Harnessing the power of gender analysis in science and engineering

Should the voices of virtual “personal assistants” such as Alexa and Siri be created to sound like women? What are the dangers of creating automotive safety standards using only male-bodied crash test dummies? Why do rodents respond differently to pain stimuli depending on the sex of the human experimenter in the lab?

The Gendered Innovations project at Stanford University is answering these questions through the creation of an open access repository of research tools for scientists and engineers. Together with their multinational team of more than 200 gender and STEM experts from around the world, Gendered Innovations develops state-of-the-art methods for gender, sex, and intersectional analysis for use across the sciences, engineering, health and medicine, and environment and provides case studies that demonstrate the innovation that arises from gender analysis.

According to Londa Schiebinger, John L. Hinds Professor of History of Science in the Department of History and a former director of the Clayman Institute (2004-2010), the benefits of providing methodological tools and training in gender, sex, and intersectionality are far-reaching. Schiebinger shared her thoughts about the current and future trajectory of this work at a recent Faculty Research Fellows program.

Schiebinger founded Gendered Innovations in 2009 to lead efforts in generating unbiased research. Through an open repository of resources, the project aims to eliminate bias in the production of knowledge and goes beyond knowledge that is explicitly about gender. Schiebinger and her colleagues stress that including gender analysis in research is essential for reducing gender bias in nearly all areas, from nanotechnology to urban design. According to the team’s 2019 article in Nature, sex and gender analysis is a catalyst for “rigorous, reproducible, and responsible science.”  

Schiebinger spoke about how the global community of research funders has taken notice of Gendered Innovations’ work, turning many of the project’s recommendations into policy guidelines. “Granting agencies can ask or require applicants to integrate sex, gender, or intersectional analysis into the design of research, or justify why it is not relevant,” said Schiebinger. The European Commission has been guided by Gendered Innovations in its requirement, since 2011, to incorporate sex and gender analysis into grant proposals across all 27 European member states.

This may be insufficient, however, to bring about the kind of change that Schiebinger envisions. Grant proposals should be the beginning of the process, not the end, a Gendered Innovations article published in Science on September 30th argues, asking funders and scientists to ensure that gender analysis is incorporated throughout all aspects of the research process, from proposal to publication. Their new article lays out a five-part framework intended to evaluate sex, gender, and diversity policy implementation for 22 funding agencies across six continents.

By designing robots to simply mirror features of society—such as the higher number of women in nursing careers compared to men—researchers and engineers run the risk of exacerbating existing social hierarchies.

Beyond their guideline development work, the impact of Gendered Innovations is most clearly demonstrated through case studies. Within the field of robotics engineering, for example, the group’s work shows how the physical and social characteristics of robots determine which people interact with them and in what ways. When receiving care from a robotic nurse, for example, patients are highly attuned to the color of plastic the machine is covered in (typically white) and its automated voice (often feminized). They may be more likely to comply with medical recommendations from a robot that corresponds with their expectation of “typical” care providers. “But there’s a danger here,” said Schiebinger, “as soon as we gender a machine, stereotypes follow.”

By designing robots to simply mirror features of society—such as the higher number of women in nursing careers compared to men—researchers and engineers run the risk of exacerbating existing social hierarchies. Through the application of Gendered Innovations guidelines, Schiebinger stressed the possibility for robots to act as tools for reconfiguring, rather than further reinforcing, society’s understanding of gender.   

Best-practice guidelines are usually implemented at the level of grant-funding agencies. But Schiebinger also spoke about extending their best-practice guidelines to peer-reviewed journals (she has contributed to the The Lancet’s progress in this regard). The team would also like to see their humanistic gender and sex research training adopted into university curricula, first at Stanford and then beyond. “If we provide the training at the undergraduate level, this could change how we do science,” Schiebinger said.

Gendered Innovations is already driving change in the way that science is done. The potential goes far beyond eliminating bias, according to Schiebinger: “Doing research right from the very beginning saves both money and lives,” she said, “and we want to harness the creative power of this type of analysis for scientific discovery.”