A comparison of lesson plans from artificial intelligence and English language teachers

This study critically examines the effectiveness of Artificial Intelligence (AI) tools, specifically ChatGPT, in the context of lesson planning for English as a Foreign Language (EFL) education in Diyarbakır, Türkiye. It explores whether AI can contribute meaningfully to the traditional lesson planning process by producing pedagogically appropriate and contextually adaptable content. A qualitatively dominant mixed-methods approach is employed, integrating quantitative evaluations of lesson plans with qualitative insights gathered from interviews, offering a comprehensive understanding of AI's potential role as teacher AI literacies continue to evolve with technological integration. The research first utilizes a detailed rubric, developed based on established academic frameworks, to evaluate the lesson plans across dimensions such as content relevance, instructional coherence, and alignment with curriculum standards. Both human instructors and ChatGPT generate lesson plans using standardized topics and structures, which are then quantitatively assessed to enable an objective comparison. The qualitative dimension of the study is conducted through semi-structured interviews with nine EFL teachers from varying educational backgrounds and experience levels. These interviews aim to capture the participants' personal and professional perspectives on the advantages and limitations of incorporating AI into lesson planning, with attention to factors like creativity, learner engagement, and contextual alignment. To maintain consistency, both human- and AI-generated lesson plans followed a standard structure, including warm-up activities, main instructional tasks, and closure components. Anonymization of the plans ensured unbiased evaluations by participating teachers using the research-developed rubric. In parallel, ChatGPT independently evaluated all lesson plans to contribute supplementary data. This dual-phase design—blending quantitative scoring with qualitative reflection—provides a holistic comparison between AI- and human-generated lesson plans. Findings revealed a nuanced understanding of AI's strengths and limitations. Quantitatively, AI-generated plans achieved slightly higher mean scores across several rubric dimensions, with a statistically significant advantage in the clarity of learning objectives (p = .007). No significant differences were observed in other areas such as structure, post-lesson activities, and time management, indicating overall comparable quality between the two groups. The qualitative findings further deepened the understanding of these results. Teachers acknowledged AI's potential to reduce workload, save time, and assist in developing instructional materials. However, they raised concerns about AI's inability to replicate emotional intelligence, cultural sensitivity, and responsiveness to real-time classroom dynamics. Human adaptability, ethical implications, unequal access to technology, and the risk of overreliance on AI were key themes in participants' critiques. Nevertheless, this study has certain limitations. Its geographic focus on Diyarbakır restricts the generalizability of the findings, and the research does not extend to the implementation of lesson plans in real classroom environments—an area that could yield deeper insights into practical effectiveness. In summary, while AI demonstrates promise as a supportive tool in lesson planning, it cannot replace the human insight and adaptability essential to effective teaching. This study contributes to the growing discourse on educational technology by offering empirical evidence and practitioner perspectives that inform balanced, thoughtful integration of AI into EFL instruction.