Shape of Educational Data

点击翻译按钮获取中文摘要

This study of the 'shape' of STEM educational assessment data will exploit recent advances in computational topology, machine learning and cognitive science. Applications of computational topology to learning theory are both new and innovative. It is anticipated that these new tools will provide critical insights into student learning of Calculus.Funded by NSF's Research on Education and Learning (REAL) program, this early concept grant for exploratory research will apply advanced topological, geometric, and Bayesian methods to analyze the shape of data generated by students taking Calculus via a massive open online course (MOOC) system (and other data sources). This project will link data analyses from three leading learning platforms in the service of better understanding student learning of Calculus. One potential high payoff is the deployment of richer analyses of test and performance data to support a recommendation system to support to students reaching their learning goals. Further, the project will seed a new community of researchers across disciplines that typically do not interact. The proposed line of work has the potential to fundamentally change the way we think about assessment data in support of mathematics learning and perhaps other STEM content.