Association between Characteristics of Impostor Phenomenon in Medical Students and Step 1 Performance.

THEORY: Impostor phenomenon (IP) is a widely recognized experience in which highly performing individuals do not internalize success. Self-doubt toward one's ability or intelligence is unexpectedly common. Prior research has associated IP with medical student perceptions, burnout, and demographic characteristics. However, understanding how student IP experiences are related to actual academic achievement could help medical schools better support student performance and resilience. Hypotheses: The purpose of this research was to examine whether there is a relationship between medical students' USMLE Step 1 scores and experiences of IP. Because medical students receive frequent, objective feedback on exam performance, we hypothesized that students with lower Step 1 scores would experience higher levels of IP.
Methods: In 2019, all M1-M4 students at the University of Louisville were invited to complete Clance's (1985) Impostor Phenomenon Scale (CIPS), a previously validated, 20-item, Likert-style scoring instrument designed to reflect respondents' IP experience. We categorized subjects into one of four levels based on CIPS scoring guidelines. For students who had completed Step 1 at the time of the survey, we conducted a Welch's ANOVA test to identify relationships between a student's level of experienced IP and Step 1 scores. We also completed an item analysis comparing individual CIPS item responses with Step 1 performance.
Results: Per the CIPS scoring guidelines, we categorized the 233 respondents as experiencing few (10.3%), moderate (47.6%), frequent (31.8%), or intense (10.3%) characteristics of IP. Nearly 90% of our sample experienced at least moderate levels of IP, with over 40% experiencing frequent or intense IP. There were no statistically significant differences among CIPS groups and mean Step 1 scores [F (3, 59.8) = 1.81, p = 0.155], and total/inter-quartile Step 1 score ranges broadly overlapped among all four IP experience levels. Within-item response patterns among high and low-scoring students also varied across individual CIPS items. Conclusions: The association between IP experience and Step 1 performance was nonlinear in our cross-sectional sample: the group having the most intense IP experiences did not have the lowest score rank on Step 1, nor did the group with the fewest IP characteristics make up our highest performing group. This, along with the broad dispersion of scores within each of the four IP levels, suggests that students' internalization of achievement and feelings of IP are not consistently aligned with their actual performance on this assessment. Response variation on individual CIPS items suggest that underlying factors may drive variation in IP and performance. These results highlight the need for additional work to identify the constructs of IP that influence medical students specifically so that medical education stakeholders may better understand IP's impact on other facets of medical school and implement the resources necessary to support individuals who experience IP.