Erzberger's dilemma: Validity in action research and science teachers' need to know

Abstract

In recent years, action research has increasingly become a part of university programs in teacher education. Courses for pre-service teachers and for teachers returning to the university for advanced degrees are requiring some form of teacher-research (Gore and Zeichner, 1991; Ross, 1987; Zelazek and Lamson, 1992). This trend can be seen in individual schools of education spread throughout the US and in major reform efforts. Action research is also being recognized as a method of in-service staff development, as can be seen in recent dissemination efforts by the Association for Supervision and Curriculum Development (ASCD) (Sagor, 1992), its inclusion in edited books on staff development (Holly, 1991), and its use as a tool in school restructuring (Sagor and Curley, 1991). While there is this growing acceptance and encouragement of action research within schools of education and in school districts, little attention has been paid to what counts as tests of validity for teacher-research. Although this question might appear to be somewhat esoteric and removed from the world of practice, it has significant implications for the way that action research is enacted. That is because how one validates research shapes the methods used. In this article I examine the issue of validity in action research and how it affects methodology. I begin with a brief history of action research. That is followed with a description of a collaborative action research project in which eight physics teachers and I have been engaged (Feldman, 1993). This leads to the statement of Erzberger's dilemma and a critical examination of what counts as validity in action research. I conclude with some implications for policy makers and for the pre-and in-service education of science teachers. Action research