The structure of problem drinking in adulthood: a confirmatory approach

The structure of problem drinking in young adults was examined using confirmatory factor analyses to test three models of the relations among measures of drinking behavior, coping functions of drinking and alcohol-related problems. Model 1 was based on the notion that the relations among the variables would be accounted for by two correlated factors, drinking behavior and alcohol problems. Underlying Model 2 was the hypothesis that a more parsimonious, single factor model could be used to represent the data. Model 3 hypothesized two orthogonal factors, a general factor and a residual problems factor, representing a reconceptualization of the structure of problem drinking.

Data were obtained from surveys of a sample of university graduates (178 women, 84 men) at two points in time and from a community sample of young adults (411 women, 347 men).

Results of the analyses indicated that, although the relations among the variables could be adequately accounted for by any of the three models tested, the data were best fit by Model 3.

Model 3 appeared to be most useful to furthering our understanding of the structure of problem drinking, one that, according to the results obtained in this work, is similar for women and men.