, R. A. (1979). Towards understanding relationships. New York/London: Academic Press.

562 PETER CLICK

Kelley, H. H. (1983). Love and commitment. In H. H. Kelley, E. Berscheid, A. Christensen, J. H. Harvey, T. L. Huston, G. Levinger, E. McClintock, L. A. Peplau, & D. R. Peterson. Close relationships. San Francisco: Freeman.

Kelley, H. H., Berscheid, E., Christensen, A., Harvey, J., Huston, T. L., Levinger, G., McClintock, E., Peplau, L. A., & Peterson, D. R. (1983). Close relationships. San Francisco: Freeman.

Levinger, G. (1976). A social psychological perspective on marital dissolution. Journal of Social Issues, 32, 21-47.

Nisbett, R., & Ross, L. (1980). Human inference: Strategies and shortcomings of social judgment. Englewood Cliffs, NJ: Prentice-Hall.

Rosenblatt, P. C. (1977). Needed research on commitment in marriage. In G. Levinger & H. L. Rausch (Eds.), Close relationships: Perspectives on the meaning of intimacy. Amherst: Univ. of Massachusetts Press.

Rusbult, C. E. (1980). Commitment and satisfaction in romantic associations: A test of the investment model. Journal of Experimental Social Psychology, 16, 172-186.

Sigall, .I., & Landy, D. (1973). Radiating beauty: Effects of having a physically attractive partner on person perception. Journal of Personnlity and Social Psychology. 28, 218- 224.

Snyder, M. (1974). The self-monitoring of expressive behavior. Journal ofPersonality and Social Psychology, 30, 526-537.

Snyder, M. (1979). Self-monitoring processes. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 12). New York: Academic Press.

Snyder, M. (1981). On the influence of individuals on situations. In N. Cantor & J. F. Kihlstrom (Eds.), Cognifion, social interaction, and personality. Hillsdale, NJ: Erlbaum.

Snyder. M. (1983). The influence of individuals on situations: Implications for understanding the links between personality and social behavior. Journal of Personality, 51, 497- 516.

Snyder, M. (1984). When belief creates reality. In L. Berkowitz (Ed.), Advances in PX- perimental social psychology (Vol. 18). New York: Academic Press.

Snyder, M., Berscheid, E., & Glick, P. (1985). Focusing on the exterior and the interior: Two investigations of the initiation of personal relationships. Journal of Personality and Social Psychology, 48, 1427-1439.

Snyder, M., & Gangestad, S. (1982). Choosing social situations; Two investigations of self-monitoring processes. Journal ofPersonality and Social Psychology, 43, 123-135.

Snyder, M., & Ickes, W. (1985). Personality and social behavior. In G. Lindzey & E. Aronson (Eds.), Handbook of social psychology (3rd ed.). Reading, MA: Addison- Wesley.

Snyder, M., & Simpson, J. A. (1984). Self-monitoring and dating relationships. Journal of Personality and Social Psychology, 47, 1281-1291.

Thibaut, J. W., & Kelley, H. H. (1959). The social psychology of groups. New York: Wiley.