Abstract
A new method called Correlated Factors Modeling (CFM) is proposed as an alternative to LISREL and LISREL-like methods of estimating structural equations. CFM employs weak mathematics. It may be appropriate when (a) available theory is weak; (b) available data resources are weak in either sample size or sampling procedure, or both; (c) researcher knowledge of structural modeling methodology is weak; or (d) a rapid, easy, preliminary look at a model, such as can be obtained from a microcomputer, is desired. The factors produced by CFM are exact linear components of their indicator variates. Path models for hypothesized causal relationships among the factors can be computed on the reported correlations among the factors by the ordinary least squares (OLS) method of conventional path analysis. Heuristic goodness-of-fit indices are available to support comparisons of models. Two examples are reported.
Get full access to this article
View all access options for this article.