This study investigates the performance of robust maximum likelihood (ML) estimators when fitting and evaluating small sample latent growth models with non-normal missing data. Results showed that the robust ML methods could be used to account for non-normality even when the sample size is very small (e.g., N < 100). Among the robust ML…

Prior studies investigating the effects of non-normality in structural equation modeling typically induced non-normality in the indicator variables. This procedure neglects the factor analytic structure of the data, which is defined as the sum of latent variables and errors, so it is unclear whether previous results hold if the source of…

Structural equation modeling (SEM) applications routinely employ a trilogy of significance tests that includes the likelihood ratio test, Wald test, and score test or modification index. Researchers use these tests to assess global model fit, evaluate whether individual estimates differ from zero, and identify potential sources of local misfit,…

We examined the effect of estimation methods, maximum likelihood (ML), unweighted least squares (ULS), and diagonally weighted least squares (DWLS), on three population SEM (structural equation modeling) fit indices: the root mean square error of approximation (RMSEA), the comparative fit index (CFI), and the standardized root mean square residual…

Evaluations of measurement invariance provide essential construct validity evidence--a prerequisite for seeking meaning in psychological and educational research and ensuring fair testing procedures in high-stakes settings. However, the quality of such evidence is partly dependent on the validity of the resulting statistical conclusions. Type I or…

Using Monte Carlo simulations, this research examined the performance of four missing data methods in SEM under different multivariate distributional conditions. The effects of four independent variables (sample size, missing proportion, distribution shape, and factor loading magnitude) were investigated on six outcome variables: convergence rate,…

The purpose of this study was to assess the psychometrics properties of the Emotional Intelligence Scale and assess the measurement invariance across elite (n = 367), amateur (n = 629), and non-athletes (n = 550). In total, 1,546 participants from various sports completed the emotional intelligence scale. Several competing models were compared…

A promising "underlying bivariate normal" approach was proposed by Jöreskog and Moustaki for use in the factor analysis of ordinal data. This was a limited information approach that involved the maximization of a composite likelihood function. Its advantage over full-information maximum likelihood was that very much less computation was…

This study illustrates an integrative psychometric framework to investigate two sources of construct-relevant multidimensionality in answers to the Self-Perception Profile for Children (SPPC). Using a sample of 2,353 German students attending Grades 3 to 6, we contrasted: (a) first-order versus hierarchical and bifactor models to investigate…

We propose a new limited-information goodness of fit test statistic C[subscript 2] for ordinal IRT models. The construction of the new statistic lies formally between the M[subscript 2] statistic of Maydeu-Olivares and Joe (2006), which utilizes first and second order marginal probabilities, and the M*[subscript 2] statistic of Cai and Hansen…

In the last decades there has been an increasing interest in nonlinear latent variable models. Since the seminal paper of Kenny and Judd, several methods have been proposed for dealing with these kinds of models. This article introduces an alternative approach. The methodology involves fitting some third-order moments in addition to the means and…

A simulation study compared the performance of robust normal theory maximum likelihood (ML) and robust categorical least squares (cat-LS) methodology for estimating confirmatory factor analysis models with ordinal variables. Data were generated from 2 models with 2-7 categories, 4 sample sizes, 2 latent distributions, and 5 patterns of category…

Actual kurtotic and skewed data and varied sample sizes and estimation methods demonstrated that normal theory maximum likelihood and generalized least square estimators were fairly consistent and almost identical. Standard errors tended to underestimate the estimator's true variation but the problem was not serious for large samples. (SLD)

Structural equation modeling (SEM) attempts to remove the negative influence of measurement error and allows for investigation of relationships at the level of the underlying constructs of interest. SEM has been regarded as a "large sample" technique since its inception. Recent developments in SEM, some of which are currently available…

The proper use of Pearson's chi-square for the analysis of contingency tables is reviewed. A 1949 article by Lewis and Burke, in which they cite nine primary sources of error in the use of chi-square, serves as the basis of the review. Those nine sources of error are re-examined in light of current research. In addition, techniques and research on…