Texture segmentation and visual search for pop-out targets. An ERP study.

"Parallel" visual search and effortless texture segmentation were believed to rely on similar mechanisms until Wolfe [Vis. Res. 32 (1992) 757] demonstrated that efficient visual search and effortless texture segmentation are not always the same thing. In a recent study, Meinecke and Donk [Perception 31 (2002) 591] varied display size in a pop-out task and found that, albeit stimulus elements and the task remained the same, different set sizes led to different processing modes. These findings indicate that it may suffice to vary set size in an otherwise unchanged pop-out task to initiate different processing which may be similar to the processing in efficient visual search and in effortless texture segmentation. In four experiments, we further investigated this issue by presenting stimulus arrays of different set sizes while recording event-related brain potentials (ERPs). We found that when display size was increased, detection performance first decreased slightly before it then increased. ERP effects were observed for the posterior N2 (N2p), the N2pc and the P3 component. All three components showed variations with set size; N2p differential amplitude effects were confined to large set sizes, whereas an N2pc was obtained for a broader set size range except for very small set sizes and the largest set size (121 elements). We interpret both the non-monotonic relationship between set size and response data and the variations of ERP components with set size as evidence in favor of different processing occurring for stimulus arrays with small and large set sizes.