Not only are women underrepresented at scientific meetings, they participate less than men in question-and-answer sessions, self-limiting their involvement and participation. But a public discussion of the problem helps.
By Amy Adams
Recently, some prominent men in science have publicly declared they wouldn’t attend scientific meetings that don’t adequately represent women, but a new study suggests the problem isn’t just representation – women also don’t participate at the same level as men, even when they are well represented.
A Stanford-led study published June 27 in the American Journal of Human Genetics reports that women asked questions at a level that fell well below their level of representation at two national genetics meetings over the course of four years.
But when the two graduate students who led the study drew attention to the problem during one of the meetings – a move that generated intense conversation among attendees – questions from women picked up.
“I think a lot of the time we say the goal is to get representation across diversity,” said Natalie Telis, who was a Stanford graduate student when she carried out the work. She and fellow graduate student Emily Glassberg collected the data and were co-first authors on the paper. “If we want to create participation, representation alone won’t get us there.”
A mathematical approach
Telis started noticing the disparity in question-asking as an undergraduate student in math. “The entire first day of a meeting, I was the only woman to ask a question,” she said. “I thought that was weird.”
Telis started thinking about the problem numerically. If women make up 10 percent of attendees, then one in 10 questions should come from women. But that wasn’t what she found. From then on, Telis made a habit of tracking women’s participation at meetings and talks.
When Telis joined the Stanford lab of Jonathan Pritchard, a professor of genetics and of biology, Glassberg noticed Telis taking notes on question-asking during meetings and grew curious.
Together, they hatched a plan to track who was asking questions at genetics meetings – a plan that eventually led Glassberg to watch hundreds of hours of videos from three years of genetics meetings. They also crowd-sourced 81 volunteers to collect data at the 2017 American Society of Human Genetics (ASHG) meeting, where Telis also gave a plenary talk on the work. Altogether, they got data on who asked questions from the ASHG annual meeting from 2014 to 2017, and the Biology of Genomes annual meeting from 2015 to 2018.
The pair also worked with Chris Gunter, an associate professor of pediatrics and human genetics at Emory University and senior author on the paper, to gather data on gender representation at ASHG overall and in the many subfields. Although the entire meeting was roughly half women, the subgroups vary dramatically.
Very consistently, and across subdisciplines with vastly different representation of women, women asked fewer questions than would be expected. Even in fields like Ethical, Legal and Social Issues in genetics, which had 67 percent representation by women, women asked only 45 percent of questions. They also found that women tended to ask more questions of female speakers, and men asked more questions of male speakers.
Telis said this finding suggests that the issue of participation isn’t just whether women feel welcomed or comfortable in a field. “I think there’s an idea that as soon as you reach proportionate ratios then the issues go away,” Telis said. “But no. We found that even when men were 33 percent of attendees, they asked 55 percent of the questions.”
An accidental experiment
At the 2015 Biology of Genomes meeting, Telis carried out an unplanned experiment. She tweeted some her findings from a session showing that although women made up 35 percent of the audience they asked 11 percent of the questions – a move that generated immediate conversation among attendees.
“It really had a big impact on the course of the meeting,” said Pritchard, who is a co-author on the paper and is also a member of Stanford Bio-X. “For a graduate student to really affect the discussion at a meeting like that was great.”
The conference organizers quickly made a rule that the first question in every session had to come from a trainee – a group that tends to be more diverse. “The effect was immediate,” Telis said.
Women immediately began to ask more questions, and that increase held steady through the 2018 meeting. Whether the increase came from the rule change, which might have made women more comfortable asking questions, or from the intense conversation highlighting the issue, the group doesn’t know.
“We wrestled with what these observations actually mean,” Pritchard said. “Question-asking is controlled by the asker, so measuring the behavior gives us insight into the comfort levels and personalities of the people in the room.”
Telis said she hopes their work shows that it is possible to improve women’s participation in STEM fields and inspires other people to measure the impact of potential interventions. “If we want to have women’s voices be heard, we need to be testing things,” she said.
Telis and Glassberg have received their PhDs and are working in industry.