Angelika Stefan

University of Amsterdam
Department of Psychology


Bayes Factor Design Analysis for the Efficient Collection of Informative Data

The diligent planning of research designs is a prerequisite for credibility and reproducibility of research results. However, frequentist power analysis, the current standard in design analysis, neglects important aspects of design quality and is not applicable to Bayesian research designs that have gained popularity during the last years. The goal of this project is to provide researchers with a comprehensive new framework for design analysis that allows planning the efficient collection of informative data, with the promise to obtain more informative results with fewer data points (on average). The research project will build on the concept of Bayes Factor Design Analysis (BFDA; Schönbrodt & Wagenmakers, 2017), a method to plan for compelling evidence in Bayesian designs. It can be roughly divided into five goals: (1) conceptually extend and streamline BFDA, (2) develop a Continuous BFDA, (3) apply BFDA to meta-analyses, (4) use BFDA to explore the efficiency of collapsing bounds in sequential designs, (5) make BFDA accessible to a broad audience. The resulting methods will provide practical researchers with a comprehensive new method to balance efficiency and informativeness of their experiments and thereby enhance the credibility and efficiency of their research.


Prof. dr. Eric-Jan Wagenmakers, PD dr. Felix Schönbrodt

Financed by

NWO Talent Grant


October 2018 – October 2022