The integration of AI tools such as ChatGPT into language learning has attracted growing interest. With features like audio chat, these tools offer more opportunities for learners to practice speaking outside the classroom, which may contribute to building learners’ confidence and reduce foreign language anxiety (FLA) caused by limited exposure to the target language. This study examines the effectiveness of using ChatGPT as a language partner outside the classroom to enhance self-confidence and reduce FLA among 90 undergraduate students in Japan. Over a 15-week semester, participants engaged in three-minute audio conversations with ChatGPT. Pre- and post-intervention questionnaires were administered to assess learners’ perceptions of ChatGPT in language learning, as well as their anxiety and confidence levels. Results showed that while learners’ anxiety when interacting with ChatGPT decreased, their classroom anxiety and confidence levels remained unchanged. However, participants reported a significant improvement in attitudes toward AI in language learning, indicating increased awareness of how to use ChatGPT effectively and a willingness to explore more functions. These findings suggest that although ChatGPT has limited impact on reducing classroom anxiety and boosting confidence, learners’ positive attitudes and motivation indicate its potential as a valuable supplementary tool for speaking practice outside the classroom.

Recent advances in generative AI have significantly influenced education, including language learning. While many students recognize the potential of AI in English learning, few actively use it. To explore further the usage of AI in English learning in self-access contexts, this study has investigated the case of one university student who actively integrated AI into his learning process. Using a qualitative approach, data was collected through in-depth interviews and a learning log. This presentation will showcase AI usage patterns, strategies, perceived benefits, and challenges. Initial findings suggest the possibility of strategic and personalized English learning using AI. Especially with regard to output opportunities for EFL learners, AI has strong potential to support autonomous learning. Through interviews, the student mentioned that he used AI to correct his diary entries, learn expressions from the corrections, and better engage in conversations in the Self-Access Learning Center with others. He found diary learning highly effective and is eager to share and recommend it to other students, demonstrating a high level of autonomy and application (Nunan, 1997). This research project should contribute to a deeper understanding of AI’s role in self-directed English learning, providing valuable insights for educators.

This research critically explores how task negotiation and interaction among learners in online learning environments contribute to developing learner autonomy. By examining the patterns of these interactions and negotiations, this study aims to uncover how online learning environment, an exchange forum in the Moodle platform, either facilitates or constrains the exercise of learner autonomy. Specifically, it investigated how students’ participation in collaborative dialogues, peer-to-peer negotiations, and teacher-student interactions affects their ability to take ownership of their learning goals and decisions. The research was conducted in the short course named Global Englishes in 10 weeks with 50 English major students. Data was collected through adapted questionnaires to measure three constructs: learner autonomy perception, interaction and negotiation behavior and then analyzed by using the correlation on Spss 27 to understand the relationship among three concepts, learners’ discussion artifacts in the online forum, and an open-ended interview at the end of this course and analyzed by thematic coding based on three emerging themes: input meaning making, learner agency and collaborative learning. This research will benefit teachers and Call researchers to understand the learning behaviors in the virtual environment apart from the traditional classroom.

As artificial intelligence becomes an integral part of education and society, students must develop critical awareness of its ethical implications and biases. AI models are not neutral; they reflect the data, values, and assumptions programmed within them—sometimes increasing biases and making ethically questionable decisions. In this session, we will explore practical strategies to help students recognize and critically engage with the ethical risks of AI. From analyzing biased AI outputs to discussing consequences of algorithmic decision-making, educators will gain concrete ideas for promoting responsible AI literacy. Attendees will leave with activities and discussion prompts designed to help students think deeply about AI’s role in shaping fairness, and accountability. Much of this information is based on a course I will be teaching this fall at a Canadian university titled “Navigating AI”.

Affinity spaces (Gee, 2004) are communities which arise through some shared interest, in which the informal learning that occurs is not constrained or directed by the imbalances of knowledge, ability, or power that underlie institutional learning. This presentation, based on a case study of agency in self-directed language learning, will explore how Hiroaki, a Japanese university student, enacted agency through his participation in an affinity space focused on members' shared interest in manga and anime. The community was internationally dispersed and primarily communicated in English via social media platforms like Twitter (X) Spaces and Line. Data from a series of semi-structured, stimulated recall interviews, as well as from learner logs and written reflections, were analyzed through thematic content analysis using both inductive and deductive coding. Despite his low English proficiency, Hiroaki successfully built and maintained relationships within the group, thereby generating multiple affordances for developing his language skills and intercultural competence. This process further enhanced his agency, supporting a view of agency as primarily relational rather than individual in nature (Burkitt, 2016). The presentation will highlight key implications for language learning pedagogy and will outline possible strategies for fostering learners’ relational agency.

While AI chatbots are increasingly integrated into language education, most studies have focused on student perceptions rather than objectively measuring linguistic improvement (Chen et al., 2023; García-Sánchez & Pérez-Paredes, 2024; Kohnke & Zou, 2023). This study presents a longitudinal analysis of chatbot interactions among Japanese university students (n=18) to examine engagement patterns, linguistic complexity, grammatical accuracy, and self-correction behaviors over an academic term. Using statistical analysis techniques, including Correlation Analysis, Paired t-tests, and ANOVA, key findings reveal that students demonstrated statistically significant improvements in mean sentence length, response expansion, and self-correction awareness (p < 0.05). However, lexical diversity remained unchanged, suggesting that students relied on familiar vocabulary despite producing longer, more structured sentences. Additionally, interaction patterns, such as total turns per session and follow-up question rates, showed no significant trends over time. These results suggest that AI chatbots can effectively encourage longer, more structured responses and increased self-correction habits, but may require additional pedagogical interventions to enhance vocabulary development. This presentation will also discuss practical implementation strategies to optimize chatbot-based learning for linguistic development.

In an attempt to explore the use of lexical bundles within the emerging cross-disciplinary domain of FinTech, this study compiles a 127,472-word corpus of 113 research articles from 30 leading journals in information technology and information systems. Through this corpus, this study identifies 110 four-word lexical bundles, 42 five-word lexical bundles, and 2 six-word lexical bundles and thereafter analyzes their forms and functions in accordance with the taxonomies devised by Biber et al. (1999, 2004) and Hyland (2008). The findings reveal that (1) most of these high-frequency bundles are not uncommon in academic writing, with some expressions unique to FinTech, (2) four-word bundles are dominated by NP-based and PP-based structures, while five-word bundles primarily consist of VP-based expressions, and (3) both four-word and five-word bundles are predominantly research-oriented and text-oriented in function. However, a closer look suggests that certain bundles exhibit cross-disciplinary linguistic features that reflect FinTech’s hybrid nature—combining discourse of finance and technology. Such findings may contribute to EAP instruction by raising students’ awareness of general academic lexis and the distinctive linguistic patterns of FinTech. This study also yields insights into how corpus-informed research can provide a more efficient means of identifying lexical trends in rapidly evolving domains.

This study explores the integration of ChatGPT into a Business English writing class with 29 Taiwanese college students. The course covered 10 business correspondence genres, including resumes, cover letters, sales letters, inquiry and reply letters, complaint letters, and business reports. Students produced two drafts for each genre: one independently and one using ChatGPT. Two Business English experts—a teacher and an industry professional—evaluated the drafts. Students also reflected on their experiences through written feedback and post-course interviews, with qualitative data analyzed using thematic analysis. Results show that while ChatGPT effectively supported formulaic writing, such as sales and inquiry letters, it struggled with personalized genres like resumes and bios. However, the tool excelled in providing scaffolding, enhancing vocabulary, and improving grammatical accuracy and tone, often outperforming human-written drafts. These findings underscore the potential of generative AI in enhancing language accuracy and professional tone. This study argues for the adoption of AI tools in EFL classrooms to better prepare students for the evolving demands of global business communication.

The three presentations in this Learner Development SIG forum will explore how technology can be used to both give feedback and also support the self-directed development of students’ language learning skills. In the first presentation, Katherine Song focuses on providing EFL students with feedback on their speech output through computer-generated transcripts of their audio or video recordings. By uploading their speeches to the learning management system (Microsoft Teams), transcripts are generated rapidly, accompanied by worksheets to guide self-evaluation and improvement plans. The presenter would appreciate an opportunity to share how this technology has been valued by the students in her Japanese university EFL classes. The second presentation, by Blair Barr, reviews student appreciation of the feedback they receive from online tests as homework. However, the presentation highlights the advantages of distributing online tests through a school’s learning management system rather than with external platforms like Google Forms. Students are significantly more likely to complete tests accessed through the school's system, suggesting that ease of access is a crucial motivator for many Japanese EFL students. Despite challenges of deadlines and question formats, the students indeed recognize the benefits of the feedback provided. In the third presentation, James Underwood introduces a classroom-based Self-Directed Learning (SDL) course supported by a website that is used in place of a Self-Access Learning Centre. Students choose a language skill, set SMART goals, follow a four-week learning plan, and track their progress using shared Google Docs with teacher feedback. All presentations underscore technology's role in empowering students with accessible tools for self-improvement and learning efficiency.

In response to the Covid-19 pandemic, many self-access learning centres (SALCs) at Japanese universities shifted their programmes online. However, such a transition put more pressure on learners to be visibly and accountably more responsible for directing their own language learning experiences. This, in turn, appears to have inadvertently encouraged many of them to see the use of technology for such purposes as a distractive rather than innovative tool. Consequently, instead of applying their language skills via the technology provided, a post-Covid-19 trend has emerged where learners are making active use of machine translation to complete self-access tasks to simply finish them and/or gain credit. Such technological practices appear symptomatic of what Bauman (2000, 2012) describes as liquid modernity—when superficiality, information-avoidance behaviour and reduced decision quality supersede consideration of the most pragmatic and desirable skills and practices for learning. To this end, this presentation will first introduce Bauman’s concept in relation to learners’ use of technology for self-access language learning. Thereafter, quantitative observation data from a case-study will be disclosed on this troubling learner practice at one Japanese university’s SALC. The talk will then conclude with a discussion of the strategies being used by staff to address the issue.

The present study aims at identifying the factors that influence learner’s adoption of machine translation (MT) for learning Japanese as a foreign language (FL). Previous studies have found various perceived factors influence language learners’ adoption of MT for FL learning (e.g., Clifford, Merschel, & Munne, 2013). However, few studies have investigated this quantitatively and considered the importance of instructing the use of MT, which may influence its adoption. The present study focused on four factors influencing the adoption of MT, 1) ethical beliefs on the use of MT, 2) perceived translation accuracy, 3) perceived usefulness, and 4) perceived ease of use, and investigated its association with learners’ behavioural intention to use MT. The questionnaire was administered to beginner learners of Japanese at an Australian university (n=50) after they received hands-on training in class on how to use MT. The analysis using partial least square path modelling revealed that the ethical beliefs and perceived ease of use had significant effects on learners’ behavioural intention to use MT while the perceived translation accuracy and perceived usefulness had little effects. The implications of the findings will also be discussed in relation to how MT can be integrated into FL teaching.

This study argues that ‘assessment for learning’ can bridge the gap between language learning practices and assessment, and address washback effects in testing. Assessment for learning is defined as teaching practices and learning for providing feedback used to improve students performance. Class blogs were offered as a means of instructing and providing lesson activity feedback for six English comprehension classes to support language learning objectives and test-taking goals. Class homepages were set up as blogs, and used by the teacher and students accessed the homepages on their smartphones. The aim of this study was to investigate if blogging provided necessary support and aided in test preparation. Data was collected from i) observations made, ii) student responses from a blog survey, and iii) a comparison of test results which were evaluated using an AI application ‘Cathaven’. These CEFR test results indicated a positive washback affect, and survey blog responses showed that assessment exercises and test preparation conducted using feedback blogs were beneficial and improved test performance. I discuss these findings and give implications associated with an alternative testing and learning ESL context with relevant literature in blogs in education, which reframes blogs in the context of contemporary language education utilizing AI.

This presentation examines the administration-led adoption of a big-brand educational support platform into the English department’s online portal at a Japanese university, implemented with minimal departmental consultation. The big brand which also provides services for other national, public and private universities aims to develop “solutions that are optimal”. The proprietary platform, designed specifically for this institution facilitates students’ textbook unit reflection activities, intended to promote metacognition, help students gain insight into their own progress, identify areas for improvement, and take a more active role in guiding their own learning. One major advantage of this platform is its capacity to manage cohorts of 3,000 plus students and automatically compile quantitative results for teachers. This feature enables educators to observe trends and identify areas of focus. However, significant drawbacks are the anonymity of responses, which prevents teachers from offering personalized feedback or additional support to individual learners’ specific needs, and the department's ability to modify platform content. Drawing from the department’s experiences, this session will discuss the impact of administrative decisions, practical considerations and the limitations imposed by anonymous data on teaching practices. Attendees will gain practical insights into balancing large-scale technology solutions with the nuanced requirement for targeted learner feedback.