Doubly Robust-Type Estimation for Covariate Adjustment in Latent Variable Modeling

Due to the difficulty in achieving a random assignment, a quasi-experimental or observational study design is frequently used in the behavioral and social sciences. If a nonrandom assignment depends on the covariates, multiple group structural equation modeling, that includes the regression function of the dependent variables on the covariates that determine the assignment, can provide reasonable estimates under the condition of correct specification of the regression function. However, it is usually difficult to specify the correct regression function because the dimensions of the dependent variables and covariates are typically large. Therefore, the propensity score adjustment methods have been proposed, since they do not require the specification of the regression function and have been applied to several applied studies. However, these methods produce biased estimates if the assignment mechanism is incorrectly specified. In order to make a more robust inference, it would be more useful to develop an estimation method that integrates the regression approach with the propensity score methodology. In this study we propose a doubly robust-type estimation method for marginal multiple group structural equation modeling. This method provides a consistent estimator if either the regression function or the assignment mechanism is correctly specified. A simulation study indicates that the proposed estimation method is more robust than the existing methods. This research was partially supported by the Ministry of Education, Science, Sports and Culture, Grant-in-Aid for Young Scientists (B), 187-30406.     

Bayesian Analysis of Structural Equation Models With Nonlinear Covariates and Latent Variables

In this article, we formulate a nonlinear structural equation model (SEM) that can accommodate covariates in the measurement equation and nonlinear terms of covariates and exogenous latent variables in the structural equation. The covariates can come from continuous or discrete distributions. A Bayesian approach is developed to analyze the proposed model. Markov chain Monte Carlo methods for obtaining Bayesian estimates and their standard error estimates, highest posterior density intervals, and a PP p value are developed. Results obtained from two simulation studies are reported to respectively reveal the empirical performance of the proposed Bayesian estimation in analyzing complex nonlinear SEMs, and in analyzing nonlinear SEMs with the normal assumption of the exogenous latent variables violated. The proposed methodology is further illustrated by a real example. Detailed interpretation about the interaction terms is presented.     

Generalized Structured Component Analysis with Latent Interactions

Generalized structured component analysis (GSCA) is a component-based approach to structural equation modeling. In practice, researchers may often be interested in examining the interaction effects of latent variables. However, GSCA has been geared only for the specification and testing of the main effects of variables. Thus, an extension of GSCA is proposed to effectively deal with various types of interactions among latent variables. In the proposed method, a latent interaction is defined as a product of interacting latent variables. As a result, this method does not require the construction of additional indicators for latent interactions. Moreover, it can easily accommodate both exogenous and endogenous latent interactions. An alternating least-squares algorithm is developed to minimize a single optimization criterion for parameter estimation. A Monte Carlo simulation study is conducted to investigate the parameter recovery capability of the proposed method. An application is also presented to demonstrate the empirical usefulness of the proposed method.     

Model-based recursive partitioning of extended redundancy analysis with an application to nicotine dependence among US adults

Extended redundancy analysis (ERA) is used to reduce multiple sets of predictors to a smaller number of components and examine the effects of these components on a response variable. In various social and behavioural studies, auxiliary covariates (e.g., gender, ethnicity) can often lead to heterogeneous subgroups of observations, each of which involves distinctive relationships between predictor and response variables. ERA is currently unable to consider such covariate-dependent heterogeneity to examine whether the model parameters vary across subgroups differentiated by covariates. To address this issue, we combine ERA with model-based recursive partitioning in a single framework. This combined method, MOB-ERA, aims to partition observations into heterogeneous subgroups recursively based on a set of covariates while fitting a specified ERA model to data. Upon the completion of the partitioning procedure, one can easily examine the difference in the estimated ERA parameters across covariate-dependent subgroups. Moreover, it produces a tree diagram that aids in visualizing a hierarchy of partitioning covariates, as well as interpreting their interactions. In the analysis of public data concerning nicotine dependence among US adults, the method uncovered heterogeneous subgroups characterized by several sociodemographic covariates, each of which yielded different directional relationships between three predictor sets and nicotine dependence.     

Parsimonious Structural Equation Models for Repeated Measures Data,with Application to the Study of Consumer Preferences

Recent research reflects a growing awareness of the value of using structural equation models to analyze repeated measures data. However, such data, particularly in the presence of covariates, often lead to models that either fit the data poorly, are exceedingly general and hard to interpret, or are specified in a manner that is highly data dependent. This article introduces methods for developing parsimonious models for such data. The underlying technology uses reduced-rank representations of the variances, covariances and means of observed and latent variables. The value of this approach, which may be implemented using standard structural equation modeling software, is illustrated in an application study aimed at understanding heterogeneous consumer preferences. In this application, the parsimonious representations characterize systematic relationships among consumer demographics, attitudes and preferences that would otherwise be undetected. The result is a model that is parsimonious, illuminating, and fits the data well, while keeping data dependence to a minimum.     

Equivalent linear logistic test models

This paper is about the Linear Logistic Test Model (LLTM). We demonstrate that there are infinitely many equivalent ways to specify a model. An implication is that there may well be many ways to change the specification of a given LLTM and achieve the same improvement in model fit. To illustrate this phenomenon, we analyze a real data set using a Lagrange multiplier test for the specification of the model. This Lagrange multiplier test is similar to the modification index used in structural equation modeling.     

Tests for stability in linear structural equation systems

Interpretations regarding the effects of exogenous and endogenous variables on endogenous variables in linear structural equation systems depend upon the convergence of a matrix power series. Convergence depends upon the eigenvalues of the structural coefficient matrix. The test for convergence developed by Jöreskog and Sörbom is shown to be only sufficient, not necessary and sufficient.     

Structural Equation Modeling in the Communication Sciences, 1995–2000

Structural equation modeling (SEM) is a viable multivariate tool used by communication researchers for the past quarter century. Building off Cappella (1975) as well as McPhee and Babrow (1987), this study summarizes the use of this technique from 1995–2000 in 37 communication‐based academic journals. We identify and critically assess 3 unique methods for testing structural relationships via SEM in terms of the specification, estimation, and evaluation of their respective structural equation models. We provide general guidelines for the use of SEM and make recommendations concerning latent variable models, sample size, reporting parameter estimates, model fit statistics, cross‐sectional data, univariate normality, cross‐validation, nonrecursive modeling, and the decomposition of effects (direct, indirect, and total).     

Integrating covariates into circumplex structures: an extension procedure for Browne’s circular stochastic process model

Circumplex structures are elements of various psychological domains. Most work focuses on assessing the circular ordering of circumplex indicators and their relationships with covariates. In this article, an extension procedure for Browne’s circumplex model is presented. Our approach models the relationships among circumplex indicators and the relationships of covariates with a latent circumplex simultaneously without affecting the circumplex indicators’ positions on the circumplex. The approach builds upon Browne’s Fourier series parameterization of a correlation function, which is used to model the latent circumplex correlation structure. It extends the shape of the correlation function to the profile of each covariate’s correlations with the circumplex. The model is specified in the framework of structural equation modeling, thereby making it possible to test various hypotheses. Procedures are presented for deriving interval estimates for the parameters that relate the covariates to the circumplex. The model is compared to other approaches for assessing the relationships of a circumplex with covariates. The results of the exemplary applications and a simulation study were in favor of the suggested model. The approach is furthermore illustrated with a real-data example, focusing on the relationships between the interpersonal circumplex and the rivalry and admiration aspects of narcissism.     

Functional Extended Redundancy Analysis

We propose a functional version of extended redundancy analysis that examines directional relationships among several sets of multivariate variables. As in extended redundancy analysis, the proposed method posits that a weighed composite of each set of exogenous variables influences a set of endogenous variables. It further considers endogenous and/or exogenous variables functional, varying over time, space, or other continua. Computationally, the method reduces to minimizing a penalized least-squares criterion through the adoption of a basis function expansion approach to approximating functions. We develop an alternating regularized least-squares algorithm to minimize this criterion. We apply the proposed method to real datasets to illustrate the empirical feasibility of the proposed method.     

A Propensity Score Adjustment for Multiple Group Structural Equation Modeling

In the behavioral and social sciences, quasi-experimental and observational studies are used due to the difficulty achieving a random assignment. However, the estimation of differences between groups in observational studies frequently suffers from bias due to differences in the distributions of covariates. To estimate average treatment effects when the treatment variable is binary, Rosenbaum and Rubin (1983a) proposed adjustment methods for pretreatment variables using the propensity score. However, these studies were interested only in estimating the average causal effect and/or marginal means. In the behavioral and social sciences, a general estimation method is required to estimate parameters in multiple group structural equation modeling where the differences of covariates are adjusted. We show that a Horvitz–Thompson-type estimator, propensity score weighted M estimator (PWME) is consistent, even when we use estimated propensity scores, and the asymptotic variance of the PWME is shown to be less than that with true propensity scores. Furthermore, we show that the asymptotic distribution of the propensity score weighted statistic under a null hypothesis is a weighted sum of independent χ² ₁ variables. We show the method can compare latent variable means with covariates adjusted using propensity scores, which was not feasible by previous methods. We also apply the proposed method for correlated longitudinal binary responses with informative dropout using data from the Longitudinal Study of Aging (LSOA). The results of a simulation study indicate that the proposed estimation method is more robust than the maximum likelihood (ML) estimation method, in that PWME does not require the knowledge of the relationships among dependent variables and covariates.     

Selecting polychoric instrumental variables in confirmatory factor analysis: An alternative specification test and effects of instrumental variables

The polychoric instrumental variable (PIV) approach is a recently proposed method to fit a confirmatory factor analysis model with ordinal data. In this paper, we first examine the small-sample properties of the specification tests for testing the validity of instrumental variables (IVs). Second, we investigate the effects of using different numbers of IVs. Our results show that specification tests derived for continuous data are extremely oversized at all sample sizes when applied to ordinal variables. Possible modifications for ordinal data are proposed in the present study. Simulation results show that the modified specification tests with all available IVs are able to detect model misspecification. In terms of estimation accuracy, the PIV approach where the IVs outnumber the endogenous variables by one produces a lower bias but a higher variation than the PIV approach with more IVs for correctly specified factor loadings at small samples.     

Generalized linear models with ordinally‐observed covariates

An ordinally‐observed variable is a variable that is only partially observed through an ordinal surrogate. Although statistical models for ordinally‐observed response variables are well known, relatively little attention has been given to the problem of ordinally‐observed regressors. In this paper I show that if surrogates to ordinally‐observed covariates are used as regressors in a generalized linear model then the resulting measurement error in the covariates can compromise the consistency of point estimators and standard errors for the effects of fully‐observed regressors. To properly account for this measurement error when making inferences concerning the fully‐observed regressors, I propose a general modelling framework for generalized linear models with ordinally‐observed covariates. I discuss issues of model specification, identification, and estimation, and illustrate these with examples.     

A general structural equation model with dichotomous,ordered categorical,and continuous latent variable indicators

A structural equation model is proposed with a generalized measurement part, allowing for dichotomous and ordered categorical variables (indicators) in addition to continuous ones. A computationally feasible three-stage estimator is proposed for any combination of observed variable types. This approach provides large-sample chi-square tests of fit and standard errors of estimates for situations not previously covered. Two multiple-indicator modeling examples are given. One is a simultaneous analysis of two groups with a structural equation model underlying skewed Likert variables. The second is a longitudinal model with a structural model for multivariate probit regressions.This research was supported by Grant No. 81-IJ-CX-0015 from the National Institute of Justice, by Grant No. DA 01070 from the U.S. Public Health Service, and by Grant No. SES-8312583 from the National Science Foundation. I thank Julie Honig for drawing the figures. Requests for reprints should be sent to Bengt Muthén, Graduate School of Education, University of California, Los Angeles, California 90024.     

A tetrad test for causal indicators

The authors propose a confirmatory tetrad analysis test to distinguish causal from effect indicators in structural equation models. The test uses "nested" vanishing tetrads that are often implied when comparing causal and effect indicator models. The authors present typical models that researchers can use to determine the vanishing tetrads for 4 or more variables. They also provide the vanishing tetrads for mixtures of causal and effect indicators, for models with fewer than 4 indicators per latent variable, or for cases with correlated errors. The authors illustrate the test results for several simulation and empirical examples and emphasize that their technique is a theory-testing rather than a model-generating approach. They also review limitations of the procedure including the indistinguishable tetrad equivalent models, the largely unknown finite sample behavior of the test statistic, and the inability of any procedure to fully validate a model specification.     

Test of structural relationships within a proposed exercise and self-esteem model

This study tested structural relationships within a model proposed to explain the manner in which self-esteem changes are associated with exercise experiences. As initially assessed by self-efficacies (EFFs) specific to physical tasks within a training program, we postulated that competence can generalize to feelings of global self-esteem (SE) through an intervening construct of perceived physical competence (PC). Three measures of EFF and two each of PC and SE were administered to 145 people in their mid- to late adulthood. Confirmatory factor analysis examined orthogonal and oblique versions of three measurement models and identified three distinct but correlated factors, each assessed by its hypothesized indicators. Alternative structural equation models were specified and tested using both normal and nonnormal estimation procedures. The proposed model provided the most parsimonious fit and explained 29% of the variance in SE. Confirmation of its structural relationships provides preliminary validity for model use in examining the manner in which exercise experiences influence levels of self-perception.     

Test of Structural Relationships Within a Proposed Exercise and Self-Esteem Model

This study tested structural relationships within a model proposed to explain the manner in which self-esteem changes are associated with exercise experiences. As initially assessed by self-efficacies (EFFs) specific to physical tasks within a training program, we postulated that competence can generalize to feelings of global self-esteem (SE) through an intervening construct of perceived physical competence (PC). Three measures of EFF and two each of PC and SE were administered to 145 people in their mid- to late adulthood. Confirmatory factor analysis examined orthogonal and oblique versions of three measurement models and identified three distinct but correlated factors, each assessed by its hypothesized indicators. Alternative structural equation models were specified and tested using both normal and nonnormal estimation procedures. The proposed model provided the most parsimonious fit and explained 29% of the variance in SE. Confirmation of its structural relationships provides preliminary validity for model use in examining the manner in which exercise experiences influence levels of self-perception.     

结构方程模型中调节效应的标准化估计

回归分析和结构方程分析中,标准化估计对解释模型和比较效应大小有重要作用。对于调节效应模型（或交互效应模型）,通常的标准化估计没有意义。虽然显变量的调节效应模型标准化估计问题已经解决,但潜变量的调节效应模型标准化估计问题复杂得多。本文先介绍回归分析中显变量调节效应模型的标准化估计,然后提出了一种通过参数的原始估计和通常标准化估计来计算潜变量调节效应模型的“标准化”估计的方法,得到的“标准化”估计是尺度不变的,说明可以用“标准化”估计来解释和比较主效应和调节效应     

Dynamic GSCA (Generalized Structured Component Analysis) with Applications to the Analysis of Effective Connectivity in Functional Neuroimaging Data

We propose a new method of structural equation modeling (SEM) for longitudinal and time series data, named Dynamic GSCA (Generalized Structured Component Analysis). The proposed method extends the original GSCA by incorporating a multivariate autoregressive model to account for the dynamic nature of data taken over time. Dynamic GSCA also incorporates direct and modulating effects of input variables on specific latent variables and on connections between latent variables, respectively. An alternating least square (ALS) algorithm is developed for parameter estimation. An improved bootstrap method called a modified moving block bootstrap method is used to assess reliability of parameter estimates, which deals with time dependence between consecutive observations effectively. We analyze synthetic and real data to illustrate the feasibility of the proposed method.     

Consistent Partial Least Squares for Nonlinear Structural Equation Models

Partial Least Squares as applied to models with latent variables, measured indirectly by indicators, is well-known to be inconsistent. The linear compounds of indicators that PLS substitutes for the latent variables do not obey the equations that the latter satisfy. We propose simple, non-iterative corrections leading to consistent and asymptotically normal (CAN)-estimators for the loadings and for the correlations between the latent variables. Moreover, we show how to obtain CAN-estimators for the parameters of structural recursive systems of equations, containing linear and interaction terms, without the need to specify a particular joint distribution. If quadratic and higher order terms are included, the approach will produce CAN-estimators as well when predictor variables and error terms are jointly normal. We compare the adjusted PLS, denoted by PLSc, with Latent Moderated Structural Equations (LMS), using Monte Carlo studies and an empirical application.