Abstract
Over the last decades, the visual search paradigm has proved to be a test bed for competing theories of visual selective attention. In a standard visual search task, participants are confronted with a display that can contain a target item amongst a variable number of distractor items, with reaction time (RT; and accuracy) being the critical performance measure(s). Interestingly, across search studies, the information necessary to decide upon the correct motor response is highly variable, for instance, information about the presence, spatial location, or, respectively, identity of the target item. Thus, the question emerges whether estimates of the time taken for (1) visual selection, (2) deciding on the motor response, and (3) executing it generalize across search tasks, or whether they are specific to the demands of a particular task set. To systematically examine this issue, we presented physically identical stimulus material in four different search task conditions (localization, detection, discrimination, and compound) and combined reaction time performance with two specific electrophysiological markers directly linkable to pre-attentive (Posterior Contralateral Negativity, PCN) and post-selective levels (Lateralized Readiness Potential, LRP) of the information-processing stream. At the behavioural level, RTs were found to be fastest for localization, slowest for compound, and intermediate for detection and discrimination responses. Electrophysiologically, this effect of task set had a strong impact on the timing and amplitude of the stimulus- and response-locked LRP (indexing motor response decisions), but not PCN (indexing visual selection), component. This dissociation clearly demonstrates that the choice of a task has no influence on the stage of pre-attentive visual coding, which mediates focal-attentional selection of the target item, whereas processes that occur after target selection are dependent on the nature of the task.