Exploring AI applications in essay-based assignments: affordances and risks

Abstract

This study examined the feasibility of employing artificial intelligence (AI) for feedback provision on essay-based assignments in a UK Higher Education setting. Although the critical role of feedback in enhancing students’ learning experiences is widely recognised, resource limitations and large student numbers often hinder its quality and timely delivery. Through in-depth interviews with four participants from a university in the UK, this research investigated AI applications in essay evaluation, utilising data from 12 AI-generated essays and their corresponding feedback. The aims of the study are to evaluate tutors’ abilities in discerning human and AI-generated essays, as well as evaluating the quality of AI-generated feedback from their perspectives. Findings showed that assessors could detect certain characteristics consistent with AI generation and noted ethical concerns regarding deviations from academic standards. Participants also acknowledged AI’s capacity for swift feedback delivery as compared to human. The results of this study help enhance our understanding of AI’s affordances and risks in assessment and feedback, particularly in the less explored university essay assignments.

More information

Identifiers

ORCID for Ying Zheng: orcid.org/0000-0003-2574-0358

Catalogue record

Export record

Altmetrics