The Global Survey of Scientists (Part 1)
Who answered the survey?
The survey was answered by 32,000 scientists, from 159 countries, of which 50% were male and 50% female.
What was the technique used to collect the answers?
We set a snowball technique, a technique where existing study subjects recruit future subjects from among their contacts. Thus the sample group is said to grow like a rolling snowball.
We used contact databases from some partnering organizations to reach students and professional scientists across the globe. Since there is not a single network or resource available to contact all students and professional scientists globally, snowball sampling was taking advantage of as many personal networks as possible.
Snowball sampling does not result in a statistically representative sample. Because of this, there are important limitations in analysis and interpretation for the data collected by the survey. Therefore, our findings only indicate trends among the individuals who responded to the survey, not the overall population.
Summarizing, our findings should not be assumed to be representative of the intended population as a whole. However, the consistency of most of our findings across disciplines, geographical zones and development level is reassuring.
In our Global Survey, some of our partners (particularly in Mathematics or Physics) had a more active network with respect to women in science or gender equality than other partners, so that the proportions of answers do not reflect the respective weight of the disciplines participating to the project in terms of number of scientists. Even so, our analysis techniques allow us to make statements about the relative experiences of men and women in multiple disciplines, working in different sectors, and studying and pursuing careers worldwide.
What are the statistical tools that you used in the analysis of the data collected with the Global Survey of Scientists?
We focus on multivariate analyses. The multivariate analyses make it possible to draw conclusions in spite of potential confounding factors, such as employment sector, discipline, geographic region, age, and more. We are still able to conclude that there are statistically significant differences in the responses of men and women after accounting for potential confounding factors. The difference between men and women is very significant in this multivariate approach.
What are the key findings of the survey?
They confirm that the Gender Gap in Science is very real, across all regions, all disciplines, and development levels. Women’s experiences in both educational and employment settings are consistently less positive than men’s.
Marie-Françoise Roy et Colette Guillopé
