Adolescent popularity: its relation to friendship characteristics and Its contagion among friends.

Abstract

Since the late 1990s, quantitative researchers have differentiated between popularity, defined as a form of status determined by the group consensus, and preference, which is based on emotional reactions of individual peers. Although a great deal of work has gone into establishing correlates and consequences of popularity, very little work has investigated how popularity interacts with particular types of social relationships. The current study aimed to investigate the relation between popularity and friendships by replicating and expanding upon early findings of Rose, Swenson, & Carlson (2004) and by providing initial findings relevant to the theory of popularity contagion. This theory, proposed in the current study, posits that popularity should spread among friends (or other relationship partners) spontaneously and regardless of behavioral changes. Data were collected annually between 6th grade and 12th grade from a total of 1062 participants as part of a larger longitudinal study of peer relationships. Peer nominations assessed adolescents' friendships as well as popularity, social preference, overt aggression, relational aggression, and prosocial behavior. Self-reports assessed friendship quality (i.e., companionship and conflict with best friends; Bukowski, Hoza, & Boivin, 1994). Replicating earlier findings of Rose et al. (2004), correlational results indicated that popular adolescents had a higher number of mutual friends, but did not seem to have friendships that lasted longer or were substantially higher in quality. Correlations between individual levels of popularity and mutual friends' levels of popularity were fairly high across all time points, indicating that popular individuals tended to have popular friends. Finally, longitudinal hypotheses generated from the theory of popularity contagion were mostly supported using path analyses, and showed (a) that individual popularity could be predicted by friends' popularity levels over time, even when controlling for stability of individual popularity; (b) that this prediction was not accounted for by behavioral contagion of aggressive or prosocial behaviors; and (c) that individual social preference generally could not be predicted by friends' preference levels over time. Implications, limitations, and directions for future research are discussed.