Problems in the Evaluation of Treatment Programs for Drunk Drivers: Goals and Outcomes

Abstract

Measures used in evaluating drunk driving treatment programs over the last 15 years are reviewed and critiqued in terms of construct validity. The concept of triangulation with increased use of self-report and collateral measures is recommended over the current reliance on the single outcome measure of rearrest. Greater attention to collecting data on processes and intermediate outcomes is also recommended to provide information for program improvement, dissemination, and identification of effective treatment models. Ambiguity concerning program goals and directions is cited as another related problem area which has led to unintended differences in program philosophies and methods.