Castledown Journals
Not a member yet
968 research outputs found
Sort by
Evaluation of automated vocabulary quiz generation with VocQGen
No full textVocQGen is an automated tool designed to generate multiple-choice cloze (MCC) questions for vocabulary assessment in second language learning contexts. It leverages several natural language processing (NLP) tools and OpenAI’s GPT-4 model to produce MCC items quickly from user-specified word lists. To evaluate its effectiveness, we used the first sublist in the Academic Word List (AWL) to generate 60 questions with VocQGen. Then we compared the quality of 60 autogenerated questions with 40 manually created ones through expert reviews and through pilot testing with 68 students. Expert review results indicate that automatically generated questions exhibit higher grammatical accuracy and clearer contexts in question stems. However, the tool occasionally produces distractors that are acceptable as correct responses. Pilot testing results show that in general the number of correct responses is higher in autogenerated questions, indicating the less challenging nature of these questions. The study concludes that manual check is still required for questions generated by VocQGen and future work should focus on improving distractor effectiveness
Generation and L2 vocabulary learning: A classroom action study on the efficacy of the generation desirable difficulty in learning L2 vocabulary
No full textThe present study investigates the use of generative strategies and the effect on second language (L2) vocabulary retention and learner confidence in vocabulary knowledge over time. Generative strategies (or generation) are one in a group of learning methods known as desirable difficulties – strategies that improve long-term learning by adding challenge to the learning process. Previous research has established the effectiveness of other desirable difficulties, such as spacing, retrieval, and interleaving, in L2 vocabulary acquisition. However, the effect generative strategies have on L2 vocabulary learning over shorter timeframes (e.g., a few days to a week) has been found to lack efficacy. Nevertheless, long-term vocabulary outcomes implementing this strategy in L2 learning remain unclear. To address this gap, this study examined the effectiveness of generation tasks used as the sole desirable difficulty for vocabulary recall over two intervals: three weeks and 15 weeks. Using quantitative methods in a classroom action research project, this study investigated 38 first-year Japanese university English language learners at the B1 to B1+ CEFR English proficiency level. The treatments included generation tasks in the form of generating novel sentences for half of the target vocabulary and a control treatment utilizing other non-generative methods for the other half through two separate lessons. Additionally, pre and post-tests were implemented to assess the perceived confidence and actual knowledge of the target vocabulary. The study`s findings on longer-term timeframes were not found to be markedly better than those of shorter timeframes and only showed slight improvement. While the study failed to demonstrate a clear benefit for the additional effort required by generation tasks, it did not identify significant costs from using generation on the specific population studied, as found in other studies. Implications for the findings and future research are included
“From zero to three hundred” - intensive acquisition techniques for the 300 most frequent content words in Welsh
No full textAround twenty thousand adult learners enrol onto Welsh language classes each year, but many do not progress beyond Mynediad level (= Entry, CEFR A1). The study reported here explored techniques to maintain early learning momentum, maximise learners’ returns for their learning efforts, and accelerate initial vocabulary acquisition. Novice learners of Welsh were challenged to learn 300 Welsh vocabulary items over a period of 50 days. They were provided with physical flash cards and information about learning techniques, and were asked to spend ten minutes per day on their learning. Tests administered immediately after and two weeks after completion of the learning period showed that acquisition success varied considerably between participants, and an end-of-study questionnaire found that the more successful learners had used highly systematic approaches to word selection and revision, and had frequently used the keyword technique of word learning. Test results also revealed that some target words were substantially less successfully acquired than others, and a list of words ranked by “learnability” was compiled. Outcomes from this study are informing curriculum revision by the National Centre for Learning Welsh, the main provider of Welsh language courses for adults.
An earlier, Welsh language paper reporting this study was published in the journal Gwerddon (Fitzpatrick & Morris, 2024)
Exploring AI integration in Japanese EFL classrooms: Insights and challenges
No full textThis study addresses a significant gap in generative artificial intelligence (GenAI) research by examining Japanese tertiary students’ perceptions of AI-enhanced English language acquisition, diverging from the Western-centric literature predominant in the field. Our longitudinal mixed-methods inquiry (N = 30) integrated multiple AI platforms (ChatGPT, DreamStudio AI, Suno AI, and Luma AI) within Vygotskian socio-cultural and Intelligence Augmentation (IA) theoretical frameworks. Thematic analysis revealed robust empirical evidence of positive engagement and instructor-AI symbiosis. The findings elucidate culturally situated responses within an educational context characterized by hierarchical pedagogical traditions. Participants appreciated AI’s role in reducing anxiety and enhancing creativity but preferred human instruction in areas requiring affective intelligence. This research contributes non-Western empirical evidence to transhuman discourse, advancing theoretical understandings of culturally responsive AI technological integration in language education pedagogy
Development and evaluation of a course integrating machine translation and Generative AI in higher education English learning
No full textThis study developed and implemented a course aimed at helping students understand the strengths and limitations of machine translation (MT) and generative artificial intelligence (GenAI), utilize these AI tools to produce higher-quality English translations, and apply them to English learning as a second language. Data from 46 students were analyzed using pre- and post-course surveys, measuring their overall course satisfaction, usefulness, and motivation. The survey results showed that most students were satisfied with the course and found it useful for English learning. Further analysis revealed that students found learning the strengths and limitations of MT, pre- and post-editing, and using ChatGPT useful for English learning. In particular, understanding the strengths and limitations of MT and creating their own prompts contributed to overall satisfaction and usefulness. Additionally, students’ self-evaluated achievements in pre- and post-editing contributed to their willingness to continue learning English even as AI develops
AI as a writing feedback companion: Student beliefs
No full textGenerative AI tools like ChatGPT and Grammarly may have a significant impact on second language learners’ perceptions of feedback in academic writing. Conducted with second-year university students in Japan, the study aimed to understand how exposure to AI influences beliefs about its role as a feedback companion. Surveys before and after AI-based tasks showed that students increasingly viewed AI as helpful for English mechanics and surface level corrections but continued to prefer feedback from teachers and peers for higher-order tasks. While 51.7% of the students expressed interest in continuing to use AI for feedback, others remained uncertain. The results suggest that AI can serve as a supplementary tool rather than a replacement for traditional feedback sources and that more education is needed to maximize its potential in enhancing writing development. Understanding students’ evolving perceptions of AI is necessary for continued integration into L2 writing classrooms
Unveiling the effect of self-regulated learning strategies on learning engagement among Chinese EFL learners in the online learning environment
No full textThe significant roles of self-regulated learning (SRL) in enhancing learning engagement have been extensively explored in previous research. However, the impact of SRL on learning engagement within technology-based online English learning environments remains underexplored. Therefore, the present study aimed to examine the effects of SRL on learning engagement, following the validation of measurement tools for online SRL and English learning engagement. This research focused on EFL learners across diverse age and gender groups within technology-mediated learning environments. In this regard, a SRL questionnaire and a learning engagement questionnaire, specifically designed for technology-based online English learning, were administered to 328 EFL learners. The results provided robust validation for the multidimensional self-report measures of SRL strategies and learning engagement in technology-based online English learning contexts. In addition, the study uncovered fluctuating changes in self-regulated learning strategies and learning engagement across different age groups and genders. Most notably, the findings revealed that a majority of SRL strategies positively influenced all three types of learning engagement. Among these, help-seeking emerged as the most influential factor in enhancing learning engagement, while environment structuring showed no significant impact on any type of engagement in the technology-based online English learning context. These findings offer valuable insights for developing learners’ self-regulation to enhance their engagement in online English learning within technology-driven environments
The effects of different types of computer-assisted corrective feedback on L2 pragmatics learning
No full textDespite the widely-acknowledged and much-investigated impact of feedback on second language acquisition, surprisingly little is known about its role in the development of L2 pragmatics. This gap may stem from the limited focus on pragmatics in second language curricula, leading to inconclusive findings on the effects of various feedback types. Additionally, the role of technology in this area is understudied, despite evidence that computer-assisted language learning (CALL) feedback positively impacts language learning. The present study addressed these gaps by examining the effects of computer assisted corrective feedback (CF) on L2 pragmatics learning in an EFL context. A quasi experimental study was conducted with three PowerPoint-based corrective feedback groups, using pre- and post-testing. The feedback types included: (1) an affective component via video-based pedagogical agent (PA), (2) high-information feedback as text, and (3) a combination of text and PA videos. Statistical analysis revealed that the group receiving feedback with only the affective component made significant learning gains and outperformed the other groups. These findings support the argument that motivation, enhanced by interaction with the PA, can improve pragmatics learning performance. This highlights the potential of affective feedback components in facilitating L2 pragmatics acquisition
Unveiling students’ experiences and perspectives of generative AI-assisted translation in a Hong Kong university: perceived benefits, limitations, and suggestions
No full textAlthough recent studies have revealed the benefits of GenAI on language education, little is known about students’ experiences and perspectives on the utilization of these tools in assisting their translation practices - one of the main ways students apply digital tools to support language learning. Therefore, the current study employed in-depth interviews to explore six Hong Kong university students’ perceived benefits and limitations of GenAI-assisted translation, and suggestions they provided for students and translation programs to use GenAI tools in translation. The findings revealed that participants identified several benefits of GenAI-assisted translation, including increased translation efficiency, high capabilities in specific domains, supportive functions for translation practices, and emotional support. Nevertheless, restricted translation performance, information leakage, time and energy requested to develop quality prompt engineering skills, and negative impacts on translators and employment were recognized. Some suggestions related to the establishment of GenAI guidelines and policies, course development, provision of GenAI-incorporated activities or materials, and supplementary GenAI training were proposed for the future development of the translation program. The study helps us have a more holistic understanding of the benefits and limitations of GenAI-assisted translation. It also provides valuable insights to translation teachers and program directors to refine the current translation courses and programs to equip students with the capabilities to respond to the rapid development of GenAI tools
Exploring Chinese EFL learners’ AI continuance use intention based on the ECM-ISC: Theoretical validation and contextual extension
No full textGrounded in the ECM-ISC, this study employs semi-structured interviews to investigate the usage experiences and continuance intention of Chinese EFL learners toward AI tools. Through systematic coding and thematic analysis, ten key influencing factors within the dimensions of perceived usefulness and confirmation were identified, covering both theoretical and localized levels. Findings indicate that learners’ satisfaction with AI tools is often implicitly expressed through nuanced emotional responses, reflecting a complex psychological negotiation among functional value, cultural identity, ethical concerns, and academic norms. The study further reveals the distinctive influence mechanisms of localized factors, such as policy support and cultural context, within the Chinese educational ecology on continuance intention. Building on the validation of ECM-ISC’s core variables, this research deepens the model’s localized interpretation, enriches understanding of AI tool usage behaviors in cross-cultural digital learning environments, and offers practical implications for localized educational technology design, teacher professional development, and policy formulation