Measures of Situation Awareness: An Experimental Evaluation
Abstract
Both explicit and implicit measures of situation awareness (SA) were evaluated in a series of experiments in order to assess their reliability and two kinds of validity: criterion and construct. In all of the experiments, subjects performed a simulated combat task in which they had to monitor the positions of enemy, friendly, and neutral objects. In addition, subjects had to attack and defend themselves against enemy objects. A memory probe procedure was used to explicitly assess two components of SA: location and identity awareness. In addition, a signal detection analysis was used to provide an implicit measure of SA. Test-retest correlations indicated that the location awareness measure was much less reliable than the others. Criterion validity was evaluated by correlating the SA measures with probability of a kill in the combat task. Although the SA metrics seemed to be fairly good predictors of kill probability, the best predictor was a measure of behavioral workload. Predictions of multiple resource theory were used to evaluate construct validity. In particular, it was predicted that difficulty in maintaining identity awareness would not affect location awareness, and this prediction was largely supported. Nevertheless, other aspects of the data seemed to contradict current versions of multiple resource theory.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 1991
- Accession Number
- ADA262732
Entities
People
- Martin L. Fracker