Reproducibility: the Wenneras and Wold Nature paper (1997)

Description

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

Steps to reproduce

Data is divided in three sheets; first is the orignal data received from Agnes Wold; second sheet is the corrections done using the assessmant sheets at the National Archive; third sheet is the suggested normalization of data.

Institutions

Categories

Funding

Licence