One of the main goals of computerized adaptive testing (CAT) is to obtain precise ability estimates with a small number of items. To achieve this goal, items are selected specifically for each examinee from a large bank. The maximum item information method (MI) is wildly used to select items during the testing session. In practice, some items are used in most test administrations, but others are rarely (if ever) selected based on the MI method. Items with a high exposure rate are undesirable for test security reasons. Items with a low exposure rate are undesirable for economic reasons. Many item exposure control strategies were proposed for (1) preventing overexposure of some items, and (2) increasing the use rate of seldom or never-selected items. In general, theses item exposure control strategies could be classified in two groups: (a) methods adding a random component to the MI method, and (b) methods based on assigning a parameter to each item to control its maximum exposure. Some commonly used exposure control strategies were considered in this paper, such as “maximum information method (MI)”, “one parameter method (1P)”, “McBride and Martin method (MM)”, “randomesque method (PA)”,”Sympson and Hetter method (SH)”, “restricted maximum information method (Rk)”,and ”progressive method (PR)”.
Simulation study one was conducted to evaluate the performance of these exposure control strategies. The results show that, (1) the best methods with regard to precision are the poorest with regard to exposure rate control. (2) MI and MM methods yielded the highest precision in ability estimation; 1P and random selection (CO) methods yielded the well controlled in exposure rate. (3) PR method increased the minimum exposure rate and reduced the number of unused items without a serious decrease in test precision, but produced a maximum rate that was too high. (4) Rk and SH methods kept the maximum rate under control and showed adequate precision, but the minimum rate remained equal to that of the MI method.
Based on the results of study one, combining the PR (progressive) and Rk (restricted) methods, denoted as progressive restricted method, would perform well in precision and exposure, and evaluate the performance of the new method by carried out simulation study two. The results show that the new method seems to perform well on precision and exposure control and no parameters have to be determined by previous simulations.