摘要：In this paper the authors argue that there are three modes of educational action research: emancipatory, practical, and knowledge generating. Furthermore, they suggest that much of action research, although predicated on notions of emancipatory research, is often not primarily emancipatory in nature. There are considerable risks involved when action research fails to adequately justify its truth claims because of a dependence on validities that primarily assess the emancipatory features of the research. Consequently, the authors propose that the various modes of action research require emphasis on different validities that are dependent on the purposes of the research. In doing this, they offer a reconceptualization of Anderson and Herr’s (1999) influential approach to validity in action research.
