Gender monitoring (part 3): What influence do confounding factors have?
Differences in success rates between women and men in SNSF funding can be explained by a variety of different direct and indirect influences. A look at confounding factors that can distort the results of analyses.
As an equal-opportunity institution, the SNSF strives to support researchers of all backgrounds. To this end, a series of regular analyses have been established to investigate the role of gender in the grading and funding of grant proposals in the SNSF’s Project funding scheme. Among other methods, these analyses make use of regression models to analyse success rates and grades and their association with gender. It is essential that appropriate confounding variables are taken into account as part of the model, in order to obtain a reliable interpretation. Potential conclusions can vary dramatically when the confounders are not properly modelled.