Three of our members - Vlasta Sikimić, Cristian Mesquida, and Daniel Lakens - have recently been awarded funding for two new research projects.
Vlasta Sikimić: Intellectual Virtues and the Future of Research with Generative AI
Intellectual Virtues and the Future of Research with Generative AI is a NWO XS project by Vlasta Sikimić that provides a normative framework for understanding and informing how scientists reason and create in the age of AI. As generative AI rapidly becomes embedded in everyday research, Sikimić confronts a pressing question: how can we maximize its potential without sacrificing critical thinking and intellectual autonomy? Through participatory research and a dynamic co-creation workshop, Sikimić will work on theory-building around the virtues of scientists and defining what virtuous reasoning with AI truly looks like in practice. The result? A cutting-edge theoretical paper and an interactive workshop designed to help researchers across disciplines engage with generative AI critically, creatively, and responsibly, shaping the intellectual future of science.
In the 4TU.CEE project Virtuous reasoning with AI, Vlasta Sikimić (PI), Karolina Doulougeri, Krist Vaesen, and Yara Daamen are launching an educational intervention designed to shape the next generation of virtuously minded engineers. At a time when AI systems influence decisions at every level of society, education faces a defining challenge: teaching students not just to use AI, but to question it. In the new course, Critical thinking in the age of AI, students will engage with intellectual justice to confront algorithmic bias and intellectual honesty to tackle transparency and accountability in AI systems. Through carefully designed learning activities, the project cultivates higher-order thinking about responsible AI use, empowering students to critically examine, ethically design, and challenge intelligent technologies. This initiative directly supports TU/e’s values by educating engineers who build new systems wisely.
Cristian Mesquida and Daniel Lakens: Reproducibility of Power Analyses in Large-Scale Replication Projects
We are pleased to announce that the proposal Reproducibility of Power Analyses in Large-Scale Replication Projects has been awarded funding under the NWO Replication Studies (2025) program. The project, led by Dr. Daniël Lakens (Eindhoven University of Technology) with co‑applicant Dr. Cristian Mesquida, will undertake a comprehensive reproduction of the statistical power analyses that underpin five major, high‑profile replication initiatives across psychology, economics, experimental philosophy, and the social and behavioural sciences.
This award supports a 24‑month 0.5 fte research effort that examines the extent to which power analyses are reproducible, and methodologically sound. Large‑scale replication projects have played a central role in diagnosing challenges to scientific credibility. Yet, the validity of their conclusions depends critically on the quality of the underlying power analyses. The funded project will reproduce 214 power analyses, independently reassessing effect‑size specifications, statistical test alignment, alpha‑level choices, software settings, and achieved sample sizes. By doing so, the team will generate updated estimates of the actual statistical power achieved in these landmark studies and will quantify the frequency and nature of mistakes in power analyses.
The project will produce a tutorial on avoiding common power‑analysis pitfalls. In addition, the team will reflect on quality‑management in large-team science projects. This award will contribute to more reliable conclusions from large‑scale replication projects, prevent common mistakes in a priori power analyses, and improve rigour in large collaborative projects.