In recent years social scientists have reviewed and evaluated a number of experiences with applied social research. There is a growing consensus that social knowledge produced for policy-action purposes should be substantively different from that produced for disciplinary purposes. Not all of the research procedures developed to produce disciplinary knowledge are applicable under the conditions of policy-action research, and scientific techniques have not been developed to produce certain kinds of needed social knowledge. The organization and administrative procedures of academic institutions have been found inadequate for the production of many types of policy-action knowledge. Nevertheless, there is still widespread agreement that academic institutions should provide the basic training of policy-action researchers.