Learners’ perceptions of feedback on writing are increasingly important, as Japanese universities incorporate more English-language curricula to deliver academic content. Using Rasch and Bayesian analyses of survey data from undergraduates (N = 247), this study explores learner views about AI-mediated feedback, underscoring the enduring importance of humans in evolving technological landscapes.
This paper investigates how international students use generative AI while composing in English to negotiate voice and agency. Drawing on qualitative case data, it argues that AI can function as a low stakes rehearsal space for multilingual writers when classroom conditions support critical evaluation, authorial control, and rhetorical decision making.
Building on earlier related conference research, this study examines how oral feedback shapes L2 revision beyond negotiated episodes. Using conference recordings, draft analyses, and interviews, it identifies how teachers select feedback, how students negotiate and apply it, and what these processes reveal about feedback uptake in contemporary, multi-sourced feedback contexts.
Generative AI can provide personalized feedback that helps learners improve their writing. However, the effectiveness of such feedback depends on students’ feedback literacy. This study investigates university students’ feedback literacy in generative AI-mediated L2 writing using an explanatory sequential mixed-methods design combining survey and stimulated recall data.
“Ecologies of Multilingual Writing” (EMW) conceptualizes SLW as an open-ended ecology in which phenomena traditionally studied separately (grammatical feedback, AI, genre, multilingualism, revision) are viewed in terms of how they interact. EMW also recognizes writing as a powerful tool for changing the world amid rapidly growing environmental crises.
This study investigated the impact of AI-assisted feedback on EFL students’ writing performance within a process-oriented framework. Fourteen Chinese-speaking English majors in Taiwan received ChatGPT-generated feedback during their writing process. Qualitative and quantitative findings indicate improved writing performance; students reported that AI feedback was particularly helpful for brainstorming and revising.
This study explores how L2 writers engage with AI-mediated support during academic writing revision. Drawing on drafts, revision histories, system logs, questionnaires, and interviews, it shows how an AI-supported Dynamic Assessment platform makes writers’ engagement visible, revealing shifts in revision behavior, meaningful feedback uptake, and developing independence in mediated revision.
Using an interactive digital map to explore writing sites at diverse institutions created by leveraging AI for data collection, we reveal trends in institutional support and requirements for multilingual writers. Via the map, users can identify trends across a range of institution types/key variables.
This presentation investigates how low-intermediate EFL students revise short narratives after receiving feedback from ChatGPT generated using a constrained prompt. Analysis of paired drafts shows that constrained AI feedback does not lead students uniformly toward extensive rewriting. Instead, students adopt suggestions selectively, producing varied patterns of revision.
Peer and AI Review + Reflection (PAIRR) offers a pedagogical model for integrating GenAI into ESL classes; students engage with peer and GenAI feedback to support their revision processes. We discuss a mixed-methods analysis of PAIRR’s impact on the linguistic diversity of students’ writing and their confidence in their writerly voice.
In AI-mediated writing classrooms, how is the humanity of teaching redefined? Drawing on EFL students’ experiences with teacher and generative AI feedback, this study shows how they strategically integrate both. Findings highlight how GAI expands access and support, while teachers sustain interpretation, judgment and human engagement in learners’ writing development.
