VARSEL: Variable Selection for Multiple-Purpose Prediction Systems in the Absence of External Criteria.
Abstract
The absence of suitable external criteria for aptitude tests is a recurrent problem for test, battery, and inventory developers in selecting items or tests for inclusion in final operational instruments. This report presents a computing algorithm developed for use when no adequate external selection criterion is available. The algorithm uses a multiple linear regression technique and an accretion variable selection process to start with a large pool of variables and select a minimum subset which can account for the domain of reliable variance measured by all variables in the pool. Two example applications are presented: (a) selection of a subset of tests from a battery of 56 tests and (b) selection of a subset of job attitude items from a pool of 348 items.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1976
- Accession Number
- ADA025328
Entities
People
- R. Bruce Gould
- Raymond E. Christal
Organizations
- Brooks Air Force Base