This article notes that much case study research focusing on educational inequalities is evaluative in character, in the sense that it draws value conclusions. Moreover, the evaluative character of these conclusions is often implicit. We argue that practical evaluation of this kind is inappropriate in research reports. We then discuss the legitimate role that values can play in case study research, notably in providing the basis for identifying important topics for inquiry and in selecting explanations from among causal factors. We (...) outline the obligations associated with this role: that the non-evaluative character of the conclusions of the research must be emphasised, and that the value assumptions used to frame descriptions and explanations must be made explicit. We conclude by briefly examining the implications of our argument for educational evaluation, for action research, and for the notion of internal critique that is central to critical research. (shrink)