[Rasch] Clinical exams & Authentic Assessments
gstone at UTNet.UToledo.Edu
Fri Dec 2 14:25:40 EST 2005
Dear Rasch Colleagues;
One of the groups with whom I am affiliated administer Clinical examinations to examinee physicians earning certification. The clinicals as this group calls them, amount to a random selection of patient charts from a list of logs kept during the year. The examiners review these charts with the candidate and assess them for clarity, process, critical thinking, etc.
Using Facets, I have structured the analyses using common judge equating. The candidates are unique, and the cases are all unique as well. For instance, even if an "appendix" or "gall bladder" case were requested of all surgeons, each case is a unique patient, and no two can be seen as alike. We've tried t consider them as similar items, but there are simply too many variables within the cases to warrant that conclusion. Common judge equating is clearly the weakest linkage, even through extensive training. Furthermore, there are few common judges to link. Out of a total of 50 examinees, for instance, a single judge may only rate 3 or 4 examinees.
I've been exploring other models of "authentic assessment" and qualitative approaches, yet few can pass the muster of legal defensibility. Any suggestions beyond those mentioned??
Gregory E. Stone, Ph.D., M.A.
Assistant Professor, Research and Measurement
University of Toledo, College of Education, Mailstop #923
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Rasch