Author(s):
Yufei Wan
Abstract:
Generative artificial intelligence (GenAI) is increasingly being used as a cognitive partner rather than merely an auxiliary writing tool. Previous research has largely examined learners’ evaluations of GenAI functions and perceived risks, whereas role perceptions in human-GenAI collaborative writing revision have received limited systematic attention. Using Q methodology, this study identifies two role-perception types among EFL learners: the Open Collaborative Type and the Prudent Autonomous Type. Both groups value integrating GenAI suggestions with their own ideas and reject delegating all revision decisions to GenAI. However, they differ in their views of feedback evaluation and uptake, revision-goal setting, and control over the revision process. These findings offer evidence on how learners understand cognitive responsibilities in GenAI-supported L2 writing revision.
Keywords:
Role perceptions; Q methodology; human-GenAI collaborative revision
Article Info:
Received: 15 Oct 2026; Received in revised form: 10 Nov 2026; Accepted: 14 Nov 2026; Available online: 21 Nov 2026
DOI:
10.22161/ijels.116.26