Anthropometric Data Reduction Using Factor Analysis and Stepwise Regression
Abstract
A two-phase approach using factor analysis and stepwise regression was used to identify the most important subset of anthropometric variable from the 1967 survey of USAF flying personnel. Factor analysis by groups was used initially to identify a subset of variables which explain the most variation within the population. Factors identified in the initial analysis were subsequently varimax rotated. Of 185 variables, 32 were selected in this way. In phase 2, stepwise regression is used to regress the remaining variables on the set identified in phase 1. In this way, the variables not selected as most important can be predicted based on the selected subset. The necessary prediction equations, as well as quality of fit indicators, are given. In all, 32 variables are identified as being the most useful subset for explaining the total variance of all 185 variables. This represents a significant reduction in the amount of data which must be collected during an anthropometric survey.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 01, 1980
- Accession Number
- ADA121488
Entities
People
- H. F. Martz Jr.
Organizations
- Texas Tech University