Strategies for measuring peer alcohol use within egocentric network analysis.

OBJECTIVE Compare different strategies for measuring perceived peer alcohol use within network research. PARTICIPANTS College students (N = 309). METHODS Collected demographic, personal alcohol use, and egocentric network data. Hierarchical linear regression analyses predicted egos' drinking behavior (AUDIT-C scores). RESULTS Egos' perceptions of nominated peers' binge-drinking frequency explained a significant portion of variance in egos' AUDIT-C scores, over and above perceptions of nominated peers' frequency/typical quantity of alcohol consumption and important demographic covariates. Using more than three items to assess perceived peer alcohol use did not offer statistically significant improvements and may add unnecessary response burden. CONCLUSIONS Utilizing the three-item approach (i.e., frequency, quantity, binge-drinking frequency) to assess perceived peer drinking explained more variance in egos' AUDIT-C scores, as opposed to one- and two-item assessments. Implementing this three-item assessment in egocentric network research can enhance programming efforts targeting reductions in binge drinking among college students within social contexts.