Is There Bias in Virtual Interviews for Emergency Medicine Residency?

Background Since the COVID-19 pandemic, residency programs have increasingly relied on virtual interviews. The impact of the interview format on interview scoring remains uncertain. As more programs offer both virtual and in-person interviews, understanding how these formats impact evaluations is increasingly essential. This study aimed to assess bias in applicant interview scores and determine whether the format (virtual vs. in-person) is associated with differences in interview scores. Methodology Data were collected on two years' worth (2023-2025) of interviewed applicants retrospectively through the Electronic Residency Application System and anonymized. After assessing baseline characteristics between virtual and in-person applicants, a linear regression analysis was performed to assess associations between several predictors and interview score. Results A total of 241 applicants were interviewed over the two years that data were collected. Of those, 130 (53.9%) were interviewed virtually and 111 (46.1%) in-person. Assessed baseline characteristics were similar between groups, other than the frequency of median standardized letter of evaluation (SLOE) scores greater than two and state of residence. Linear regression accounted for 27.3% of the variance in interview scores (indicating moderate explanatory power). Golden Humanism and Alpha Omega society membership, as well as female gender and SLOE scores, were positively correlated with interview scores. Interview format (in-person vs. virtual) was not a significant predictor of interview score. Conclusions There was no difference in interview scores between in-person and virtual interviewees when controlled for other factors.