Fewer women than men are employed in science, technology, engineering, and mathematics (STEM) careers. At the top science and technology universities in the United States, less than 25% of the faculty are women (MIT: 22.1%, Georgia Tech: 24.5%, Cal Tech: 20.7%, https://www.chronicle.com/interactives/faculty-diversity). A commonly held belief is that this gender discrepancy in STEM fields exists because boys and men have an innate, natural talent for math that women and girls lack. In our paper, “Gender similarities in the brain during mathematics development,” we put this public belief to the test by comparing the neural activity of 104 boys and girls while they watched video clips about early mathematics concepts (e.g., counting and addition). We predicted that if innate, biological differences are a root cause of gender differences in STEM career representation, there should be gender differences in brain function in early childhood.

We designed our study with three important considerations in mind:

1. In adults, we cannot disentangle innate, biological influence from sociocultural ones. Children have less cultural experience than adults so we need to test for gender differences in children. The younger the children, the more we can minimize sociocultural influence, which allows us to get closer to capturing “innate” potential for STEM ability. In our study, children ranged in age from 3 to 10 years with a median age of 5.5 years.
2. Previous work (including our earlier work published last year) has found that young children perform similarly on tasks that tap into foundational math concepts. However, even if performance is equivalent, boys and girls could still be using different strategies and different neural process to complete the tasks. This study extends the previous work comparing math performance by testing whether the biological processes that support mathematical thinking are the same across children or whether there are early gender differences that could contribute to gender gaps in STEM career representation later in life. 
3. Traditional statistical tests such as t-tests, linear regressions, and ANOVAs allow researchers to determine that two or more groups are different, but they do not provide evidence for “sameness” when differences are not identified (instead the findings are considered inconclusive). To quantify evidence for gender similarities, we used two types of statistics: Tests of Equivalence and Bayes Factors.

In our analyses, we compared the “maturity” of children’s brains by correlating children’s temporal patterns of neural activity with adults’ patterns of neural activity while the adults watched the same videos as the children during fMRI scanning. This type of approach is called an “intersubject correlation analysis”. Greater similarity between a child’s neural activity and adults’ neural activity indicates higher “neural maturity”. We tested for both gender similarities and gender differences in neural maturity.

Overall, we found very little evidence of differences and instead found widespread gender similarities! There were equivalent levels of neural maturity throughout the brain, and the variability in neural maturity related to math ability in the same regions for boys and girls. We also quantified “neural similarity” among children and found that children’s neural activity was as similar to children from their same gender as it was to children of different genders. There were no regions where girls were more similar to girls than boys and no regions where boys were more similar to boys than girls. Together, these results show that boys and girls are using their brains in the same way when they are thinking about math and the same brain regions are important for math development in all children. 

Given the broad similarities between boys and girls, gender differences observed in STEM career representation are unlikely to originate from early childhood differences in the brain. Rather, sociocultural influences such as parent or teacher encouragement in math and science or structural systems for career placement and retention might account for the disparity in STEM careers. Ensuring that children and adults receive the same opportunities and encouragement to participate in STEM will be important for minimizing the influence of sociocultural factors on the gender gap in career representation.

Reference

Kersey, A.J., Csumitta, K.D., Cantlon, J.F. "No intrinsic gender differences in children’s earliest numerical abilities” npj Science of Learning 4, Article Number: 19 (2019)

The original article is freely available here.