One rule for one, and a different rule for another: The case of the Parental Rules about Alcohol Questionnaire

Abstract

Background. The parental rules about alcohol questionnaire (Van der Vorst et al., 2005, 2006) uses 10 items to assess how strictly adolescents believe the rules set by their parents about drinking are. An increasing body of literature has attested to the importance of rule setting in the prevention of problematic alcohol use among adolescents. A recent study proposed a two-factor solution in place of the hypothesized unidimensional one, with factors assessing non-normative, and normative rules. Methods. The present study used five waves of data to examine the structure of the scale, and how well it relates to a measure of heavy episodic drinking (HED). Participants in Waves one to four {10,954–9,383} were substantively more numerous than those at wave five (N = 2,332). Results. Confirmatory Factor Analyses did not support either the ten-item hypothesized model, nor the proposed two-factor solution. Results of exploratory factor analyses all pointed to a one factor solution. Using Modification Indices, we obtained a good-fitting, five-item unidimensional model in Waves one to four. At wave five, a good fitting unidimensional model was obtained with the dropping of a further item. Scores on this shortened scale were internally consistent, correlated highly with scores on the original ten-item version, and correlated to a similar degree as the original 10-item measure, with scores on a HED measure. Conclusion. Further work is required in assessing the properties of this scale across cultures and samples before definitively determining that two factors best represent parental rules.