Validating Mastery Learning: Assessing the Impact of Adaptive Learning Objective Mastery in Knewton Alta

摘要

Adaptive courseware products implementing mastery learning pedagogy must determine when each student reaches mastery. Such determinations are often made in real time, in order to inform student progress, but the validity of algorithmically determined mastery typically can only be assessed by examination of later student performance. This paper examines the impact of platform-determined mastery on future quiz and assignment preparedness in the context of Knewton alta. With simple controls for overall student initial ability, platform-wide results indicate that students achieving mastery (as calculated by Knewton's Proficiency Model) outperform students who do not, with largest future performance gains seen by students with lowest initial ability levels.