FROM TEACHER TO TOOL: A COMPARATIVE ANALYSIS OF HUMAN-GENERATED AND AI-GENERATED FEEDBACK ON ESL LEARNERS' GRAMMAR
Abstract
The increasing integration of Artificial Intelligence (AI) in language education has significantly transformed feedback practices in ESL writing. This study examines the comparative effectiveness of human and AI feedback in improving grammatical accuracy among higher secondary ESL learners in Pakistan. Using a quantitative quasi-experimental pre-test–post-test design, 100 students were divided into two groups: one receiving teacher feedback and the other receiving AI-based feedback through tools such as ChatGPT and Grammarly. The study focused on three common grammatical problem areas: articles, prepositions, and verb forms. Data were collected through writing tasks and analyzed by comparing error frequencies before and after the intervention. The findings reveal that both feedback types significantly contributed to improved grammatical accuracy. However, human feedback proved more effective in enhancing learners’ comprehension and long-term application of grammatical rules, while AI feedback facilitated immediate correction and increased opportunities for practice. The findings have important implications for teachers, learners, and educational policy in technology-enhanced learning environments.
Downloads
References
The increasing integration of Artificial Intelligence (AI) in language education has significantly transformed feedback practices in ESL writing. This study examines the comparative effectiveness of human and AI feedback in improving grammatical accuracy among higher secondary ESL learners in Pakistan. Using a quantitative quasi-experimental pre-test–post-test design, 100 students were divided into two groups: one receiving teacher feedback and the other receiving AI-based feedback through tools such as ChatGPT and Grammarly. The study focused on three common grammatical problem areas: articles, prepositions, and verb forms. Data were collected through writing tasks and analyzed by comparing error frequencies before and after the intervention. The findings reveal that both feedback types significantly contributed to improved grammatical accuracy. However, human feedback proved more effective in enhancing learners’ comprehension and long-term application of grammatical rules, while AI feedback facilitated immediate correction and increased opportunities for practice. The findings have important implications for teachers, learners, and educational policy in technology-enhanced learning environments.
