Abstract
Four measures of college student drug use are investigated: frequency of cannabis use, time as a user, range of polydrug experimentation and self-reported adverse consequences of use. Distributed properties of the measures are presented, with particular problems in the frequency of use criterion. Intercorrelations and factor analyses support the relative independence of the criteria, with substantial overlap only between range and time. Multiple regression analyses show that discrete sets of social psychological variables constituted relatively high and stable prediction of all criteria except frequency. Canonical correlation analyses suggest patterns of criteria interpreted as "normal" and "abusive" drug-using behavior. Implications for measurement development and for a multi-dimensional conception of drug use are discussed.
Get full access to this article
View all access options for this article.
