A PRIMER FOR SOFT MODELING
is a guide to the 'soft modeling' approach to structural equation modeling that relies on a computer application strategy. The theoretical as well as practical requirements for soft modeling (partial least squares estimation procedures) are different from those of other modeling procedures because the basic assumptions about data are less stringent.
This fundamental difference enables path models to be analyzed with soft modeling techniques that are rejected by many popularly-used programs. Because soft modeling procedures facilite analysis of data with less stringent measurement requirements, it represents a modeling system that can be used by most social and behavioral scientists.

Written by authors who teach structural equation modeling, as well as undergraduate statistics, this book presents soft modeling in ways that students and researchers cand readly comprehend, demonstrating the applicability of this new modeling approach to social science data.
As a practical guide to latent variable path analysis, the basic 'how-to's' of modeling are explained - how to configure research questions, how to prepare data, how to construct an LVPLS computer run, how to interpret the results, and how to present the findings. Basic modeling concepts are presented in an accurate yet non-technical manner through-out the text, and the non-mathematical reader has been kept in mind.