In August 2023, I attended the European Ornithologists’ Union (EOU) conference, facilitated by the BOU member attendance grant. While I was at the conference, I couldn’t help but reflect on the participation of under-represented groups in scientific conferences, including the active participation of women (or, more broadly speaking, female-presenting people) when asking questions after the talks. In fact, after several discussions with other female colleagues, I became increasingly aware that women ask fewer questions than men in conferences. This phenomenon has been described in a few scientific papers and news articles, across multiple scientific areas (e.g. Telis et al. 2019) and countries (Carter et al. 2018). Worryingly, this trend does not seem to be explained by demographic inertia, as younger male scientists also ask more questions than younger female scientists (Hinsley et al. 2017; Carter et al. 2018). This imbalance in visibility helps increase internalised stereotypes in people from under-represented groups and may impair how younger scientists perceive their possibility of success in academia (Hinsley et al. 2017).
During the EOU conference, together with João Gameiro, Inês Catry, Fábio Marcolin and Teresa Catry, we collected data on the number of questions asked by female- and male-presenting attendees, as well as the presenting gender of the person asking the first question after the talk. We also recorded the presenting gender of the speaker and of the moderator of the session, the approximate number of audience members in the session, and finally the career stage of the presenter (“early-career” if the speaker was between MSc and post-doc stages, and “senior” if the speaker was a senior scientist, lecturer, or professor). Whilst it was not my intention to turn this into an accurate scientific analysis, I couldn’t help myself and ended up running simple univariate generalised linear models to try to explain what influences the number and the proportion of questions asked by female-presenting attendees. The gender variables were recorded in a binary way (female or male-presenting), which I acknowledge increases the visibility problem for non-binary people and for people whose real gender does not correspond to their presenting gender. In future conferences, this error could be minimised by including the pronouns of the speakers in the conference program, which could also potentially facilitate the interactions during the conference.
Figure 1: Who attended EOU 2023 and who asked the questions? A) Gender of the attendees of EOU 2023. This data was collected through a questionnaire during the opening session of the conference. B) Presenting gender of all speakers of the EOU 2023, gathered from the conference abstract book. C) Number of questions asked by female and male-presenting attendees during the 76 conference talks our team attended (70 talks with questions asked). D) Number of first questions asked by female and male-presenting attendees during the 76 conference talks our team attended (70 talks with questions asked). In all analyses that required identifying the gender of the person visually, it was not possible to account for non-binary people or for people whose gender does not match their presentation. These figures illustrate a widely acknowledged problem: despite the number of attendees being similar between genders, male-presenting attendees ask more questions in scientific conferences than female-presenting attendees.
Overall, both the total number of attendees and the total number of speakers were very close to “female-male” gender equality, but the number of people that described themselves as “non-binary / transgender / other” was still very small (Figure 1). In total, 23 sessions were chaired by female-presenting scientists, 18 by male-presenting, and 10 sessions were chaired by a combination of both genders. There was, however, a bias in the plenary speakers: 5 out of 7 speakers were female-presenting. In total, we recorded the questions of 76 talks (of which 70 had questions asked), and we found a clear bias towards male-presenting attendees: 74% of the questions were asked by male-presenting people, and at about two-thirds of the talks, male-presenting attendees initiated the discussion by asking the first question. At 89% of the talks, male-presenting attendees asked at least one question, while at only 53% of talks female-presenting attendees asked at least one question.
Our statistical analysis shows that female-presenting attendees asked more questions when the speakers were female-presenting as well (Figure 2), highlighting the importance of gender representation when considering the outline of speakers in a conference. Also, female-presenting attendees asked proportionally more questions when there were smaller audiences and in talks presented by early-career researchers. Importantly, both the number and the proportion of questions asked by female-presenting attendees sharply increased when the first question was asked by another female-presenting audience member. This was true independently of the number of questions asked in the session, although sessions with larger number of questions had more questions by female-presenting attendees. However, on all these analyses we did not account for the gender-structure of the attendees in the room, which should be corrected in future studies.
Figure 2: What influences the number and proportion of questions by female-presenting attendees? A-B) Model predictions and 95% confidence intervals of the relationship between the number of questions by female-presenting attendees (per session) and A) the presenting gender of the person asking the first question, and B) the presenting gender of the speaker. C-D) Model predictions and 95% confidence intervals of the relationship between the proportion of questions asked by female-presenting attendees (per session) and C) the career stage of the speaker, and D) the number of people in the audience. These relationships show that female-presenting people are more likely to ask more questions when a female-presenting person asks the first question, when the speaker is female-presenting speaker and an early-career researcher (MSc to post-doc) and when there are fewer people in the room.
So, what can we do to change these statistics? As conference organisers and attendees, it is our responsibility to reflect on how we can create conferences that are equally safe for all people, regardless of their gender, skin colour, nationality, sexual orientation or physical ability, amongst other sources of discrimination. The first, and easiest, step is for all of us to read the literature on the topic, raising awareness and trying the solutions recommended by the experts. Secondly, we need to ensure that our conferences are as egalitarian as possible, and that people from all under-represented groups and career stages are given equal opportunities to present their work. Thirdly, when selecting the person who asks the questions, moderators of sessions should consider these biases and take steps to ensure that under-represented groups are given equal opportunities to ask questions. In our analysis (as well as in Carter et al. 2018) there was a clear pattern: if female-presenting people asked a question first, other female-presenting attendees were more likely to ask questions afterwards. Therefore, we can try to minimise the biases by giving people from under-represented groups the chance to initiate the discussions.
We also need to create spaces where people from under-represented groups feel safe to ask questions. In conferences, women described feeling intimidated or anxious that their questions were not good enough (Carter et al. 2018; Jarvis et al. 2022), so perhaps making questions anonymous (e.g., through a conference app) could potentially lead to more people from under-represented groups asking questions, increasing their confidence. While our results and other papers show that women ask more questions if there are more questions being asked, increasing the amount of question time at a conference may not necessarily increase the number of questions asked by women (Carter et al. 2018). On a larger scale, building a more egalitarian and kinder society is bound to empower people from under-represented groups, which is likely to increase representation in all fields of society, including in the number of people that get their voices heard during conferences. As for myself, in future conferences I will take advantage of talks with smaller audiences, presented by early-career female-presenting scientists, to gain the courage to raise my hand and ask a question.
The European Ornithologists Union provided this supportive comment on Marta’s research
“It is tempting to feel satisfied when the proportion between men and women is about equal among the attendees and speakers on a conference. The survey conducted by Marta Acácio demonstrates that we have more work to do before having true equality between the genders. One of the take-home messages from Marta’s results is that the willingness to ask questions from female attendants was dependent on the gender of the one that asked the first question during the question session. That is something to keep in mind when we are chairing sessions – we should be mindful of people from under-represented groups and encourage them to ask the first question to promote equal
opportunities. Another thing to consider is how we all, as attendees or speakers, react to questions and those that ask them. All questions should be asked out of a genuine interest in knowing the answer, thus all questions should be treated with respect. However, sometimes we might tend to forget this and, maybe unintentionally, respond in an inappropriate way. It is easy to understand that you will think twice before asking a question again if you felt that your last question was not taken seriously. Thus, we can all add to the environment in which everyone can feel that their questions are important. Thanks Marta Acácio for drawing our attention to this inequality and we hope that the
gender balance between people asking questions will be more equal at the next conference.”
On behalf of EOU and the local organising committee in Lund,
I want to thank my colleagues who helped collecting data for this blog post: João Gameiro, Inês Catry, Fábio Marcolin and Teresa Catry, and to Arne Hegemann for sending the data on the gender of the attendees of the conference. I also want to thank Claire Buchan, Jenny Gill and Charlie Russell for their valuable insights. Finally, thank you to the BOU for the member attendance grant.
Carter, A.J., Croft, A., Lukas, D. & Sandstrom, G.M. 2018. Women’s visibility in academic seminars: Women ask fewer questions than men. PlosOne 13:9. VIEW
Hinsley, A., Sutherland, W.J. & Johnston, A. 2017. Men ask more questions than women at a scientific conference. PlosOne 12:10. VIEW
Jarvis, S.N., Ebersole, C.R., Nguyen, C.Q. & Kray L.J. 2022. Stepping up to the mic: Gender gaps in participation in live question-and-answer sessions at academic conferences. Psychological Science 33:11. VIEW
Telis, N., Glassberg, E.C., Pritchard, J.K. & Gunter C. 2019. Public discussion affects question asking
at academic conferences. The American Journal of Human Genetics 105. VIEW
Hammond, C. 2023. Why are women less likely to ask questions in public? BBC Article VIEW