A Comparative Evaluation of a Conditional Median-Based Bayesian Growth Curve Modeling Approach with Missing Data

被引:0
作者
Tang, Dandan [1 ]
Tong, Xin [1 ]
Zhou, Jianhui [1 ]
机构
[1] Univ Virginia, Dept Psychol, Charlottesville, VA 22904 USA
基金
美国国家科学基金会;
关键词
Full information maximum likelihood; longitudinal data; missing data; robust median-based Bayesian approach; two-stage robust estimation; MAXIMUM-LIKELIHOOD;
D O I
10.1080/10705511.2025.2519209
中图分类号
O1 [数学];
学科分类号
0701 ; 070101 ;
摘要
Longitudinal data are essential for studying within-subject change and between-subject differences in change. However, missing data, especially with nonnormal distributions, remains a significant challenge in longitudinal analysis. Full information maximum likelihood estimation (FIML) and a two-stage robust estimation (TSRE) are widely used for handling missing data, but their effectiveness may diminish with data skewness, high missingness rate, and nonignorable missingness. Recently, a robust median-based Bayesian (RMB) approach for growth curve modeling (GCM) was proposed to offer a novel way to handle nonnormal longitudinal data, yet its effectiveness with missing data has not been fully investigated. This study fills this gap by using Monte Carlo simulations to systematically evaluate RMB's performance relative to FIML and TSRE. It is shown that, in general, RMB GCM is a reliable option for managing ignorable and nonignorable missing data in various data distributional scenarios. An empirical example further demonstrates the application of these methods.
引用
收藏
页数:10
相关论文
共 44 条
[21]   TESTING HYPOTHESES FOR THE GROWTH CURVE MODEL WHEN THE DATA ARE INCOMPLETE [J].
LEEPER, JD ;
WOOLSON, RF .
JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION, 1982, 15 (2-3) :97-107
[22]   Performance of Estimators for Confirmatory Factor Analysis of Ordinal Variables with Missing Data [J].
Lei, Pui-Wa ;
Shiverdecker, Levi K. .
STRUCTURAL EQUATION MODELING-A MULTIDISCIPLINARY JOURNAL, 2020, 27 (04) :584-601
[23]   Bayesian Approaches for Missing Not at Random Outcome Data: The Role of Identifying Restrictions [J].
Linero, Antonio R. ;
Daniels, Michael J. .
STATISTICAL SCIENCE, 2018, 33 (02) :198-213
[24]   THE UNICORN, THE NORMAL CURVE, AND OTHER IMPROBABLE CREATURES [J].
MICCERI, T .
PSYCHOLOGICAL BULLETIN, 1989, 105 (01) :156-166
[25]  
Muthen L.K., 2017, MPLUS USERS GUIDE ST
[26]   A Longitudinal Analysis of Risk and Protective Factors of Bias-Based Bullying Victimization Among Adolescents [J].
Parodi, Katharine B. ;
Holt, Melissa K. ;
Aradhya, Pooja ;
Green, Jennifer Greif ;
Merrin, Gabriel J. .
JOURNAL OF INTERPERSONAL VIOLENCE, 2025,
[27]  
Plummer M., 2007, P 3 INT WORKSHOP DIS, DOI DOI 10.1111/J.2517-6161.1996.TB02070.X
[28]   Analysis of longitudinal studies with missing data using covariance structure modeling with full-information maximum likelihood [J].
Raykov, T .
STRUCTURAL EQUATION MODELING-A MULTIDISCIPLINARY JOURNAL, 2005, 12 (03) :493-505
[29]   Patterns of mean-level change in personality traits across the life course: A meta-analysis of longitudinal studies [J].
Roberts, BW ;
Walton, KE ;
Viechtbauer, W .
PSYCHOLOGICAL BULLETIN, 2006, 132 (01) :1-25
[30]   lavaan: An R Package for Structural Equation Modeling [J].
Rosseel, Yves .
JOURNAL OF STATISTICAL SOFTWARE, 2012, 48 (02) :1-36