AI MyData: Fostering Middle School Students’ Engagement with Machine Learning through an Ethics-Infused AI Curriculum

• DOI: 10.1145/3702242
• Publication Date: 2024-11-08T12:43:34Z
• Authors: Ismaila Temitayo ...; Fred Martin; Ruizhe Ma; Joseph E. Gonzales; Vaishali Mahipal; Solomon Sunday Oy...; Jarkko Suhonen; Markku Tukiainen
• Keywords: Student Engagement

As initiatives on artificial intelligence (AI) education in K-12 learning contexts continues to evolve, researchers have developed curricula among other resources promote AI across grade levels. Yet, there is a need for more effort regarding curriculum, tools, and pedagogy, as well assessment techniques popularize at the middle school level. Drawing prior work, we created original curriculum activities with innovative use of existing technology, new computational teaching tool, series approaches assessments evaluate students’ engagement resources. Our called MyData comprises elements ML data science infused ethical orientation. In this paper, describe novel further discuss how engaged students critiquing dilemmas. We gathered from two pilot studies conducted Northeast United States, one Artificial Intelligence Afterschool (AIA) program, virtual summer camp. The AIA program was carried out local public four aged 12 13; consisted eleven two-hour sessions. camp sessions over consecutive days, eighteen 15. facilitated both programs hands-on plugged unplugged activities. method capturing included artifact collection, structured interviews, written assessments, pre- post-questionnaire tapping participants’ dispositions about its societal implication. Participant artifacts, survey, observation, analysis tasks completed revealed that children improved their knowledge AI. addition, units accompanying study successfully participants, even without related concepts. also found an indication introducing ethics adolescents will help development ethically responsive citizens. results indicate lessons establishing links personal lives (e.g., letting choose personally meaningful datasets) implications using interactive tools were particularly valuable promoting integration subject domains settings. Based these results, our findings, identify limitations, propose future work.