Linear structural equations with latent variables

An interdependent multivariate linear relations model based on manifest, measured variables as well as unmeasured and unmeasurable latent variables is developed. The latent variables include primary or residual common factors of any order as well as unique factors. The model has a simpler parametric structure than previous models, but it is designed to accommodate a wider range of applications via its structural equations, mean structure, covariance structure, and constraints on parameters. The parameters of the model may be estimated by gradient and quasi-Newton methods, or a Gauss-Newton algorithm that obtains least-squares, generalized least-squares, or maximum likelihood estimates. Large sample standard errors and goodness of fit tests are provided. The approach is illustrated by a test theory model and a longitudinal study of intelligence.This investigation was supported in part by a Research Scientist Development Award (KO2-DA00017) and a research grant (DA01070) from the U. S. Public Health Service.     

Extending latent class solutions to other variables

A procedure is derived for estimating the latent parameters of items not initially included in a latent class solution, on the assumption that the relations between those additional items and the original ones are accounted for by the same latent structure. A chi-square test is proposed for evaluating the relatedness of the additional items to the latent structure. The extension is generalized to associate continuous outside variables with the original solution, and this is accompanied by the suggestion that a simple analysis of variance be used to assess the results. Finally, this latent structure extension is compared with the Dwyer extension of factor analysis.I am grateful to J. J. Mellinger for clarifying discussions of the statistical portions of this paper.Opinions expressed herein are the author's, not the Army's.     

A general class of latent variable models for ordinal manifest variables with covariate effects on the manifest and latent variables

Previous work on a general class of multidimensional latent variable models for analysing ordinal manifest variables is extended here to allow for direct covariate effects on the manifest ordinal variables and covariate effects on the latent variables. A full maximum likelihood estimation method is used to estimate all the model parameters simultaneously. Goodness‐of‐fit statistics and standard errors are discussed. Two examples from the 1996 British Social Attitudes Survey are used to illustrate the methodology.     

Randomization-based inference about latent variables from complex samples

Standard procedures for drawing inferences from complex samples do not apply when the variable of interest cannot be observed directly, but must be inferred from the values of secondary random variables that depend on stochastically. Examples are proficiency variables in item response models and class memberships in latent class models. Rubin's multiple imputation techniques yield approximations of sample statistics that would have been obtained, had been observable, and associated variance estimates that account for uncertainty due to both the sampling of respondents and the latent nature of. The approach is illustrated with data from the National Assessment for Educational Progress.This research was supported by Grant No. NIE-G-83-0011 of the Office for Educational Research and Improvement, Center for Education Statistics, and Contract No. N00014-88-K-0304, R&T 4421552 from the Cognitive Sciences Program, Cognitive and Neural Sciences Division, Office of Naval Research. It does not necessarily reflect the views of either agency. I am grateful to R. Darrell Bock for calling my attention to the applicability of multiple imputation to the assessment setting; to Albert Beaton and Eugene Johnson for enlightening discussions on the topic; and to Henry Braun, Ben King, Debra Kline, Gary Phillips, Paul Rosenbaum, Don Rubin, John Tukey, Ming-Mei Wang, Kentaro Yamamoto, Rebecca Zwick, and two anonymous reviewers for comments on earlier drafts. Example 4 is based on the analysis of the 1984 National Assessment for Educational Progress reading survey, carried out at Educational Testing Service through the tireless efforts of too many people to mention by name, under the direction of Albert Beaton, Director of NAEP Data Analyses. David Freund, Bruce Kaplan, and Jennifer Nelson conducted additional analyses of the 1984 and 1988 data for the example.     

Relating latent class membership to external variables: An overview

In this article we provide an overview of existing approaches for relating latent class membership to external variables of interest. We extend on the work of Nylund-Gibson et al. (Structural Equation Modeling: A Multidisciplinary Journal, 2019, 26, 967), who summarize models with distal outcomes by providing an overview of most recommended modeling options for models with covariates and larger models with multiple latent variables as well. We exemplify the modeling approaches using data from the General Social Survey for a model with a distal outcome where underlying model assumptions are violated, and a model with multiple latent variables. We discuss software availability and provide example syntax for the real data examples in Latent GOLD.     

Observed and latent variables: Discussion of a basic LISREL paradigm

The aim of the present paper is to discuss a general paradigm of LISREL: that observed psychologocal variables are causally determined by non-observed, latent variables. An example from the LISREL manual is used to show that this basic assumption is not generally meaningful. The possibility of understanding the theoretical variables of LISREL in an instrumentalist way is also considered. But instrumentalism, it is concluded, does not provide a tenable way of defending the LISREL paradigm. For LISREL conceives latent variables as causes of observed variables, and this is not compatible with an instrumentalist understanding of theoretical terms.     

Predicting group-level outcome variables from variables measured at the individual level: a latent variable multilevel model

In multilevel modeling, one often distinguishes between macro-micro and micro-macro situations. In a macro-micro multilevel situation, a dependent variable measured at the lower level is predicted or explained by variables measured at that lower or a higher level. In a micro-macro multilevel situation, a dependent variable defined at the higher group level is predicted or explained on the basis of independent variables measured at the lower individual level. Up until now, multilevel methodology has mainly focused on macro-micro multilevel situations. In this article, a latent variable model is proposed for analyzing data from micro-macro situations. It is shown that regression analyses carried out at the aggregated level result in biased parameter estimates. A method that uses the best linear unbiased predictors of the group means is shown to yield unbiased estimates of the parameters.     

Using phantom and imaginary latent variables to parameterize constraints in linear structural models

The most widely-used computer programs for structural equation models analysis are the LISREL series of Jöreskog and Sörbom. The only types of constraints which may be made directly are fixing parameters at a constant value and constraining parameters to be equal. Rindskopf (1983) showed how these simple properties could be used to represent models with more complicated constraints, namely inequality constraints on unique variances. In this paper, two new concepts are introduced which enable a much wider variety of constraints to be made. The concepts, phantom and imaginary latent variables, allow fairly general equality and inequality constraints on factor loadings and structural model coefficients.During the preparation of this article, it was discovered that another researcher, Jack McArdle, had concurrently and independently discovered some of the techniques reported here. While he has chosen not to publish his research, I wish to acknowledge his work. I would like to thank Art Woodward for telling me about sort-of simple structure.     

Factor analysis with (mixed) observed and latent variables in the exponential family

We develop a general approach to factor analysis that involves observed and latent variables that are assumed to be distributed in the exponential family. This gives rise to a number of factor models not considered previously and enables the study of latent variables in an integrated methodological framework, rather than as a collection of seemingly unrelated special cases. The framework accommodates a great variety of different measurement scales and accommodates cases where different latent variables have different distributions. The models are estimated with the method of simulated likelihood, which allows for higher dimensional factor solutions to be estimated than heretofore. The models are illustrated on synthetic data. We investigate their performance when the distribution of the latent variables is mis-specified and when part of the observations are missing. We study the properties of the simulation estimators relative to maximum likelihood estimation with numerical integration. We provide an empirical application to the analysis of attitudes.     

Building an identifiable latent class model with covariate effects on underlying and measured variables

In recent years, latent class models have proven useful for analyzing relationships between measured multiple indicators and covariates of interest. Such models summarize shared features of the multiple indicators as an underlying categorical variable, and the indicators' substantive associations with predictors are built directly and indirectly in unique model parameters. In this paper, we provide a detailed study on the theory and application of building models that allow mediated relationships between primary predictors and latent class membership, but that also allow direct effects of secondary covariates on the indicators themselves. Theory for model identification is developed. We detail an Expectation-Maximization algorithm for parameter estimation, standard error calculation, and convergent properties. Comparison of the proposed model with models underlying existing latent class modeling software is provided. A detailed analysis of how visual impairments affect older persons' functioning requiring distance vision is used for illustration.This work was supported by National Institute on Aging (NIA) Program Project P01-AG-10184-03 and National Institutes of Mental Health grant R01-MH-56639-01A1. Dr. Bandeen-Roche is a Brookdale National Fellow. The authors wish to thank Drs. Gary Rubin and Sheila West for kindly making the Salisbury Eye Evaluation data available. We also thank the Editor, the Associate Editor, and three referees for their valuable comments.     

Parents' teaching strategies with their children: The effects of parental and child status variables

The intent of the study was to examine the effects of parental (mother and father) and child (gender and communicative status) status variables on the teaching strategies used by parents in a paper-folding task. The linguistic and nonverbal strategies of 120 parent-child groupings were analyzed during paper-folding tasks. Parents' strategies were coded according cognitive demand and directiveness. Results indicated that strategies varied as a function of children's communicative status. Parents were less directive and more demanding of nonhandicapped children compared to handicapped children. Parents' strategies seemed to be determined by children's ability to sustain discourse. Results are discussed in terms of Vygotsky's notion of the zone of proximal development.This research was partially supported by a grant from the U.S. Office of Education, Office of Special Education (Grant No. G007902000) to I. Sigel and A. McGillicuddy-DeLisi at Educational Testing Service, Princeton, and from the Biomedical Research Support Grant Program, Division of Research Resources, N.I.H. (Grant No. BRSG 507RR07025-17) to the first two authors. We acknowledge the helpful comments of two anonymous reviewers.     

Attitudinal variables as estimates of behavior: A theoretical examination of the attitude-action controversy

Discusses the current status of the controversy about the relationship between individual's attitudes and subsequent behavior. It is suggested that one step toward the resolution of the existing controversy is the construction of a formal theory, specifying and integrating variables other than attitude into the framework. To this end, a model (in propositional form) is advanced focusing on attitude, centrality and salience as determinants of action. The model was partially tested through reanalysis of survey data dealing with attitudes and behaviors of 221 individuals with regard to halting air pollution. The data confirmed that subjects for whom the attitude object was central exhibited a higher level of attitude-behavior consistency than subjects for whom the attitude object was of low centrality. It is argued that the use of the proposed model clarifies relationships which otherwise might incorrectly be interpreted and provides a practical logic for determining when a comparison may be reliable and when one should beware of probable measurement difficulties.