• Students reported better understanding of writing with screencast assignments. • Students reported technology gains from screencast and screen recording assignments. • Students reported screencast and screen recording assignments were not complicated. • Blending spontaneous speech with the writing process helped students. • Students may feel self conscious when recording their screens and voices. Inexperienced writers often resist meaningful revision, which underscores the need for pedagogical approaches that foster deeper engagement. This study explores the use of student-led screen recordings and screencasts as pedagogical tools to promote students’ ownership and confidence in their writing processes. Our study surveyed 76 student writers in First-Year Writing classrooms to investigate this approach. The findings suggest that these assignments are easy to use, focus writers’ attention on the writing process, and leverage learning opportunities afforded by the transmodal blends of writing, video, and speech. Specifically, students reported more benefits from screencast assignments that allowed them to blend spontaneous speech into the writing process. Additionally, students reported that their technology skills improved after completing either the screencast or screen recording assignment. One downside was that students tended to feel self-conscious when recording their screens and voices. Overall, these student-led assignments are worth exploring in composition classrooms as they can lead to a deeper, more hands-on understanding of the writing process.

• Postgraduates engaged behaviorally, affectively, and cognitively with GenAI feedback. • Postgraduates dealt with ChatGPT primarily as a tool for refining their proposals, not for generating content. • Postgraduates demonstrated agency by actively questioning, annotating, and negotiating feedback. • Postgraduates engaged in diverse affective responses, ranging from appreciation to frustration. As Generative artificial intelligence (GenAI) tools such as ChatGPT are becoming increasingly integrated into English as a Foreign Language (EFL) academic writing context, learners’ engagement with AI-generated feedback remains insufficiently examined. This case study investigated how four Chinese EFL postgraduates joining a course in a Malaysian university engaged with ChatGPT feedback while revising their academic research proposals. The study triangulated screen recordings, pre- and post-revision drafts, and stimulated recall interviews. Participants displayed a range of behavioural strategies, including accepting, questioning, rejecting suggestions, annotating visually, and seeking external validation. Affective responses ranged from appreciation and curiosity to doubt and frustration, particularly when feedback appeared conflicting or imprecise. Cognitively, learners applied various strategies such as evaluating, comparing, negotiating feedback, and regulating its use. Yet, they showed differing levels of engagement, shaped by individual perceptions and writing intentions. Importantly, participants regarded ChatGPT as a tool for linguistic refinement rather than content generation. Overall, the findings revealed that learners did not passively receive feedback but interacted with it in agentive and critical ways. The study highlights the interplay among these three dimensions of engagement and the importance of individual differences when evaluating the pedagogical potential of GenAI-generated feedback in academic writing.

• AI can offer useful writing feedback when used combined with peer review. • AI and peer responses were often similar and mutually reinforcing. • When AI and peer responses differed, the perspectives were often complementary. • Evaluating AI feedback fostered student agency and AI literacy. Cycles of drafting and revising are crucial for student writers' growth, and formative assessment plays an important role. However, many teachers lack the time or resources to provide feedback on drafts. While research suggests that AI feedback is high enough quality to be used for draft feedback, especially when assignment-specific criteria are used (Steiss et al., 2024), it must be used in a human-centered process. AI has the potential to reduce educational equity gaps in writing support (Warschauer et al., 2023), but when narrowly implemented, technologies can deepen divides (Stornaiuolo, et al., 2023). Peer and AI Review + Reflection (PAIRR) combines peer review best practices with AI review in an approach that emphasizes student agency and reflection. Using a mixed methods approach, this study examined student perceptions of AI utility in the context of peer review. Results indicate that AI tools offer useful feedback when combined with peer review. Students found the similarity between AI and peer feedback reassuring, while also valuing their complementary perspectives. Moreover, by evaluating AI outputs, students developed AI literacy, gaining familiarity with AI feedback's affordances and limitations while learning ethical ways to use AI in their writing processes.

This article explores teaching writing with generative AI as critical play where students and teachers engage in an ethically dialectical and aleatory game with generative AI. I qualitatively surveyed 24 writing teachers about how they teach writing with generative AI as well as its advantages and disadvantages. I discovered that teachers used generative AI to teach about the ethics of generative AI's design and rhetorical use to avoid plagiarism. Teachers also critically played with generative AI to teach the writing process of invention, drafting, revision, and editing. Specifically, the critical, dialectical interplay of human and machine invents in aleatory and emergent ways, creating moments of epiphany for students and teachers within the writing process for invention, drafting, revision, and editing while the real time pace of generative AI democratizes education, making writing and teaching more accessible for them.

• The informality of video essay narration engendered ideation in drafting and script writing for students. • Students felt more responsible and personally invested in their arguments when they narrated and dramatized them in the video making process. • While students admitted that they tended to “gloss over” written drafts when revising, the video making process prompted students to be more self-motivated in the revising process, enabling them to evaluate and develop their arguments. • Unlike oral presentations, as students viewed their video essays as audience members, they could more clearly discern if their arguments lacked coherence or depth. This self-evaluation resulted in students taking the initiative to revise their final written assignments. Although multimodal assignments have increasingly been incorporated into academic writing curricula, research into their impact on student writing remains limited. This study, conducted at a Singaporean university, required students to transform a written essay draft into a video essay and then revise their draft into a written essay assignment. By comparing students’ initial drafts and their final submissions, and analysing interviews and reflective journals, we identified significant benefits stemming from the transmediation between written and multimodal text. Specifically, we found that 1) transmediation enabled students to self-evaluate their writing as they repeatedly listened to their voiceovers, found concrete visuals to illustrate their ideas, and edited their work to fit the concise video format; 2) students broke with habitual, less useful revision practices as they were freed from the conventional and grammatical concerns of written academic text and narrated their arguments colloquially in their voiceovers; 3) students exhibited an improved awareness of audience and medium; and 4) students were more enthusiastic with the course due to the novelty of the multimodal assignment. These findings suggest that including a video essay assignment during the drafting process can serve as an effective tool in advancing students’ abilities to evaluate their own academic writing.

• This study explores GenAI's role in multimodal composition, including Adobe Firefly and DALL·E. • GenAI reshapes the composition stages of invention, designing, and revising. • Despite its limitations, GenAI offers alternative solutions to wicked problems. • Post-GenAI use, students critically revise and iterate their compositions. • The study contributes to future research and teaching of AI-assisted composition. This study explores the integration of generative artificial intelligence (GenAI) design technologies, including Adobe Firefly and DALL·E, into the teaching and learning of multimodal composition. Through focus group discussions and case studies, this paper demonstrates the potential of GenAI in reshaping the various stages of the composition process, including invention, designing, and revising. The findings reveal that GenAI technologies have the potential to enhance students’ multimodal composition practices and offer alternative solutions to the wicked problems encountered during the design process. Specifically, GenAI facilitates invention by offering design inspirations and enriches designing by expanding, removing, and editing the student-produced design contents. The students in this study also shared their critical stance on the revision process by modifying and iterating their designs after their uses of GenAI. Through showcasing both the opportunities and challenges of GenAI technologies, this paper contributes to the ongoing scholarly conversations on multimodal composition and pedagogy. Moreover, the paper offers implications for the future research and teaching of GenAI-assisted multimodal composition projects, with the aim of encouraging thoughtful integration of GenAI technologies to foster critical AI literacy among college composition students.

There have been few studies examining the variation that exists within modes of feedback: for example, comparing how electronic text feedback created using Google Docs differs from electronic text feedback created using Microsoft Word or how audiovisual feedback created using TechSmith Capture differs from audiovisual feedback created using Screencast-O-Matic. However, the programs that instructors use to create feedback have different affordances, meaning that even within a single mode, the feedback students receive on their writing can vary significantly. To better understand the variation that exists within a single mode, this study investigates how affordances of Canvas Speedgrader, Google Docs, and Turnitin GradeMark impacted electronic text feedback.Based on analysis of 131 feedback files created using the 3 programs, in conjunction with 5 student surveys, and 2 instructor interviews, the study provides insights into how instructor written commentary (location, form, type, focus, and mitigation) varied by program and how participants perceived of feedback provided through the 3 programs. The study...s primary finding is that the affordances of the programs used to create electronic text feedbackresulted in significant differences ininstructorcommentary and instructor and student perceptions of feedback.