Paris (AFP)

Scientists may profess rigor, when it comes to assessing a woman's ability to become a research director, they do not escape gender stereotypes without necessarily being aware, reveals Monday a French CNRS study .

Women remain underrepresented in scientific research. At the CNRS, the average percentage of women among the research fellows was 38.1% at the end of 2017 and 29.2% among the research directors.

The concept of science remains "much more strongly associated with the masculine than with the feminine at the majority of the scientists" as in the general population, "because of automatisms stored in our memory since the childhood", explains to the AFP Isabelle Regner, researcher at Aix-Marseille University and co-author of the study.

The team made up of researchers in social and cognitive psychology managed to bring "behavioral proof" while "many people think that this question of gender stereotypes is now settled," says Pascal Huguet, the other author of the study published in Nature Behavior.

The researchers studied for two years 40 juries (about twenty people each) to evaluate applications for CNRS Director positions in various disciplines.

"This is the first time that a research institution has agreed to carry out a study of this type on its practices, in a real-life situation and all over the scientific spectrum." The CNRS took a risk ", considers Pascal Huguet, Research Director CNRS at Clermont Auvergne University.

In the first year, juries, who had been forewarned from this study, passed "implicit association tests", carried out on a computer during their break time. They were asked questions about the possible causes of under-representation of women in scientific disciplines. Was it a question of competence? Motivational? Family constraints? Discrimination? At the end of the year, the CNRS communicated to the team the number of women and men chosen by these juries during the competitions.

The second year, the same juries worked as usual, without having in mind that the team of psychologists would study, in real life, their possible gender bias.

"The second year, the bias is there," notes Isabelle Régner.

"We have shown that juries that explicitly downplay the existence of gender discrimination" (about one in two juries) "are more sensitive to implicit and automatic stereotypes and tend to recruit fewer women", says she. And in this case, the more the implicit stereotypes are strong, the less women are promoted.

"On the other hand, juries who explicitly consider that we must continue to pay attention to gender discrimination, manage not to be influenced in their choice by implicit stereotypes even if they are important," she continues.

In his eyes, the results of the study show "what to do", since juries that recognize a risk of discrimination "manage to control the impact of implicit biases on their decision". "We have the solution," says the researcher, emphasizing the importance of training.

© 2019 AFP