不完全双向列联表中缺失机制的不可识别性和可测试性

IF 0.6 Q4 STATISTICS & PROBABILITY

Communications for Statistical Applications and Methods Pub Date : 2021-05-31 DOI:10.29220/CSAM.2021.28.3.307

Yousung Park, Seung Mo Oh, Tae Yeon Kwon

{"title":"不完全双向列联表中缺失机制的不可识别性和可测试性","authors":"Yousung Park, Seung Mo Oh, Tae Yeon Kwon","doi":"10.29220/CSAM.2021.28.3.307","DOIUrl":null,"url":null,"abstract":"We showed that any missing mechanism is reproduced by EMAR or MNAR with equal fit for observed likelihood if there are non-negative solutions of maximum likelihood equations. This is a generalization of Molenberghs et al. (2008) and Jeon et al. (2019). Nonetheless, as MCAR becomes a nested model of MNAR, a natural question is whether or not MNAR and MCAR are testable by using the well-known three statistics, LR (Likelihood ratio), Wald, and Score test statistics. Through simulation studies, we compared these three statistics. We investigated to what extent the boundary solution affect tesing MCAR against MNAR, which is the only testable pair of missing mechanisms based on observed likelihood. We showed that all three statistics are useful as long as the boundary proximity is far from 1.","PeriodicalId":44931,"journal":{"name":"Communications for Statistical Applications and Methods","volume":" ","pages":""},"PeriodicalIF":0.6000,"publicationDate":"2021-05-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Non-identifiability and testability of missing mechanisms in incomplete two-way contingency tables\",\"authors\":\"Yousung Park, Seung Mo Oh, Tae Yeon Kwon\",\"doi\":\"10.29220/CSAM.2021.28.3.307\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"We showed that any missing mechanism is reproduced by EMAR or MNAR with equal fit for observed likelihood if there are non-negative solutions of maximum likelihood equations. This is a generalization of Molenberghs et al. (2008) and Jeon et al. (2019). Nonetheless, as MCAR becomes a nested model of MNAR, a natural question is whether or not MNAR and MCAR are testable by using the well-known three statistics, LR (Likelihood ratio), Wald, and Score test statistics. Through simulation studies, we compared these three statistics. We investigated to what extent the boundary solution affect tesing MCAR against MNAR, which is the only testable pair of missing mechanisms based on observed likelihood. We showed that all three statistics are useful as long as the boundary proximity is far from 1.\",\"PeriodicalId\":44931,\"journal\":{\"name\":\"Communications for Statistical Applications and Methods\",\"volume\":\" \",\"pages\":\"\"},\"PeriodicalIF\":0.6000,\"publicationDate\":\"2021-05-31\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Communications for Statistical Applications and Methods\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.29220/CSAM.2021.28.3.307\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q4\",\"JCRName\":\"STATISTICS & PROBABILITY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Communications for Statistical Applications and Methods","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.29220/CSAM.2021.28.3.307","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"STATISTICS & PROBABILITY","Score":null,"Total":0}

引用次数: 0

摘要

我们发现，如果存在最大似然方程的非负解，任何缺失的机制都可以通过EMAR或MNAR重现，并且具有与观测似然相等的拟合。这是Molenberghs等人(2008)和Jeon等人(2019)的概括。然而，随着MCAR成为MNAR的嵌套模型，一个自然的问题是，MNAR和MCAR是否可以通过使用众所周知的三种统计量，LR(似然比)，Wald和Score测试统计量进行测试。通过仿真研究，我们比较了这三种统计数据。我们研究了边界解在多大程度上影响MCAR对MNAR的测试，MNAR是基于观察到的似然的唯一可测试的缺失机制对。我们表明，只要边界接近度远离1，这三个统计量都是有用的。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

Non-identifiability and testability of missing mechanisms in incomplete two-way contingency tables

We showed that any missing mechanism is reproduced by EMAR or MNAR with equal fit for observed likelihood if there are non-negative solutions of maximum likelihood equations. This is a generalization of Molenberghs et al. (2008) and Jeon et al. (2019). Nonetheless, as MCAR becomes a nested model of MNAR, a natural question is whether or not MNAR and MCAR are testable by using the well-known three statistics, LR (Likelihood ratio), Wald, and Score test statistics. Through simulation studies, we compared these three statistics. We investigated to what extent the boundary solution affect tesing MCAR against MNAR, which is the only testable pair of missing mechanisms based on observed likelihood. We showed that all three statistics are useful as long as the boundary proximity is far from 1.

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

Communications for Statistical Applications and Methods STATISTICS & PROBABILITY-

CiteScore

0.90

自引率

0.00%

发文量

期刊介绍： Communications for Statistical Applications and Methods (Commun. Stat. Appl. Methods, CSAM) is an official journal of the Korean Statistical Society and Korean International Statistical Society. It is an international and Open Access journal dedicated to publishing peer-reviewed, high quality and innovative statistical research. CSAM publishes articles on applied and methodological research in the areas of statistics and probability. It features rapid publication and broad coverage of statistical applications and methods. It welcomes papers on novel applications of statistical methodology in the areas including medicine (pharmaceutical, biotechnology, medical device), business, management, economics, ecology, education, computing, engineering, operational research, biology, sociology and earth science, but papers from other areas are also considered.