Reproducibility: the Wenneras and Wold Nature paper (1997)

Aim: test of the reproducibility of a high-profile study published by Nature (1997). With access to the original applications and the assessment sheets from the Swedish Medical Research Council, stored at the National Archives, we are able to undertake a thorough examination of the data used by W&W. Our analysis uncovers several errors in the dataset, a disregard for data heterogeneity, unexplained, questionable methodological steps and lack of rigor in categorizations. Additionally, after a normalization procedure, we find that the relationship between gender and assessment is disadvantageous for men, not for women as claimed by W&W