Effects of ChatGPT on Korean EFL Learners’ Main-Idea Reading Comprehension via Top-Down Processing

Abstract

This study quantitatively and qualitatively investigated the effects of ChatGPT on top-down processing in English reading comprehension among Korean English as a foreign language (EFL) learners. Participants were divided into experimental and control groups of 20 individuals each, completed pretest, immediate posttest, and delayed posttest sessions, and received explicit main idea identification instructions for Finding the Main Idea (FMI) test items from the Korean College Scholastic Aptitude Test. However, only the experimental group had access to ChatGPT. The results showed a statistically significant improvement in top-down processing skills in the experimental group. Additionally, a qualitative analysis of interviews with learners and ChatGPT-human interaction data revealed responses to ChatGPT’s use in English reading comprehension, emphasizing its potential benefits and challenges. These findings highlight the significance of balancing technological integration with pedagogical instruction to optimize learning experiences in the Korean EFL context.

Ähnliche Arbeiten