Skip to ContentSkip to Navigation
About us Faculty of Behavioural and Social Sciences Psychology Research Units Developmental Psychology

Back to Bayesics: Solving the Reproducibility Crisis in Biomedicine

Developmental Psychology, Psychometrics and Statistics

Large parts of the biomedical literature are difficult to replicate. A major culprit in this reproducibility crisis is the over-reliance on inappropriate statistical evidence criteria in both basic biomedical research and clinical trials (van Ravenzwaaij & Ioannidis, 2017, 2019). This project develops tools for better evaluating statistical evidence in medical research to combat the reproducibility crisis. First, state-of-the-art Bayesian statistical techniques are developed and made available to clinicians in a user-friendly form. Second, these cutting-edge techniques are applied to data from biomedical research and clinical trials in order to diagnose false positives. By identifying those studies for which the reported statistical evidence is inconclusive, a targeted replication program can take place. Third, tailor-made decision criteria are developed for each of five large drug families by combining data from medical experts in a wisdom-of-the crowd analysis. Fourth, through close ties with national and international drug regulators, this project aims to lead to real change in the way statistical evidence is evaluated in the decision to endorse new drugs and treatments.

Researchers and partners

Behavioural and Social Sciences, Psychology
University of Groningen, outside of Behavioural and Socials Sciences
Partners outside of the University of Groningen
  • John Ioannidis, Departments of Medicine, of Health Research and Policy, and of Statistics and Meta-Research Innovation Center at Stanford (METRICS), Stanford University

Education

Selection of BA and MA student projects
  • Development of Bayesian techniques for survival analysis designs
  • A comparison of efficacy of psychoactive and physiological drugs
  • A comparison of strength of evidence of FDA trial data and corresponding published papers
  • An overview of policy for endorsing new drugs and biologics in different countries

Results

Publications
  • van Ravenzwaaij, D. & Ioannidis, J. P. A. (2017). A Simulation Study of the Strength of Evidence in the Recommendation of Medications Based on Two Trials with Statistically Significant Results. PLoS One, 12, e0173184.
  • Monden, R., Roest, A. M., van Ravenzwaaij, D., Wagenmakers, E.--J., Morey, R., Wardenaar, K. J. & de Jonge, P. (2018). The Comparative Evidence Basis for the Efficacy of Second-Generation Antidepressants in the Treatment of Depression in the US: A Bayesian Meta-Analysis of Food and Drug Administration Reviews. Journal of Affective Disorders, 235, 393–398.
  • van Ravenzwaaij, D., Monden, R., Tendeiro, J., & Ioannidis, J. P. A. (2019). Bayes Factors for Superiority, Non-Inferiority, and Equivalence Designs. BMC: Medical Research Methodology, 19, 71.
  • van Ravenzwaaij, D., & Ioannidis, J. P. A. (in press). True and False Positive Rates for Different Criteria of Evaluating Statistical Evidence from Clinical Trials. BMC: Medical Research Methodology.
  • Linde, M., & van Ravenzwaaij, D. (2019). baymedr: An R Package for the Calculation of Bayes Factors for Equivalence, Non-Inferiority, and Superiority Designs. Manuscript submitted for publication.
Successful grant proposals and other achievements
  • NWO Vidi grant to Don van Ravenzwaaij (016.Vidi.188.001)

Main research theme

Human Development, Health, and Well-being across the Life Span

More information

Last modified:29 March 2021 10.18 a.m.
printView this page in: Nederlands