LAUSR

INTRODUCTION We had 3 aims for this study: (1) to explore the relative impact of perceived drinking group norms versus campus drinking norms on university students' heavy alcohol use, (2) to examine how students' overestimation of their drinking group norms predicts individual heavy alcohol use, while controlling for actual group drinking, and (3) to test if the interaction between overestimation and actual group drinking predicts increased student drinking. Further, we adopted a longitudinal design to tease apart within- and between-person effects in the aforementioned relationships. METHODS University students (N = 118, Mage, 19.40, SD = 1.49, 60.2% women) were recruited in their peer drinking groups and all group members completed 3 online surveys in two-month intervals. Overestimation was calculated as the difference between students' perceptions of their drinking groups' HED and the actual reported HED of group members. RESULTS As expected, results demonstrated notable overestimation of group HED. Further, key results of multilevel growth curve modeling demonstrated that at time points when university students overestimated their drinking groups' HED more than they usually do (i.e., more than their average), they increase in their own HED. Similar within-person results were not found for campus drinking norms or actual group HED and the interaction between overestimation and actual group HED was not significant. CONCLUSIONS Findings emphasize the importance of incorporating the peer drinking group as a reference group in personalized normative feedback interventions.