Quantitative measurement of perceived social distance through individual differences scaling

Abstract

The research was an attempt to demonstrate the efficacy of Individual Differences Scaling (INDSCAL) in developing a quantitative spatial model of perceived social relationships in groups. INDSCAL is based on the assumption that individuals perceive stimuli along a common set of dimensions, but that these individuals may differentially weight these dimensions in making their judgments. It was hypothesized that the scaling model would result in a multidimensional space of geometric distances between group members which accurately reflected the empirical judgments of social distance. Results are given concerning the relationship of personal characteristics to the dimensions of social perception and the individual weighting of these dimensions. A validation of the model was also attempted. Two groups engaged in sensitivity training at the university were used for study and data was collected in two steps. First, all group members filled out, at the beginning group sessions, a biographical inventory, the Guilford-Zimmerman Temperament Survey, and the Allport-Vernon-Lindzey Study of Values. Secondly, each group member made a paired comparison judgment of social distance between all members in his group at the end of approximately 13.5 hours of meeting time. In addition, each member ranked his fellow members in terms of their closeness to him. The paired comparison data was analyzed by the Carroll and Chang Individual Differences Scaling computer program and various relationships were analyzed between the scaling solutions and the personal characteristic variables and "closeness" rankings. The results support the underlying assumption of the scaling model since the multidimensional space of social distances accounted for a large proportion of the variance of the original distance judgments. The geometric distances between individuals in the multidimensional space had a strong correspondence with sociometric rankings made by the individuals in the smaller group, indicating that the scaling approach accurately reflected the social distances in the group. The attempts to interpret the dimensions of social perception and obtain correlates of individual dimension weights were encouraging but not conclusive. The implication of INDSCAL for sociometric measurement, interpersonal attraction, and social perception were discussed.