An Analytic Approach to Estimating the Generalizability of Tank Crew Performance Objectives

Abstract

Generalizability is an important consideration in item selection, if test performance is expected to be predictive of performance on a larger domain. In this study, analytic means were used to estimate the generalizability of performance objectives which were candidates for inclusion in a tank gunnery test. Cluster analyses were conducted to identify subsets or famities of objectives in the domain which were homogeneous in terms of behavioral elements. Then indexes of generalizability were computed to describe how generalizable a given objective was to others in its family. These analyses provided a useful framework for applying the criterion of generalizability during item selection. The data base suggests including at least one objective from each of the subdomains that were identified. Given relatively limited resources, for example, the most generalizable objective from each of the eight major famities would be included in the gunnery test. Given fewer constraints, the most generalizable objective from each of the 17 family or constituent subfamilies would be selected. Although multiples of eight or 17 objectives represent the ideal, a partial sampling across families or subfamilies is also possible, and may even become necessary when other factors such as criticality, existing range facilities and ammunition costs are considered.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 1977
Accession Number
ADA077939

Entities

People

  • Andrew M. Rose
  • G. Gary Boycan

Organizations

  • U.S. Army Research Institute for the Behavioral and Social Sciences

Tags

Communities of Interest

  • Human Systems

DTIC Thesaurus Topics

  • Ammunition
  • Combat Readiness
  • Databases
  • Frequency
  • Guns
  • Inclusions
  • Machine Guns
  • Military Research
  • Plastic Explosives
  • Simulations
  • Social Sciences
  • Test And Evaluation
  • Training
  • Weapons

Readers

  • Computational Modeling and Simulation
  • Marksmanship and Weaponry.
  • Molecular Genetics