Iterative Refinement of an AIS Rewards System

Abstract

Gamification-based reward systems are a key part of the design of modern adaptive instructional systems and can have substantial impacts on learner choices and engagement. In this paper, we discuss our efforts to engineer the rewards system of Kupei AI, an adaptive instructional system used by elementary and middle school students in afterschool programs to study English and Mathematics. Kupei AI’s rewards system was iteratively engineered across four versions to improve student engagement and increase progress, involving changes to how many points were awarded for success in different activities. This paper discusses the design changes and their impacts, reviewing the impacts (both positive and negative) of each generation of re-design. The end result of the design was improved learning and more progress for students. We conclude with a discussion of the implications of these findings for the design of gamification for adaptive instructional systems.