Problem-driven research in mathematics education

How does one get started with research in mathematics education? What are "good" research problems to investigate? Answers to these questions largely depend on the research tradition one selects, and on the main issues driving the research (e.g. theory? problems?). Often, we are unable to determine the initial impetus of a research project, because researchers usually present finished polished results and say little about the ways in which the projects started and developed. This paper is an attempt to uncover the often messy early stages of several research studies driven by didactical problems. Some of these projects were successfully completed, some are in progress and others are just starting and may end up stillborn. This paper presents a classification of the possible points of departure of research studies, a discussion of some heuristics, which may be helpful when choosing and pursuing research, and an analysis of some examples, including some methodological issues.