Influence of Implementation of Implicit Bias Training and Standardized Behavioral Questions on Diversity and Hiring of Internal Medicine Residents



Journal Title

Journal ISSN

Volume Title



Background: Implicit bias is a modifiable root cause of discrimination and is common in hiring. The business literature provides well-supported evidence of minorities experiencing implicit bias-based disadvantages during employment interviews. Minimizing implicit bias and increasing workforce diversity is essential for better organizational dynamics, job retention and satisfaction, and performance which can be mitigated by implicit bias training and standardized behavioral questionnaires (SBQ). Implicit bias has not been explored within the hiring process of internal medicine (IM) residents. Purpose: This study assesses the impact of implicit bias training and SBQ on IM resident selection as measured by ACGME® milestone evaluation scores, in-training examination (IM-ITE®) scores, and distributions of ethnicity and gender. Research Method: A quasi-experimental study used six years of IM resident applications and performance archival data. The independent variable consisted of two levels: 1) no implicit bias training and SBQ (Cohort 1, n = 124) and 2) implicit bias training and SBQ (Cohort 2, n = 120). Cohorts were compared to assess differences in residents' knowledge and professionalism as demonstrated by IM-ITE® scores and ACGME® milestone scores using Independent-samples Mann-Whitney U tests. Differences in pre and post-intervention distributions of resident ethnicity and gender were assessed using Pearson Chi-Square tests. Results: There were no statistically significant ethnicity differences across cohorts (χ2 = 1.17, p = 0.76). Residents of Asian descent accounted for 49% of selected candidates, followed by Whites (24%), URM (17%), and others (10%). Gender proportions differed significantly, with Cohort 2 having a higher proportion of males (χ2 = 5.61, p. = 0.02). Cohort 1 had significantly higher mid-year (M = 151.55) and end-of-year (M = 140.32) scores compared to Cohort 2 mid-year (M = 88.71) and end-of-year (M = 99.85) scores (U = 3385.0, p < 0.001). However, Cohort 1 declined end-of-year evaluation scores, whereas Cohort 2 showed an increase in end-of-year evaluation scores. Cohort 1 also had significantly higher scores in professionalism both mid-year (M = 151.38) and end-of-year (M = 146.68) compared to Cohort 2 mid-year (M = 89.62) and end-of-year (M = 92.32) scores (U = 3494.0, p < 0.001). Cohort 1 showed a decline in end-of-year professionalism scores, whereas Cohort 2 showed an increase. Cohort 2 scored significantly higher on IM-ITE® scores (M = 134.36) compared to Cohort 1 (M = 105.52, U = 8863.0, p = 0.001). Cohort 1 also had a significantly higher ITE rank (M=133) than Cohort 2(M=112.34). Conclusion: Implicit bias workshop training and SBQ are not associated with higher composite milestone scores or individual professionalism scores. However, implicit bias training and SBQ are associated with increases in the selection of male residents and higher resident IM-ITE® scores.



Implicit, bias, Standardized behavioral questions, Internal medicine residents, Medical education, Residency, Implicit bias training