Order-restricted hypothesis tests for nonlinear mixed-effects models with measurement errors in covariates

被引:0
作者
Zhang, Yixin [1 ]
Liu, Wei [2 ]
Wu, Lang [3 ]
机构
[1] Univ Technol & Sci, Sch Math Sci, Hefei, Peoples R China
[2] York Univ, Dept Math & Stat, Toronto, ON, Canada
[3] Univ British Columbia, Dept Stat, Vancouver, BC, Canada
来源
CANADIAN JOURNAL OF STATISTICS-REVUE CANADIENNE DE STATISTIQUE | 2024年 / 52卷 / 04期
基金
加拿大自然科学与工程研究理事会;
关键词
Constrained hypothesis tests; longitudinal data; measurement error; multiple imputation; power; MULTIPLE-IMPUTATION; INFERENCE; WINBUGS;
D O I
10.1002/cjs.11812
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
Order-restricted hypothesis testing problems frequently arise in practice, including studies involving regression models for longitudinal data. These tests are known to be more powerful than tests that ignore such restrictions. In this article, we consider order-restricted tests for nonlinear mixed-effects models with measurement errors in time-dependent covariates. We propose to use a multiple imputation method to address measurement errors, since this approach allows us to use existing complete-data methods for order-restricted tests. Some theoretical results are presented. We evaluate our proposed methods via simulation studies that demonstrate they are more powerful than either a competing naive method or a two-step approach to testing hypotheses. We illustrate the use of our proposed approach by analyzing data from an HIV/AIDS study. Les tests d'hypoth & egrave;ses avec restrictions d'ordre sont fr & eacute;quents en pratique, notamment pour les mod & egrave;les de r & eacute;gression sur donn & eacute;es longitudinales. Ces tests sont reconnus comme & eacute;tant plus puissants que les tests ignorant de telles restrictions. Les auteurs de cet article s'int & eacute;ressent aux tests avec restrictions d'ordre pour les mod & egrave;les non lin & eacute;aires & agrave; effets mixtes en pr & eacute;sence d'erreurs de mesure sur les covariables d & eacute;pendantes du temps. Ils proposent d'utiliser l'imputation multiple pour traiter ces erreurs, une approche permettant d'appliquer les m & eacute;thodes existantes de tests avec restrictions d'ordre aux donn & eacute;es compl & egrave;tes. Quelques r & eacute;sultats th & eacute;oriques sont pr & eacute;sent & eacute;s. Par des & eacute;tudes de simulation, ils d & eacute;montrent que leurs m & eacute;thodes sont plus puissantes qu'une approche na & iuml;ve ou une proc & eacute;dure en deux & eacute;tapes pour tester les hypoth & egrave;ses. Enfin, les auteurs illustrent leur d & eacute;marche sur des donn & eacute;es d'une & eacute;tude sur le VIH/SIDA.
引用
收藏
页数:15
相关论文
共 27 条
[1]   A Unified Approach to Measurement Error and Missing Data: Overview and Applications [J].
Blackwell, Matthew ;
Honaker, James ;
King, Gary .
SOCIOLOGICAL METHODS & RESEARCH, 2017, 46 (03) :303-341
[2]  
Carroll RJ, 2006, BIOMETRICS, V62, P75, DOI 10.1111/j.1541-0420.2005.00400.x
[3]   Multiple-imputation for measurement-error correction [J].
Cole, Stephen R. ;
Chu, Haitao ;
Greenland, Sander .
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY, 2006, 35 (04) :1074-1081
[4]  
Dong X., 2021, J STAT RES, V55, P175
[5]  
FULLER WA, 1987, MEASUREMENT ERROR MO
[6]  
Higgins KM, 1997, J AM STAT ASSOC, V92, P436
[7]   Hierarchical Bayesian methods for estimation of parameters in a longitudinal HIV dynamic system [J].
Huang, Yangxin ;
Liu, Dacheng ;
Wu, Hulin .
BIOMETRICS, 2006, 62 (02) :413-423
[8]  
Johnson N. L., 1995, Continuous Univariate Distributions, V2
[9]   Overview and evaluation of various frequentist test statistics using constrained statistical inference in the context of linear regression [J].
Keck, Caroline ;
Mayer, Axel ;
Rosseel, Yves .
FRONTIERS IN PSYCHOLOGY, 2022, 13
[10]   LARGE-SAMPLE SIGNIFICANCE LEVELS FROM MULTIPLY IMPUTED DATA USING MOMENT-BASED STATISTICS AND AN F-REFERENCE DISTRIBUTION [J].
LI, KH ;
RAGHUNATHAN, TE ;
RUBIN, DB .
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1991, 86 (416) :1065-1073