Selecting and Representing Information Structures for Battlefield Decision Systems
Abstract
The objective of this research was to identify and apply experimental techniques for locating and evaluating data patterns and informational 'chunks' which are meaningful to the battlefield commander in his decision-making tasks. The identification of meaningful chunks of information is useful in specifying criteria for the development of decision-aiding algorithms that search for, classify, and display information. Two main tasks were used for collecting data about informational chunks: reconstruction and copying. In the reconstruction task an experimental run consisted of a successive number of trails. In each trial the participant first viewed a battlefield map scenario for either ten seconds or one minute; then it was removed and he was asked to reconstruct it. In general, the experts' performance was superior to that of the novices on accuracy of reconstruction both for structured scenarios representing likely battlefield situations and unstructured scenarios representing situations unlikely to occur on a real battlefield. Both the IPT and sequential techniques revealed that the experts' chunks were tactically meaningful with high frequency, while the novices' chunks were not. The basic element of a chunk for the expert was the tactical relation between two or more battlefield units rather than the battlefield unit itself. However, even the novice chunked by relating symbols in some meaningful way.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1979
- Accession Number
- ADA071117
Entities
People
- Albert N. Badre
Organizations
- Georgia Tech