Evaluating AI-Generated Lesson Plans for English Language Teaching: A Rubric-Based Comparative Analysis of MagicSchool.ai and ChatGPT
cilt.4, sa.1, ss.25-46, 2026 (Hakemli Dergi)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 4 Sayı: 1
- Basım Tarihi: 2026
- Sayfa Sayıları: ss.25-46
- İstanbul Üniversitesi-Cerrahpaşa Adresli: Evet
Özet
Artificial intelligence (AI) tools are increasingly being adopted in educational contexts, yet empirical evidence regarding their capacity to generate pedagogically rigorous English Language Teaching (ELT) lesson plans remains limited. This study addresses that gap through a structured, rubric-based comparative evaluation of two widely used AI platforms—MagicSchool.ai and ChatGPT—as lesson planning tools for ELT. Using a standardised teaching scenario (A2-level, university preparatory English, 18 students, 80-minute session on air pollution), the researchers applied a multi-dimensional evaluation rubric grounded in constructive alignment (Biggs, 1996), Bloom's Taxonomy (Anderson & Krathwohl, 2001), the Common European Framework of Reference for Languages (CEFR), and the Learning Design specification (Koper & Olivier, 2004). Findings reveal that both tools generated structurally functional lesson plans within seconds, each with distinct affordances: MagicSchool.ai demonstrated implicit alignment with Rosenshine's Principles of Instruction, while ChatGPT produced scaffolded, dual-session plans with higher internal coherence and resource specificity. Both tools exhibited limitations in differentiation depth, digital media integration, and time allocation. The study proposes a practical prompt engineering framework—the SCOPE protocol (Scenario, Context, Objectives, Pedagogy, Extensions)—to enhance AI output quality. Implications for teacher training, institutional AI governance, and ELT professional development are discussed.