Background essential to the proper use of results of step 1 and step 2 of the USMLE

Abstract

This first of the four-part set of articles published in this issue provides general information and concepts about the Step 1 and Step 2 examinations of the United States Medical Licensing Examination (USMLE) needed for the proper use of these tests' results in three general, related, non-licensure applications: for the evaluation of the examinees' levels of academic achievement, for the evaluation of educational programs the examinees have experienced, and for the selection of examinees into residency programs. Several aspects of the tests are discussed: (1) their original and continuing purpose (which is to assess certain qualifications required for licensure of physicians); (2) their content; (3) their format and the emphases (relative to knowledge or skills testing) that different formats can give, and the concept that the ways the same content areas are formatted in a test will produce different results; (4) test administration, with a discussion of "speeded" versus "power" tests and the organization of test items by difficulty and by (or not by) topic; (5) reliability, with a discussion of standard error and the importance of understanding measurement error in order to discern real differences in scores; (6) scoring, with definition of anchor scores; and (7) interpretation, with discussions of pass/fail criteria in the past, present, and future of the USMLE tests. The authors close by saying that to interpret test performance properly, it is important not only to bear in mind the ideas in the present article but also to carefully review an actual test or at least a representative sample of test questions.