On the Growth of Knowledge in Public Administration

The issue of proper research and theory development in public administration has been raised again by Howard E. McCurdy and Robert E. Cleary in their article, Why Can't We Resolve Research Issue in Public Administration? They express concern for lack of adequate research being done at dissertation level and cite research findings suggesting that very few of recent dissertations meet criteria that conventionally define careful, systematic study in social sciences. These criteria include purpose, validity, testability, causality, topical importance, and cutting edge significance. By not meeting these criteria, they feel that current dissertation research is not advancing knowledge in our field.' Following their prescriptions for theory building could lead to conclusion that histories, descriptions of administrative experiences, reports of action research projects, political theories, philosophical analyses, and social critiques will not contribute significantly to growth of knowledge in public administration. This type of research normally does not satisfy criteria of validity, testability, and causality. Nevertheless, this type of research has contributed significantly to our knowledge of public administration. Although Cleary and McCurdy may recognize historical importance of descriptions and critiques for generating ideas about public administration, they claim: A field that promotes descriptions and critiques still needs research. . . . and Someone has to publicly expose descriptions and critiques to standards of scientific verification before they become 'usable knowledge.' They allow that case studies can contribute to verification of concepts or critiques, provided that they are consciously used to do so, especially in combination with other cases or studies, but they remind us about validity problems of studies in general. 2 Cleary and McCurdy advocate a mainstream social science approach: belief that the social sciences differ in degree and not in kind from more well established natural sciences, and that best way to achieve scientific success is to emulate logic and methodology of natural sciences.3 This is evidenced by their adherence to criteria of validity, testability, and causality; their call for testing of ideas generated by descriptions and critiques; and by their appeal to Kerlinger's Foundations of Behavioral Research to identify criteria for quality research in behavioral sciences.4 But serious questions have been raised about