This study investigates the self-initiated use of real-time speech-to-text translation (STTT) tools among Thai university students in English language classrooms, and how they perceive these tools to support their comprehension, motivation, engagement, and participation. Adopting a mixed-methods research design, data were collected from 224 students across various faculties. Quantitative findings indicated a two-construct model comprising comprehension and engagement/participation, each validated through exploratory and confirmatory factor analyses. Students reported that STTT tools supported their comprehension, particularly in vocabulary recognition, teacher clarity, and lecture following, and were associated with active classroom engagement, increased confidence, reduced hesitation, and more collaborative participation. Both scales demonstrated excellent internal consistency (α = .95), with overall favorable perceptions (M=4.19 and SD=0.572). Thematic analysis of open-ended responses suggested that students were motivated by perceived gains in comprehension, vocabulary acquisition, increased confidence, and greater classroom participation. However, reported challenges included translation inaccuracies, technological limitations, accent-related recognition issues, and internet instability. Grounded in cognitive load theory and sociocultural theory, this research highlights the nuanced interplay between learner agency, assistive technology use, and classroom interaction. The findings contribute to the underexplored area of STTT use in low-resource language settings and offer actionable insights for educators and policymakers aiming to integrate assistive technologies into inclusive pedagogical practices. This underscores the transformative potential of STTT tools to bridge linguistic gaps and promote more equitable and effective English language learning experiences in EFL contexts. Keywords: speech-to-text; machine translation; autonomous learning; assistive technology; language learning
UNESCO (2023) has called for educators to use pedagogical approaches that prioritize human agency and a responsible interaction between humans and generative artificial intelligence (GenAI). The present study, therefore, sought to design and implement an innovative 5-stage GenAI-based project with Sports Science English for Specific Purposes (ESP) students. Specifically, the objective of the project was for students to create multimodal texts for promotional purposes while developing both critical thinking skills and vocabulary range. Upon completing the project, 42 students participated in a survey, analyzed with both quantitative and qualitative methods, to gauge the impact of the project. Regarding critical thinking skills, 95.2% of students revealed they learned it was important to use these skills when using chatbots, emphasizing the need to verify GenAI-generated content and external sources a chatbot uses. Concerning vocabulary learned, 85.7% of the students, irrespective of their self-perceived level, reported learning useful topic-based and general vocabulary, thus strongly implying generalized intentional and incidental vocabulary learning. The study’s main conclusion is that GenAI-based projects, when designed to primarily assess student critical reasoning, can create optimal conditions for vocabulary acquisition and critical thinking skill development. This approach can also have significant implications for GenAI-related teaching practice and student assessment. Keywords: Critical thinking; Vocabulary learning; GenAI; Project-based learning; ESP; Multimodal text creation
Despite their use in the classroom remaining controversial, generative AI chatbots that utilize LLMs have revolutionized all aspects of learning, formal and informal, and are proving remarkably useful in a second-language setting even if not primarily designed for foreign language learning. Across the globe, students’ and faculty members’ written English has been lifted to a new level as the widely accessible AI text-generation tools help non-native speakers produce natural-sounding idiomatic texts. The role of advanced LLM-powered chatbots in English language teaching has sparked a wave of research in recent years and the present study joins these efforts. It reports on an experiment which examined the effectiveness of ChatGPT use for students’ writing tasks, featuring 47 students of different ESL proficiency levels. Findings show that ChatGPT represents a beneficial intervention in ESL writing, particularly for lower-proficiency users, most notably improving users’ grammar, vocabulary, organization, and coherence. Our results suggest that ChatGPT use may increase writing time, though this requires further confirmation. The study thus explores the pedagogical implications of ChatGPT and similar chatbots for ESL writing instruction, and both aligns with and challenges some of the research conducted so far, proposing a subsequent longitudinal study to explore the long-term L2 gains of using chatbots as writing scaffolds. Keywords: ESL writing; ChatGPT; artificial intelligence; higher education; generative AI; English for tourism
The rapid growth of artificial intelligence (AI) has impacted diverse sectors, including language education, highlighting the critical importance of user acceptance and key factors that facilitate AI adoption. This study employs a quantitative approach to investigate the factors influencing English as a Foreign Language (EFL) students’ willingness and objections to using AI in English language learning, utilizing the AIDUA framework. An online survey was sent to EFL Vietnamese students, and 1,118 valid responses were analyzed using SmartPLS 4.0.9.8. The findings revealed that although all the factors in primary appraisal significantly influenced students’ willingness to accept AI, they had little impact on students’ objections. Surprisingly, while affective attitude significantly influences students’ willingness to use AI for English learning, it also triggers emotional resistance, revealing a complex emotional dynamic. Conversely, cognitive attitude marginally softens hesitance yet exerts a more substantial influence on students’ willingness. The study provides insights into the application of AI in English learning, highlighting the cognitive and affective aspects of students’ learning experiences. Keywords: AIDUA model; AI acceptance; cognitive and affective appraisal; English learning; PLS-SEM
This study explores the use of ChatGPT in an Academic English course at a Spanish Polytechnic university, focusing on critical thinking, writing skills, and academic integrity. It proposes a workshop demonstrating how Artificial Intelligence tools can be integrated into the classroom routine. Involving 101 engineering students, ChatGPT was used to write cover emails for a professional communication assignment. Students used comparative judgment to rank their peers’ emails from high to low quality. Their evaluations were analysed using the Legitimation Code Theory (LCT), with a focus on the Specialization dimension, examining how knowledge and personal attributes are valued within this specific context. The findings highlight the effectiveness of peer assessment in writing tasks where the use of ChatGPT is explicitly permitted as a resource. Students demonstrated the ability to differentiate quality in AI-assisted writing and provided structured evaluations of their peers’ work. The study emphasizes the need for balancing AI use with fostering originality and critical thinking. LCT provides a useful framework for understanding students’ perceptions of quality in AI-mediated writing tasks, offering insights into the integration of AI tools in educational settings.
