Refining EFL Essay Writing with Multimodal GenAI Tools

Identifier:  imarina:9633119

Handle: https://hdl.handle.net/20.500.11797/imarina9633119

Authors:  Rebolledo R; Veas C; Gisbert M

Abstract:
This study presents a structured didactic intervention that uses multimodal Generative Artificial Intelligence (GenAI) tools to enhance English essay writing. It was implemented among 50 second-and fourth-year teacher-training students at a Chilean university. After an initial diagnostic activity, the students took part in six structured sessions that combined traditional teaching with various AI tools: ChatGPT for self-assessment, Claude for outlining, Perplexity for brainstorming, Gemini for thesis development, NotebookLM for expanding ideas, and a custom GPT for building paragraphs. Only the first session included some quantitative analysis, comparing student self-ratings with AI feedback, while the rest focused on reflections collected through online forms. A manual and AI-supported thematic analysis found that students valued quick, personalized feedback that helped them with grammar, organization, and overall clarity. Learners at lower levels used AI for basic support, whereas more advanced students used it to strengthen arguments. Some worries emerged, including becoming too dependent on AI, losing originality, and facing technical issues. Finally, the study suggests that AI can enhance writing and self-awareness when used alongside explicit instruction and training. Future research should look into improving student retention and integrating AI tasks into regular coursework.