Abstract
Some writers state that experimental research mustform the core of all evaluation. Others counter that experrmental studies tend not to be used, and urge reliance on more subjective and process-oriented research methods. If we accept the premise that the "hard-science" and "anthropological" approaches are truly paradigms of evaluatron research, the resultant issue is which approach to use m specific cases. The present report rejects the "competing paradigms" thesis, and argues that there is a solid middle ground. It suggests why that ground should be used, and considers how the two approaches can be mixed to surt particular research settings .