Abstract
Social psychologists have often used the technique of forming hierarchies in small groups by taking all the possible dyadic combinations of persons in a group and establishing dominance in each dyad. These dyad dominance ranks are then used as the basis for generalizing to dominance rankings when the people are brought together in one large group. This study addressed the methodological problem that occurs when both dyadic subjects are not strangers to the procedures and each other. Type 1 dyadic encounters, in which both subjects were strangers to each other and the procedures, yielded equal probabilities of either dominating. Type 3 dyadic encounters, in which both subjects were strangers to each other but had been in one previous dyad and, therefore, were familiar with the laboratory and the procedures, also yielded equal probabilities of either dominating. In Type 2 dyadic encounters, in which one person had been in a dyad already and the other had not, however, the experienced subject was more likely to dominate.