Researchers who undertake the evaluation of social action programs face several hazards, among which is the politicization of evaluation. Several alternatives in research methodology are suggested, as well as implications from the evaluations themselves. (DM)