Expanding the Bayesian structural equation, multilevel and mixture models to logit, negative-binomial, and nominal variables

Asparouhov, Tihomir; Muthén, Bengt

doi:10.1080/10705511.2021.1878896

Cited by 15 publications

(12 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Lee et al, 2007;Lüdtke et al, 2020b;Q. Zhang & Wang, 2017), regression models with discrete metrics and nonnormal continuous outcomes (Asparouhov & Muthén, 2021b; M. C. Lee & Mitra, 2016;Lüdtke et al, 2020b), latent variable models (Keller & Enders, 2021; S.-Y. Lee & Shi, 2000;Lüdtke et al, 2020b;Merkle & Rosseel, 2018;Palomo et al, 2007), models for missing not at random processes (Du, Enders, et al, 2022), auxiliary variable models (M. J.…”

Section: Factored Regression Specificationsmentioning

confidence: 99%

“…The variation for multicategorical nominal variables instead uses category-specific latent variables that are ultimately expressed as a set of latent difference scores, much like a dummy coding scheme (Aitchison & Bennett, 1970;Carpenter & Kenward, 2013;Enders, 2022;Goldstein et al, 2009). Recent work has extended the latent response framework to count and other types of outcomes (Asparouhov & Muthén, 2021b;Demirtas, 2017;Polson et al, 2013).…”

Section: Categorical Variablesmentioning

confidence: 99%

Section: Bayesian Estimationmentioning

confidence: 99%

See 2 more Smart Citations

Missing data: An update on the state of the art.

Enders¹

2023

Psychological Methods

115

102

View full text Add to dashboard Cite

The year 2022 is the 20th anniversary of Joseph Schafer and John Graham's paper titled "Missing data: Our view of the state of the art," currently the most highly cited paper in the history of Psychological Methods. Much has changed since 2002, as missing data methodologies have continually evolved and improved; the range of applications that are possible with modern missing data techniques has increased dramatically, and software options are light years ahead of where they were. This article provides an update on the state of the art that catalogs important innovations from the past two decades of missing data research. The paper addresses topics described in the original paper, including developments related to missing data theory, full information maximum likelihood, Bayesian estimation, multiple imputation, and models for missing not at random processes. The paper also describes newer factored regression specifications and missing data handling for multilevel models, both of which have been a focus of recent research. The paper concludes with a summary of the current software landscape and a discussion of several practical issues.

show abstract

Section: Factored Regression Specificationsmentioning

confidence: 99%

Section: Categorical Variablesmentioning

confidence: 99%

See 1 more Smart Citation

Missing data: An update on the state of the art.

Enders¹

2023

Psychological Methods

115

102

View full text Add to dashboard Cite

show abstract

“…Latent Profile Analysis was used to identify the subgroups of mental wellbeing status, with MPLUS version 8.5 (Asparouhov and Muthén, 2021). Five items of the MHI-5 scale as a continuous variable were selected to perform LPA (Rhemtulla et al, 2012).…”

Section: Discussionmentioning

confidence: 99%

Mental wellbeing among urban young adults in a developing country: A Latent Profile Analysis

Nguyen

Dam

et al. 2022

Front. Psychol.

View full text Add to dashboard Cite

IntroductionThis study aimed to explore the mental wellbeing profiles and their related factors among urban young adults in Vietnam.MethodsA cross-sectional study was conducted in Hanoi, which is the capital of Vietnam. There were 356 Vietnamese who completed the Mental Health Inventory-5 (MHI-5) questionnaire. The Latent Profile Analysis (LPA) was used to identify the subgroups of mental wellbeing through five items of the MHI-5 scale as the continuous variable. Multinomial logistic regression was used to determine factors related to subgroups.ResultsThree classes represented three levels of MHI-5 score, which included “Poor mental health,” “Fair mental health,” and “Good mental health,” were, respectively, 14.3, 46.6, and 39.0%. Compared to a low household economy, participants with an average household economy had 2.11 and 4.79 times higher odds of being in a good mental health class relative to fair and poor mental health classes. Respondents with more than two acute symptoms had 3.85 times higher odds of being in a good mental health class relative to a poor mental health class, as compared to those without acute symptoms. Regarding the measurement of the Perceived Social Support Scale (MSPSS), people having support from their family had 1.80 and 2.23 times higher odds of being in classes of fair and good mental health relative to the poor mental health class; and participants having friend support also had 1.87 times higher odds of being in a good mental health class compared with the fair mental health class, as the MSPSS score increased by 1 unit. People with Rosenberg’s self-esteem scale increased by 1 score, those who had 1.17, 1.26, and 1.47 times higher odds of being in a good compared to fair mental health class, fair compared to poor mental health class, and good compared to poor mental health class, respectively.ConclusionOur findings were given to promote a new classification method for mental health screening among the general population. The current findings could be used as evidence to develop policies and plans that focus on encouraging early screening for mental health problems among the general young population in the future.

show abstract

“…Finally, while we illustrate using a probit model to model the binary response, the factored regression specification does allow the use of a logistic model for f (X). As of Blimp 3.0 [11], a binary logistic model can be estimated using the Pólya-Gamma specification [30,31].…”

Section: Mediation Models As Factored Regressionsmentioning

confidence: 99%

An Introduction to Factored Regression Models with Blimp

Keller

2021

Psych

View full text Add to dashboard Cite

In this paper, we provide an introduction to the factored regression framework. This modeling framework applies the rules of probability to break up or “factor” a complex joint distribution into a product of conditional regression models. Using this framework, we can easily specify the complex multivariate models that missing data modeling requires. The article provides a brief conceptual overview of factored regression and describes the functional notation used to conceptualize the models. Furthermore, we present a conceptual overview of how the models are estimated and imputations are obtained. Finally, we discuss how users can use the free software package, Blimp, to estimate the models in the context of a mediation example.

show abstract

Expanding the Bayesian structural equation, multilevel and mixture models to logit, negative-binomial, and nominal variables

Cited by 15 publications

References 19 publications

Missing data: An update on the state of the art.

Missing data: An update on the state of the art.

Mental wellbeing among urban young adults in a developing country: A Latent Profile Analysis

An Introduction to Factored Regression Models with Blimp

Contact Info

Product

Resources

About