Transforming AI Chatbots for a Brainstorming Teaching Technique of Process Writing

Abstract

This study investigated the impact of AI-driven brainstorming tools on process writing instruction and students' writing outcomes in the context of third-year Thai university students. A mixed-methods approach was employed to examine the effectiveness of AI-driven brainstorming tools, foreign English lecturers' preferences and comments towards AI-generated brainstorming results, students' preferences for specific AI chatbots, and the perceived challenges and facilitative factors experienced by students. The findings revealed that the intervention group using AI tools significantly outperformed the conventional group on two out of three assignments (People: p = .002; Things: p < .001), with ChatGPT emerging as the most popular AI chatbot (78.8%). Foreign English lecturers acknowledged the AI chatbots' strengths but preferred students' brainstorming results. Overreliance on AI for idea generation was identified as the most concerning challenge (M = 4.62), while enhanced creativity (M = 4.53) and increased idea generation (M = 4.51) were the most appreciated facilitative factors. The study demonstrates the potential of AI-driven brainstorming tools to revolutionize process writing instruction and highlights the importance of striking a balance between leveraging AI benefits and fostering students' independent thinking and creativity skills.