Alcohol Involvement and the Five-Factor Model of Personality: A Meta-Analysis

• Published in: Journal of drug education Vol. 37; no. 3; pp. 277 - 294
• Main Authors: Malouff, John M, Thorsteinsson, Einar B, Rooke, Sally E, Schutte, Nicola S
• Format: Journal Article
• Language: English
• Published: United States Baywood Publishing Company, Inc 01.01.2007
• Subjects: Age; Age Factors; Alcohol Drinking - psychology; Alcoholism; Alcoholism - psychology; Case Studies; Cross-Sectional Studies; Data Analysis; Drinking; Effect Size; Factor Analysis; Humans; Longitudinal Studies; Measures (Individuals); Meta Analysis; Models, Psychological; Personality; Personality Traits; Profiles; Sex; Sex Factors
• ISSN: 0047-2379
• DOI: 10.2190/DE.37.3.d

The purpose of this meta-analysis was to quantify the relationship between the Five-Factor Model of personality and alcohol involvement and to identify moderators of the relationship. The meta-analysis included 20 studies, 119 effect sizes, and 7,886 participants. Possible moderators examined included: five-factor rating type (self vs. other); study time-frame (cross sectional vs. longitudinal); sample type (treatment vs. non-treatment); type of alcohol involvement measure used; gender of the participants; and age of the participants. The meta-analysis showed alcohol involvement was associated with low conscientiousness, low agreeableness, and high neuroticism, a personality profile that: a) fits on the low end of a superordinate personality dimension that has been called self-control; and b) makes treatment difficult. Several significant moderators of effect size were found, including the following: studies of individuals in treatment for alcohol problems showed a more negative pattern of personality traits than did other studies; crosssectional studies, but not longitudinal studies, showed a significant effect for agreeableness, perhaps suggesting that low agreeableness may have a different causal link to alcohol involvement from the other factors; mixed-sex samples tended to have lower effect sizes than single-sex samples, suggesting that mixing sexes in data analysis may obscure effects. (Contains 6 tables.)