Blending Gen AI-Generated and Textbook-Based Materials in ELT Grammar Instruction: A Case Study

Abstract

This study investigates the implementation of Gen AI-assisted grammar instruction through the integration of Gen AI-generated learning materials and textbook-based instruction in an English as a Foreign Language (EFL) context. The instructional design combined a Gen AI tool for generating grammar exercises and speaking prompts, Quizizz as a formative assessment platform, and English for Nusantara Grade VIII as the core textbook. The study employed a qualitative case study approach involving a single Grade 8 junior high school student, referred to as Student 1, observed across five instructional meetings. Each session focused on progressive grammar topics: Simple Present Tense, Simple Past Tense, Past Continuous Tense, and Past Perfect Tense. Data were collected through pre-test and post-test assessments administered via Quizizz, classroom observation notes, and teacher reflective journals. The findings indicate a consistent improvement in grammar accuracy, increasing from 60% in the baseline assessment to 100% in the final stage. The results also show a shift in learner engagement from low participation during evening sessions to high responsiveness during morning sessions. In addition, error analysis reveals a gradual reduction in grammatical inaccuracies, particularly in subject-verb agreement, auxiliary verb usage, and tense-form selection. The study concludes that Gen AI-assisted instruction, when integrated with textbook-based learning and guided by teacher mediation and scaffolding, can substantially enhance grammar learning outcomes in EFL private tutoring contexts.