关键词: 计算机化自适应测验, 选题策略, 等级反应模型

Abstract: The initial purpose of comparing item selection strategies was to increase the efficiency of tests. As studies continued, however, it was found that increasing the efficiency of item bank using was also an important goal of comparing item selection strategies. These two goals often conflicted. The key solution was to find a strategy with which both goals could be accomplished.
The item selection strategies for graded response model in this study included: the average of the difficulty orders matching with the ability (ADMA); the medium of the difficulty orders matching with the ability (MMDA); maximum information (MI); A stratified (average) (ASA); and A stratified (medium) (ASM). The evaluation indexes used for comparison included: the bias of ability estimates for the true (Bias); the standard error of ability estimates (Se); the average items which the examinees have administered (Aiea); the standard deviation of the frequency of items selected (Sdf); and sum of the indices weighted (Siw). Using the Monte Carlo simulation method, we obtained some data and computer iterated the data 20 times each under the conditions that the item difficulty parameters followed the normal distribution and even distribution. The results were as follows:

Table 1.
The result of item difficulty parameters following normal distribution
strategies Bias Se Aiea Sdf Siw
ADMAMMDAMI ASA ASM 0.2989 (5)0.2668 (2)0.2638 (1)0.2764 (4)0.2711 (3) 0.4218 (4)0.3541 (3)0.4361 (5)0.3313 (2)0.3181 (1) 26.7391 (4)27.5874 (5)22.6463 (1)22.9780 (2)23.0762 (3) 72.8872 (4)58.0541 (2)127.3360 (5)67.7299 (3)58.0313 (1) 3.28003.70763.18513. 75703.9545

Table 2.
The result of item difficulty parameters following even distribution
strategies Bias Se Aiea Sdf Siw
ADMAMMDAMI ASA ASM 0.2423 (3)0.2372 (2)0.1646 (1)0.2709 (4)0.2745 (5) 0.2857 (3)0.2850 (2)0.2022 (1)0.3220 (4)0.3260 (5) 29.6080(5)28.7977(4)24.1035(3)22.0724(1)22.2843(2) 38.2600(3)43.2547(4)141.7506(5)38.0857 (2)33.5079 (1) 3.00832.94463.15213.11543.2103

The results indicated that no matter difficulty parameters followed the normal distribution or even distribution. Every type of item selection strategies designed in this research had its strong and weak points. In general evaluation, under the condition that items were stratified appropriately, A stratified (medium) (ASM) had the best effect

Key words: CAT, item selection strategy, Graded Response Model