대학 영어 수업에서 AI 기반 적응형 학습이 TOEIC 학업성취도와 자기 주도적 학습 능력에 미치는 영향

Abstract

Objectives This article aims to validate the efficacy of AI-adaptive learning (AI-AL) in college English courses. It achieves this aim by examining the influence of AI-AL on learner's academic achievements in TOEIC and self-perception of self-directed learning abilities. This investigation involves analyzing TOEIC exam outcomes, encompassing key English education components like listening, grammar, vocabulary, and reading.
 Methods We computed descriptive statistics using SPSS 25.0 and performed a paired t-test on the pre- and post- exam results of an online mock TOEIC for 61 students enrolled in Sections A and B of AI English 2 at H University. To explore the change in learners’ self-perception of their academic achievements and improved self-directed learning skills, we examined 56 selected reflective journals. Additionally, we conducted two rounds of expert Delphi assessments to validate the content.
 Results First, the analysis of the online exams showed a modest increase of 5.246 points in the average score of the listening comprehension section, although no statistically significant difference was observed. Secondly, the average score of the reading comprehension section exhibited a significant increase of 12.459 points, indicat-ing a statistically notable difference. Thirdly, upon analyzing students' reflective journals using factors employed to assess self-directed learning ability, it was found that “desire to learn” had the highest frequency (96.4%). This was followed by “self-reflection” (91.1%), “attributing results to their efforts” (66.1%), “persistence in learning” (62.5%), “goal setting” (48.2%), “self-management” (46.4%), “identifying learning resources” (35.7%), and “selecting learning strategies” (30.4%).
 Conclusions This study demonstrates that AI-AL positively influences learners’ proficiency in TOEIC RC and self-perception of self-directed learning abilities. This research is particularly relevant in today's era of hy-per-connectivity and personalized learning, as it substantiates the potential of AI-AL in catering to individual lear-ners' needs and learning pace, while offering practical solutions to enhance its effectiveness. The results of this research are expected to be useful as foundational materials for colleges looking to integrate AI-AL programs into their curriculum.