A Systematic Review of Artificial Intelligence in K-12 Science Education：Themes, Effectiveness, and Challenges

• In Event: AERA Poster Session Wednesday 11:45am
  In Poster Session: ATL Poster Session 1: AI Integration and Student Outcomes in K–12 Settings

Abstract

This systematic review analyzes 23 empirical studies (2018–2025) on AI applications in K-12 science education. Findings reveal AI enhances academic performance (e.g., personalized learning paths), scientific literacy (e.g., AI visualization for complex concepts), and higher-order cognition (e.g., critical thinking via generative AI). Key domains include AI-driven curriculum design, adaptive assessment systems, and interdisciplinary skill development. Challenges persist: (1) Technical limitations in complex task scoring (e.g., scientific argumentation); (2) Equity gaps in resource-limited settings; (3) Ethical risks (data privacy, algorithmic bias). Future research must prioritize longitudinal studies, low-cost AI tools, and ethical frameworks. The study underscores AI’s transformative potential while advocating for equitable, teacher-supported implementation.

Author

• Haifeng Wu, East China Normal University