Estimating Standardized SEM Parameters Given Nonnormal Data and Incorrect Model: Methods and Comparison
The DOI matches its Crossref record; 47 of the 50 references it cites were found in a registry or an index.
Abstract
When both model misspecifications and nonnormal data are present, it is unknown how trustworthy various point estimates, standard errors (SEs), and confidence intervals (CIs) are for standardized structural equation modeling parameters. We conducted simulations to evaluate maximum likelihood (ML), conventional robust SE estimator (MLM), Huber–White robust SE estimator (MLR), and the bootstrap (BS). We found (a) ML point estimates can sometimes be quite biased at finite sample sizes if misfit and nonnormality are serious; (b) ML and MLM generally give egregiously biased SEs and CIs regardless of the degree of misfit and nonnormality; (c) MLR and BS provide trustworthy SEs and CIs given medium misfit and nonnormality, but BS is better; and (d) given severe misfit and nonnormality, MLR tends to break down and BS begins to struggle.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Citations
Cites
R: A Language and Environment for Statistical Computing
R Core Team (2025) · doi:10.32614/r.manuals · Cited by 352,181
An Introduction to the Bootstrap
Efron, B., & Tibshirani, R. J. (1994) · doi:10.1201/9780429246593 · Cited by 40,500
lavaan : An R Package for Structural Equation Modeling
Rosseel, Y. (2012) · Journal of Statistical Software · doi:10.18637/jss.v048.i02 · Cited by 27,955 · cited 2×
“…literature, and is the SE given by the estimators MLM and MLMV in Mplus (Muthén & Muthén, 2012), and by MLM, MLMV, and MLMVS in the lavaan package (Rosseel, 2012) in R (R Core Team, 2017).1 However, as we review in the next section and Yuan and Hayashi (2006) explained in detail, this type of SE…”
MethodAn Introduction to the Bootstrap
Efron, B., & Tibshirani, R. J. (1993) · doi:10.1007/978-1-4899-4541-9 · Cited by 21,230
Structural Equations with Latent Variables.
Clogg, C. C., & Bollen, K. A. (1991) · Contemporary Sociology · doi:10.2307/2072165 · Cited by 17,442
Structural Equations with Latent Variables
Bollen, K. A. (1989) · Wiley series in probability and statistics · doi:10.1002/9781118619179 · Cited by 17,277 · cited 1×
“W (e.g., factor loading, structural coefficient) can be standardized as bUW ¼ βUW ðσU=σW Þ, where σU and σW are the modelimplied standard deviations of U and W in the unstandardized metric (e.g., Bollen, 1989, p. 349).”
MethodLikelihood Ratio Tests for Model Selection and Non-Nested Hypotheses
Vuong, Q. H. (1989) · Econometrica · doi:10.2307/1912557 · Cited by 5,929 · cited 1×
“Without assuming normality or correct model, the sampling behaviors of θ̂ can be described as (Gourieroux et al., 1984; Vuong, 1989; White, 1982): ffiffiffi n p ðθ̂ θ Þ”
Method(1967) · Project Euclid (Cornell University) · Cited by 4,720
Maximum Likelihood Estimation of Misspecified Models
White, H. (1982) · Econometrica · doi:10.2307/1912526 · Cited by 4,098
Maximum Likelihood Estimation of Misspecified Models
White, H. (1983) · Econometrica · doi:10.2307/1912004 · Cited by 3,674 · cited 3×
“Another type of robust SE, occasionally known as the Huber–White SE (due to Huber, 1967; White, 1982), has a more general form and is asymptotically robust to both nonnormality and model misfit.”
BackgroundMethod
Research works printed but not linked: 7
- Bentler P. M., 1985, Efficient estimation via linearization in structural models
- 10.1111/bmsp.1984.37.issue-2 · doi:10.1111/bmsp.1984.37.issue-2
- 10.1111/bmsp.1991.44.issue-2 · doi:10.1111/bmsp.1991.44.issue-2
- 10.1037/0033-2909.105.2.317 · doi:10.1037/0033-2909.105.2.317 · Paper page
- Muthén L. K., 2012, Mplus 7 user’s guide
- Nelsen R. B., 2006, Introduction to copulas
- 10.1111/bmsp.1984.37.issue-1 · doi:10.1111/bmsp.1984.37.issue-1
Cited by
Appropriate Reliance on AI Advice: Conceptualization and the Effect of Explanations
Schemmer, M., Kuehl, N., Benz, C., Bartos, A., & Satzger, G. (2023) · Proceedings of the 28th International Conference on Intelligent User Interfaces · doi:10.1145/3581641.3584066 · Cited by 208
“Therefore, we use the MLR estimator [39].”
Simione, L., Vagni, M., Gnagnarella, C., Bersani, G., & Pajardi, D. (2021) · Frontiers in Psychology · doi:10.3389/fpsyg.2021.683684 · Cited by 161
“Parameters of both regressions and SEM were estimated by means of bootstrapping over 1,000 samples, because it is considered the best method to make a model fitting robust to non-normal data (Mooney and Duval, 1993; Lai, 2018).”
MethodIslam, T., Ahmad, S., & Ahmed, I. (2022) · Review of Managerial Science · doi:10.1007/s11846-022-00547-3 · Cited by 162
“2020, p. 289) and estimate model fit values (Lai 2018) of data that is free of missing values (Byrne 2010), outliers (Kline 2005) and have normal distributions (Finney and DiStefano 2013).”
BackgroundMethodResultInfluentialMaas, J., Schoch, S., Scholz, U., Rackow, P., Schüler, J., Wegner, M., & Keller, R. (2021) · Social Psychology of Education · doi:10.1007/s11218-020-09605-8 · Cited by 158
“Multivariate normality was not affirmed in our data, therefore we estimated the model with robust maximum likelihood (MLR; Lai 2018).”
MethodSavalei, V., & Rosseel, Y. (2021) · Structural Equation Modeling A Multidisciplinary Journal · doi:10.1080/10705511.2021.1877548 · Cited by 116
Metadata
- Authors
- Lai, K.
- Affiliations
- University of California, Merced (US)
- Corresponding author
- Lai, Keke
- ORCID
- Lai, Keke
- Venue
- Structural Equation Modeling: A Multidisciplinary Journal
- Publisher
- Informa UK Limited
- Published
- 2018-01-24
- Volume
- 25
- Issue
- 4
- Pages
- 600-620
- ISSN
- 1070-5511 (ISSN-L)
- 1532-8007
- OpenAlex
- W2793883315
- Cited by
- 190
- References
- 50
- Citations by year
- 2026: 30 · 2025: 24 · 2024: 37 · 2023: 33 · 2022: 23 · 2021: 25 · 2020: 12 · 2019: 4 · 2018: 2
- FWCI
- 8.57
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Structural Equation Modeling A Multidisciplinary Journal (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
- model misspecification, non-normal data, bootstrap standard errors, maximum likelihood estimation, confidence intervals, estimation bias, finite sample size
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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Estimating Standardized SEM Parameters Given Nonnormal Data and Incorrect Model: Methods and Comparison. https://app.cochat.ai/papers/10.1080/10705511.2017.1392248
- The paper
- Lai, K. (2018). Estimating Standardized SEM Parameters Given Nonnormal Data and Incorrect Model: Methods and Comparison. Structural Equation Modeling: A Multidisciplinary Journal. https://doi.org/10.1080/10705511.2017.1392248
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.