Validating a dynamic model of decision making on an air traffic control conflict detection task

Human factors research is responsible for the development of models of decision making performance that allow for the analysis of performance on tasks. Such models have the potential to benefit industrial contexts by assisting with measurement, design, interventions and evaluations of systems. Signal Detection Theory is widely used to analyse decision making on industrial tasks. However, it is limited by the fact that it does not account for response times and speed-accuracy tradeoffs. Thus, Dynamic Signal Detection (DST) theory has been proposed as a means to overcome the crucial limitations of the former model. The model estimates 3 parameters – discriminability, bias and threshold. However, the DST model has not yet been validated. Thus, this experiment aimed to test the validity of the DST model on an Air Traffic Control conflict detection task, where 51 participants completed the task under varying conditions of biased instructions, base rates and speed versus accuracy emphases. The 3 DST parameters were calculated and analyses determined that the bias and threshold parameters respond to these experimental manipulations in ways that are consistent with the theoretical perspective of DST. Inconsistent results pertaining to the discriminability parameter are discussed along with implications of the current results and future directions for research