Interrelationships of Six Measures of Interpersonal Attraction

Abstract

Several models of human social interaction have attempted to relate a number cf verbal and nonverbal behaviors along a common dimension of interpersonal approachavoidance. Few studies, however, have attempted to measure more than one or two of these behaviors within a single experiment. The present study assessed the intercorrelations of two written and four behavioral indicators of interpersonal attraction, including a measure of actual physical approach. Thirty-six male and 36 female undergraduates interacted in randomly assigned o p posite-sexed pairs over a two-way closed circuit television system that allowed the simultaneous recording of each individual's verbal and nonverbal behaviors. The number of seconds each subject spent smiling, talking, and looking in the direction of the other during their initial 1-min television-mediated encounter was assessed by two independent raters viewing videotapes of each session. Afterwards, subjects were asked to fill out an anonymous questionnaire that assessed, among other things, the extent to which each subject found his partner physically attractive and how much he liked the other (using two 5-point bipolar adjective scales). Then each pair of subjects was led to a debriefing room and asked to be seated on a 5.5-ft. couch. The number of inches each subject sat from the center of the couch (subjects always sat on opposite sides of center) was recorded as the tendency to approach the other physically. The total number of seconds each subject spent smiling, talking, and looking in the direction of the other (averaged across raters), together with the physical distancing dam and questionnaire responses were submitted to a principle components (varimax rotation) factor analysis. Separate analyses were conducted for male and female subjects. For males two orthogonal factors with eigenvalues greater than one emerged from the factor analysis. The first factor was composed of the four behavioral indicators of attramion (rs = -.23 to .47). The second factor was comprised of the remaining two self-report measures (s = .74). For females three factors emerged with eigenvalues greater than one. The first was composed of female speaking, smiling, and looking time (ss = .25 to .45). The second factor was comprised of the two self-report measures ( r = 3 3 ) . The final factor was the distance measure existing essentially in isolation. In general males who tended to talk more to their female partners during their first encounter also tended to smile and look more and later tended to sit closer than those who talked less. The same pattern was noted for females' behavior toward the males, with the exception that females tended to choose a seating distance independent of their earlier talking, smiling, and looking behaviors. The latter finding reflected a tendency for those who entered the debriefing room first (generally the female) to allow partners maximum freedom in setting the final seating distance by sitting toward one end of the couch. For both males and females the measures of liking and physical attractiveness, while positively correlated, showed little relationship to the behavioral measures. This finding questions the appropriateness of using the and behavioral measures employed here as substitute measures of a unidimensional attraction construct. While here may be occasions when all six indicators vary in unison, in other situations they either measure different aspects of social relationships or are differentially sensitive to orher influences. Replication is needed.