Social science research has traditionally sought to incorporate a process of self criticism within its methodologies with variable degrees of success. Despite attempts at integration, questions about the assumptions researchers bring to their research have only too often been considered separately from the treatment of data or the development of theory. By referring to a recent study at Sussex University on which the authors collaborated, we should like to illustrate one example of constructive reflexivity, of how self ciiticism and the questioning of assumptions can become central, rather than peripheral to both methodology and the analysis of data. We were concerned with how our respective interests in education were reflected in differences in our practice of Illuminative methodology.1 C.M.'s interest and research experience was in how students learn in university settings; while A.H. was more concerned with course development in higher education. Our reflexive analysis took place after we had collected the data. It had two main outcomes. First, it gave us a deeper understanding of the data itself. Secondly, this led us to a reassessment of the role of Illuminative methodology as a means to course development. The differences between the experiences we brought to the research were the starting point for our reflections. A.H. had several years teaching experience in secondary and higher education. The emphasis in C.M.'s work was on the study of how students developed as they went through their degree course; how they came to understand what was expected of them; how they changed their thinking about their subjects; the kinds of problems they encountered in their learning; and how this was related to their teachers' views of the course. This research involved interviewing students and staff so as to see things through their eyes and to avoid imposing a preconceived structure upon the context being studied. While taking an M.A. course in curriculum development, A.H. had carried out a very small-scale study of a sample of Sussex geography students which had raised a range of issues about the geography courses. When A.H. had the opportunity of staying at Sussex for a further period as a research fellow, his suggestion that he should follow up these issues by a longer study was welcomed by the geographers. C.M. had already made a three-year research study of a group of social science students, majoring in geography, at another university. A.H. and C.M. shared an interest in how students learn, and an enthusiasm for Illuminative methodology; C.M. having used it for several years in different research studies; A.H. had initially warmed to it as a teacher, and now had practical experience of it through the M.A. course.