Date of Award


Document Type


Degree Name

Master of Arts (MA)




The intent of this investigation was to determine what judgment strategies are actually employed when judging the similarity of individuals, and the relationship of similarity estimates to a rated attraction index. The Stone-Coles (1970) multidimensional scaling technique was utilized. The input was based on the estimates of interperson similarity made by 16 pledges of a sorority. Each pledge also ranked the attractiveness of the other 15 pledges.

The four factor-dimensions which emerged from the judgmental data, similarity estimates, were interpreted as "dating conservatism-liberalism," "sociability," "non-dependability-dependability," and "dominance-submission". Only the "sociability" and "dominance-submission" dimensions were found to be linearly related to the rated attraction index. The more sociable, extroverted individuals and the more dominant individuals were better liked by all Ss regardless of their own positions on these same factor-dimensions. However, those Ss loading high on dating conservatism and those loading low on this dimension showed a preference for others with similar loadings. A similar finding was observed for the dependability dimension, showing a relationship between similarity on these dimensions and attraction.

Other findings presented were: (1) individuals who were seen as being more similar to all those in the group were ranked as being more attractive by the group; (2) those pairs whom the group saw as being most similar to each other ranked each other as more attractive; (3) for most judges, those individuals who the judge saw as being most similar to herself were also seen by the judge as more attractive.

A comparison was drawn between methodologies and results of Hogan and Mankin's (1970) investigation and the present investigation. Both Studies found personal interaction styles that were preferred across subjects. However, the present investigation also found a relationship between personality similarity and attraction which Hogan and Mankin purport to disprove. The difference in findings was interpreted partially as a result of the different methods used to measure similarity. The judgment dimensions used in the present investigation arose from Ss' estimations of overall similarity between pairs. The dimensions used by Hogan and Mankin were determined by the investigators prior to the data gathering and were based on a personality inventory. It was concluded that those dimensions which emerge from similarity judgments are more defensible as measurements of similarity in a similarity-attraction study than those selected to constitute an a priori definition of similarity.