...
首页> 外文期刊>Applied Psychological Measurement >Effects of Unequal Ability Variances on the Performance of Logistic Regression, Mantel-Haenszel, SIBTEST IRT, and IRT Likelihood Ratio for DIF Detection
【24h】

Effects of Unequal Ability Variances on the Performance of Logistic Regression, Mantel-Haenszel, SIBTEST IRT, and IRT Likelihood Ratio for DIF Detection

机译:能力差异不均对Logistic回归,Mantel-Haenszel,SIBTEST IRT和IRT似然比进行DIF检测的性能的影响

获取原文
获取原文并翻译 | 示例
   

获取外文期刊封面封底 >>

       

摘要

Differential item functioning (DIF) of items has become an important issue in test fairness and equity in large-scale assessments. DIF occurs when subgroups of test takers have equal trait levels (on the construct the test is intended to measure, such as ability) but differ in their probabilities of a correct response (Roussos & Stout, 1996). DIF items may threaten the validity of test scores for subgroups (Borsboom, 2006; Maller, 2001) and can mislead researchers about group differences (Woods, 2008). Thus, it is critical to identify DIF items. Numerous parametric and nonparametric methods have been proposed for detecting DIF (Holland & Wainer, 1993; Furlow, Ross, & Gagne, 2009; Rivas, Gabriel, Stark, & Chernyshenko, 2009; Shih & Wang, 2009; Woods, 2009), and a lot of simulation studies have been done to examine the performance of these methods to flag DIF items. However, among these studies, there is little attention to the effects on DIF detection methods of the difference in ability variance between two groups. The empirical study of Bielinski and Davison (1998) and the simulation study of Monahan and Ankenmann (2005) confirmed that the effect of difference in ability variance between reference and focal groups is strong in DIF detection. However, Monahan and Ankenmann's study focused only on the Mantel-Haenszel chi-square test (Monahan & Ankenmann, 2005), whereas Bielinski and Davison's study focused only on the likelihood ratio test (Bielinski & Davison, 1998). Thus, the aim of this study is to examine how the difference in ability variance between reference and focal groups affects four commonly used DIF detection methods: (a) logistic regression modeling (Zumbo, 1999), (b) Mantel-Haenszel chi-square (MH test) (Holland & Thayer, 1988; Mantel & Haenszel, 1959), (c) simultaneous item bias test (SIBTEST; Shealy & Stout, 1993), and (d) likelihood ratio test (Thissen, Steinberg, & Wainer, 1988, 1993).
机译:项目的差异项目功能(DIF)已成为大规模评估中测试公平性和公平性的重要问题。当应试者的子群具有相同的特征水平时(在测试要测量的结构上,例如能力),但其正确回答的概率却不同时,就会发生DIF(Roussos&Stout,1996)。 DIF项目可能会威胁子组测试成绩的有效性(Borsboom,2006年; Maller,2001年),并且可能误导研究人员关于组间差异(Woods,2008年)。因此,识别DIF项目至关重要。已经提出了许多用于检测DIF的参数和非参数方法(Holland&Wainer,1993; Furlow,Ross,&Gagne,2009; Rivas,Gabriel,Stark,&Chernyshenko,2009; Shih&Wang,2009; Woods,2009),以及已经进行了许多模拟研究来检验这些方法标记DIF项目的性能。但是,在这些研究中,很少关注两组之间的能力差异对DIF检测方法的影响。 Bielinski和Davison(1998)的经验研究以及Monahan和Ankenmann(2005)的模拟研究证实,参考组和焦点组之间的能力方差差异对DIF检测的影响很大。但是,Monahan和Ankenmann的研究仅关注于Mantel-Haenszel卡方检验(Monahan&Ankenmann,2005),而Bielinski和Davison的研究仅关注似然比检验(Bielinski&Davison,1998)。因此,本研究的目的是检验参考组和焦点组之间的能力差异如何影响四种常用的DIF检测方法:(a)Logistic回归建模(Zumbo,1999),(b)Mantel-Haenszel卡方(MH检验)(Holland&Thayer,1988; Mantel&Haenszel,1959),(c)同​​时项目偏倚检验(SIBTEST; Shealy&Stout,1993),以及(d)似然比检验(Thissen,Steinberg,&Wainer, 1988,1993)。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号