MATEMATİK ÖĞRETMEN ADAYLARININ CHATGPT İLE MODELLEME ETKİNLİĞİ OLUŞTURMA SÜREÇLERİNİN İNCELENMESİ

Abstract

This thesis examines the processes of prospective mathematics teachers in designing model-eliciting activities (MEAs) using ChatGPT-4o and how they guide artificial intelligence in line with their pedagogical decisions. Structured with a holistic single case study design, one of the qualitative research methods, the study group consisted of six prospective mathematics teachers selected through criterion sampling at a state university. Data were collected through interaction documents of participants with ChatGPT and audio recordings of semi-structured interviews, followed by content analysis within the framework of MEA principles. Findings revealed that participants critically evaluated the initial drafts produced by ChatGPT—which were generally traditional or closed-ended—and performed systematic interventions to align them with MEA principles. Participants specifically emphasized the principles of reality and model construction in their designs, developing open-ended activities that support student autonomy by utilizing contemporary social contexts such as fertility rates, carbon footprint, and earthquakes. It was determined that ChatGPT served multidimensional roles as an idea partner, content editor, critical reflection tool, and conceptual learning supporter throughout the process. Consequently, the ChatGPT-supported MEA design process was identified as an iterative and dialogic journey that enhances the pedagogical reasoning skills and technology integration competencies of prospective teachers. Despite receiving intensive support from ChatGPT, prospective teachers maintained their pedagogical autonomy and positioned the AI as a "thinking partner".