Publication:

ChatGPT vs EFL pre-service and in-service teacher’s lesson planning

Resumen en inglés

The increasing use of artificial intelligence (AI) and language models, such as OpenAI’s ChatGPT, has impacted educational settings, particularly language teaching. This research, emerging from a critical literature review, highlights the need to explore AI’s application in lesson planning, especially in second language teaching in a Colombian context. The study compared AI-generated lesson plans with those created by four EFL pre-service teachers and two in-service teachers and explored their perceptions of AI in lesson planning. Using a qualitative descriptive design, data was collected over six months through documentary analysis and interviews. The analysis included 21 lesson plans: ChatGPT (n = 3), pre-service teachers (n = 12), and in-service teachers (n = 6), using a checklist and documentary analysis to identify common and different patterns and structures in lesson planning (e.g., warmups, activities, assessments) and methodologies. The findings showed that the lesson plans created by pre-service teachers, in-service teachers, and ChatGPT shared a similar structure and components such as Context or lesson details, Objectives, Warm-up, Pre-task, Task, Post-task, Closure, Resources, Time, Evaluation, and Homework. As to methodologies, ChatGPT typically implements skills-based and PPP approaches, whereas teachers implement TBLT. Regarding teachers’ perceptions of AI in lesson planning, pre-service and in-service teachers recognized the benefits of AI in lesson planning, especially for efficiency and organization. However, teachers expressed concerns about AI’s lack of flexibility, personalization, and the risk of over-dependence. While pre-service teachers were more open to using AI, in-service teachers emphasized the need for human adaptability.