Catch Them Before They Fall: Predicting Early-Grade Student Struggle During Educational Game Play

• In Event: Technology, Instruction, Cognition and Learning SIG. Assessment With Technology: Diverse Data, Collection, and Use

Abstract

In this paper, we describe methods and data analyses that allow game designers to identify when students are struggling in an educational game. The game, developed by Twin Cities PBS, was designed to support children in grades K-2 to learn about map navigation and interpreting map icons. During beta testing of the game, researchers collected log data from game play of 122 K-2nd students, along with in-person observation and screen capture video data. Researchers used the three sources of data and data mining techniques to build a predictive model that will detect student struggle during game play, and help designers refine support for students such as hints or other interventions. We describe the game, data collection process and predictive modeling.

Authors

• Mingyu Feng, WestEd

• Elizabeth M. McCarthy, WestEd

• Melissa Cheung, Center for the Collaborative Classroom