The Limits of Positivism in Social Work Research

This paper considers the argument that research on social work should be more rigorously scientific in its methods, using the work of Brian Sheldon as representative of this position. It is argued that a wholly positivist approach is inadequate on epistemological grounds, and that the procedures characteristic of this approach, particularly the experimental method, are unlikely to prove generally feasible or useful. Evaluation research in other fields is used to illustrate this argument, and it is suggested that social work is far from unique in its under-use of research findings. A possible alternative, that social work should be evaluated by artistic criteria, is briefly examined and found to be problematic. The paper concludes with an argument that research in future should be more concerned with process, and more open, participative and pragmatic in style.