Gersten Jonker

110   Chapter 5 Assessment scores Each station had a specific checklist following the chronology of the consultation, steps in skill execution, or expected actions and important steps in case management. Assessors scored each itemon a three-point scale (done adequately – done incompletely or not timed well – done insufficiently or omitted). Every assessment element (e.g. station) yielded a summary score, calculated as the awarded points as percentage of the maximum points. From the element summary scores ameanmodality score couldbe calculated for CBDs, forOSCEs, and for simulations. On CBDs, assessor-observed OSCEs, and simulations, assessors evaluated participant performance with a global rating score (GRS), in addition to checklist scores. This score was a holistic, gestalt, expert evaluation of overall performance on a three-point scale (does not meet expectations – borderline – meets expectations) in answer to the question “How did the participant’s performance compare to the expected level of a PNIT with six months of clinical experience?”. We chose the arbitrary level of six months to have assessors think of a graduate with substantial but delineated experience and to make a match with the actual experience of the PNIT comparison group. We ran 25 highly similar test sessions, consisting of two separate half-days. In one test session, all participants would take the same test, either as pre-test or as post-test. Participants’ pre-test and post-test content was never identical. Students attended the pre-test in the first month of their final year and the post-test near the end of the year. Recruited as recent graduates, PNITs could not take a pre-test at the start of their final year. Thus, PNITs attended the test once and, in the comparison, this is regarded as a post-test. All participants received feedback on their performance in the tests. Ethical approval The ethical reviewboard of the Netherlands Association for Medical Education approved the study (NERB file 369, November 2014). All participants gave written informed consent. Analysis Modality scores (i.e. knowledge score, mean CBD score, mean OSCE score, and mean simulation score) are expressed in percent points and are compared as group means with SD and differences with 95% CI in rounded percent points.