The mass influx of English-speaking “TikTok refugees” to Xiaohongshu (also known as RedNote) has not only reshaped Xiaohongshu into a dynamic space for intercultural exchange but has also unlocked a wealth of opportunities for informal language learning within an integrated, accessible, and engaging translingual and transcultural digital ecosystem. Building upon the theoretical construct of language learning affordances (Barton & Potts, 2013; van Lier, 2004), this article examines how the emergent features of Xiaohongshu, catalyzed by digital mass migration, can be harnessed as valuable resources to support informal language learning. Focusing on a sample of 40 English-speaking content creators who self-identified as TikTok Refugees on Xiaohongshu, we gathered data comprising 120 bilingual posts and their associated commenting areas through hashtag and keyword tracking (Sloan & Quan-Haase, 2017). A qualitative content analysis of the collected data revealed three key patterns of affordances: naturalistic language communication through platform-mediated authentic interactions and role-reversal dynamics; communitybuilding learning features through collaborative atmosphere and mentorship networks; and identity development processes oriented toward a “global village” imaginary. While our analysis focuses on Chinese EFL (English as a foreign language) learners, the findings have broader implications for understanding how unexpected digital migration can transform social media platforms into vibrant spaces for bilateral language learning, including both English and Chinese as second languages. The article finally points out strategies for pedagogically leveraging these affordances as bridging activities (Thorne & Reinhardt, 2008) across formal and informal contexts.
2025-10-10
Extensive reading (ER) has long been a foundational construct in second language acquisition (SLA) theory, due to its ability to foster implicit linguistic knowledge and improve fluency. Rooted in the principles of providing learners with abundant, comprehensible, and self-selected input, its implementation has often faced practical challenges related to resources and assessment. This article introduces the key construct of ER in the tech-driven, AI-powered era. Drawing on the seminal work by Day and Bamford (1998) as well as recent research on the primacy of input-based learning, we argue that AI does not render ER obsolete; rather, it offers vast opportunities to scale, personalize, and enhance ER programs. AI-powered tools can address traditional implementation challenges by facilitating content curation, providing adaptive scaffolding, and enabling more efficient monitoring.
In recent years, Q methodology has become increasingly popular in applied linguistics and language education research, as a surge of Q studies published in these two fields could be spotted especially in the past five years. Nonetheless, most of these Q studies only present the procedures of Q in a descriptive manner without discussing or justifying the theoretical underpinnings or rationale of Q in depth largely due to the word limit of journals or their empirical focuses. To date, Q still remains contentious in academia, and the debate on Q’s theoretical underpinnings and practical application are still ongoing. Therefore, there is a dire need for more articles to offer a theoretical illustration of Q and to justify several key steps during implementing Q in applied linguistics and language education, which is the primary aim of this article. In this article we elucidated the historical background of Q and its form of inference as well as the research paradigm. In addition, we also addressed some key practical issues in applying Q, including its study design, factor extraction and factor rotation. We hope that this article serves as a resource or reference to novice Q researchers in applied linguistics and TESOL as well as other related areas and contribute to the hot debate on Q.
2026-07-16
The significance and malleability of learner engagement in second language (L2) learning have prompted L2 researchers to investigate its predictors, such as emotions, cognition, and learning attitudes. However, no prior study has examined the roles of growth language mindset and intrinsic motivation in engagement, particularly in the context of L2 pragmatics learning. To address this lacuna, this study aimed to explore the extent to which growth pragmatic mindset could predict cognitive pragmatic engagement, either directly, indirectly through intrinsic motivation, or both. A total of 262 Indonesian first- and second-year non-English major students consented to participate by completing a questionnaire measuring the three constructs. Partial least squares structural equation modelling (PLS-SEM) analysis revealed that while growth pragmatic mindset did not directly predict cognitive pragmatic engagement, it had a significant indirect effect on intrinsic pragmatic motivation. A significant relationship was also evident between intrinsic motivation and cognitive engagement. Growth pragmatic mindset was found to indirectly predict cognitive engagement through intrinsic motivation, with a large effect size. These findings are discussed in light of mindset meaning system theory and self-determination theory, and pedagogical implications are proposed based on the results.
The prevalent use of AI technology in education has sparked versatile innovations for pedagogical enhancement. Teachers possess unique and irreplaceable qualities that are crucial to education. In the context of L2 students’ genre-based writing, it remains an open question how GenAI and human teachers’ feedback can complement each other to enhance students’ learning outcomes. This study compares 14 feedback reports from teacher markers who utilised AI-recommended action points to varying degrees with the original AI-generated feedback report of a sample book review, along with data from a focus group interview and 12 reflective journals. It explores how teachers and Generative AI collaboratively construct feedback through an interactive AI-assisted platform tailored for genrebased writing. This study highlights the importance of teacher engagement and teacher agency in AI-enabled settings. The results reveal the complementary strengths of Generative AI (GenAI) and teachers in feedback practices. However, its effectiveness relies on teachers’ critical engagement and familiarity with technology. A balanced approach is essential to maximise benefits and address challenges in AI-assisted feedback. Based on our findings, a theoretical model illustrating teachers’ behavioural engagement with GenAI feedback has been derived to evaluate the effectiveness of AIassisted feedback. This model highlights teachers’ engagement with the GenAI feedback based on their agency, influenced by their capacity and willingness to perceive and shape the affordances of GenAI feedback. This research contributes to the ongoing conversation on the future of feedback delivery in an AI-driven world.
This paper introduces the construct of feedback literacy, charts its development and highlights key research findings from applied linguistics. The concept of feedback literacy originated in higher education research and has been enthusiastically taken up in L2 writing and English for Academic Purposes. Feedback literacy involves working with feedback information to enhance performance or ongoing learning. Undergraduate students need student feedback literacy; their instructors need complementary teacher feedback literacy to scaffold student feedback literacy; and university scholars themselves need academic feedback literacy to publish research, manage peer review and enhance their teaching. Applied linguistics research on feedback literacy has focused on various aspects, such as peer feedback and written corrective feedback, and also started to investigate less heavily researched sub-topics, such as feedback seeking and exemplars. The interplay between automated writing evaluation, generative AI as a feedback source and student feedback literacy is addressed. The article concludes with a discussion of limitations and critiques of feedback literacy research and suggests some further research directions.
This research explores the dynamics in perceptions of L2 English learners in task-based speaking interactions with human interlocutors in comparison to AI (chatbot “Doubao”) communication partners over seven weeks. Findings from three rounds of interviews indicated enthusiastic initial adoption of the AI talk pal, driven by its ability to stimulate output, its ease of access, and its lower affective barriers to communication. However, with continued use, their fervent zeal had cooled to being “lukewarm” and finally “cool” due to AI’s lack of social presence, speech prosody, responsive interaction, and appropriate feedback, whereas interaction with real communication partners was consistently preferred for their authentic nature. In light of AI’s limitation, learners adapted to develop prompt literacy which enabled them to deliberately employ AI for particularised tasks, such as increasing fluency and acquiring vocabulary. Based on the findings, a dynamic model for AI acceptance in L2 task interaction is proposed to advance our understanding of the role of AI in taskbased learning and teaching. It suggests that a hybrid pedagogy, which sequences AI and human interaction to leverage their complementary strengths, is essential for effective task-based language learning.
The influence of Artificial Intelligence (AI) technologies in different areas of English as a foreign language (EFL) education has been widely explored in recent years. However, the emotional consequences of adopting such tools in at-risk education have remained underexplored, to date. To address the gap, our study drew on the control-value theory (CVT) to uncover the role of AI adoption in reducing 35 at-risk undergraduate EFL learners’ absenteeism and boredom. The data were collected from online interviews and a narrative frame. Thematic analysis of the data revealed five ways in which AI integration could regulate both constructs of absenteeism and boredom. Particularly, it was found that by ‘offering personalized learning’, ‘increasing classroom engagement’, ‘providing instant and tailored feedback’, ‘diversifying learning materials’, and ‘granting agency and autonomy’, AI tools affected learners’ absenteeism and boredom. These findings suggest that AIbased professional development could empower EFL teachers to more effectively support at-risk learners’ emotional and behavioral needs. The study contributes to our understanding of emotional aspects of AI adoption in educational domains.
This paper explores the role of artificial intelligence (AI) in feedback on second language (L2) writing, focusing on both Automated Writing Evaluation (AWE) and emerging Generative AI (GenAI) tools. It examines AI-generated feedback across five dimensions: accuracy, relevance, empathy, learner engagement, and educational impact. The analysis shows that AI is effective in supporting surfacelevel features such as grammar, vocabulary, and clarity, and offers advantages in terms of immediacy and accessibility. However, it is less reliable in addressing deeper aspects of writing, including argumentation, disciplinary conventions, and audience awareness. Student perspectives suggest that AI feedback is valued for revision support but does not replace the depth and interaction of teacher feedback. The paper argues that AI is most effective when used in combination with teacher guidance, supporting both efficient feedback practices and meaningful learning.
This study presents a task-based needs analysis (TBNA) for English for International Communication (EIC) majors at a university in Thailand, with the aim of providing an empirical foundation for taskbased language teaching (TBLT) course design in the Thai EFL context. Using a six-cycle Delphi methodology, data were triangulated from multiple sources and methods: document analysis of job placement records; semi-structured and structured interviews with graduates and workplace managers; a means analysis of institutional constraints; a confirmatory survey administered to 70 in-service graduates; and an analysis of target discourse (ATD) based on authentic workplace interactions. Results revealed that EIC graduates are predominantly placed in customer service positions across the hospitality, retail, and recreation sectors, and that eight task types are critical across these workplace domains, with giving directions, responding to emails, and providing reception services rated most highly. Successful task performance was conceptualized primarily in terms of effective communication and pragmatic competence rather than grammatical accuracy or syntactic complexity. An ATD of the task ‘Giving Directions’ identified the discourse structure and linguistic strategies characterizing prototypical task performance in Thai workplaces. The findings provide a principled approach for developing needs-based instruction, criterion-based assessments, and pedagogic tasks for additional language learners with specific needs.
Over the past few decades, second language (L2) listening assessments have become an integral component of language teaching and learning. However, there have been few systematic reviews examining the validity, reporting practices, and test specifications of existing L2 listening assessments. To gain a comprehensive understanding of L2 listening assessments, we reviewed 126 studies (comprising 160 listening assessments) published between 1997 and 2021. We focused on test specifications, test characteristics, and test validation by adopting established frameworks for test specification alongside the argument-based validation (ABV) framework. The studies were coded using a coding scheme organized into six variable categories, including administrative information, test information, textual features, participant information, validity inferences, and results. The results showed that standardized listening assessments were widely used by L2 listening researchers, and that test specifications across studies were often heterogeneous. In addition, the majority of studies provided limited information on test specifications, thereby limiting the replicability of findings across different contexts. Finally, the distribution of validity evidence across the six ABV inferences was markedly uneven: the generalization inference was the most frequently examined (56.88%), whereas no study provided evidence for the domain description or utilization inferences, and 38.75% of assessments (62 out of 160) presented no validity evidence of any kind. Implications for L2 listening assessment and test development are discussed.
8-72 characters
1 day ago
Your contribution has been approved successfully
12 hours ago
Your submission has failed to be reviewed; Failure reason: the
