Optimizing Narrative for Different Learners Within an AI Literacy Middle School Game

Artificial intelligence increasingly shapes how people obtain information, communicate, and make decisions, yet many students have limited opportunities to understand how these systems work or affect individuals and communities. Classification, the process of grouping items according to shared attributes, underlies technologies such as recommendation systems, facial recognition tools, spam filters, and social media platforms. The project will investigate how stories and games can make these concepts accessible and meaningful to middle school learners with varied interests, experiences, and approaches to learning. It will serve the national interest by expanding access to foundational artificial intelligence education and preparing young people to participate thoughtfully and help design a society increasingly influenced by automated systems. A freely available learning game, co-designed with educators, students, game designers, and artificial intelligence researchers, will provide classrooms across the United States with an engaging educational resource. The project will also produce broader design knowledge that can guide the development of future educational games across science, technology, engineering, and mathematics topics. The research will examine how five dimensions of narrative design influence student outcomes: choice structure, character richness, narrative richness, learning-agent competency, and setting realism. Students will complete brief measures of their values and interests, and the game will collect detailed interaction data to determine which narrative configurations best support learning, engagement, and enjoyment for different learner profiles. The research will combine embedded assessment, educational data mining, large-scale gameplay analysis, and computational optimization to evaluate many narrative configurations more efficiently than traditional experiments. During the first phase, educators, students, researchers, and designers will co-design and pilot the game to ensure that it addresses classroom needs and presents artificial intelligence concepts accurately. During the second phase, the game will be released for broad classroom use, and the resulting data will be used to model relationships among learner characteristics, narrative designs, and educational outcomes. During the third phase, the project will test whether matching students with narrative configurations predicted to support their profiles produces better outcomes than baseline configurations. The project will contribute a freely available artificial intelligence literacy game, evidence-based principles for designing educational narratives, and reusable research methods and infrastructure for studying adaptive digital learning environments. This project was funded by the STEM K-12 program, which supports fundamental, applied, and translational research on STEM learning in both formal K-12 education and informal learning environments. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.