At an academic conference some years ago, Michèle Belot remembers talking with a participant who was convinced she had authored a research paper that wasn’t hers. He’d confused her with another female scholar, an experience she said is familiar to many colleagues.
Such incidents – plus awareness of her own imperfect memory – inspired Belot, a professor in the Department of Economics, to investigate systemic biases in the way we remember people, since this could influence social networks important to career advancement.
In new research focused on academia, Belot finds that being a woman or racial minority can help someone stand out and be remembered when few others look like them. But they are more likely to be confused in settings where others share the same attributes.
“Minority attributes can help memory, but they also lead to confusion,” said Belot, who holds appointments in the ILR School and College of Arts and Sciences. “There is this double edge to it: If I am the only woman or racial minority navigating social networks, that might actually be helpful to being remembered. But as soon as there are others, that’s not going to be the case anymore.”
Belot is the co-author with Marina Schröder, economics professor at Leibniz University Hannover in Germany, of “,” published March 22 in the Journal of Behavioral and Experimental Economics.
The authors believe they are the first to provide evidence of bias in our ability to remember professionally relevant information about people, which could contribute to discrimination. If one isn’t remembered based on gender or race, it could hurt their chances of being recruited for a job or invited to collaborate or give a talk, for an academic, or equivalent opportunities in other businesses.
Belot and Schröder began their study in the field with a pair of prominent economics conferences held in the U.S. and Scotland, where respectively 35% and 20% of presenters were women, and 11% and 16% were nonwhite.
One month after each of the events, the researchers followed up with conferencegoers to see how well they remembered who had presented what. Nearly 90 study participants were asked to match pictures of presenters with titles of their research papers, given four choices. Then, seeing a presenter’s photo, they were asked to write that person’s name and institution.
The results showed that female presenters were much more likely than male presenters – by 14 percentage points – to be correctly matched to a photo, but study participants couldn’t remember their names or institutions any better.
The scholars controlled for measures of experience and productivity that might make a presenter better known, such as their institution’s rank or publications in top journals, and if they shared the same field, gender or ethnicity as the study respondents.
The field data limited by the conferences’ small number of nonwhite presenters, so the researchers also conducted controlled experiments online to test a larger sample. Nearly 400 study participants viewed photos from a database randomly paired with titles of economics papers, then tried to match the names and titles in multiple choice questions.
The results confirmed those from the field. Women were remembered better if they were a minority, but not if the choices included multiple women. Women and nonwhite people were more likely to be confused with others of the same gender or race, particularly when there were more to choose from.
“Gender and race are attributes that people encode very quickly about others,” Belot said. “But we demonstrate that they are not able to recall the exact person very well. They are more likely to confuse these people with others who share the same attributes.”
The researchers said the findings are consistent with prior research on memory proposing that that people are categorized by minority attributes and “blended together” with others sharing them.
Belot said many academic institutions now are more systematic about maintaining lists of job candidates and potential speakers to avoid reliance on social networks. Personally, after conferences and other networking events, Belot takes notes about whom she has met to reduce the possibility of cognitive bias limiting her memory.
“We do not know the implications of these biases for people’s careers,” the authors conclude, “but given the importance of recall in network formation, we conjecture these effects may not be small.”
