Incorporating the Sympson-Hetter Exposure Control Method into the a-Stratified Method with Content Blocking

Previous research (e.g., Leung, Chang, & Hau, 2001; Yi & Chang, 2001) has shown that the astratified method with content blocking (STR_C) performed well in content balancing constrained computerized adaptive tests (CATs). It effectively reduced the exposure rates of highly discriminating items, produced more balanced item usage within a pool, and maintained measurement precision. This method, however, does not have a mechanism that can fix (prespecify) the maximum item exposure rate. In the current study, the Sympson-Hetter exposure control procedure (Sympson & Hetter, 1985) was incorporated into STR_C (denoted STR_CSH). Under content balancing constraint conditions, this study investigated the effectiveness of STR_C-SH by comparing the STR_C-SH procedure with the original STR_C method and the maximum information selection method with Sympson-Hetter exposure control (MI-SH). This study also examined the effects of the test length, the number of strata, and the pre-specified maximum item exposure rate on the performance of each method. The results of this study have shown that STR_C-SH is an effective item exposure control method for CATs with practical content constraints. Incorporating the Sympson-Hetter Exposure Control Method into the a-Stratified Method with Content Blocking