The implementation of research design in studies on substance abuse education is discussed in this meta-analysis. Findings indicate that the quality of research design is constrained by negative feedback from evaluation stakeholders. This finding is applied to an evaluation of the Here's Looking at You, 2000 drug education program currently underway in California. A conclusion is that the evaluation enterprise, particularly when it concerns the method of subject assignment, is influenced by various mutually antagonistic social and political factors that result in methodological compromise. Seemingly contradictory recommendations for researchers are to adhere to more rigorous evaluation methods and to maintain realistic expectations. A goal is the development of a randomization process that is robust against indifference, incompetence, and corruption. Four statistical tables are included. (12 references) (LMI)