Unsupervised and semi-supervised clustering methods to identify and refine participant experience levels in educational research

Abstract

The progression from novice to disciplinary expert is a longstanding area of inquiry in educational research. Studies investigating such progressions have often resorted to participants' self-assessments or other qualitative indicators as a starting point to define experience. But does a participant's estimated experience coincide with metrics derived from their conceptual understanding of a discipline? Using data extracted from over 150 concept maps, we first demonstrate that disciplinary experience is a reliable variable to explain differences in conceptual understanding across a highly diverse learners' population. Through a comparison of unsupervised and semi-supervised models, we then motivate clustering participants into three distinguished experience levels, and support such a classification performed in other studies of educational research. By analysing cluster composition, we also identify discrepancies between the perceived and predicted experience levels of the study participants. Lastly, for studies processing participants data through network analysis, we present insights into statistically significant metrics that can characterise each experience level, and advocate for the use of node-level metrics in such studies.