Introduction: It is difficult for learners to perform accurate self-assessments. This difficulty may be exaggerated in unskilled learners, a phenomenon termed the Dunning-Kruger Effect (Dunning & Kruger, 1999). Learners with the least amount of knowledge or skill may paradoxically be more likely to evaluate themselves favorably compared with their peers. This phenomenon is particularly relevant in medicine where we rely on self-directed learning not only in many of our undergraduate and postgraduate programs, but in guiding the pursuit of continuing medical education. The objectives of this study are to 1) determine whether the Dunning-Kruger Effect is present in medical education settings, 2) to determine the quality of studies in this area, and 3) to determine how this effect, if present, could influence approaches to the learner in difficulty. Methods: This is a review of the literature. PubMed databases were searched for all relevant articles. Included studies reported self-assessment of medical trainees or staff and comparison with an external rating. Studies were identified using select keywords and MeSH terms. Only studies published in English were included. No publication date limits were adopted. The Medical Education Research Study Quality Instrument (MERSQI) was used to assess study quality. Both authors independently abstracted data and rated study quality. Results: Eighty-six articles were identified in the PubMed search. On abstract review, 45 studies were found to meet criteria for further full article review. Studies were variable in setting and approach to self and external assessment. Criteria were not met for pooled analyses/meta-analysis. Results are presented as a summary of findings with special consideration of findings based on level of training (undergraduate, postgraduate, staff clinician). Conclusion: This review summarizes the current literature on the Dunning-Kruger Effect in medical education and provides an assessment of the quality of studies in this area to date. The potential relevance of the Dunning-Kruger Effect in medical education is discussed as are implications for interventions to support the learner in difficulty. Additional study in this area is indicated, in particular given the significant upcoming changes to postgraduate medical education in Canada in the era of Competence By Design (CBD).
