Analyzing teacher candidates' arguments on AI integration in education via different chatbots

Abstract

The burgeoning role of Artificial Intelligence (AI) in education prompts crucial discussions regarding its implications for teaching and learning. This qualitative study probes the argumentative perspectives of 118 teacher candidates from Iğdır University on the integration of AI into educational practices. Employing Toulmin's (1958) model, we analyzed their arguments, which encompass claims, evidence, warrants, backings, rebuttals, and conclusions, to ascertain their stance on AI's pedagogical integration. Utilizing four distinct AI chatbots—GPT-4, Gemini AI, Claude 3 Haiku, and Mistral AI—the research deciphers thematic undercurrents within these dimensions. Moreover, a novel methodological contribution is made through 'negative space exploration', focusing on the unmentioned themes to identify latent biases and assumptions in the argumentation. The study's dual analytical approach, combining AI-driven theme identification and negative space exploration, resulted in an enriched understanding of the content. Key findings suggest a nuanced perception among participants: while AI chatbots are acknowledged for enhancing educational efficiency and enabling personalized learning, concerns regarding diminished human interaction, potential erosion of critical thinking skills, and ethical use persist. The analyses also highlight the need for a balanced AI implementation that supports, not supplants, traditional educational methods. This research contributes to the ongoing debate on effective AI integration in education and calls for responsible pedagogical adoption of AI technologies.