Express attentional re-engagement but delayed entry into consciousness following invalid spatial cues in visual search

Abstract

Background: In predictive spatial cueing studies, reaction times (RT) are shorter for targets appearing at cued locations (valid trials) than at other locations (invalid trials). An increase in the amplitude of early P1 and/or N1 event-related potential (ERP) components is also present for items appearing at cued locations, reflecting early attentional sensory gain control mechanisms. However, it is still unknown at which stage in the processing stream these early amplitude effects are translated into latency effects.

Methodology/principal findings: Here, we measured the latency of two ERP components, the N2pc and the sustained posterior contralateral negativity (SPCN), to evaluate whether visual selection (as indexed by the N2pc) and visual-short term memory processes (as indexed by the SPCN) are delayed in invalid trials compared to valid trials. The P1 was larger contralateral to the cued side, indicating that attention was deployed to the cued location prior to the target onset. Despite these early amplitude effects, the N2pc onset latency was unaffected by cue validity, indicating an express, quasi-instantaneous re-engagement of attention in invalid trials. In contrast, latency effects were observed for the SPCN, and these were correlated to the RT effect.

Conclusions/significance: Results show that latency differences that could explain the RT cueing effects must occur after visual selection processes giving rise to the N2pc, but at or before transfer in visual short-term memory, as reflected by the SPCN, at least in discrimination tasks in which the target is presented concurrently with at least one distractor. Given that the SPCN was previously associated to conscious report, these results further show that entry into consciousness is delayed following invalid cues.

Figure 1. Stimulus sequence in valid, invalid, and no-target trials.

One four-alternative discrimination speeded response was required on each trial (except in no-target trials) as to the location of the gap in the target-colored square. Colors were equiluminant red, green, blue and yellow in the actual experiment. The target-colors were counterbalanced between participants. In this example, target-colors are red and yellow.

Figure 2. Grand-average event-related potential (ERP) waveforms time-locked to the target display onset at PO7 and PO8 sites for valid, invalid, and no-target trials.

Contralateral and ipsilateral were defined in relation to the cued location.

Figure 3

A) Contralateral minus ipsilateral difference waves time-locked to target display onset at PO7/PO8 sites in valid, invalid, and no-target trials. Contralateral and ipsilateral were defined relative to the cued location in no-target trials, but were redefined in respect to the target location in valid and invalid trials. B) Contralateral minus ipsilateral corrected difference waves for valid and invalid trials. The corrected valid difference wave was obtained by subtracting the no-target difference wave from the valid difference wave, whereas the invalid corrected difference wave was obtained by summing the no-target difference wave and invalid difference wave. Shown are the 10 Hz low-pass filtered waveforms.

Figure 4. Scatterplots of the A) N2pc latency effect and RT effect, and of the B) SPCN latency effect and RT effect.

Individual N2pc and SPCN latencies were recovered from the jackknife values according to the formula presented in the Methods section.

Figure 5. Scalp distributions of the electrical potentials measured during the N2pc (230–270 ms for both the valid and invalid conditions) and SPCN (360–400 ms in the valid condition and 410–450 ms in the invalid condition) time windows.

The scalp distributions were calculated on the basis of the corrected contralateral minus ipsilateral difference waves used to calculate the N2pc and SPCN, and are thus symmetrical about the midline.