Have Multilevel Models Been Structural Equation Models All Along?

  title={Have Multilevel Models Been Structural Equation Models All Along?},
  author={Patrick J. Curran},
  journal={Multivariate Behavioral Research},
  pages={529 - 569}
  • P. Curran
  • Published 1 October 2003
  • Economics
  • Multivariate Behavioral Research
A core assumption of the standard multiple regression model is independence of residuals, the violation of which results in biased standard errors and test statistics. The structural equation model (SEM) generalizes the regression model in several key ways, but the SEM also assumes independence of residuals. The multilevel model (MLM) was developed to extend the regression model to dependent data structures. Attempts have been made to extend the SEM in similar ways, but several complications… 

Implementing Restricted Maximum Likelihood Estimation in Structural Equation Models

This article shows how REML estimation can be implemented in SEM, and two empirical examples on latent growth model and meta-analysis are used to illustrate the procedures implemented in OpenMx.

Applications of Multilevel Structural Equation Modeling to Cross-Cultural Research

This work first showed how Muth�n's MSEM procedure could be applied to a real data set on cross-cultural research, and demonstrated that results on the individual level were quite stable even when using small individual-level sample sizes, whereas the group-level parameter estimates and their standard errors were affected unsystematically by varying individual- level sample sizes.

Using SAS PROC CALIS to fit Level-1 error covariance structures of latent growth models

This article demonstrates the use of SAS PROC CALIS to fit various types of Level-1 error covariance structures of latent growth models (LGM) and provides a tutorial on the SAS syntax for modeling the growth of a manifest variable and therowth of a latent construct.

People are variables too: multilevel structural equations modeling.

The article uses confirmatory factor analysis (CFA) as a template to explain didactically multilevel structural equation models (ML-SEM) and to demonstrate the equivalence of general mixed-effects

The Use of Mixed Models for the Analysis of Mediated Data with Time-Dependent Predictors

It is found that when carefully constructed, the LMM adequately models mediated exposure effects that change over time in the presence of mediation, even when the data arise from an SEM.

From Data to Causes II: Comparing Approaches to Panel Data Analysis

This article compares a general cross-lagged model (GCLM) to other panel data methods based on their coherence with a causal logic and pragmatic concerns regarding modeled dynamics and hypothesis testing, including the GCLM approach proposed.

A structural equation modeling approach for modeling variability as a latent variable.

Drawing upon recent developments in structural equation modeling, the current study presents an analytical framework for addressing research questions in which, rather than focusing on means, it is

Extending Latent Basis Growth Model to Explore Joint Development in the Framework of Individual Measurement Occasions

This work proposes a novel specification for LBGMs that allows for (1) unequally-spaced study waves and (2) individual measurement occasions around each wave, and demonstrates that the proposed parallel LBGM can capture the underlying developmental patterns of these two abilities.

Right-sizing statistical models for longitudinal data.

A 3-step approach is proposed in which models are evaluated based on the number and patterning of variance components prior to selection of better-fitting growth models that explain both mean and variation-covariation patterns.



Estimating Multilevel Linear Models as Structural Equation Models

It is demonstrated that a broad class of MLMs may be estimated as structural equation models (SEMs), and within the SEM approach it is possible to include measurement models for predictors or outcomes, and to estimate the mediational pathways among predictors explicitly, tasks which are currently difficult with the conventional approach to multilevel modeling.


Busemeyer and Jones (1983) and Kenny and Judd (1984) proposed methods to include interactions of latent variables in structural equation models (SEMs). Despite the value of these works, their methods

Generalized multilevel structural equation modeling

Maximum likelihood estimation and empirical Bayes latent score prediction within the GLLAMM framework can be performed using adaptive quadrature in gllamm, a freely available program running in Stata.

Contextual analysis through the multilevel linear model.

An extended empirical example was presented with the primary purpose of illustrating the use of the statistical methodology in conjunction with a meaningful substantive problem not to carry out a critical test of the underlying substantive theory and not to demonstrate the general superiority of the proposed estimation procedure.

Using SAS PROC MIXED to Fit Multilevel Models, Hierarchical Models, and Individual Growth Models

This paper is written as a step-by-step tutorial that shows how to fit the two most common multilevel models: (a) school effects models, designed for data on individuals nested within naturally occurring hierarchies (e.g., students within classes); and (b) individual growth models,designed for exploring longitudinal data (on individuals) over time.

Using covariance structure analysis to detect correlates and predictors of individual change over time

Recently, methodologists have shown how two disparate conceptual arenas—individual growth modeling and covariance structure analysis—can be integrated. The integration brings the flexibility of

A structural equations modeling approach to the general linear mixed model

This chapter presents a method based on a structural equation modeling (SEM) framework for estimating models representing specific instances of the general linear mixed model (GLMM); the authors state their choice of the GLMM as a starting point because it can be considered the basis of thegeneral multilevel modeling approach.

Finite-Mixture Structural Equation Models for Response-Based Segmentation and Unobserved Heterogeneity

Two endemic problems face researchers in the social sciences e.g., Marketing, Economics, Psychology, and Finance: unobserved heterogeneity and measurement error in data. Structural equation modeling

Putting the individual back into individual growth curves.

An Mx script for fitting individual data vector-based LGC models is provided and issues regarding the interpretation of within- and between-persons effects of time that parallel those associated with centering of predictor variables in MLM are illustrated.

Generalized Latent Variable Modeling: Multilevel,Longitudinal, and Structural Equation Models

To fully grasp some of the more complex aspects of sensitivity analysis, additional references are needed to supplement this text, and the technical material seemed overly complicated and lacking sufficient explanation.