Educators as Clinicians: Small Data for Education Research

Abstract

Much work in computing education research focuses on large-scale data collection and analyses, bringing “big data” approaches to bear on the educational research agenda. Drawing on lessons from the medical research community, we argue that the work of many computing education researchers is more akin to that of a medical clinician than an experimental researcher. Education researchers working in a small-class setting will often not be able to exercise the experimental controls necessary for large-scale, statistically-driven research. In this setting, educational researchers must work through the ambiguity and complexity of their classes to respond to the specific needs of their students in much the same way that clinicians respond to the specific needs of their patients. Small-data approaches tailored specifically to such environments can help educators measure their effectiveness when controlled experiments are not an option. As such, we describe a model for “small data” approaches in computing education research and demonstrate a case study where such an approach has been used effectively to analyze curricular changes.