On Structural Equation Modeling with Data that are not Missing Completely at Random
The DOI matches its Crossref record; 44 of the 45 references it cites were found in a registry or an index.
Abstract
A general latent variable model is given which includes the specification of a missing data mechanism. This framework allows for an elucidating discussion of existing general multivariate theory bearing on maximum likelihood estimation with missing data. Here, missing completely at random is not a prerequisite for unbiased estimation in large samples, as when using the traditional listwise or pairwise present data approaches. The theory is connected with old and new results in the area of selection and factorial invariance. It is pointed out that in many applications, maximum likelihood estimation with missing data may be carried out by existing structural equation modeling software, such as LISREL and LISCOMP. Several sets of artifical data are generated within the general model framework. The proposed estimator is compared to the two traditional ones and found superior.
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
“The mean values of the skewness and kurtosis were smaller than the prescribed levels (skewness, 2.0 and kurtosis, 7.0), indicating no significant problems regarding multivariate normality of the data [35].”
From The Mediating Role of Generative AI Self-Regulation on Students’ Critical Thinking and Problem-Solving (2024) · cites this paper as [35]
Citations
Cites
Maximum Likelihood from Incomplete Data Via the EM Algorithm
Dempster, A. P., Laird, N. M., & Rubin, D. B. (1977) · Journal of the Royal Statistical Society Series B: Statistical Methodology · doi:10.1111/j.2517-6161.1977.tb01600.x · Cited by 50,444
Statistical Analysis With Missing Data
Lahiff, M., Little, R. J. A., & Rubin, D. B. (1989) · Journal of the American Statistical Association · doi:10.2307/2289883 · Cited by 17,265
RUBIN, D. B. (1976) · Biometrika · doi:10.1093/biomet/63.3.581 · Cited by 9,925
Journal of The American Statistical Association
Fienberg, S. E. (2005) · Encyclopedia of Statistical Sciences · doi:10.1002/0471667196.ess1317 · Cited by 6,125
BMDP Statistical Software 1981.
Bubolz, T. A., & Dixon, W. J. (1984) · Journal of the American Statistical Association · doi:10.2307/2288308 · Cited by 5,221
(1976) · Annals of Economic and Social Measurement · Cited by 5,035
Assessing Reliability and Stability in Panel Models
Wheaton, B., Muthen, B., Alwin, D. F., & Summers, G. F. (1977) · Sociological Methodology · doi:10.2307/270754 · Cited by 3,116
Self-citationA General Approach to Confirmatory Maximum Likelihood Factor Analysis
Jöreskog, K. G. (1969) · Psychometrika · doi:10.1007/bf02289343 · Cited by 2,489
Marginal Maximum Likelihood Estimation of Item Parameters: Application of an EM Algorithm
Bock, R. D., & Aitkin, M. (1981) · Psychometrika · doi:10.1007/bf02293801 · Cited by 2,339
Quasi-likelihood functions, generalized linear models, and the Gauss—Newton method
WEDDERBURN, R. W. M. (1974) · Biometrika · doi:10.1093/biomet/61.3.439 · Cited by 2,122 · cited 2×
“In terms of the Wheaton et al. (1977) example, an increased value of any of the socioeconomic background indicators is taken to yield an increased propensity not to attrit (a high s* value), while an increased value of the time one alienation indicators is taken to yield a decreased propensity not to attrit (i.”
BackgroundMethod
Research works printed but not linked: 5
- Little, 1983, Incomplete data in sample surveys, Vol. 2: Theory and bibliographies
- Dixon, 1983, BMDP Statistical Software
- Olsson, 1978, Selection bias in confirmatory factor analysis
- Marini, 1980, Sociological Methodology
- Lawley, 1943, A note on Karl Pearson's selection formulae, Proceedings of the Royal Society Edinburgh, Section A (Mathematics and Physics Section), 62, 28
Cited by
Items 801–881 of 881
Investigating the Role of Appearance-Based Factors in Predicting Sunbathing and
Use, T. S., & Cafri, G. (2008)
Gender Differences in Broad and Narrow Ability Dimensions
Rosén, M. (2016) · Methodology of educational measurement and assessment · doi:10.1007/978-3-319-43473-5_4
Hall, B. J. (2011) · OhioLink ETD Center (Ohio Library and Information Network)
Yoo, J. E., French, B. F., & Maller, S. J. (2007)
The Impact of the use of internet-based technologies on new product development performance
de Grosbois, D. (2007) · Carleton University's Institutional Repository (MacOdrum Library, Carleton University) · doi:10.22215/etd/2007-06555
Item Parameter Drift in a Time-Varying Predictor
Lee, H. (2017) · Applied Measurement in Education · doi:10.1080/08957347.2017.1391259
“The criterion values were based on Muthén, Kaplan, and Hollis (1987) in which relative bias less than ±.10 to ±.15 might be considered negligible.”
MethodHoang, T.-M. H. (2017) · Illinois Digital Environment for Access to Learning and Scholarship (University of Illinois at Urbana-Champaign)
A Probability Based Framework for Testing the Missing Data Mechanism
Lin, J. C.-H. (2013) · eScholarship (California Digital Library)
Neilands, T. B., Binson’s, D., & Woods, B. (2014)
Liao, K. Y. (2008) · Iowa State University Digital Repository (Iowa State University) · doi:10.31274/rtd-180813-16567 · Cited by 1
“Muthén, Kaplan, and Hollis (1987) and Wothke (2000) indicated that this 38 method is efficient and produces parameter estimates that are less biased than the previous methods of managing missing data such as the pairwise or listwise deletion of missing cases.”
MethodYalçın, A., & Ülbeği, İ. D. (2022) · Akademisyen Kitabevi eBooks · doi:10.37609/akya.2006
Examining Nonnormal Latent Variable Distributions for Non-Ignorable Missing Data
Liu, C. (2021) · Applied Psychological Measurement · doi:10.1177/0146621621990753
“Several kinds of MNAR modeling have been studied such as confirmatory factor analysis (Muthén et al., 1987), structural equation modeling (Bacci & Bartolucci, 2015), latent class models (List et al., 2019), multi-group models (Rose et al., 2015), or item response theory (IRT; Holman & Glas, 2005).”
MethodDesign, measurement, and analysis considerations and evaluations in intensive longitudinal method
Lim, J. X. (2021) · DR-NTU (Nanyang Technological University) · doi:10.32657/10356/146267
مقارنة معاملات ثبات درجات الاختبار في ظل مجموعة من الاشتراطات: دراسة محاکاة مونت کارلو
زارع, ن. م. س. (2021) · المجلة التربوية لکلية التربية بسوهاج · doi:10.21608/edusohag.2021.180584
Parent-Taillon, É., Bouffard, T., & Maltais, C. (2021) · Revue québécoise de psychologie · doi:10.7202/1078918ar
Lee, S., & Whittaker, T. A. (2021) · Structural Equation Modeling A Multidisciplinary Journal · doi:10.1080/10705511.2021.1956320
Avcu, A. (2021) · Trakya Eğitim Dergisi · doi:10.24315/tred.789832
(2014) · Wiley eBooks · doi:10.1002/9781118634646.refs
General Analytic Considerations
Wright, A. G. C., & Hallquist, M. N. (2020) · Cambridge University Press eBooks · doi:10.1017/9781316995808.034
O'Connor, S. E., Blais, C., Mésidor, M., Talbot, D. C., Poirier, P., & Leclerc, J. (2022) · SSRN Electronic Journal · doi:10.2139/ssrn.4052643
Editors, J. (2008) · Journal of Modern Applied Statistical Methods · doi:10.22237/jmasm/1209616320
(2015) · Elsevier eBooks · doi:10.1016/b978-0-12-801342-7.00019-8
Rosén, M., Hansen, K. Y., & Wolff, U. (2016) · Methodology of educational measurement and assessment · doi:10.1007/978-3-319-43473-5_1
Shadish, W. R., & Luellen, J. K. (2005) · Wiley StatsRef: Statistics Reference Online · doi:10.1002/9781118445112.stat06672
Structural Equation Modeling: Nonstandard Cases
Marcoulides, G. A. (2014) · Wiley StatsRef: Statistics Reference Online · doi:10.1002/9781118445112.stat06477
Shadish, W. R., & Luellen, J. K. (2005) · Encyclopedia of Statistics in Behavioral Science · doi:10.1002/0470013192.bsa031
Sintomas depressivos em escolares na cidade de São Paulo: um estudo de análise fatorial
Maia, A. P. F. (2022) · Digital Library of Theses and Dissertations (Universidade de São Paulo) · doi:10.11606/t.5.2022.tde-17042023-143133
Ponce, F. P., Gómez, F., Arias, V. B., & Cumsille, P. E. (2023) · PsyArXiv (OSF Preprints) · doi:10.31234/osf.io/9sbw7 · Cited by 1
McAdam, C. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4610024
Sun, R., Abouarghoub, W., & Demir, E. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4641164
A metabolic trade-off impacting childhood language development and body growthPreprint
Bouton, S., Chevallier, C., Cissé, A. H., Heude, B., & Jacquet, P. O. (2023) · doi:10.31234/osf.io/p2kge
German translation and validation of the Interprofessional Facilitation Scale
Kerry, M. J., Paignon, A., Conti, J. W., Sy, M., & Huber, M. (2023) · Journal of Interprofessional Care · doi:10.1080/13561820.2023.2287024
“The average-relative parameter bias (.01) value indicates that the impact of ignoring the multidimension-ality by using unidimensional IPFS scores was negligible (Muthén et al., 1987).”
BackgroundOh, V. Y. S. (2024) · International Journal of Psychology · doi:10.1002/ijop.13218
“FIML provides the gold standard approach to addressing missing data even under conditions where the data is MNAR, which is typical of large-scale public datasets such as this (Enders & Bandalos, 2001; Muthén et al., 1987).”
Grady, K. L., Burns, J. L., Allen, L. A., Stehlik, J., Teuteberg, J. J., McIlvennan, C. K., Kirklin, J. K., Beiser, D. G., Lindenfeld, J. M., Denfeld, Q. E., et al. (2024) · The Journal of Cardiovascular Nursing · doi:10.1097/jcn.0000000000001129
Ajayi, O., & Fakunmoju, S. B. (2024) · Technium Business and Management · doi:10.47577/business.v9i.11772 · Cited by 1
Jeong, J., & Sacco, P. (2024) · Addictive Behaviors Reports · doi:10.1016/j.abrep.2024.100570
Heritage Languages and Linguistic Diversity: Psycholinguistic and Sociolinguistic Perspectives
Muthyala, U. (2024) · Springer eBooks · doi:10.1007/978-981-97-9354-9_3
Sunderland, M., Halladay, J. E., & Chapman, C. (2024) · Journal of Psychopathology and Behavioral Assessment · doi:10.1007/s10862-024-10181-w
“Values greater than 0.7 on the ECV and less than 0.1 on ARPB indicate a strong general factor and the suggest that factor loadings estimated in a unidi-mensional model may approximate well to loadings estimated when considering multidimensionality (Rodriguez et al., 2016b; Muthen et al., 1987).”
InfluentialFakunmoju, S. B. (2025) · International Journal of Social Science Research and Review · doi:10.47814/ijssrr.v8i1.2484 · Cited by 1
Mou, N. L., Dufour, G. K., & Jackson, D. L. (2025) · Journal of Sexual Aggression · doi:10.1080/13552600.2025.2471770 · Cited by 1
“Such a stringent assumption is unlikely to be met in real world data (Muthén et al., 1987), particularly in research on sensitive topics such as sexual violence (Giroux et al., 2020; Salazar et al., 2014; Testa & Cleveland, 2017) and substance use (Messiah et al., 2011; Rendina et al., 2016).”
Accuracy in Invariance Detection With Multilevel Models With Three Estimators
Finch, W. H., Demir, C., French, B. F., & Vo, T. T. (2025) · Applied Psychological Measurement · doi:10.1177/01466216251325644
“For statistically signi fi cant effects, bias was used to judge parameter recovery accuracy, with values less than 0.10% – 0.15% considered small (M´uthen et al., 1987).”
InfluentialLee, S., & Whittaker, T. A. (2025) · Multivariate Behavioral Research · doi:10.1080/00273171.2025.2478071
“Acceptable RB falls within ±0.10, with values close to zero indicating better accuracy (Muth � en et al., 1987).”
InfluentialSkaalvik, E. M., & Skaalvik, S. (2025) · Creative Education · doi:10.4236/ce.2025.164034 · Cited by 2
Oh, V. Y. S., Hartanto, A., Ho, M. R., & Tong, E. M. W. (2025) · Personality and Social Psychology Bulletin · doi:10.1177/01461672251337120
“…of previous research, we utilized full-information maximum likelihood (FIML) procedures to address missing data (Enders & Bandalos, 2001; Muthén et al., 1987)—this approach provides the gold standard method for addressing missing data which outperforms alternative approaches even under…”
Germo, G. R., Shane, J., & Farruggia, S. P. (2025) · Children and Youth Services Review · doi:10.1016/j.childyouth.2025.108402
Schlochtermeier, D., Blankers, M., Peen, J., van den Berg, E. M., van Beers, E., Melisse, B., Koenders, J., Goudriaan, A. E., de Jonge, M., Dekker, J., et al. (2025) · Eating and Weight Disorders - Studies on Anorexia Bulimia and Obesity · doi:10.1007/s40519-025-01767-8
“An ARPB lower than 0.10–0.15 is acceptable [18, 28] .”
InfluentialFactors contributing to the decline in problematic mobile phone use among Chinese emerging adults
Wu, Z., Ma, M., Bian, T., Li, Y., Chen, H., & Cao, J. (2025) · Addictive Behaviors · doi:10.1016/j.addbeh.2025.108480
Regularized Cross-Sectional Network Modeling with Missing Data: A Comparison of Methods
Falk, C. F., & Starr, J. (2025) · Multivariate Behavioral Research · doi:10.1080/00273171.2025.2551373
“…violations (e.g., Arbuckle, 1996; Cai & Lee, 2009; Enders & Bandalos, 2001; Enders & Bandalos, 2001; Graham, 2003; Graham et al., 1996; S.-Y. Lee, 1986; Muth � en et al., 1987; Savalei, 2010; Savalei & Bentler, 2005; Savalei & Rosseel, 2022; Yuan, 2009; Yuan & Savalei, 2014; Yuan et al., 2015).”
Reliability and Validity of the Korean Version of the Suicide Rumination Scale (K-SRS)
Hur, W., Rogers, M. L., & Shim, E. (2025) · Korean Journal of Stress Research · doi:10.17547/kjsr.2025.33.3.163
Validation of the Parental Worries About Racial Profiling Scale Across Black and Latine Parents
Christophe, N. K., Stevenson, H. C., Kiang, L., Livas, G., Desmarais, A., Winn, D.-M. C., Jones, S. C. T., & Anderson, R. E. (2025) · Race and Social Problems · doi:10.1007/s12552-025-09465-x
“…‘acceptable’ amount of bias in parameter estimates may change depending on the measure and context in which it is used, an ARPB of less than 10–15% is commonly regarded to be satisfactory and pose little risk to altering later statistical conclusions (Muthén et al., 1987; Rodriguez et al., 2016).”
Modeling the Multidimensional Nature of Impulsivity and its Relation to Functional Outcomes
Knežević, B. (2013) · Scholarship at UWindsor (University of Windsor) · Cited by 1
Wen, F.-H., Prigerson, H. G., Chou, W.-C., Huang, C.-C., Hu, T.-H., Chiang, M. C., Chuang, L.-P., & Tang, S. T. (2022) · Figshare · doi:10.6084/m9.figshare.21456154
Sachmpazidi, D., Verostek, M., Turpen, C., & Petrella, J. (2025) · Physical Review Physics Education Research · doi:10.1103/mt34-y14m · Cited by 1
Li, C., Cheng, G., He, S., Xie, X., Gang, T., Jiang, N., Min, X., Shi, Y., Li, R., Zhou, T., et al. (2022) · Figshare · doi:10.6084/m9.figshare.20056558
Wen, F.-H., Prigerson, H. G., Chou, W.-C., Huang, C.-C., Hu, T.-H., Chiang, M. C., Chuang, L.-P., & Tang, S. T. (2022) · Figshare · doi:10.6084/m9.figshare.21456151
Wen, F.-H., Prigerson, H. G., Chou, W.-C., Huang, C.-C., Hu, T.-H., Chiang, M. C., Chuang, L.-P., & Tang, S. T. (2022) · Figshare · doi:10.6084/m9.figshare.21456157
Hox, J. J., van Buuren, S., & Jolani, S. (2015) · Emerald Group Publishing Limited eBooks · doi:10.1108/978-1-68123-329-120251004
Samiotis, A., Carmichael, J. J., Wong, D. K., Gould, K. R., Johnston, L., Ponsford, J. L., & Spitz, G. (2025) · Brain Impairment · doi:10.1071/ib25139
Cella, D. F., Vallow, S., Bermann, G., McDonald, J., Ngerano, G. N., Linton, S., Arenson, E., Maitra, S., Yen, G., Galipeau, N., et al. (2026) · Journal of Patient-Reported Outcomes · doi:10.1186/s41687-026-00996-4
“[15]) is 15.0% in magnitude, this is considered sufficient evidence of essential unidimensionality to support reporting the FACIT-Fatigue total score [34].”
Influentialder Nat, M. S.-V., van der Wilt, F. M., Meeter, M., & van der Veen, C. (2026) · Social Development · doi:10.1111/sode.70050 · Cited by 1
Frashëri, E., & Teneqexhi, M. (2026) · International Journal of Assessment Tools in Education · doi:10.21449/ijate.1791291
Menon, V., Balasubramanian, I., Rogers, M. L., Grover, S., Lakdawala, B. M., Ranjan, R., Sarkhel, S., Nebhinani, N., Kallivayalil, R. A., Raghavan, V., et al. (2026) · Asian Journal of Psychiatry · doi:10.1016/j.ajp.2026.104897
Clarifying perceived task help receipt in the workplace: Construct definition and scale development
Law, D., Nishioka, M., Brown, D. J., & Beck, J. W. (2026) · Applied Psychology · doi:10.1111/apps.70079 · Cited by 1
Wu, D., Zhu, Y., Qiu, Z., Wang, J., Zhang, Y., & Li, J. (2026) · BMC Pediatrics · doi:10.1186/s12887-026-06703-y
“…indicates minimal influence of multidimensionality, 0.10–0.15 is generally acceptable, and > 0.15 may indicate problematic parameter distortion [82] ω (and ω S for subscales) > 0.70 indicates acceptable reliability [83] ω H > 0.80 indicates that the total score primarily reflects the general…”
InfluentialOgresta, J., Kramarić, M., & Tarle, A. (2026) · Current Psychology · doi:10.1007/s12144-026-09517-4
“The percentage of uncontami - nated correlations (PUC) was 0.70, and the average relative parameter bias (ARPB) was 0.06, whereas the acceptable values to use a scale as unidimensional are between 0.10 and 0.15 (Muthén et al., 1987 ).”
InfluentialNoordzij, A. S. Y., Terluin, B., van der Wouden, J. C., van der Horst, H. E., Essery, R., Yardley, L., Maarsingh, O. R., & van Vugt, V. A. (2026) · Journal of Patient-Reported Outcomes · doi:10.1186/s41687-026-01109-x
“According to Muthén et al., a relative bias of no more than 10–15% may not be serious in most contexts [22].”
Kost, D., Runge, J. M., Kopperud, K. H., Nerstad, C. G. L., & Seljeseth, I. M. (2026) · Journal of Occupational and Organizational Psychology · doi:10.1111/joop.70132
A general teamwork measure’s use in interprofessional education: item response theory analysis
Kerry, M. J., Schwärzler, P., Feusi, E., & Huber, M. (2026) · Journal of Interprofessional Care · doi:10.1080/13561820.2026.2702993
Factors Impacting Treatment Decisions in Caregivers of Autistic Children
Harris, T. G., Record, J. M., & Smith, S. D. (2026) · Behavioral Sciences · doi:10.3390/bs16081280
“Given that a missing data percentage of 10% or less is considered inconsequential, no method to handle missing data for demographic or study variables was used besides pairwise deletion or WLS-based estimators for ordinal variables (Bennett, 2001; Enders & Bandalos, 2001; B. Muthén et al., 1987).”
Liang, P., Zhang, X., Qiu, Z., Zhang, R., Zhang, H., & Bian, Y. (2026) · Research on Child and Adolescent Psychopathology · doi:10.1007/s10802-026-01492-5
“This approach was considered preferable to listwise deletion, which would have reduced statistical power and may have introduced additional bias by excluding participants with partially observed longitudinal data (Muthén et al., 1987).”
Lee, S., Cobb, C. L., & Kim, S. (2026) · Multivariate Behavioral Research · doi:10.1080/00273171.2026.2699709
Klapp, T., Algurén, B., Sevdalis, V., & Klapp, A. (2026) · European Journal of Psychology of Education · doi:10.1007/s10212-026-01183-1
“With FIML, a likelihood function for each individual in the data set is estimated using observed information from the variables in the model, such as grades, gender, and parental education (Muthén et al., 1987).”
Fang, J., Jiang, N., Wei, Y., & Zhang, Y. (2026) · Frontiers in Psychology · doi:10.3389/fpsyg.2026.1847679
“FIML has been shown to perform favorably compared with alternative missing-data approaches and may yield less biased parameter estimates even when missingness deviates from complete randomness (Muthen et al., 1987).”
Shared values and social support as predictors of teachers’ motivation and well-being
Skaalvik, E. M., & Skaalvik, S. (2026) · Social Psychology of Education · doi:10.1007/s11218-026-10310-1
“Missing data were handled using maximum likelihood (ML) estimation, which provides more reliable estimates and less bias than listwise or pairwise deletion (Little & Rubin, 1989; Muthén et al., 1987).”
Latent variable modelling of personality-health associations : measures, models and extensions
Hagger‐Johnson, G. (2008) · OpenGrey (Institut de l'Information Scientifique et Technique)
Metadata
- Authors
- Muthén, B., Kaplan, D., & Hollis, M.
- Affiliations
- University of California, Los Angeles (US)
- ORCID
- Muthén, Bengt
- Venue
- Psychometrika
- Publisher
- Cambridge University Press (CUP)
- Published
- 1987-09
- Volume
- 52
- Issue
- 3
- Pages
- 431-462
- ISSN
- 0033-3123 (ISSN-L)
- 1860-0980
- OpenAlex
- W2065262588
- Cited by
- 883
- References
- 45
- Citations by year
- 2026: 22 · 2025: 30 · 2024: 27 · 2023: 37 · 2022: 53 · 2021: 37 · 2020: 35 · 2019: 37 · 2018: 28 · 2017: 31 · 2016: 29 · 2015: 36 · 2014: 39 · 2013: 39 · 2012: 33
- FWCI
- 6.04
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Psychometrika (publishedVersion)
- RePEc: Research Papers in Economics (submittedVersion)
- 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 › Applied Mathematics › Statistical and numerical algorithms
- Keywords
- structural equation modeling, missing data, missing completely at random, latent variable models, missing data mechanisms, maximum likelihood estimation, selection bias, measurement invariance, listwise deletion, pairwise deletion, LISREL
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.