Design and Implementation of an Educational Game for Teaching Artificial Intelligence to High School Students

Abstract

Artificial intelligence (AI) is becoming ubiquitous in our daily lives. In more and more fields, Al systems are transforming how knowledge is constructed, discoveries are realized, and how solutions are developed and tested. These changes have profound implications for the future workforce and citizenry. Yet, learning Al remains a niche subject largely reserved for advanced post-secondary educational contexts. While there is growing attention to broadening Al educational opportunities and, especially, to providing learning experiences for younger students, relatively little is currently known about how to most effectively provide Al education to K-12 (kindergarten through 12th grade) students. In this paper, we discuss the design and present findings from an implementation study of an educational game for high-school Al education called ARIN-561. Drawing on an integrated analysis of gameplay log data, pre/post knowledge, and disposition surveys for nearly 1,000 high school students, we present findings on the efficacy of the educational game and its constituent activities in advancing Al learning goals. We explore possible interactions between learning outcomes, incoming math knowledge, prior gaming experience, and other factors that can inform future learning design and shed light on what can position youth for success in game-based Al learning experiences.