Speech Shadowing while Driving: On the Difficulty of Splitting Attention Between Eye and Ear

Abstract

We investigated the role of cross-modal links in spatial attention in modulating the efficiency of dual-task performance. The difficulty of combining speech shadowing with a simulated driving task was modulated by the spatial location from which the speech was presented. In both single- and dual-task conditions, participants found it significantly easier to shadow one of two auditory streams when the relevant speech was presented from directly in front of them, rather than from the side. This frontal speech advantage was more pronounced when participants performed the demanding simulated driving task at the same time as shadowing than when they performed the shadowing task alone. These results demonstrate that people process auditory information more efficiently (with a lower overall dual-task decrement) when relevant auditory and visual stimuli are presented from the same, rather than different, spatial locations. These results are related to recent findings showing that there are extensive cross-modal links in spatial attention, and have clear implications for the design of better user interfaces.