Speaker introductions at major meetings may strengthen gender bias
CHICAGO — An analysis of speaker introductions at a major international oncology conference revealed patterns that may strengthen gender bias, according to findings presented at ASCO Annual Meeting.
When men introduced speakers, they more frequently used a professional address for males than females, and they also were more likely to introduce female speakers by first name only.
“How a speaker is introduced affects the audience’s perception of that speaker,” Narjust Duma, MD, chief hematology/medical oncology fellow at Mayo Clinic in Rochester, Minnesota, and a HemOnc Today Next Gen Innovator, told HemOnc Today. “Several studies have suggested the audience is most likely to pay attention and believe the data presented if the speaker is given a proper introduction. If we are not introducing women and men the same way, we are feeding into the existing gender bias in medicine.”
The unconscious use of gender-subordinating language, as well as variations in the forms of address, can reinforce gender bias, according to study background.
Duma and colleagues had noticed during ASCO Annual Meeting in 2018 that female speakers often were introduced by first name only but male speakers were not.
The investigators — who received an ASCO Employee Endowed Merit Award for their abstract — had observed similar patterns at other international oncology conferences.
Duma and colleagues conducted a retrospective observational study in which they reviewed video-archived speaker introductions from ASCO Annual Meeting in 2017 and 2018.
Mixed-gender coders evaluated frequency of professional address, defined as professional title followed by full or last name.
Researchers used multivariable logistic regressions to identify factors associated with the form of address.
Investigators reviewed 2,511 videos, 781 of which met inclusion criteria.
The majority of speakers were non-Hispanic white (65%), held MD or MD-PhD degrees (92%), and held associate or full professor positions (60%).
Results showed introducers were significantly less likely to use professional address for female speakers than male speakers (61% vs. 81%; P < .001). Female speakers also were more likely to be introduced by first name only (17% vs. 3%; P < .001).
When men delivered introductions, they were significantly less likely to use professional address for female speakers than male speakers (53% vs. 80%; P < .01). They also were more likely to address female speakers by first name only (24% vs. 7%; P < .01).
Investigators observed no difference in professional address when women delivered introductions. Female introducers were more likely than male introducers to refer to all speakers either with professional title and last name, or professional title followed by full name.
Multivariable regression analyses that accounted for gender, race, degree and academic rank showed male speakers were more likely than female speakers to receive a professional address (OR = 2.67; 95% CI, 1.81-3.94). Female gender was a predictor for nonprofessional form of address (OR = 9.5; 95% CI, 4.38-20.62).
“We were hoping our observations and anecdotes were isolated events,” Duma told HemOnc Today. “We were surprised that this was observed on a larger scale and that, despite controlling for several factors — including academic ranking, degree and type of sessions — women were still more likely to be introduced by first name only or not be introduced by their professional title.”
The findings have considerable ramifications, Duma said, particularly as more women enter the oncology and hematology workforces.
“It is our duty to welcome them with equal opportunities and reduced bias, and to give the new generation the proper tools to advance,” she said.
Increasing awareness of these patterns is an important first step, Duma said.
“Disseminating our findings will help introducers be aware of these disparities and introduce all speakers equally,” she said. “Additionally, we aim [to encourage] the creation of introduction guidelines for all large conferences, where the professional title is mandatory to all introductions.”
The study also revealed that black speakers — regardless of gender — were less likely to receive professional address than non-Hispanic white speakers (OR = 0.1; 95% CI, 0.01-0.53).
Data about race and ethnicity were assigned based on available information and coder discretion, making them subject to bias and limiting accuracy.
Given the limited number of minorities who presented at the meetings — for example, researchers identified fewer than 15 black presenters — Duma said she and her colleagues were surprised to see the differences in professional address based on race or ethnicity reach statistical significance,
The investigators hope to improve methods of collecting information on the race and ethnicity of speakers and introducers to better understand if other forms of implicit bias besides gender bias may have contributed to these patterns and differences.
“We all have implicit or unconscious bias as the result of media influence, where we grew up and other factors,” Duma said. “We cannot eliminate implicit bias, but we can work hard to reduce them.” – by Mark Leiser
Duma N, et al. Abstract 10503. Presented at: ASCO Annual Meeting; May 31-June 4, 2019; Chicago.
Disclosures: Duma reports no relevant financial disclosures. Please see the abstract for all other authors’ relevant financial disclosures.