Relationship Between Multiple Regression, Path, Factor, and LISREL Analyses

Abstract

A basic knowledge of multiple regression concepts permits further understanding of path, factor, and lisrel analyses. Specifically, standardized partial regression coefficients (beta weights) as applied in path, factor, and lisrel analyses are presented. The multivariable methods have in common the general linear model and are the same in several respects. First, they identify, partition, and control variance. Second, they are based upon a linear combination of variables. And third, the linear weights can be computed based on standardized partial regression coefficients.