DIF值和样本量对SIBTEST检测方法的影响研究  

Effects of DIF Magnitude and Sample Size on Robustness and Power of SIBTEST Procedure

作　　者：朱乙艺[1] 韦小满[1] 

机构地区：[1]北京师范大学

出　　处：《中国考试》2012年第9期9-14,共6页journal of China Examinations

摘　　要：本模拟研究的目标组和参照组的题目作答反应数据是通过三参数逻辑斯蒂克模型来产生的,通过模拟数据探讨了DIF值和样本量对SIBTEST方法的检测效果的影响。研究结果表明:在一定的DIF值和样本量条件下,SIBTEST方法的检验力和I型错误随着样本量和DIF值的增大而增大;当用SIBTEST方法对实测数据进行DIF检测时,1000～2000的样本量是比较合适的;当用模拟数据进行SIBTEST方法的研究时,DIF值和样本量的选择需要满足一定的限制条件。In this simulation study, the effects of six DIF magnitudes and six sample sizes on robustness and power of SIBTEST DIF detection procedure were studied. Item response data were generated under the three parameter logistic model for focal and reference group subjects. The findings indicate that： under the condition of certain DIF magnitude and sample size, the power and Type I error of SIBTEST procedure increase with DIF magnitude and sample size; the proper sample size for SIBTEST method is between 1000 and 2000 when detecting real data; the choice of DIF magnitude and sample size for SIBTEST procedure is not free when carrying out simulation study.

关 键 词：DIF值 样本量 SIBTEST方法 项目功能差异 

分 类 号：O212.2[理学—概率论与数理统计]