Investigating the Applicability of the Normalized Programming State Model to BlueJ Programmers
Brad Richards, Ayse Hunt · 2018
It has long been a goal of educators to accurately identify at-risk students early enough in the term to intervene, and the increasing availability of programming process data and learning analytics tools has brought education researchers closer to realizing that objective. The Normalized Programming State Model (NPSM), for example, a recent approach that considers students' editing and testing behaviors in addition to compilations, has been shown to predict a student's overall course grade with 36-67% accuracy [14] depending on the number of assignments analyzed.