Construção e validação de escalas para medida de crenças e de percepção de controle do consumo de bebidas alcoólicas

Abstract

This study aimed to expand the knowledge about the role of cognitive factors related to drinking behavior, through the validation of psychosocial measures. The main objectives were: (a) to construct and to validate, for Brazilian samples, a scale of perceived benefits and barriers to prevent the consumption of alcoholic beverages; (b) to adapt and to validate a version of the DRIE scale (Drinking Related Internal-External Locus of Control Scale) for Brazilian samples; (c) to investigate the relationship between the variables of the proposed model. So, the project was developed into two phases, both using the AUDIT to measure the alcohol consumption pattern in the last year. At the first phase a sample of 66 problem drinkers, in treatment or not, was interviewed in order to identify the perceived benefits and barriers related to avoid the consumption of alcohol beverages, being these variables theoretically based on the Health Belief Model. The results of this phase permitted the construction of the items for a pilot version of the scale. The original DRIE scale was translated to Portuguese and back translated to English, and the following steps were the content and semantic analysis of both scales and a pilot test with them. In the second phase of the study, related to the empirical validation, the pilot version of the scales was applied to 433 public workers from a federal institution of higher education. The data were analyzed by using the SPSS for Windows, in order to evaluate the factor structure and the reliability of the instruments. The factor analyses pointed the 19 items to maintain in the final version of the Scale of Perceived Benefits and Barriers to Avoid Alcohol Drinking, organized into two second-order factors (perceived benefits and perceived barriers), both with satisfactory internal consistency. The Brazilian version of the DRIE scale, after factor analyses, remained with 35 items divided into two factors (internality and externality), with very satisfactory internal consistency too. The results of correlation and comparison between groups were mostly consistent with theoretical expectations. The problem-drinkers group perceived significantly less benefits and more barriers to avoid the alcoholic beverage, and showed to be more external in the locus of control of drinking behavior than the non drinkers group. Positive correlations were found between the AUDIT score and the perceived barriers, and negative correlation were found between alcohol consumption and perceived benefits. Concerning to the locus of control orientation, the AUDIT score was just positively correlated with Externality. Multiple regression analysis showed that the proposed model predicted significantly the consumption of alcohol in the last 12 months and explained 16% of the total variance of the dependent variable. It can be suggested that the validated instruments can contribute for the development of Brazilian researches, and to plan the prevention and intervention programs, since they are also useful to apply in non-alcoholics population.