Theoretical Development for Identifying Underlying Internal Processes. Volume 2. Modifications to Hierarchical Factor Analysis: Positive Manifold (POSMAN) Rotations.
Abstract
Factor Analytic (FA) methods have as their objective the derivation of a parsimonious and meaningful set of independent factors to explain a matrix of interrelated variables. Unfortunately, the presence of higher-level factors will obscure the results of traditional factor analytic methods and defeat this objective. The purpose of Hierarchical Factor Analysis (HFA) is to detect the presence of and to seperate higher-level (i.e., general and subgeneral) factors from lower-level factors. A higher-level factors will obscure the results of traditional factor analytic methods and defeat this objective. The purpose of Hierarchical Factor Analysis (HFA) is to detect the presence of and to separate higher-level factor is one which, while independent of all lower-level factors, does affect all variables also affected by certain lower-level factors. While HFA is an extremely powerful tool for accomplishing this purpose, it has not been widely used by the factor analytic community. In part, this is because the method for accomplishing HFA is somewhat complex and not broadly understood. Also, the necessary procedures for HFA have not been available among the more frequently used computerized statistical analytic packages. Recently, Wherry, Sr. (1984) published a discussion of the HFA method, originally developed in 1969 by Wherry, Sr. and Wherry, Jr. Their technique was the first to incorporate a totally objective, computerized procedure for accomplishing HFA. This paper presents several recent modifications to the HFA method.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1986
- Accession Number
- ADA176349
Entities
People
- R. J. Wherry Jr.