European Journal of Social Psychology, zyxw Vol. 11, zyx 267-283 (1981) zy Social comparison on two dimensions A. VAN KNIPPENBERG zyx University of Groningen zyxw H. WlLKE University of Gmningen N. K. DE VRlES University of Groningen Abstract In this experiment social comparison on two task dimensions has been studied. The subjects received bogus feedback on their performance on two tests, one allegedly measuring creativity, the other alertness. By means of this feedback four relative position conditions were induced: scoring high on both tests, scoring low on both tests, scoring high on creativity and low on alertness, and vice versa. Anticipating either a co-operative or a competitive game, the subjects indicated their preference for one or the other task dimension and also expressed their preference for a comparison person. It was found that, as predicted, the subjects preferred and valued comparison dimensions on which they occupied a favourable position. With respect to preference for a comparison person, compensatory choices were obtained: on the dimension on which the subject performed well, they preferred an inferior other. On the dimension on which they performed poorly, they preferred a superior other. This pattern of choices was found in co-operation as well as in competition. Finally, upward preference was stronger in co-operative than in competitive conditions, particularly on the dimension on which the subject’s own score was low. These and other results were discussed in relation to theoretical social comparison notions. zyxw INTRODUCTION Social comparison behaviour has mainly been studied as a function of the subject’s position on one dimension. Most of the work in this field focused on the choice of comparison persons given relative performances on a task, or given opinions on some issue. Very little work has been devoted to comparison behaviour involving two or more comparison dimensions. Yet, in everyday