Evaluation of a Game-Based Personalized Learning System

Modern classroom settings require integrating many students of varying backgrounds and varying levels of classroom performance into the same educational process. Ideally, each student should receive personalized support that is tailored to their specific learning style. However, with limited resources and time available to educators and teaching facilities, personalized support is often infeasible. To address this issue, this project focuses on a learning system that uses artificially intelligent agents to provide students with personalized feedback and support. To further engage students, the system is built on top of an existing narrative game environment called Gridlock. Gridlock provides students with a narrative game experience that focuses on creating a traffic light controller to teach students the basics of sequential digital logic design, a core component in both Computer Engineering and Computer Sciences. Gridlock was chosen as it already implements several meta-cognitive strategies designed to promote student learning and student self-reflection, thus giving a solid foundation to build the learning support system on top of. This paper reports preliminary results from early testing and continued development of the Gridlock system. In testing the game system, students in Introduction to Digital Systems courses and Computer Architecture courses at Rowan University utilized the game as a supplementary tool to assist them with lab work. The overall goal of the improved game system is to improve student comprehension and classroom results. Additionally, the finished system is planned to be fully automated, requiring no intervention from instructors or researchers. Assessments of the effectiveness of the game system will be shown through the following: 1. Student game performance. 2. Student performance on content tests related to the game content. 3. Student lab work performance. 4. Student surveys.