Our response to Caliendo and Kyle's (1996) editorial, entitled "Establishing the Theoretical Frame," has a decade-long gestation period. As we were considering how we might frame our response, we each recalled one Sunday afternoon in the winter of 1986 when we were both working back at the office in our jobs as curriculum writers in a large Australian school board. In our spare time we were writing Master's dissertations. John was conducting an investigation of the impact of a curriculum intervention on high school students' science process skills using a pretest/posttest experimental design involving over 800 students. Bill was undertaking a closeup account of the work of a teacher-friend, looking at the problem of how teachers know what they know.
Taking some time out from our work that afternoon and over a cup of coffee we soon found ourselves talking about our respective dissertations and what counts as rigor in educational research. John-with a background in the sciences and well-schooled in the scientific research tradition-was concerned about how Bill might be able to justify his findings in terms of the norms of validity and reliability. What could an N of 1 study possibly say about the larger population and how could Bill study the work of a friend and still maintain a credible arm's length relationship? Bill-with a background in literary theory-argued that his study of the practical knowledge of one person was concerned to understand the particular, not to generalize beyond the case. He said that questions about reliability and validity made sense in the research tradition of John's study, but he also said that he was not very interested in the kinds of questions that may be answered by an experimental study of 800 students. John argued that large samples and experimental designs are needed in order to be more certain about what to do to improve teaching. Bill countered by saying that teaching was by nature uncertain and that what was needed was a better understanding of these uncertainties and a different kind of research tradition.
Following that Sunday afternoon, we have had many such conversations about educational research. Two years later, we found ourselves together again as doctoral students in a class