After proposing principles and guidelines for the use of research findings in the development of public policy, this essay presents five recommendations for the use of research in policy formulation/recision. First, the research base must be solid, which implies quality of design, thorough peer review, and replication. Second, the sampling of units in the various studies that form the research base must be such as to allow generalizations to the populations affected by the policy decisions in question. Third, there should be a means of estimating the potential harmful effects that may result from policy influenced by social research; i.e., would the effectiveness of the policy in the targeted area produce negative consequences in another area? Fourth, the research should provide clear direction for policymakers "either for immediate action or for considering alternative approaches to problems" (Weiss and Bucuvalas, 1980); the research should show where and how to make changes or redirect thinking on current philosophy or guiding principles of operation. Fifth, social scientists should never be open, persistent advocates of their research products. Since few initial positive findings hold up in replication by others, researchers should make clear the limitations of their findings and not confuse what is good for society with what is good for their own careers. 39 references