# Latent Variable Models: An Introduction to Factor, Path, and Structural Analysis

# Latent Variable Models: An Introduction to Factor, Path, and Structural Analysis

## Synopsis

The author utilizes path diagrams to explain the underlying relationships in multiple-latent-variable models. He also provides an appendix on elementary matrix algebra.

The book is not closely tied to a particular computer program or package; however, special attention is paid to two leaders in the field (LISREL and EQS). Users should have access to a latent-variable model-fitting program on the order of LISREL, EQS, CALIS, AMOS, Mx, RAMONA, or SEPATH, and an exploratory factor analysis package such as those in SPSS or SAS. In some places, a matrix manipulation facility such as that found in MINITAB, SAS, or SPSS would be useful.

## Excerpt

This book is intended as an introduction to an exciting growth area in social science methodology--the use of multiple-latent-variable models. Psychologists and other social scientists have long been familiar with one subvariety of such modeling, factor analysis--more properly, exploratory factor analysis. In recent decades, confirmatory factor analysis, path analysis, and structural equation modeling have come out of specialized niches and are making their bid to become basic tools in the research repertoire of the social scientist, particularly the one who is forced to deal with complex real-life phenomena in the round: the sociologist, the political scientist, the social, educational, clinical, industrial, personality or developmental psychologist, the marketing researcher, and the like.

All these methods are at heart one, as I have tried to emphasize in the chapters to follow. I have used earlier versions of this book in teaching graduate students from psychology and related disciplines, and have found the particular approach used--via path diagrams--to be effective in helping not-too- mathematical students grasp underlying relationships, as opposed to merely going through the motions of running computer programs. In some sections of the book a certain amount of elementary matrix algebra is employed; an appendix on the topic is provided for those who may need help here.

In the interests of accessability, I have tried to maintain a relatively informal style, and to keep the main text fairly uncluttered with references. The notes at the end of each chapter are intended to provide the serious student with a path into the technical literature, as well as to draw his or her attention to some issues beyond the scope of the basic treatment.

The book is not closely tied to a particular computer program or package, although there is some special attention paid to LISREL and EQS. I assume that most users will have access to a latent-variable model-fitting program on the order of LISREL, EQS, CALIS, AMOS, Mx, RAMONA, or SEPATH, and an exploratory factor analysis package such as those in SPSS or SAS. In some places, a matrix manipulation facility such as that in MINITAB, SAS, or SPSS would be helpful. I have provided some introductory material but have not tried to tell students all they need to know to run actual programs-- such information is often local, ephemeral, or both. The instructor should expect . . .