This study examined the effects of creating English picture books using generative artificial intelligence (AI) on Korean high school students’ reading and writing skills, AI literacy, and self-efficacy. Forty-five students were divided into two groups and participated in tasks that included selecting a character from English-speaking cultures, generating images using Bing Image Creator, drafting and revising stories with ChatGPT, and creating audiobooks with ClovaDubbing. Reading and writing skills were evaluated using pre- and post-tests, and AI literacy as well as affective factors, including selfefficacy, were measured through surveys. The results indicated a significant improvement in students’ writing skills and self-efficacy, whereas reading skills did not demonstrate statistically significant progress. The study underscores the potential of generative AI as a tool for enhancing writing skills, AI literacy, and self-efficacy in language learning. However, it also emphasizes the need for further pedagogical efforts to design instructional strategies that effectively improve reading skills. These findings offer practical guidance for integrating generative AI into EFL education to enhance language learning and AI literacy.

The purpose of the current study is to present a class model of SQ3R instruction, which consists of 5 stages of reading strategy, Survey, Question, Read, Recite, Review for foreign language learning and to investigate its effects on learners’ self-efficacy and their reading comprehension. Five reading steps of reading strategy were instructed associated with skill practices, annotating, note-taking and summary outlining. A total of 23 students in a college academic reading class participated in the study. Students’ self-selected academic reading materials were used for the course. Reading self-efficacy survey and the academic reading part of IELTS practice test were conducted before and after the instruction. Student products such as annotated texts and reaction writing as well as portfolios were collected and analyzed. The survey and academic reading test results indicated that the learners’ reading self-efficacy and their reading ability were improved and that the students seemed to have gained more confidence in academic reading.

The present study investigated which facets of perceived self-efficacy (PSE) in L2 reading are significantly related to L2 reading proficiency (L2RP), which type of linguistic knowledge feeds into PSE, and how they are related to L2RP when considered together. Participants (n = 95) were college students from two universities in Seoul. Four subcomponents of PSE were identified for investigation: text-based PSE, general PSE, PSE in linguistic knowledge, and PSE in authentic reading. The result of stepwise multiple regression analysis showed that the general PSE whose items reflect dimensions of social comparative influences and perceived controllability over environments was the only significant predictor of L2RP (R2 = 17.7%). For the relationships between linguistic knowledge and PSE, vocabulary knowledge (VK) was shown to be the only significant predictor of PSE when considered together with grammar knowledge (GK) and L2RP (R2 = 22.9%), while VK and GK were significant predictors of L2RP (R2 = 69.4%). PSE was not found to make an independent contribution to L2RP when considered with linguistic knowledge.