A Study of Variables to Help Predict Navigator Training Success and Classification

Abstract

This study attempted to build statistical models which could possibly be used to help predict an individual's final outcome of pass or fail from Undergraduate Navigator Training (UNT). Also, the research tried to develop models to help place a potential trainee into one of the three training tracks at UNT. The variables studied were test scores from a computerized testing device which measured psychomotor and cognitive skills of individuals entering training. The Air Force Officers Qualifying Test (AFOQT) scores were also variables considered. Data was from 317 trainees from the years 1988 to 1990. Discriminant analysis was applied in an effort to place an individual accurately into one of the groups of pass or fail and into one of the three tracks based on his/her scores. Logistic regression was performed on the binary response of pass/fail to give models which would predict probability of passing using the test scores. Factor analysis was used to explore the underlying dimensions of all the variables. Some variables were found to be important predictors. Although models were built from the analyses, the study could use more data on individuals who failed from training.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 01, 1991
Accession Number
ADA238354

Entities

People

  • Gary A. Hagler

Organizations

  • Air Force Institute of Technology

Tags

Communities of Interest

  • Air Platforms
  • Biomedical
  • Human Systems

DTIC Thesaurus Topics

  • Air Force
  • Data Science
  • Data Sets
  • Databases
  • Discriminant Analysis
  • Discriminate Analysis
  • Factor Analysis
  • Flight Training
  • Human Resources
  • Information Processing
  • Information Science
  • Pilots
  • Regression Analysis
  • Statistical Analysis
  • Students
  • Trainees
  • Training

Readers

  • Computational Modeling and Simulation
  • Psychometric Testing or Psychological Assessment.