2023, Vol.24, No.

Modern English Education 2023 KCI Impact Factor : 1.55

pISSN : 1598-0782 / eISSN : 2586-6141
https://journal.kci.go.kr/meeso

Artificial intelligence (AI)-based tools for teaching and learning English as a foreign language have become increasingly functional, leading to the growing use of AI translators and chatbots in fields of English education and research. This study investigated the potential of English-to-Korean machine translators (MTs) as media of interactive language learning to support Korean EFL learners’ reading comprehension development. Using translation data from research articles on English passives involving NP movement, this study evaluated how well two popular MTs could translate English passive sentences into Korean. This study analyzed MTs’ output in two ways: (i) whether the product’s meaning was accurate, and (ii) what type of Korean passive the product was. First, it calculated rates of semantically accurate and grammatically appropriate translations, both actives and passives produced by the two MTs. Second, it calculated rates of three different types of Korean passives produced by MTs. Results demonstrated that currently available English-to-Korean MTs could produce semantically correct translations at high rates in the form of Korean passives. Active voice constructions also showed significant rates. These findings suggest that MTs have strong potential as tools for English reading pedagogy.

Since the early 1990s, effects of testing on teaching and learning have drawn increased attention. However, most studies in second language assessment have been quantitative. Although a few studies were qualitative, they lacked a conceptual or theoretical framework, thus failing to provide a comprehensive picture of washback. This is also true in the Korean context. An operational model is needed to effectively investigate washback. This paper reviews prior research studies that have introduced seminal washback hypotheses and empirically driven operational models. These studies have identified significant macro- and micro factors such as curriculum, perception, beliefs, students’ and teachers’ previous experiences, and tests that can impact teaching and learning. Based on a review of existing operational models, this paper proposes Engeström’s activity theory as a new and comprehensive model of washback that can be adopted in any context to analyze high-stake language examinations. These findings have theoretical and practical implications for qualitative washback research.

The aim of this study was to investigate effects of English classes using AI PengTalk on English pronunciation ability of elementary students and affective factors of English pronunciation. Forty-two participants (5th graders) were divided into experimental (n = 21) and control (n = 21) groups. Their pronunciation scores of pre-and post-pronunciation tests focusing on evaluating segmentals, intonation, stress and rhythm, and speech rate were quantitatively analyzed. Data from a questionnaire about affective factors of English pronunciation including a journal writing were quantitatively and qualitatively analyzed. Findings of this study indicated that: 1) there were statistically significant differences in English pronunciation abilities between the two groups, 2) there were also statistically significant differences in affective factors such as self-confidence, interests, and attitudes, and 3) students indicated in their responses to open-ended questions and their selfevaluation journal that AI PengTalk classes helped them speak English well and enhanced their self-confidence, interests, and positive attitudes in their pronunciation. These results suggest that AI PengTalk need to be actively used in schools as a good tool for developing students’ English pronunciation abilities and that primary school teachers should focus on their pronunciation teaching for promoting students’ self-confidence, interests, and attitudes towards pronunciation learning.