Abstract
Purpose: Contrast thresholds in noise for face identification were measured in conditions of accurate cueing of the spatial location, and in which the location was correctly cued only 50% of the time. Our previous study1 had shown that accurate cueing increases contrast sensitivity for face identification. We investigate whether this increase is due to a change in sampling efficiency and/or equivalent noise levels. Thresholds were measured for a single task and a divided attention task. Methods: Contrast thresholds (79% correct responses) were obtained in the presence of zero noise, and externally added visual noise. Attention was directed using the Posner spatial cueing method. An arrow cue (100 ms) preceded the signal display (60 ms). Cue validity was 100%, 50% valid or ‘neutral’. The face image was presented on either the left side or the right side of the fixation point. In the single task experiment, only the ‘face image’ was presented. In the divided attention task, a ‘distracter face’ was presented in the location not occupied by the signal. Sampling efficiency and equivalent noise values were calculated, based on data from an ‘ideal observer’. Results: For both tasks, the increased sensitivity with 100% was due to a significantly (p<0.05) higher sampling efficiency compared to 50% cueing. The presence of a ‘distracter face’ also produced a lower sampling efficiency compared to a single task. Equivalent noise levels were not significantly different (p>0.05) in any of the viewing conditions. Discussion: Human observers are relatively more inefficient in uncertain conditions (50% cueing) than in conditions without uncertainty (100%). In the divided attention task, the presence of a distracter decreased efficiency further, even in the 100% cue conditions, demonstrating a further inability to fully utilise the cueing information whilst showing no change in equivalent noise level. 1. Pardhan S, Tiippana K, Näsänen R, Bhudia M. VSS (2003)
RDF Applied Sciences grant (SP)