With our interest to improve our education in computer science, an understanding of how students learn about CS concepts, how different concepts are understood, as well as the conditions for learning, become important issues. A better understanding of our students and their learning gives us a strong tool in our efforts to develop teaching. There is an increasing awareness of the usefulness of theoretically sound research approaches: it opens for generalisations of results, it invites comparison between researchers, methods and results, and at the same time it makes the limits of the research visible. As examples on initiatives that have lately been taken to promote a conscious use of relevant research approaches, can be mentioned the bootstrapping project [13], the special issue on import and export of Computer Science Education (to appear), as well as papers offering overviews of the current use of certain approaches ([4], [8]) and attempts to verbalize models for a successful research process ([5], [6], [12]).These initiatives do not advocate the primacy of a certain approach over others. This openness is well-grounded, since "a particular approach offers certain perspectives on a research question, and, in this way, enables the researcher to study [these] aspects of learning, while other aspects, that are not in focus using the selected approach, become unclear or 'blurred' (Berglund, submitted for review). Thus, the selection an approach is closely intertwined with the research question under investigation.In this panel, the theoretical foundations for four different research approaches will be described, and examples of research performed within each of these approaches will be given. The examples will serve to illuminate which kinds of results that can be offered by a particular approach, and thereby illustrate its use.