Mapping the Territory

Computer science (CS) education research is an emergent area and is still giving birth to a literature. While scholarly and scientific publishing go back to Philosophical Transactions (first published by the Royal Society in England in 1665), one of our oldest dedicated journals Computer Science Education was established less than two decades ago, in 1988. Growth which has led to the emergence of CS education research as an identifiable area has come from various places. Some from sub-specialist areas: the Empirical Study of Programming (ESP) and Psychology of Programming Interest Group (PPIG) series of workshops; some from the major practitioner conference which have included research papers—the Innovation and Technology in CS Education (ITiCSE) conference and the SIGCSE Symposium (now in its 36 th year) run by the Special Interest Group in Computer Science Education of the ACM. Following on from these starting points, there has been a burgeoning growth of publications appearing, perhaps opportunistically, in diverse locations, such as the IEEE Frontiers in Education (FiE) and the American Society for Engineering Education (ASEE) conferences, the ACM OOPSLA etc. Simultaneously within this same time there have emerged a number of CS education research groups within academic institutions. Despite this growth—and because of it—we are struggling to find the shape and culture of our literature. The task is difficult not only because the literature is distributed (there is no CS education research conference or publication) but also because our researchers and writers come from many established fields of scholarship and research—at least from education, psychology, computer science, technology and engineering. We have different intellectual traditions, and different conceptual frames, not to mention methodological differences and different reporting and citation styles. What gets published from these different disciplinary areas, in diverse venues, are papers of very different types. However, in a simple-minded way, whatever their tradition, they can be thought of as having two components: a dimension of rationale, argumentation or " theory " , and a dimension of empirical evidence. If we think of these dimensions as plotting a space, then four quadrants can be defined. On the top left, we have papers that have lots of argument, but little empirical evidence