The Impact of Omitting Confounders in Parallel Process Latent Growth Curve Mediation Models: Three Sensitivity Analysis Approaches

IF 2.5 2区心理学 Q2 MATHEMATICS, INTERDISCIPLINARY APPLICATIONS Structural Equation Modeling: A Multidisciplinary Journal Pub Date : 2023-05-19 DOI:10.1080/10705511.2023.2189551

Xiao Liu, Zhiyong Zhang, Kristin Valentino, Lijuan Wang

{"title":"The Impact of Omitting Confounders in Parallel Process Latent Growth Curve Mediation Models: Three Sensitivity Analysis Approaches","authors":"Xiao Liu, Zhiyong Zhang, Kristin Valentino, Lijuan Wang","doi":"10.1080/10705511.2023.2189551","DOIUrl":null,"url":null,"abstract":"AbstractParallel process latent growth curve mediation models (PP-LGCMMs) are frequently used to longitudinally investigate the mediation effects of treatment on the level and change of outcome through the level and change of mediator. An important but often violated assumption in empirical PP-LGCMM analysis is the absence of omitted confounders of the relationships among treatment, mediator, and outcome. In this study, we analytically examined how omitting pretreatment confounders impacts the inference of mediation from the PP-LGCMM. Using the analytical results, we developed three sensitivity analysis approaches for the PP-LGCMM, including the frequentist, Bayesian, and Monte Carlo approaches. The three approaches help investigate different questions regarding the robustness of mediation results from the PP-LGCMM, and handle the uncertainty in the sensitivity parameters differently. Applications of the three sensitivity analyses are illustrated using a real-data example. A user-friendly Shiny web application is developed to conduct the sensitivity analyses.","PeriodicalId":21964,"journal":{"name":"Structural Equation Modeling: A Multidisciplinary Journal","volume":"47 4","pages":""},"PeriodicalIF":2.5000,"publicationDate":"2023-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Structural Equation Modeling: A Multidisciplinary Journal","FirstCategoryId":"102","ListUrlMain":"https://doi.org/10.1080/10705511.2023.2189551","RegionNum":2,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"MATHEMATICS, INTERDISCIPLINARY APPLICATIONS","Score":null,"Total":0}

引用次数: 0

Abstract

Parallel process latent growth curve mediation models (PP-LGCMMs) are frequently used to longitudinally investigate the mediation effects of treatment on the level and change of outcome through the level and change of mediator. An important but often violated assumption in empirical PP-LGCMM analysis is the absence of omitted confounders of the relationships among treatment, mediator, and outcome. In this study, we analytically examined how omitting pretreatment confounders impacts the inference of mediation from the PP-LGCMM. Using the analytical results, we developed three sensitivity analysis approaches for the PP-LGCMM, including the frequentist, Bayesian, and Monte Carlo approaches. The three approaches help investigate different questions regarding the robustness of mediation results from the PP-LGCMM, and handle the uncertainty in the sensitivity parameters differently. Applications of the three sensitivity analyses are illustrated using a real-data example. A user-friendly Shiny web application is developed to conduct the sensitivity analyses.

查看原文

微信好友朋友圈 QQ好友复制链接

本刊更多论文

忽略混杂因素对平行过程潜在生长曲线中介模型的影响:三种敏感性分析方法

摘要平行过程潜在生长曲线中介模型(parallel process latent growth curve mediation models, PP-LGCMMs)常用来通过中介的水平和变化纵向考察治疗对结果水平和变化的中介作用。在经验PP-LGCMM分析中，一个重要但经常被违反的假设是，治疗、中介和结果之间的关系中没有遗漏的混杂因素。在本研究中，我们分析了忽略预处理混杂因素如何影响PP-LGCMM的中介推断。利用分析结果，我们开发了PP-LGCMM的三种灵敏度分析方法，包括频率分析方法、贝叶斯方法和蒙特卡罗方法。这三种方法有助于研究关于PP-LGCMM中介结果鲁棒性的不同问题，并以不同的方式处理敏感性参数的不确定性。最后以实际数据为例说明了三种灵敏度分析方法的应用。开发了一个用户友好的Shiny web应用程序来进行灵敏度分析。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文去求助

来源期刊

Structural Equation Modeling: A Multidisciplinary Journal 数学-数学跨学科应用

CiteScore

8.70

自引率

11.70%

发文量

审稿时长

>12 weeks

期刊介绍： Structural Equation Modeling: A Multidisciplinary Journal publishes refereed scholarly work from all academic disciplines interested in structural equation modeling. These disciplines include, but are not limited to, psychology, medicine, sociology, education, political science, economics, management, and business/marketing. Theoretical articles address new developments; applied articles deal with innovative structural equation modeling applications; the Teacher’s Corner provides instructional modules on aspects of structural equation modeling; book and software reviews examine new modeling information and techniques; and advertising alerts readers to new products. Comments on technical or substantive issues addressed in articles or reviews published in the journal are encouraged; comments are reviewed, and authors of the original works are invited to respond.