Estimation of scientific productivity with a hierarchical Bayesian model

A new measure for the scientific productivity of researchers is introduced. Through a Bayesian hierarchical model, the productivity of Italian researchers is reconstructed and used to understand whether within Italian academics gender difference can be assessed. The new measure takes into account not only the quantity and quality of the scientific pro- duction but also the time that elapses between two successive research products, obtaining a productivity curve for each male and female scientist. The results show a persistence of women in the lower levels of the university career even if their scientific productivity is greater than that of their male colleagues.