Advanced Personnel Testing Project
Abstract
A new experimental test battery, known as the Advanced Personnel Testing (APT) battery, was investigated in two contexts relating to the Armed Services Vocational Aptitude Battery (ASYAB). Interim results of these investigations are documented in an American Psychological Association Conference paper (first author Goff) and a briefing given to OASD (Sawin). However, the results repofled in this paper are based on the full sample collected. Both investigation contexts were large n samples of Air Force basic recruits who took all tests. One context (n = 9,325) examined how the best factor structures for the APT and ASVAB tests (modeled separately) were related to each other when assessed in the same latent structure model. Of special interest was a factor (Fact Learning) present in APT but not in ASVAB. A second context involved assessing the incremental validity of APT relative to ASVAB for prediction of technical school grade: in a large-flow school (i.e., security police, n =2,270). While the ASYAB predicted technical school grades more strongly than the APT battery, latent structure analyses, on correlations corrected for range-restriction (on the ASVAB), indicated at !east some unique prediction of grades by the APT Fact Learning factor (r = 11; z = 4.3). Standard regression analyses on the same correlations also indicated a significant incremental r (r = .014; F(12,2247) = 10.5) for APT, with fact-learning and skill-learning tests providing most of the increments r. This work is legacy work from a closed work unit. There is no active selection research at the Air Force Research Laboratory.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 01, 2001
- Accession Number
- ADA399062
Entities
People
- Ginger Nelson Goff
- Jim Earles
- Linda Sawin
- Scott R. Chaiken