Gender, status, and self-discounting of expertise: A field experiment in scientific peer review

• In Event: 3546 - Sociology of Science

Abstract

Organizations often rely on experts to evaluate new projects and ideas. When experts evaluate ideas collaboratively, they must evaluate not only the idea, but also ``meta-evaluate'' the judgment of other experts. Here, we report a field experiment on the role that epistemic and social groups play in how experts meta-evaluate. After 277 faculty members from 26 US medical schools completed independent reviews of project proposals in biomedicine, we exposed them to artificial scores attributed to anonymous ``other reviewers,'' and enabled them to update initial assessments. Artificial scores were randomly generated and ascribed to reviewers in the same or different discipline. We found that disciplinary categories did not affect whether reviewers chose to update, but social categories did. Men updated their scores 13% less often than women, and very high-status reviewers updated 24% less often than others. Women updated 8% less for every 10% increase in their representation in the subfield. The experiment strongly suggests that outcomes of expert deliberation depend not only on the quality of individuals' judgment, but on the stereotyped meta-evaluations experts make of themselves relative to perceived others. The process by which expert panels arrive at judgments may thus be just as important as their composition.

Authors

• Misha Teplitskiy, University of Michigan

• Hardeep Ranu, Harvard Medical School

• Gary Gray, Harvard Medical School

• Eva Guinan, Harvard Medical School

• Karim R. Lakhani, MIT Sloan School of Management