So far, researchers have compared those in and not in close relationships, or those in relationships that are close to various degrees, using measures of degree of closeness (e.g., Aron, Aron, & Smollan, 1992; Berscheid, Snyder, & Omoto, 1989). But in all of these approaches, the existence of a close relationship, the characteristics of its members, or the circumstances under which the relationship developed are not subject to experimental manipulation.
This article presents a practical methodology for creating closeness in an experimental context, so that whether or not a subject is in a relationship, the particular pairings of individuals in the relationship, and the circumstances under which the relationship develops all become manipulated independent variables.