Reliability and validity of a high-profile peer review study: Probing Wennerås and Wold’s data in <i>Nature</i>

Abstract This article revisits the widely cited study by Wennerås and Wold (1997), which reported evidence of gender bias and nepotism in the peer review process for Swedish medical research fellowships. Their study has had substantial scholarly and policy influence, often cited as a foundational demonstration of systemic bias in science evaluation to the disadvantage of women. By accessing and reanalyzing the original data set and reconstructing the analytical model used by Wennerås and Wold, we reproduce their findings while identifying key methodological inconsistencies and overlooked complexities. Applications to program types were excluded from their analysis without justification. With the use of normalization techniques, we reassess gender disparities in reviewer scores and demonstrate that much of the reported gender bias can be attributed to disciplinary and programmatic variation rather than applicant sex per se. Moreover, we introduce a network-based analysis of applicant success rates. We conclude that while Wennerås and Wold’s original study remains a milestone in exposing bias in science, aspects of its empirical foundation merit reconsideration. Our findings contribute to ongoing debates about the reproducibility of influential social science studies and the persistent challenges in measuring fairness in research evaluation. The broader political context within Sweden is also outlined.