Performance-Based Testing and Success in Naval Advanced Flight Training.

Abstract

Roughly 5% of student naval aviators fail the advanced phase of flight training. At this stage of training, the Navy has spent between $300,000 and $1,000,000 per student. Any reduction in this attrition rate through prior screening would be of great economic benefit to the Navy. Computer-based performance tests developed at the Naval Aerospace Medical Research Laboratory were assessed to determine whether they could augment the present medical screening standards and thereby help identify potential failures in advanced flight training. A weak statistical relationship exists between a dual-task performance test, accession source, college major, an aptitude test, and success in advanced flight training. Discriminant analysis was employed to find a linear composite score of these variables that could be used to classify a student as a probable pass or fail in advanced flight training. For example, the model presented in this report reduced failures by 50% at the cost of rejecting roughly 20% of those students who eventually passed. A Bayesian analysis of the success rate parameter showed that this particular model significantly improved the present selection system.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Nov 01, 1992
Accession Number
ADA260838

Entities

People

  • David J. Blower

Organizations

  • Naval Aerospace Medical Research Laboratory

Tags

Communities of Interest

  • Biomedical

DTIC Thesaurus Topics

  • Attrition
  • Biomedical Research
  • Classification
  • Composite Materials
  • Computers
  • Data Science
  • Databases
  • Discriminant Analysis
  • Flight Training
  • Information Science
  • Instructors
  • Performance Tests
  • Probability
  • Psychological Tests
  • Students
  • Task Performance And Analysis
  • Training

Readers

  • Aviation Science / Aeronautics.
  • Psychometric Testing or Psychological Assessment.
  • Regression Analysis.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • Space