A simulation study compared three methods of estimating parameters within structural equation models (SEM) with polytomous variables. These methods appear in three SEM computer software packages: (1) LISREL (Joreskog and Sorbom, 1996) with PRELIS (Joreskog and Sorbom); (2) EQS (Bentler, 1995); and (3) the new Mplus (Muthen and Muthen, 1998). The…

A Monte Carlo study is reported that shows the comparative performance of alternative approaches under deviations from their respective assumptions in the case of structural equation models with latent variables with attention restricted to point estimates of model parameters. The conditional polychoric correlations method is shown most robust…

The relationship between structural equation modeling (SEM) and canonical correlation analysis (CCA) is illustrated. The representation of CCA in SEM may provide interpretive information not available from conventional CCA. Hierarchically, the relationship suggests that SEM is a more general analytic approach. (SLD)

Shows how to extend the inferential test of B. Shipley (2000), which is applicable to recursive path models without correlated errors, to a class of recursive path models that includes correlated errors. Discusses when the extended model is and is not superior to classical structural equation modeling. (SLD)

Tested empirically the applicability of a method developed by S. Mattson for generating data on latent variables with controlled skewness and kurtosis of the observed variables. Monte Carlo simulation results suggest that Mattson's method appears to be a good approach to generate data with defined levels of skewness and kurtosis. (SLD)

Discusses the difference in noncentrality parameters of nested structural equation models and their utility in evaluating statistical power associated with the pertinent restriction test. Asymptotic confidence intervals for that difference are presented. These intervals represent a useful adjunct to goodness-of-fit indexes in assessing constraints…

Provides a discussion of an alternative two-stage least squares (2SLS) technique to include interactions of latent variables in structural equation models. The method requires selection of instrumental variables, and rules for selection are presented. An empirical example and Statistical Analysis System programs are presented. (SLD)

Explains how the Gibbs sampler can be applied to obtain a sample from the posterior distribution over the parameters of a structural equation model. Presents statistics to use to summarize marginal posterior densities and model checks using posterior predictive p-values. (SLD)

Outlines a method for comparing completely standardized solutions in multiple groups. The method is based on a correlation structure analysis of equal-size samples and uses the correlation distribution theory implemented in the structural equation modeling program RAMONA. (SLD)

Adapts robust schemes to mean and covariance structures, providing an iteratively reweighted least squares approach to robust structural equation modeling. Each case is weighted according to its distance, based on first and second order moments. Test statistics and standard error estimators are given. (SLD)

Structural equation modeling was used to test the discriminant validity of three cognitions--thinking of quitting, intent to search, and intent to leave--in 242 sales workers. Results demonstrating a different relationship among the three cognitions were used to revise a 1977 model. (SK)

Extends the LISREL model to incorporate fixed covariates at both the measurement and the structural equations of the model, establishing a Bayesian procedure with conjugate type prior distributions. Illustrates the efficiency of the algorithm and presents a goodness of fit statistic for assessing the proposed model. (SLD)

Provides advice on writing a research paper when structural equation models are being used in empirical work. Focuses on what information should be reported and what can be deleted without much loss of judgment about the quality of the research and validity of conclusions. (SLD)

Describes some applications of latent growth curve models in the context of structural equation modeling using data from P. Trickett and F. Putnam (1993) on the physical height of abused (n=77) and nonabused (n=75) adolescent girls. Presents power calculations for the ability of the different models to discern the growth of the abuse sample from…

Derives approximations to the distributions of goodness-of-fit indexes in structural equation modeling with the assumption of multivariate normality and slight misspecification of models. Also derives an approximation to the asymptotic covariance matrix for the fit indexes by using the delta method and develops approximations to the densities of…