Although social policy makers and researchers assume that research is a major part of the information base upon which policy decisions are made, social policy is by nature incompatible with the organizational structure and time requirements for decision making in an administrative system. Whereas social policy decisions are timed in accordance with legislative sessions, social research tends to be a lengthy process and often does not provide results in time to be of use for policy making. There are, however, other less direct ways in which social research affects policy. One way in which policy can be used is to legitimate a challenge to existing policy. An example of this type of use is the widespread citation of the 1966 U.S. Office of Education Report "Equality of Educational Opportunity" by civil rights groups, plaintiffs in court cases, and in school board deliberations. Another way in which research is useful to policy is in answering procedural questions about how to increase the effectiveness of policies in force. In general, research on procedural questions will be more effective when it is done in-house, communicated directly and privately to those in authority, and performed on a continuous monitoring basis. Research on policy issues, on the other hand, should be done by an outside group which focuses on one particular issue and disseminates research results to all interested parties. The conclusion is that social research is less frequently used by an administrative body with authority to legitimate a policy than by external groups, without authority, to challenge existing policy. (Author/DB)