Biological, psychological and sociocultural factors discriminating between alcoholic and non-alcoholic women
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Research on alcoholism has primarily focused on men, and the results are assumed to apply to women. In addition, empirical studies on alcoholism in general, and female alcoholism in particular have traditionally been unidimensional. Research with women alcoholics tends to employ small samples and no comparison groups. In this research, alcoholism in women is viewed as a multifactorial problem which must be investigated through both a theoretical and an empirical approach which recognizes the complexity of the problem. Riegel's dialectic theory provided the conceptual framework of this study. Six variables, number of alcoholic relatives, number of health-related problems, depression, self-esteem, femininity and the meeting of significant others' expectations were chosen from the literature on female alcoholism to represent three of the dimensions in dialectic theory and the three most commonly held theories of alcoholism: biological, psychological, and sociocultural. Sixty women, under 90 days sober, and in treatment for alcoholism, were compared with sixty nonalcoholic women, randomly selected from subjects who had participated in the Women's Identity Development Project, and who had responded to a followup questionnaire. Preliminary analysis showed significant differences between the groups on marital and employment statuses and income, which were considered a consequence of the alcoholism, and education which was used as a covariate. When a MANCOVA was employed, femininty was found to be nonsignificant and was dropped from further analysis. Both depression and self-esteem were significant for the covariate. An ANCOVA indicated that the main effect for alcoholism was also significant. Therefore, five variables were entered into a stepwise discriminant analysis, Wilks' method. Self-esteem failed to reach the tolerance level at step 5 and was dropped from further analysis. The four remaining variables accounted for 72% of the variance. Calculation of a shrinkage formula yielded a corrected multiple R^2 of .70. Depression was shown to have the highest discriminant capability, followed by health-related problems, number of alcoholic relatives, and others' expectations. A classification analysis, using the function, discriminated between the two groups with 94% accuracy. These findings support the importance of employing a multifactorial approach, both theoretically and empirically, to the study of female alcoholism. In addition, the findings have implications for early intervention, education, prevention and treatment.