This is a corpus-based study to compare stance features in excellent L2 engineering writing rated by English teachers and a GenAI tool. From the lens of stance features, the findings reveal the limitations of using GenAI to assess second language writing.
This paper reports on the development of a new rubric used to score two assessments for English learners in K-12 U.S. public schools. During development, researchers sought input from multiple groups, including educators and experienced trained raters, and concluded with an empirical validation study of the new rubric.
This qualitative case study examines four L2 writing feedback teacher-researchers, exploring how research shapes their feedback beliefs/practices, noting alignments/misalignments, and highlighting their role as research-practice bridges.
This design study traces a three-year, iterative implementation of critical GenAI literacy in second language writing programs. Moving from a short unit to an integrated curricular model, the study highlights evolving frameworks, instructor and student tensions, and the complex negotiations required to incorporate GenAI ethically and effectively at program level.
Critical AI literacy (CAIL), extending beyond AI literacy, empowers L2 writers to navigate ethical, cultural, and linguistic tensions inherent in human-AI interaction. This qualitative study of 12 Chinese EFL learners using ChatGPT explores CAIL enactment, influenced by proficiency and context, highlighting varied patterns and implications in AI-mediated writing.
This study examines how graduate EFL students’ AI-mediated autonomy and AI dependence develop across three summary-writing tasks. Using questionnaires, AI interaction logs, and student texts, it explores changes in AI use and identifies trajectories suggesting increasing autonomy, persistent dependence, or mixed patterns.
This study examines how multilingual undergraduate writers enact genre awareness when engaging with AI-generated feedback in a genre-based writing course. Drawing on ecological and sociocultural perspectives, it analyzes uptake decisions and justificatory discourse to show how genre awareness mediates writers’ regulation of AI suggestions in socially situated writing contexts.
This presentation examines how an ESL graduate teaching assistant Zahra enacted her agency through the critical use of generative artificial intelligence (GenAI) in teaching freshman composition at a U.S. university. The presentation draws on longitudinal data collected from the first year of Zahra’s graduate study and examines three teaching-related critical incidents in which GenAI use played a significant role in facilitating writing and communication.
This presentation offers an in-depth tour of a new AI-powered teaching application developed in The College called VibeCheck. The presenters will describe the problem that informed the application’s development, demonstrate how the application’s design addressed this problem, provide success stories from the initial rounds of BETA testing, and project how the application’s infrastructure makes future developments possible.
This mixed-methods study systematically investigated university students’ motivations and perceptions of GenAI-enhanced L2 French writing. Results indicated high motivation, with significantly higher autonomy and competence than relatedness, and positive perceptions across different French proficiency and writing grades. Findings highlight critical digital literacy and community-focused strategies for GenAI-enhanced writing.
This study examines Singapore English teachers’ beliefs and practices in classroom writing assessment and the role of technology, including AI. Based on questionnaire responses from 76 teachers, findings show a hierarchy of strong formative assessment beliefs, moderate implementation, and low technology use, revealing gaps between assessment beliefs, classroom practices, and technology integration.
In this presentation, I examine the side effects of teacher feedback in L2 writing. A synthesis of literature reveals that feedback may negatively impact students’ writing quality and feedback engagement, while leading to teachers’ stress and burnout. Based on these findings, implications for feedback practices and further investigation are proposed.
This qualitative multiple-case study examines how GenAI tools reshape TOEFL writing teachers’ professional authority in high-stakes test-preparation contexts in China. Drawing on epistemic cognition, teacher professional identity, and accountability, it suggests that GenAI redistributes authority as teachers negotiate expertise and responsibility when interpreting algorithmic feedback and standardized rubric alignment.
Despite AI reshaping writing assessment, fairness remains a matter of human judgment as well as careful measurement. This study uses cross-classified modeling of an operational L2 writing placement test to examine prompt comparability and argues that writing teachers’ interpretive and ethical roles remain central in identifying and addressing inequitable task effects.
This study examines how L2 writing feedback conferences with human teachers and AI tools differentially affect second-language learners’ basic psychological needs and task engagement. Using a between-subject randomized experiment, we demonstrate a statistically significant group difference in psychological needs and task engagement and also address how conferencing mode impacts revision quality and knowledge transfer.
This scoping review synthesizes 27 studies (2012–2025) on L2 student feedback literacy, mapping research contexts and designs. Findings reveal geographic and participant concentration and a predominance of qualitative approaches. The review highlights gaps in diversity across participants and settings and calls for greater methodological transparency, rigor, and conceptual development.
This study examines how L2 writing tutors’ metacognitive experiences shape tutoring practices in a multilingual university writing centre. Analysis of interviews, observations, and stimulated recall from six tutors identifies tutor-centred, tutee-centred, and learning-centred experiences that dynamically inform directive, interactive, and facilitative approaches, highlighting tutoring as contextually negotiated expertise.
This study examines the ecological transformation of ESL writing in the AI era. Drawing on Barthes’ notion of neutral writing, I theorize algorithmic silence as a threat to authorial agency. The paper proposes a pedagogy of ecological voice that helps teachers support students’ presence and agency in AI-mediated writing environments.
This paper argues that learners may use AI in ways that reinforce a form-based orientation, bypassing communicative competence development. It proposes the teacher as an expander of experiences and beliefs, cultivating struggles and reflections that underlie communicative competence—not easily provided by current AI-human dynamics, which predominantly accommodate learners’ beliefs rather than challenge them.
Drawing on a year-long study of Hong Kong primary classrooms, this study identifies a three-dimensional model of metacognitive instruction and shows how teachers’ structural, cognitive, and relational practices enhance learners’ metacognitive knowledge, strategic awareness, and metacognitive experience in L2 writing.
This presentation shares findings from a case study that examines how L2 writing teacher (L2WT) expertise is conceptualized and developed at a training practicum for teaching L2 writing at the college level. It discusses the context-dependent nature of L2WT expertise and its implications for L2WT preparation.
This study examines how pre-service teachers in South Korea used generative AI while writing English lesson plans. Drawing on a cognitive writing model, it analyzes AI use across stages and perceptions of its usefulness. Finding show reliance on AI for translation and idea generation, highlighting implications for ESP writing instruction and AI literacy.
This study examines how narrative ownership is negotiated in an EFL picture book writing project integrating hand-drawn, digital, and AI-generated illustrations. Grounded in task-based writing pedagogy, it explores how students navigate authorship, creative agency, and technological mediation, highlighting the teacher’s role in preserving human storytelling in AI-augmented writing environments.
This paper presents the application of the AcOnHE e-tivity framework to an online L2 academic writing course at a Ukrainian university. Through qualitative case study analysis, it maps the course design onto Salmon’s five-stage model, identifying implications for writing teachers and course designers working in online and disrupted learning environments.
This study examines affective and interactional dynamics during young learners’ L2 writing assessment. Using multimodal interaction analysis from test administrations with 57 children aged 5–6, we show how student interruptions and test administrator responses shape agency and engagement. Results demonstrate writing assessment as a relational and interactionally mediated process.
This presentation examines the use of myProse, an AI-supported analytic writing environment, in two first-year writing courses at a multilingual U.S. branch campus. Analysis of student reflections shows that analytic AI can support diagnostic awareness, authorial control, and critical AI literacy when instructor-mediated and embedded within process-oriented writing pedagogy.
This study examines how pre-service EFL teachers learn to compose in AI-mediated multimodal environments within a technology-enhanced language teaching course. Using social semiotic and New Literacies perspectives, it analyzes students’ artifacts and reflections to explore shifts in authorship, design practices, and pedagogical understanding, highlighting implications for preparing future writing teachers in the age of AI.
