Heuristic Classification in a Simulated Air Threat Task
Abstract
Fast and frugal heuristics have proved effective models of human judgment for certain kinds of problems. To further explore the value of this approach, two experiments investigated the decision procedures used by human subjects to perform a cue-based classification task in a simulated air threat assessment task. Threat assessment is the classification of aircraft on the basis of sensor data that can be likened to probabilistic cues. Subjects learned to classify simulated aircraft using four probabilistic cues then classified test sets designed to contrast predictions of several heuristics, including the Take-the-Best-for-Classification (TTB-C) and Pros Rule developed specifically for the threat classification task. Results indicated that a proportion of subjects could be classified using TTB-C and another significant proportion as using the less frugal Pros Rule. No subject was observed to respond as predicted by a Bayesian strategy. Despite predictions that time pressure and perceived uncertainty of cues would affect how many subjects employed TTB-C, no effect of these variables was observed. These results suggest that it is possible to model multi-attribute decision tasks like threat assessment with fast and frugal heuristics but no single heuristic is a general model for the simulated threat assessment task.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 2005
- Accession Number
- ADA596443
Entities
People
- David J. Bryant
Organizations
- Defence Research and Development Canada