Exploring the Influence of AI-based Feedback on Students' Psychological Ownership Towards Their Written Texts: The Moderating Role of Feedback Modality
How can AI-powered feedback help students improve without taking away their sense of ownership over their work, motivation, and responsibility?
Duration: January 2026 – December 2026
Status: Ongoing
Educational Level: Tertiary Level
Topic: Artificial Intelligence AI, Digital Skills & Literacy
Keywords: Psychological Ownership, Feedback, Writing Tasks, AI
Initial Situation
Recent advancements in Generative Artificial Intelligence (GenAI) have led to the development of diverse solutions to automate the feedback process in educational settings. Built on Large Language Models (LLMs), such tools can support students with their writings by giving critical, corrective, and adaptive feedback based on theoretical frameworks. Further, the availability of such tools allows students to receive one-on-one personalized feedback asynchronously when they need it without the involvement of their lecturer, ensuring timeliness – a critical quality of feedback. However, while the efficiency and scalability of LLM-based feedback tools are clear, their impact on students’ psychological ownership of their work remains underexplored. Psychological ownership refers to the feeling that something is “mine”, characterized by a sense of personal investment and responsibility towards an object or outcome. In educational contexts, fostering psychological ownership over one’s work has been associated with increased motivation, deeper engagement, and improved learning outcomes. When students revise their texts based on external feedback, particularly when that feedback is automated, questions arise about the degree to which they still feel connected to, and take ownership of, their revised work. Different modalities of feedback may vary in their influence on students’ sense of ownership over their revised texts.
Objectives
The goal of this project is to explore psychological ownership in the interaction with technology-driven feedback tools for educational contexts. We aim to design, evaluate, and refine an LLM-driven feedback mechanism for the Rflect tool to provide guidance to researchers and developers on how to develop feedback tools that empower students’ agency and sense of ownership. The objectives are 1) evaluation of three feedback modalities in a classroom setting and 2) design principles for integration of LLM-driven feedback in an educational context that safeguards psychological ownership.
Method
The project applies an HCI and Information Systems methodology combining learner-centred design, prompt engineering, and a longitudinal field experiment. In Phase 1, student interviews inform the design of feedback principles grounded in self-regulated learning and psychological ownership. Three AI feedback strategies are conceptualised and implemented through tailored prompt engineering within an existing reflective writing tool. In Phase 2, a six-week classroom field experiment tests these strategies with students from different education backgrounds. Writing quality, feedback use, self-efficacy, and psychological ownership are measured over time using quantitative analysis.
Planned Translation
A translation of the project results into educational practice is anticipated through direct implementation within the institutions of the project leaders. As the experiment will be embedded in existing writing courses, findings can be quickly evaluated in authentic classroom settings and iteratively refined with Rflect. This setup allows for immediate feedback from both students and educators, facilitating the development of practical, context-sensitive recommendations. Furthermore, the collaboration with our partner Rflect enables straightforward adaptation and scaling, laying the groundwork for broader application across Swiss educational institutions after initial validation.
This project advances research on GenAI in education by providing novel evidence on how different AI feedback strategies affect not only writing quality but also students’ psychological ownership, motivation, and deep learning. It delivers design principles for AI feedback that balance performance gains with student agency. The findings from this project are aimed to be presented in scientific papers in IS journals, but also educational technology and HCI. For Swiss higher education, the project enables the scalable integration of adaptive, learner-centred feedback into reflective writing through an existing edtech tool, directly benefiting thousands of students.
