Enhancing Teacher Professional Development through Knowledge Building and Generative AI-Based Assessment for STEM Lesson Design

• In Event: AERA Poster Session Friday 11:45am
  In Poster Session: AERA Division K Section 6 In-Service Educators - Poster Session 3

Abstract

This study explores the application of knowledge building (KB) pedagogical principles to guide the design of STEAM lesson plans and examines the potential of generative artificial intelligence (GAI) for automated lesson plan assessment. The research was conducted in a teacher professional development program focused on knowledge building. Participants included 55 teachers who completed the training program. The study examined the reliability and validity of ChatGPT-based assessments using different structural prompts, and applied ChatGPT to assess the effectiveness of the teachers’ lesson plans. Results indicated that COPs significantly enhanced the reliability and validity of ChatGPT's assessments. Furthermore, participants designed lesson plans more closely aligned with KB principles and incorporated instructional activities that support sustained idea improvement.

Authors

• Cing-Sin Lin, National Chengchi University

• Guo-Tsai Hung, National Taichung University of Science and Technology

• Chih-Hsuan Chang, National Kaohsiung University of Science and Technology

• Mei-Ju Chen, National Tsing Hua University

• Huang Yao Hong, National Chengchi University