The effects of experimental abstinence from alcohol on abstract thinking were investigated. Subjects were 185 female and male University of Nebraska-Lincoln undergraduates enrolled in lower division psychology courses. Instruments included Shipley Institute of Living Scale, Numerical Reasoning Test, Symbolic Reasoning Test, Michigan Alcoholism Screening Test, a demographic questionnaire, and a
... [Show full abstract] drinking log. Two experimental groups and two control groups were created. Subjects in the first experimental condition (n = 55) took a pretest, abstained from alcohol for a month, then posttested. Subjects in the second experimental condition (n = 55) abstained for a month, took the pretest, returned to their normal drinking for a month, then posttested. Subjects in the social drinking control condition (n = 55) and the abstaining control condition (n = 20) took tests but did not modify their drinking during the intertest interval. All subjects recorded their daily drinking for the 90 day course of the study.^ Statistical analyses were based on a 4 (group membership) x 2 (gender) repeated measures multivariate analyses of variance design (MANOVA). Other analyses employed univariate analyses of variance (ANOVAS), analyses of covariance (ANCOVAS), and Pearson correlation procedures.^ Previous findings of modest inverse correlations between usual number of drinks and Shipley Abstraction score were partially replicated. Main effects associated with group membership and practice emerged as predicted. Number of drinks reported consumed during the abstinence period correlated inversely with Numerical Reasoning performance for females. Persistence of drinking into the abstinence period correlated inversely with Numerical Reasoning performance for males. Social drinking subjects who abstained from alcohol between tests improved on Numerical Reasoning; other subjects did not. Hypotheses predicting similar findings across other dependent variables were not supported. Post hoc analyses failed to support a dose-threshold effect to account for these findings. Differential Numerical Reasoning performance apparently did not reflect a speed-mediated process. Findings were discussed in the context of current research and suggestions for future research were offered.