TITLE Computerized Adaptive Testing Simulations Using Real Test

A considerable amount of data on computerized adaptive testing (CAT) has been conducted using simulated data. However, most researchers would agree that simulations may not fully reflect the reality of examinee performance on a test. This study used maximum likelihood procedures to investigate the accuracy and efficiency of examinee ability recovery using real test taker responses (n=969) to 127 items from the Law School Admission Test (LSAT). It was hoped that through control of item parameters and examinee abilities based on a realistic test, insights would be gained on practical issues involved in an operative computerized adaptive LSAT. The 127 items were found to be sufficient to conduct the CAT sessions for virtually all test takers at the 3 accuracy levels. The highest accuracy level seemed to be necessary to recover the original paper-and-pencil ability estimates of the test takers. The average number of items used in the CAT sessions ranged from 13 for the lowest accuracy level to 37 for the highest accuracy level. This study represented the first CAT simulation that used real test taker responses. (Contains 4 figures, 6 tables, and 11 references.) (Author/SLD) Reproductions supplied by EDRS are the best that can be made from the original document