Shifts of visuospatial attention to invisible (metacontrast-masked) singletons: Clues from reaction times and event-related potential

In the current study, we tested whether a masked and, thus, invisible singleton-cue captures atten-tion in a stimulus-driven manner or in a top-down contingent manner. The manual RT (Reaction Time) capture effect with the invisible singleton--cue decreased substantially when a match between the singleton-cue and the top-down controlled set of searched-for target features was also decreased. By contrast, with the PCN (Posterior Contralateral Negativity), an electrophysiological measure of the capture of visuospatial attention by the invisible singleton-cue, no significant decrement was observed. Taken together, the results support the assumption that an invisible singleton-cue can capture attention in a stimulus-driven manner, and that different delays in the deallocation of attention (i.e., attention is deallocated more efficiently from acue that does not match the top-down controlled set than from a cue that does match the same set) account for the weaker manual RT capture effect with a set- -nonmatching invisible singleton-cue.