Publication details
Who Graded Me? Exploring Students' Perceptions of AI vs. Human Assessment
| Authors | |
|---|---|
| Year of publication | 2026 |
| Type | Appeared in Conference without Proceedings |
| Citation | |
| Description | This poster presents the findings of an experimental study which explored how university students perceived feedback generated by artificial intelligence (AI) compared to feedback provided by human evaluators. Conducted within a bachelor’s course on academic writing in Educational Sciences, the study examines students' perceptions of the usefulness of feedback and their ability to distinguish between AI- and human-generated assessments. Throughout the course, students submit weekly writing tasks that are randomly evaluated by either AI or human evaluators, unaware of the source of their feedback. Alongside these assignments, students are asked to reflect on their assumptions about the source of the feedback, and how useful they perceive the feedback to be. The preliminary analysis aimed to clarify whether students could tell whether the feedback they received was generated by AI or humans, and whether the perceived usefulness differed by source. Using mixed-effects models on repeated ratings, students demonstrated an above-chance ability to discriminate between AI- and human-generated feedback that was greater than chance, selecting higher 'AI' categories when the true source of the feedback was AI. In contrast, perceived usefulness did not differ between AI- and human-generated feedback. |