Striving for Sparsity
On Exact and Approximate Solutions in Regularized Structural Equation Models
Bibliographic Data
| ID | 21641697 |
|---|---|
| Authors | Jannik H Orzek (0000-0002-3123-2248, Humboldt-Universität zu Berlin, corresponding author), Manuel Arnold (0000-0002-1623-2565, Humboldt-Universität zu Berlin), Manuel C Voelkle (0000-0001-5576-8103, Humboldt-Universität zu Berlin) |
| Year | 2023 |
| Volume | 30 |
| Issue | 6 |
| Pages | 956-973 |
| Publication date | 2023-11-02 |
| Open Access | Yes |
| Type | ARTICLE |
| Venue | Structural Equation Modeling: A Multidisciplinary Journal (JOURNAL) |
| Journal identifiers | ISSN: 1070-5511 • E-ISSN: 1532-8007 |
| Publisher | Informa UK Limited (PUBLISHER • GB) |
| DOI | 10.1080/10705511.2023.2189070 |
| OpenAlex | W4376128199 |
| Language | EN |
| Citations received | 1 |
| References cited | 49 |
Regularized structural equation models have gained considerable traction in the social sciences. They promise to reduce overfitting by focusing on out-of-sample predictions and sparsity. To this end, a set of increasingly constrained models is fitted to the data. Subsequently, one of the models is selected, usually by means of information criteria. Current implementations of regularized structural equation models differ in their optimizers: Some use general purpose optimizers whereas others use specialized optimization routines. While both approaches often perform similarly, we show that they can produce very different results. We argue that in particular, the interaction between optimizer and selection criterion (e.g., BIC) contributes to these differences. We substantiate our arguments with an empirical demonstration and a simulation study. Based on these findings, we conclude that researchers should consider specialized optimizers whenever possible. To facilitate the implementation of such optimizers, we provide the R package lessSEM
Implementation · Machine learning · Mathematical optimization · Model selection · Overfitting · Structural equation modeling · Computer Science · Mathematics · Mental Health Research Topics · Qualitative Comparative Analysis Research · Statistical Methods and Inference · Artificial Intelligence · Theoretical Computer Science
An Introduction to Statistical Learning
Ggplot2
Regularization and Variable Selection Via the Elastic Net
Ridge Regression
Generalized Network Psychometrics
An SEM approach to continuous time modeling of panel data
The Adaptive Lasso and Its Oracle Properties
A new look at the big five factor structure through exploratory structural equation modeling.
Variable Selection via Nonconcave Penalized Likelihood and its Oracle Properties
Regularization Paths for Generalized Linear Models via Coordinate Descent
Estimating the Dimension of a Model
Regression Shrinkage and Selection Via the Lasso
On the Performance of Different Regularization Methods in Bifactor-(S-1) Models with Explanatory Variables—Caveats, Recommendations, and Future Directions
A Comparison of Regularized Maximum-Likelihood, Regularized 2-Stage Least Squares, and Maximum-Likelihood Estimation with Misspecified Models, Small Samples, and Weak Factor Structure
Penalized Least Squares for Structural Equation Modeling with Ordinal Responses
Regularized Estimation of the Nominal Response Model
| Citations per year | 1 |
|---|---|
| Citation span | 2026 - 2026 (1) |
| Highly cited | No |