The Board of Registry of the American Society of Clinical Pathologists has been researching the psychometric properties of computer-adaptive testing (CAT) for the past 3 years, and has demonstrated that CAT is a makes reliable pass/fail certification decisions. Item selection is tailored to examinee ability and is made from Rasch-calibrated item pools distributed across difficulty and content areas. A pass/fail point has been established on the item pool scale. For a clear pass or fail, the stopping rule requires that the current estimated examinee ability measure be 1.5 S.E.s above or below the pass/fail point. This ensures that the pass/fail decision is made with over 90% confidence. Tailoring the test to the examinee's ability and using this stopping rule allows the administration of fewer items, while increasing the percent of examinees who clearly pass or fail, compared to a conventional paper-and-pencil test.
This summer the Board will conduct CAT examinations at a "model" site. After their CAT examination, examinees will also be asked to complete a satisfaction survey. Those who meet all requirements will be awarded certification. If CAT continues to be reliable psychometrically and also proves acceptable to the examinees, the next step will be to develop strategies for national implementation.
Computer-Adaptive Testing, M Lunz … Rasch Measurement Transactions, 1991, 5:1 p. 137
The URL of this page is www.rasch.org/rmt/rmt51m.htm
Website: www.rasch.org/rmt/contents.htm