在等级反应模型(GRM)下考察改进的最大优先级指标(MMPI)中引入曝光因子的选题策略与其他选题策略之间的优劣.蒙特卡洛模拟中采用定长CAT,其中能力估计采用贝叶斯期望后验估计(EAP).研究结果表明,MMPI与曝光因子相结合的选题策略,在保证精度的情况下,很大程度上解决了曝光率不均匀问题.
The study puts forward a new item selection strategy named MMPI with exposure factor under the graded response model, and Monte Carlo simulation is implemented to compare the superiority between the new item selection strategy and others. The simulation uses fixed-length test with the expected a posterior estima- tion(EAP) for ability estimation. The result shows that the new item selection strategy largely solves the issue of non-uniform exposure under ensuring the accuracy.