Know Your Options –Analyzing Human Decision Making in Dynamic Task Environments with State-Space Methods

Abstract

A recurring problem when trying to analyze human decisions in dynamic task environments is that decisions are evaluated with an incomplete understanding of available choices. While simulators can be used to record actions actually taken by the operator, adequate information on available alternative strategies is often missing. Especially the consequences, costs and utilities of different “what-if” solutions remain unknown. This complicates the analysis of human performance in various ways, particularly the understanding of human errors (asking why a decision did in fact make sense to the operator at the time of operation) and the judgement of operators’ situation awareness in terms of ability to predict the possibly upcoming system states. To deal with these problems the contribution introduces state-space based methods which evaluate human decisions in relation to the specific solution spaces. An application example and empirical data from an air traffic control simulation is investigated where subjects have to avoid traffic conflicts with appropriate control clearances. The analysis reveals for each dynamic situation, which traffic conflicts were solvable with which solution strategies (e.g. changes of speed, direction or altitude for one or more aircraft) and what was actually implemented. Situation Awareness tests are used to find out which of the potential solutions are correctly identified by the individual operators.