Abstract
The Rasch model is one of several item response theory measurement models that provides generality of measurement and linearity of scale. The former is a result of the separation of test item and person measures. Item difficulties are adjustedfor the effects of the particular group of examinees tested, and examinee scores are adjusted for the effects of the particular items in the test. Linearity of scale is accomplished by expressing item difficulty and person ability in a natural log odds unit. After several years of study, the National Board of Medical Examiners (NBME) now uses the Rasch model to calibrate items, build and maintain calibrated item banks, equate scores, and monitor the consistency ofexaminee item response patterns. The model is also being used by the NBMEin the study ofpatient management problems (PMP) examinations, in standard-setting studies, and in development of computer-based examinations.
Get full access to this article
View all access options for this article.