An Investigation of Learning Behaviors and Thematic Data Analysis using Generative Artificial Intelligence in Elementary Students’ Group-Based Cell Pizza Activity

Abstract

Teaching complex biological concepts like cell structure to elementary students is challenging due to the abstract nature of the content and the limitations of traditional didactic methods. Creating hands-on, constructive learning environments is crucial for fostering engagement and scientific literacy. This research aimed to investigate elementary students’ learning behaviors through participation in the Cell Pizza activity, a hands-on collaborative science learning approach, utilizing thematic analysis and Generative Artificial Intelligence (Generative AI) for systematic data coding and synthesis. The population consisted of elementary students from grades 4 to 6 at Tha Khlong 1 Municipal School, and the sample of 75 students was selected through purposive sampling based on participation in the science camp program. The research instruments included observation forms, video recordings, student worksheets, and Cell Pizza evaluation rubrics. Quantitative findings showed that the majority (70%) accurately identified cell types and labeled all organelles correctly. Prominent learning behaviors included accurate scientific vocabulary usage, collaborative learning, visual referencing, and creative representation using tangible materials. However, students encountered confusion regarding complex organelle names, particularly smooth and rough endoplasmic reticulum. The integration of Generative AI effectively reduced analysis time and enhanced precision in natural language processing, although limitations remained regarding deep contextual interpretation. Therefore, combining Generative AI with human analysis is recommended to ensure optimal accuracy and interpretative quality.