There is a fundamental lack of understanding concerning the relationship between the fidelity of driving simulators (extent to which simulators replicate reality) and validity (extent to which drivers behave as they would in equivalent real-world situations). For distraction research, knowledge on how drivers prioritise primary driving tasks over secondary tasks can be a potential indicator of simulator validity. Theoretical propositions, consequently established in on-road research, show that higher primary task demands result in increases in time pressure and forward spatial-scene uncertainty, prompting drivers to return their vision to the road ahead in a predictable fashion. This paper addresses whether drivers within simulators of varying fidelities exhibit the attention behaviour and time-sharing strategies predicted by theoretical and empirical research. Twenty-four drivers drove in two (low and medium-fidelity) simulators following a simple rural-road scenario comprising straight and curved-road driving, whilst performing a series of visual search tasks using an in-vehicle display. Results showed that drivers exhibited the predicted visual attention behaviours in both simulators. Nevertheless, the low-fidelity simulator was associated with reduced lane keeping performance, due primarily to physical, rather than psychological differences in the driving experience. It is concluded that driving simulators of varying fidelity can successfully be employed in distraction research. (Author/publisher)
Abstract