Spotlight: Improving Writing Through AI Feedback
VWBPE 2025 Spotlight
Presented: April 05, 2025
Presenter: Zeppelin123 Resident (Dr. Laura Gilmour, PhD, Psychological studies in education)
Categories: Research and Studies
Attendees of this presentation will delve into a comparative study that evaluates the efficacy of Artificial Intelligence (AI) software versus instructor feedback in the revision process of academic writing. The study features approximately 60 second-year developmental psychology students from an Urban Canadian Undergraduate University, who engage in a structured three-phase draft revision of a discussion paper. Initially, students will create drafts without any feedback. They will then be randomly assigned to groups, where they will receive feedback either from an instructor or through an AI system designed for self-assessment.
The presentation will showcase how each group's drafts evolve over the phases and the impacts of each feedback method on student writing. An independent research assistant will grade each draft using a specific rubric focusing on organization and content. Results analyzed through a repeated measures two-way ANOVA will be discussed to highlight significant differences and insights on the effectiveness of AI and human feedback in educational settings.
Objectives:
Describe an ongoing study on AI and self-editing in a university psychology class.
Explore ethics and safeguards for student AI use and collecting data from current students.
Accessibility: SpeakEasy HUD
Tags: artificial intelligence, self-correction, ethics
