Produtividade científica e desigualdade de gênero: evidências nas áreas de matemática e estatística

Abstract

Since the 1970s, research productivity grants awarded by the National Council for Scientific and Technological Development (CNPq) have been considered one of the main programs for promoting science in Brazil. Until 2024, these grants were divided into three categories (Senior, Category 1 and Category 2), with Category 1 further subdivided into four distinct levels (A, B, C and D). In this context, the present study analyzed the scientific output of researchers in the field of Mathematics and Statistics, aiming to assess whether it is possible to establish objective criteria to classify them as grant recipients — including the specific category/level — or as non-recipients. To achieve this, logistic regression was applied to data extracted from the Lattes Platform, considering variables such as the number of student supervisions, published articles, participation in academic events, years since obtaining a Doctor of Philosophy degree (Ph.D.), among others. Moreover, taking into account the systemic barriers faced by women in scientific careers, the study also sought to investigate gender differences in the distribution of researchers, academic productivity and classification produced by the model. In this process, nonparametric tests were applied to characterize the profile of researchers according to productivity, grant level (for those receiving research grants), and gender, as well as to preselect variables for the statistical model. The analysis revealed that men publish more prestigious articles, while women stand out in mentoring undergraduate students and writing book chapters — activities that are generally less recognized. Furthermore, the findings show that women’s academic contributions are evaluated more rigorously, requiring higher performance to achieve the same level of recognition as their male peers. The study reinforces the need for more equitable policies in Brazilian science.