Effects of Artificial Intelligence Chatbot-Integrated Instruction on Science-Positive Experiences and Attitudes toward Artificial Intelligence: Focusing on the Unit of Stellar Types and Evolution

Abstract

This study examined the effects of artificial intelligence (AI) chatbot–integrated instruction with embedded scaffolding on high school students’ science-positive experiences and attitudes toward AI in the Earth science unit on stellar types and evolution. To this end, the researchers developed an instructional program grounded in the Exploratory Scientific Data Analysis Inquiry Model and designed an AI chatbot to support students’ inquiry processes and conceptual understanding. The participants were 138 11th-grade students, and a quasiexperimental one-group pre-test–post-test design was employed. The results indicated that the intervention significantly improved students’ science-positive experiences across all subdomains – affective experience in science learning, science self-concept, motivation for science learning, science career aspirations, and attitudes toward science. In addition, students’ attitudes toward AI significantly improved across cognitive, affective, and behavioral dimensions. Qualitative findings revealed that the AI chatbot’s immediate feedback and hints reduced students’ perceived cognitive burden in science learning and encouraged more active engagement in self-directed problem solving and collaborative inquiry. Students also came to perceive AI as a useful tool for learning, career exploration, and everyday problem-solving.