Analyzing Properties of Dyads Determinants of Similarity of Marijuana Use in Adolescent Friendship Dyads

Similarit between members of dyads has been studied for a variety of attributes, but not much is known about how similarity varies as a function of dyad charac teristics One reason for this is the existence of methodological problems that have restricted the analy ses of similarity. In this article, a multinomial logistic regression method is presented which allows the examination of similarity as a function of multiple independent variables. As an applic ation of the method, data on similarity of marijuana use among 1,367 adolescent frienciship dyads are analyzed Intensity of friendship bond is the only significant predictor of variability on similarity.