Bayesian structural equation modeling: A more flexible representation of substantive theory
The DOI matches its Crossref record; 43 of the 43 references it cites were found in a registry or an index.
Abstract
This article proposes a new approach to factor analysis and structural equation modeling using Bayesian analysis. The new approach replaces parameter specifications of exact zeros with approximate zeros based on informative, small-variance priors. It is argued that this produces an analysis that better reflects substantive theories. The proposed Bayesian approach is particularly beneficial in applications where parameters are added to a conventional model such that a nonidentified model is obtained if maximum-likelihood estimation is applied. This approach is useful for measurement aspects of latent variable modeling, such as with confirmatory factor analysis, and the measurement part of structural equation modeling. Two application areas are studied, cross-loadings and residual correlations in confirmatory factor analysis. An example using a full structural equation model is also presented, showing an efficient way to find model misspecification. The approach encompasses 3 elements: model testing using posterior predictive checking, model estimation, and model modification. Monte Carlo simulations and real data are analyzed using Mplus. The real-data analyses use data from Holzinger and Swineford's (1939) classic mental abilities study, Big Five personality factor data from a British survey, and science achievement data from the National Educational Longitudinal Study of 1988.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“This traditional approach depends on the unrealistic assumptions of zero cross-loadings [30] and limits the researchers’ ability to investigate cross-loadings and residual covariance parameters [31].”
Citations
Cites
Gelman, A., Carlin, J. B., Stern, H. S., & Rubin, D. B. (2003) · doi:10.1201/9780429258480 · Cited by 27,276
Alternative Ways of Assessing Model Fit
Browne, M. W., & Cudeck, R. (1992) · Sociological Methods & Research · doi:10.1177/0049124192021002005 · Cited by 27,200 · cited 1×
“BSEM with its adjoining posterior predictive p value model test is similar in spirit to the frequentist conceptualization of close fit (M. W. Browne & Cudeck, 1993).”
ResultInference from iterative simulation using multiple sequences
Gelman, A., & Rubin, D. B. (1992) · Statistical Science · doi:10.1214/ss/1177011136 · Cited by 16,627 · cited 1×
“(A8) (Appendix continues) Convergence is determined using the Gelman–Rubin convergence diagnostic (Gelman & Rubin, 1992; Gelman et al., 2004).”
MethodTesting Structural Equation Models.
Clogg, C. C., Bollen, K. A., & Long, J. S. (1995) · Social Forces · doi:10.2307/2580595 · Cited by 8,211
Markov Chain Monte Carlo in Practice
(1995) · Chapman & Hall/CRC eBooks · doi:10.1201/b14835 · Cited by 7,757
(2010) · Cited by 7,034
Analysis of Incomplete Multivariate Data
Booth, D. E. (2000) · Technometrics · doi:10.1080/00401706.2000.10486013 · Cited by 5,627 · cited 1×
“Schafer (1997) gave a statistical discussion from a missing data and multiple imputation perspective, whereas Enders (2010) gave an applied discussion of these same topics.”
BackgroundIntroduction to Statistical Mediation Analysis
MacKinnon, D. P. (2012) · doi:10.4324/9780203809556 · Cited by 5,618 · cited 1×
“An example is an indirect effect a b in a mediation model (MacKinnon, 2008).”
BackgroundHarman, H. H. (1960) · Cited by 5,148
Exploratory Structural Equation Modeling
Asparouhov, T., & Muthén, B. (2009) · Structural Equation Modeling: A Multidisciplinary Journal · doi:10.1080/10705510903008204 · Cited by 2,665
Self-citationPOSTERIOR PREDICTIVE ASSESSMENT OF MODEL FITNESS VIA REALIZED DISCREPANCIES
Gelman, A., Meng, X., & Stern, H. S. (1996) · Cited by 2,639
Casella, G., & George, E. I. (1992) · The American Statistician · doi:10.1080/00031305.1992.10475878 · Cited by 2,364 · cited 1×
“Statistical overview articles include Gelfand, Hills, Racine-Poon, and Smith (1990) and Casella and George (1992).”
BackgroundStructural Equation Modeling: Foundations and Extensions
Kaplan, D. E. (2000) · University of Maribor digital library (University of Maribor) · Cited by 1,894
Bayesian Data Analysis, 2nd edn
Brix, A. (2005) · Journal of the Royal Statistical Society Series A: Statistics in Society · doi:10.1111/j.1467-985x.2004.00347_4.x · Cited by 1,767
Model modifications in covariance structure analysis: The problem of capitalization on chance.
MacCallum, R. C., Roznowski, M., & Necowitz, L. B. (1992) · Psychological Bulletin · doi:10.1037//0033-2909.111.3.490 · Cited by 1,632 · cited 1×
“Modification indices pertain to freeing only one parameter at a time, and a long sequence of model modification is often needed, running the risk of capitalizing on chance (see, e.g., MacCallum et al., 1992).”
BackgroundAn Overview of Analytic Rotation in Exploratory Factor Analysis
Browne, M. W. (2001) · Multivariate Behavioral Research · doi:10.1207/s15327906mbr3601_05 · Cited by 1,140 · cited 10×
“Bayes clearly gives a simpler pattern than ML EFA for these data. This is achieved by shrinking the cross-loadings toward their zero prior means. The degree of shrinking that is possible while still obtaining reasonable model fit is gauged by the posterior predictive p value. It should be emphasized that cross-loading…”
BackgroundMethodInfluentialMarsh, H. W., Muthén, B., Asparouhov, T., Lüdtke, O., Robitzsch, A., Morin, A. J. S., & Trautwein, U. (2009) · Structural Equation Modeling: A Multidisciplinary Journal · doi:10.1080/10705510903008220 · Cited by 1,093
Self-citationLatent Variable Modeling in Heterogeneous Populations
Muthén, B. O. (1989) · Psychometrika · doi:10.1007/bf02296397 · Cited by 1,008 · cited 1×
“The special case of intercept noninvariance can be handled by letting the grouping variables be covariates that influence the factors, also referred to as MIMIC (multiple-indicators, multiple-causes) modeling (see, e.g., Muthén, 1989).”
BackgroundSelf-citationWinBUGS user manual version 1.4
Spiegelhalter, D. J. (2003) · Medical Entomology and Zoology · Cited by 983
Illustration of Bayesian Inference in Normal Data Models Using Gibbs Sampling
Gelfand, A. E., Hills, S. E., Racine-Poon, A., & Smith, A. F. M. (1990) · Journal of the American Statistical Association · doi:10.1080/01621459.1990.10474968 · Cited by 969
The Role of Constructs in Psychological and Educational Measurement
(2001) · Routledge eBooks · doi:10.4324/9781410607454 · Cited by 871
Yuan, Y., & MacKinnon, D. P. (2009) · Psychological Methods · doi:10.1037/a0016972 · Cited by 637 · cited 1×
“Overview articles of an applied nature and with a latent variable focus include Scheines, Hoijtink, and Boomsma (1999); Rupp, Dey, and Zumbo (2004); Yuan and MacKinnon (2009); and Kaplan and Depaoli (in press).”
BackgroundA comparison of Bayesian and likelihood-based methods for fitting multilevel models
Browne, W. J., & Draper, D. (2006) · Bayesian Analysis · doi:10.1214/06-ba117 · Cited by 603
Doing Bayesian Data Analysis: A Tutorial with R and BUGS
Barry, J. (2011) · Europe’s Journal of Psychology · doi:10.5964/ejop.v7i4.163 · Cited by 563 · cited 1×
“Gelman, Carlin, Stern, and Rubin (2004) provided a good general statistical description, whereas Kruschke (2010) and Lynch (2010) gave somewhat more introductory accounts.”
BackgroundIntroduction to Applied Bayesian Statistics and Estimation for Social Scientists
(2007) · Springer eBooks · doi:10.1007/978-0-387-71265-9 · Cited by 555
Structural equation modeling: a Bayesian approach
(2007) · Choice Reviews Online · doi:10.5860/choice.45-1510 · Cited by 441
Marsh, H. W., Nagengast, B., & Morin, A. J. S. (2013) · Developmental Psychology · doi:10.1037/a0026913 · Cited by 422
Testing differences between nested covariance structure models: Power analysis and null hypotheses.
MacCallum, R. C., Browne, M. W., & Cai, L. (2006) · Psychological Methods · doi:10.1037/1082-989x.11.1.19 · Cited by 385 · cited 1×
“In assessing differences between models, MacCallum, Browne, and Cai (2006) also argued against exact fit as being of limited empirical interest given that it is never true in practice.”
MethodBayesian Estimation and Testing of Structural Equation Models
Scheines, R., Hoijtink, H., & Boomsma, A. (1999) · Psychometrika · doi:10.1007/bf02294318 · Cited by 272
Subjective and Objective Bayesian Statistics: Principles, Models, and Applications
Press, S. J. (2002) · Wiley Series in Probability and Statistics · doi:10.1002/9780470317105 · Cited by 248
Cudeck, R. A., & O'Dell, L. L. (1994) · Psychological Bulletin · doi:10.1037//0033-2909.115.3.475 · Cited by 246 · cited 1×
“A cross-loading of 0.1 is considered to be of little importance, a cross-loading of 0.2 is considered to be of some importance, and a cross-loading of 0.3 is considered to be of importance (see also Cudeck & O’Dell, 1994).”
BackgroundInfluentialSubjective and objective Bayesian statistics : principles, models, and applications
Press, S. J. (2003) · Wiley-Interscience eBooks · Cited by 166
To Bayes or Not to Bayes, From Whether to When: Applications of Bayesian Methodology to Modeling
Rupp, A. A., Dey, D. K., & Zumbo, B. D. (2004) · Structural Equation Modeling A Multidisciplinary Journal · doi:10.1207/s15328007sem1103_7 · Cited by 148 · cited 1×
“Overview articles of an applied nature and with a latent variable focus include Scheines, Hoijtink, and Boomsma (1999); Rupp, Dey, and Zumbo (2004); Yuan and MacKinnon (2009); and Kaplan and Depaoli (in press).”
BackgroundPost-Processing Posterior Predictive p Values
Hjort, N. L., Dahl, F. A., & Steinbakk, G. H. (2006) · Journal of the American Statistical Association · doi:10.1198/016214505000001393 · Cited by 128
Measurement From a Hierarchical Point of View
Gustafsson, J. (2001) · Routledge eBooks · doi:10.4324/9781410607454-12 · Cited by 76 · cited 1×
“Factor analyses of these data have been described, for example, by Harman (1976, pp. 123–132) and Gustafsson (2002).”
MethodA random change point model for assessing variability in repeated measures of cognitive function
Dominicus, A., Ripatti, S., Pedersen, N. L., & Palmgren, J. (2008) · Statistics in Medicine · doi:10.1002/sim.3380 · Cited by 57
Muthén, B., Muthén, L. K., & Asparouhov, T. (2010) · Cited by 56
Self-citationIndicators for Monitoring Mathematics and Science Education: A Sourcebook
Shavelson, R. J., McDonnell, L. M., & Oakes, J. (1989) · RAND Corporation eBooks · doi:10.7249/r3742 · Cited by 48
Subjective and Objective Bayesian Statistics: Principles, Models, and Applications
Short, T. H. (2005) · Journal of the American Statistical Association · doi:10.1198/jasa.2005.s11 · Cited by 48
Gilks, W. R., Richardson, S. T., & Spiegelhalter, D. J. (1995) · Chapman & Hall/CRC eBooks · doi:10.1201/b14835-26 · Cited by 31 · cited 1×
“Examples of the former include image analysis (see, e.g., Green, 1996), and examples of the latter include random change-point analysis (see, e.g., Dominicus, Ripatti, Pedersen, & Palmgren, 2008).”
MethodTEACHER'S CORNER: Examining Identification Issues in Factor Analysis
Hayashi, K., & Marcoulides, G. A. (2006) · Structural Equation Modeling A Multidisciplinary Journal · doi:10.1207/s15328007sem1304_7 · Cited by 29 · cited 1×
“In frequentist analysis, freeing all cross-loadings in a CFA model such as Table 1 leads to a nonidentified model because the m2 restrictions, where m is the number of factors, necessary to eliminate indeterminacies are not present (see, e.g., Hayashi & Marcoulides, 2006).”
BackgroundBayesian Analysis of Multivariate Probit ModelsPreprint
Chib, S., & Greenberg, E. (1996) · RePEc: Research Papers in Economics · Cited by 27
Markov chain Monte Carlo in Practice
Green, P. J. (1996) · Cited by 3
Cited by
El análisis factorial exploratorio de los ítems: una guía práctica, revisada y actualizada
Lloret-Segura, S., Ferreres-Traver, A., Hernández-Baeza, A., & Tomás-Marco, I. (2014) · Anales de Psicología · doi:10.6018/analesps.30.3.199361 · Cited by 1,616
Marsh, H. W., Morin, A. J. S., Parker, P. D., & Kaur, G. (2014) · Annual Review of Clinical Psychology · doi:10.1146/annurev-clinpsy-032813-153700 · Cited by 1,801
“We view the two approaches as complementary: Increasing knowledge based on ESEM provides a basis for specifying priors in BSEM, but additional research that juxtaposes these approaches is needed (for further discussion, see Muthén & Asparouhov 2012).”
MethodBayesian statistics and modelling
van de Schoot, R., Depaoli, S., King, R., Kramer, B., Märtens, K., Tadesse, M. G., Vannucci, M., Gelman, A., Veen, D., Willemsen, J., et al. (2021) · Nature Reviews Methods Primers · doi:10.1038/s43586-020-00001-2 · Cited by 1,176
MplusAutomation: An R Package for Facilitating Large-Scale Latent Variable Analyses in M plus
Hallquist, M. N., & Wiley, J. F. (2018) · Structural Equation Modeling: A Multidisciplinary Journal · doi:10.1080/10705511.2017.1402334 · Cited by 871
“The creators of M plus regularly make substantive contributions to the SEM literature (e.g., Asparouhov & Muthén, 2014; Muthén & Asparouhov, 2012) and consistently implement these new methods in M plus .”
MethodA Gentle Introduction to Bayesian Analysis: Applications to Developmental Research
van de Schoot, R., Kaplan, D., Denissen, J., Asendorpf, J. B., Neyer, F. J., & van Aken, M. A. G. (2014) · Child Development · doi:10.1111/cdev.12169 · Cited by 779
Metadata
- Authors
- Muthén, B., & Asparouhov, T.
- Affiliations
- Muthén & Muthén (United States) (US)
- Venue
- Psychological Methods
- Published
- 2012-09
- Volume
- 17
- Issue
- 3
- Pages
- 313-335
- ISSN
- 1082-989X (ISSN-L)
- 1939-1463
- OpenAlex
- W2157335584
- Cited by
- 1,562
- References
- 43
- Citations by year
- 2026: 148 · 2025: 140 · 2024: 136 · 2023: 118 · 2022: 135 · 2021: 169 · 2020: 124 · 2019: 127 · 2018: 103 · 2017: 99 · 2016: 74 · 2015: 77 · 2014: 60 · 2013: 32 · 2012: 12
- FWCI
- 76.81
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Psychological Methods (publishedVersion)
- PubMed (publishedVersion)
- Topics
- Social Sciences › Decision Sciences › Management Science and Operations Research › Psychometric Methodologies and Testing
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Bayesian Inference
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Distribution Estimation and Applications
- Keywords
- Bayesian structural equation modeling, confirmatory factor analysis, structural equation modeling, system identification, cross-loadings, residual correlation, model misspecification, posterior predictive checking, model modification, latent variable measurement, Big Five personality traits, science achievement
- MeSH terms
- Bayes Theorem, Factor Analysis, Statistical, Humans, Markov Chains, Monte Carlo Method, Neuropsychological Tests, Personality Tests, Models, Statistical, Likelihood Functions
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.