From Topic to Thesis: Critical Engagement with Generative AI in Thesis Preparation
Year: 2026
Type: conference-paper
Abstract: Numerous engineering students experience anxiety when choosing their thesis topic and initiating the writing process. Uncertainty about viable topics and inexperience in research design often result in procrastination. Furthermore, reliance on general web searches instead of academic sources, compromises their work quality. Copy-paste practices or delegating writing to Generative Artificial Intelligence (GenAI) may appear as attractive shortcuts but lead to inadequate understanding and learning. Responsibility for the thesis text and the responsible use of sources always rests with the author. This study focuses on a stage in engineering education where students should integrate technical skills with the preparation of an academic report. It addresses the challenges students face in topic selection, information-gathering, critical thinking, and academic writing. The research problem was "How to facilitate the early stages of thesis writing through critical engagement with GenAI?" In a research skills course taken before the thesis, two interventions were introduced: The first encouraged students to use GenAI to assist in ideating and refining their thesis topic. The second involved examining an article on biases in information gathering. Inductive content analysis was applied to reflections on these interventions, and pre-and post-questionnaires provided quantitative data on changes in critical thinking practices. In this mixed-methods action research, students found using GenAI as a thesis tutor rewarding and identified several areas for improvement in their information-gathering approaches. Statistical analyses showed positive changes in students’ critical thinking practices.
Authors: Outi T. Virkki, Usman Naeem, Arne Styve
Language: English
Subfield: Education
Field: Social Sciences
Domain: Social Sciences
SDG: Quality education
Open Access status: hybrid
OpenAlex ID: W7166543886
