Meeting women who have achieved success in science or learning about their stories may not be sufficient to challenge stereotypes about women in STEM. The researchers therefore investigated whether a more effective approach might be to embody the Polish Nobel laureate Maria Skłodowska-Curie in virtual reality. They found that this experience weakened automatic associations linking science with men.
Although the number of women in science is steadily increasing, women still make up only one-third of the global scientific community and are more likely than men to leave scientific careers. The probability that a woman will continue publishing 10 years after beginning a research career is approximately 49 percent, compared with 54 percent for men, according to research from Adam Mickiewicz University in Poznań.
Despite decades of efforts to promote equality, women remain underrepresented in STEM fields, encompassing science, technology, engineering, and mathematics. In artificial intelligence research, according to UNESCO data, women account for only 20 percent of specialists.
One reason why women may be underrepresented in STEM fields is the tendency to consciously and unconsciously associate science with men. Previous research has suggested that exposure to successful women in science can weaken such stereotypes, but these effects are often small and short-lived. Simply seeing a photograph of a female scientist, hearing her story, or meeting an inspiring female researcher may not be enough to produce lasting changes in deeply rooted associations.
A team of researchers from New York University Abu Dhabi—Marta Beneda, Julia Spielmann, Amna Ramadhan, Domna Banakou, and Andrea Vial—therefore decided to examine whether something more could be done: rather than simply observing a female scientist, could participants temporarily embody Maria Skłodowska-Curie? The researchers published the results of the experiment in Computers in Human Behavior (DOI: 10.1016/j.chb.2026.109146).
“We employed a phenomenon known as virtual embodiment. This refers to a psychological process in which a person immersed in virtual reality begins to experience the virtual body as their own and feels that they are controlling its movements. In a suitably designed virtual environment, the brain can integrate information from different senses so that the embodied virtual avatar temporarily becomes part of the self,” explained Dr. Marta Beneda.
This experiment extends a line of research led by Dr. Domna Banakou demonstrating that the characteristics of a virtual body experienced as one’s own can influence perception, implicit social attitudes, cognitive processes, and behavior. This phenomenon is related, among other things, to the so-called Proteus effect.
Dr. Beneda explained that, in psychology, the Proteus effect refers to the tendency to adopt characteristics associated with the virtual avatar one embodies. The virtual body is therefore not merely a digital “costume”; it can become a source of information about oneself and influence one’s behavior and way of thinking. In the context of the present study, embodying a female scientist may lead participants to incorporate aspects of this identity into their own self-concept.
The study employed three types of avatars. Some participants embodied Maria Skłodowska-Curie, others embodied Albert Einstein, while the remaining participants embodied an avatar representing a person of a similar age and the same gender as themselves who was not associated with science. The researchers hypothesized that embodying Maria Skłodowska-Curie, a famous female scientist, would weaken automatic associations linking science with men compared with embodying Albert Einstein, who represents the stereotypical association between the categories “men” and “science.” The control avatar, meanwhile, allowed the researchers to test whether any changes resulted specifically from embodying a female scientist or reflected the broader experience of virtual embodiment.
All participants “entered” a virtual laboratory modeled on an early twentieth-century scientific workspace. Participants could see their virtual bodies in a mirror, and their movements were represented by the avatar in real time. In the virtual laboratory, they completed science-related tasks. In one task, they organized sets of numbers and equations by touching the appropriate elements with their virtual hands. In another, they generated different uses for a particular object, engaging in divergent thinking.
A total of 253 participants completed the experiment. At the beginning of the study, immediately after the virtual experience, and one week later, participants completed a measure of implicit stereotypes assessing automatic associations between gender and science.
As predicted, participants who embodied Maria Skłodowska-Curie showed an overall reduction in the strength of automatic associations linking men with science. This effect was particularly pronounced among women, whose associations linking science with men weakened both immediately after the experience and one week later. In contrast, embodying Albert Einstein did not produce any significant changes in these associations.
Interestingly, an unexpected effect was also observed among men who embodied the control avatar. In this group, automatic associations linking science with men also weakened, and the effect persisted one week later. The researchers suggest that the immersive and novel nature of the experience itself may have contributed to changes in participants’ existing associative patterns.
Dr. Julia Spielmann commented on the findings: “Viewing photographs or watching informational programs about outstanding female scientists places people in the position of observers. The virtual embodiment we used changes this relationship. Such an experience may increase psychological closeness to the role model and strengthen the association between ‘women’ and ‘science.”
It is not possible to conclude whether the observed effects would persist for a longer period of time, such as several weeks or months. Nor can they be equated with changes in behavior, such as greater willingness to hire women in science or support their careers.
“Stereotypes are only one of the factors affecting women’s scientific careers. Institutional structures, how achievements are evaluated, access to positions and funding, organizational norms, and the possibility of balancing work with private life also matter. Changing automatic associations may be one element of a broader strategy, but it will not by itself remove these barriers,” concluded Dr. Andrea Vial. (PAP)
