Five ways to get a grip on grouped self-assessments of competence for program evaluation

Self-assessments conducted by individuals when taken together (grouped) provide valid and accurate measures of learning outcomes of the group. This is useful for program evaluation. Grouped self-assessments are simple to understand and construct, easy to implement, relatively accurate, and do not require extensive and complex pre-post testing measures. However, group self-assessments have the potential to be misused. To examine how group self-assessments have been used in medical education, we conducted a search of journal articles published in 2017 and 2018 from eight prominent medical education journals. Twenty-seven (n = 27) articles that used self-assessments for program evaluation were selected for data extraction and analysis. We found three main areas where misuse of self-assessments may have resulted in inaccurate measures of learning outcomes: measures of "confidence" or "comfort", pre-post self-assessments, and the use of ambiguous learning objectives. To prevent future misuse and to build towards more valid and reliable data for program evaluations, we present the following recommendations: measure competence instead of confidence or comfort; use pre-test self-assessments for instructional purposes only (and not for data); ask participants to do the post-intervention self-assessments first followed by retrospective pre-intervention self-assessments afterwards; and use observable, clear, specific learning objectives in the educational intervention that can then be used to create the self-assessment statements.