When asked to draw a scientist, children often reproduce common stereotypes about who scientists are and what they do. However, new research, which I led, shows that these stereotypes have changed over time, at least within the United States. My study, which was published March 20 in Child Development, finds that U.S. children now draw female scientists more often than ever before.

In the 1960s and 1970s, one landmark study asked nearly 5,000 elementary school children to draw a scientist. Their artwork almost exclusively depicted men, often with lab coats, working indoors with lab equipment. Of those nearly 5,000 drawings, only 28 depicted a female scientist, which were all drawn by girls. Not a single boy drew a woman.

Those findings were striking to me. But they also made me wonder: how have children’s stereotypes changed over time? Women have made substantial gains in educational attainment and employment since the 1960s, especially in science fields. For instance, women earned 19 percent of U.S. chemistry bachelor’s degrees in 1966, compared to 48 percent in 2015.

Female scientists are also now more often depicted in children’s media. One content analysis found evidence for this change in the popular magazine Highlights for Children. Women and girls were 13 percent of images of people in the magazine’s science feature stories in the 1960s, compared to 44 percent in the 2000s. Other research has found that women and girls were 42 percent of scientist characters in popular children’s television programs in 2006.

Given these cultural changes, I wanted to know if children’s images of scientists have also become more gender diverse over time. I therefore started to quantitatively review Draw-A-Scientist studies for a graduate course that social psychologist Alice Eagly taught (who is a study co-author). At that time, I was blissfully unaware of how large and complex that research project would be. I was in my first year of Northwestern’s psychology Ph.D. program, and now five years later, this study has just come out.

Science teachers and education researchers have often used the Draw-A-Scientist Test because it’s easy to administer. I therefore had a wealth of data to review. My colleague Kyle Nolla, Ph.D. student in psychology at Northwestern, and I developed procedures to systematically search for studies and extract their data. This process involved filtering over 3,000 references from databases such as Google Scholar. In the end, we found 78 U.S. Draw-A-Scientist studies, including over 20,000 children, that met our inclusion criteria.

Our results suggested that U.S. children’s stereotypes linking science with men have weakened over time. Less than 1 percent of children drew a female scientist in the original Draw-A-Scientist study, conducted between 1966 and 1977. In contrast, 28 percent of children drew a female scientist on average in subsequent studies, conducted between 1985 and 2016.

Statistical analyses showed robust evidence for change across five decades of data collection. In addition, both girls and boys drew female scientists more often over time, though girls drew female scientists much more often than boys overall.

Even within studies conducted in the 1980s and onward, children tended to depict female scientists more often over time. Based on a statistical model, 22 percent of children drew a female scientist in 1985 on average, compared to 34 percent in 2016. However, the change within these later decades was tentative. The shift from 1985 to 2016 did not meet conventional levels of statistical significance. In other words, U.S. children have clearly depicted female scientists more often since the 1960s and 1970s, but the evidence for change within later decades was less clear.

Our results and other studies counter some researchers’ claims that the sex difference in science and math engagement “has remained stable for decades.” Nevertheless, despite change over time, women remain a minority in several science fields. Children may learn about such imbalances through multiple sources such as television shows, textbooks, and in-person interactions with scientists. Consistent with this hypothesis, U.S. children still draw more male than female scientists by a ratio of about two to one in recent years.

Our study shed light on how children acquire these stereotypes. My co-author, developmental psychologist David Uttal, helped us think about changes across children’s age. Some gender stereotypes emerge as early as age 2 and 3. However, we reasoned that the stereotype of scientists as male should emerge somewhat later because preschoolers’ knowledge of scientists is likely limited.

Consistent with this reasoning, children drew roughly equal proportions of male and female scientists when they started kindergarten around ages 5 and 6. However, the tendency to draw male scientists increased strongly with age during elementary school and middle school. In addition, older children drew scientists with lab coats and eyeglasses more often than younger children. Hence, children likely learn multiple stereotypes about scientists as they mature, not just those about gender.

Teachers and parents should therefore be aware that elementary school and middle school is a critical period when students start forming stereotypes about scientists. Children should be exposed to diverse examples of scientists that go beyond the typical dead, White, male scientists usually presented in classrooms. Encouragingly, our results provide evidence that children’s stereotypes respond to changes in their environments.