The neural processing fate of target and nontarget singleton stimuli was investigated in a series of visual search tasks. The first experiment showed that the ERPs elicited by nontargets defined in the same feature dimension as targets were identical to those of targets until a relatively late divergence in the P3 time range. The second experiment showed that increased stimulus duration allowed slightly faster attentional selection: The ERPs of targets and nontargets now diverged earlier at the N2pc component, although nontargets still elicited a reliable N2pc, which was indicative of the processing of features of these stimuli. It furthermore seemed that task difficulty did not modulate the observed differences between target and nontarget processing. The third experiment investigated the impact of stimulus–response mapping as well as target probability. The former did not modulate the observed differences, and while the latter modulated absolute ERP amplitude, it again did not change the overall pattern of results. No evidence was found in these experiments for differential processing of targets and nontargets defined in the same feature dimension in the time range of the P2 component or before. In a final experiment, targets were compared with nontargets defined in the same or another feature dimension, and for the latter nontargets a clearly much earlier locus of divergent processing was observed, starting at the P2. The N2pc to these nontargets was also strongly suppressed. The relatively late locus of attentional selection between targets and nontargets defined in the same feature dimension suggested that early attentional processes cannot yet fully distinguish between specific within-dimension features.

Electrophysiological measurements were taken from observers performing a visual search within a single feature dimension, and between multiple dimensions. The N2pc to selected singleton target stimuli (rightward tilted lines) was increased when targets varied between feature dimensions, compared to the N2pc to the same rightward tilted line targets, in a condition in which targets varied only between feature values within the same dimension. The anterior and posterior N2 were not reliably modulated, but P3(b) amplitude was higher for singleton present trials that varied between dimensions than for those that varied within. The ERP elicited by singleton absent trials showed reduced P3 amplitude in the between-dimension condition. The electrophysiological modulations were accompanied by increased reaction times in the between-dimension condition, on both singleton present and absent trials. The results suggested that visual target detection is affected by early dimension-specific weighting of the current attentional task set. Furthermore, exhaustively searching multiple feature dimensions to determine the absence of a target incurs dimensional switching costs, possibly at a later stage.

Recently, Yeshurun and Levy (Psychol Sci 14:225–231, 2003 ) have provided evidence for the notion that visual attention impairs the temporal resolution of the visual system. Specifically, the detection of a temporal gap within a visual stimulus was impaired when a cue directed attention towards the spatial location of the stimulus. As this negative cueing effect is important to constrain theories about visual attention, we further investigated this novel effect and assessed whether it truly reflects an attentional effect. Experiment 1 examines whether the negative cueing effect is due to local temporal interference, and Experiments 2 and 3 investigate whether it reflects a luminance confound. The complete pattern of results argues against these alternatives and thus further strengthens the conclusion of Yeshurun and Levy (Psychol Sci 14: 225–231, 2003 ).