The present study uses data from a meta-analytic archive of prospective longitudinal studies (N= 3206) to examine the association between negative affect, emotionality, and unconventionality on the use, misuse, and abuse of marijuana. For each of the three constructs, variables were divided into two categories--a "trait," which refers to a personality characteristic or attitudes and beliefs, and a "behaviour," which refers to something that a subject does. A total of 63 reports from 40 studies provided effect sizes on the bivariate relationship of one or more of these six categories with current or later marijuana use, misuse, or abuse. Pooling partially redundant estimates for independence reduced the dataset from 358 estimates to 93 aggregated effect sizes for the cross-sectional data included in the archive, and from 478 estimates to 73 aggregated effect sizes for the longitudinal data. The effect sizes obtained from the longitudinal data were modest, none above 0.20. Indeed, only those for unconventionality-trait and emotionality-trait and marijuana use were of sufficient magnitude and reliably to warrant attention. Of the behavioural constructs assessed using the cross-sectional data, only unconventionality was associated with marijuana use and misuse. The emotionality-behaviour construct and both unconventionality constructs were most strongly associated with marijuana abuse. The conceptual and methodological limitations of the study along with the implications of its findings for prevention are briefly discussed. (Author/publisher)
Abstract