Skip to main content

Teaching and assessing reflective writing in generative AI-mediated higher education: Vignettes and provocations

PDF

Abstract

This Pecha Kucha presentation offers a situated account of teaching and assessing in the evolving context of generative AI-mediated higher education. It aims to open up space for contemplation and dialogue around the complexities of the current teaching and learning landscape, where students and educators alike are grappling with new norms, and much grace and understanding needs to be extended on both sides as each learns from, and with, the other how to better navigate the complex nuances and permutations of AI-mediated academic work.

Drawing on experiences from a postgraduate academic writing course that I teach, I share a selection of pedagogical vignettes and provocations emerging from my marking of the end-of-semester reflection assignments for the course. This assignment, requiring students to highlight key insights gained from the course and to reflect on these with reference to course discussions and their own scholarly practices, is one that I have greatly enjoyed marking over the years. However, vastly different patterns of engagement have been emerging from the students in recent semesters, and it seems evident that the affordances of generative AI, in changing how students tackle academic assignments, are at the same time disrupting some of higher education’s long-held assumptions about what constitutes learning, writing, and ownership.

The pedagogical encounters highlighted include: students asking my own course chatbot to reflect on the course for them, reflections being flagged by AI-detection software as being potentially entirely AI-generated, diverse understandings regarding what constitutes one’s “own work” where AI has been used to assist in the writing process, and students unquestioningly accepting AI’s reformulations as “better” than their own (even when they were not). Encountering these has caused frustration, but I also see in these examples deeper epistemological and pedagogical tensions. Literacy scholars, particularly those working within an “Academic Literacies” perspective (e.g. Jones et al., 1999; Lea & Street, 1998; Lillis, 2003), have long argued that reading and writing are “social practices.” They are, in other words, situated within, influencing and being influenced by, the social and cultural contexts in which they take place. The educational landscape of today, with its myriad AI-assisted possibilities, raises questions of whether, how, and to what extent, our conceptions about “learning,” “writing,” and “ownership” may need recalibration. In fact, scholars have begun to recognise that current writing practices resulting in hybrid texts co-constructed by humans and AI are making it difficult, if not impossible, to identify distinct boundaries between AI-generated output and a student’s own intellectual efforts (e.g. Eaton, 2023; Wu et al., 2026).

Higher education is in a season where excitement, curiosity, and unease about generative AI’s impact on academic work and learning sit in tension. Even as one who has actively sought to engage students in discussions of how we can productively and legitimately use generative AI in our writing, I found myself having unexpected reactions (of relief) when I encountered assignments with linguistic infelicities of the sort that suggested their writers had not used generative AI. The way forward, this presentation suggests, requires an intentional commitment to prioritise pedagogical approaches that foreground human connection – both in terms of cultivating a renewed intellectual and emotional connection of students with their own work, and in terms of building connection between students and instructors through open dialogue and compassionate, nuanced approaches to teaching and assessing.

Keywords

Generative AI (GenAI), higher education, academic writing, student reflection, teaching, human centred learning.