Free
Article  |   June 2015
Human feature-based attention consists of two distinct spatiotemporal processes
Author Affiliations
  • Daniela Gledhill
    Department of Human Neurobiology, Center for Cognitive Science, University of Bremen, Bremen, Germany
    d.gledhill@uni-bremen.de
  • Cathleen Grimsen
    Department of Human Neurobiology, Center for Cognitive Science, University of Bremen, Bremen, Germany
    cgrimsen@uni-bremen.de
  • Manfred Fahle
    Department of Human Neurobiology, Center for Cognitive Science, University of Bremen, Bremen, Germany
    Division of Optometry & Visual Science, The Henry Wellcome Laboratories for Vision Sciences, City University, London, UK
    mfahle@uni-bremen.de
  • Detlef Wegener
    Department of Theoretical Neurobiology, Center for Cognitive Science, University of Bremen, Bremen, Germany
    wegener@brain.uni-bremen.de
Journal of Vision June 2015, Vol.15, 8. doi:10.1167/15.8.8
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Daniela Gledhill, Cathleen Grimsen, Manfred Fahle, Detlef Wegener; Human feature-based attention consists of two distinct spatiotemporal processes. Journal of Vision 2015;15(8):8. doi: 10.1167/15.8.8.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

In human and nonhuman primates, goal-directed behavior requires the selection of relevant pieces of information from the multitude of simultaneous sensory inputs. Feature-based attention (FBA) plays a crucial role in this selection by improving the neuronal representation of an attended stimulus feature. Of particular interest for understanding the neuronal mechanisms behind FBA is the processing fate of spatially unattended stimuli, either sharing the attended feature attribute or belonging to the attended or to a nonattended feature dimension. Using a wide range of cue/stimulus combinations, we investigated event-related potentials from the human brain, recorded under conditions of different feature attention but constant visual stimulation. We found that neural processing of visual stimuli sharing the dimension or the attribute of the attended target is associated with two distinct spatiotemporal processes, particularly prominent during the selection negativity period. Dimension-based modulation of neural signals first emerged over frontal electrode sites, and temporally preceded and accompanied attribute-specific FBA effects at occipital, parieto-occipital, and parietal electrodes. The findings suggest a process of FBA that not only increases responses of those neurons particularly tuned to the attended attribute but also modulates activity in the cortical module that is selective for the feature dimension to which the attended attribute belongs.

Introduction
In everyday life, stimulus features such as color or motion might be of different importance for the current behavioral goal: In sports, exact representation of motion direction and speed might be substantial to successfully hitting the ball, while in other situations color can be more relevant, as when driving a car through a city with many traffic lights. Feature-based attention (FBA) describes the neuronal process by which a specific stimulus feature receives enhanced processing, thus resulting in improved behavioral performance such as increased accuracy and reduced detection times (Carrasco, 2011; Maunsell & Treue, 2006). 
On a neuronal level, FBA has been described as a selective response enhancement of those neurons that are tuned to the attended feature attribute. For example, recordings in monkey motion-sensitive area MT have shown that direction-selective neurons with receptive fields away from the attended location were significantly more active if the attended motion direction matched the preferred motion of these neurons (Martinez-Trujillo & Treue, 2004; Treue & Martínez Trujillo, 1999). Because this feature-specific modulation is independent of the spatial focus of attention, the results indicate a globally effective mechanism facilitating the representation of the attended feature and thereby supporting selection of the relevant visual information independent from its spatial coordinates. Imaging experiments have confirmed and extended this view in the human visual cortex by showing a globally effective processing enhancement in response to the attended feature not only for the motion domain but also during color and orientation processing (Liu, Larsson, & Carrasco, 2007; Saenz, Buracas, & Boynton, 2002; Serences & Boynton, 2007). This global, FBA-dependent processing enhancement is behaviorally effective, as shown psychophysically by both dual-task experiments and measurements of reaction times and aftereffects (Liu & Mance, 2011; Saenz, Buracas, & Boynton, 2003; Wegener, Ehn, Aurich, Galashan, & Kreiter, 2008), even for task-irrelevant stimuli (White & Carrasco, 2011). 
In the context of these studies, the term “feature” usually refers to a particular feature attribute, such as upward motion, the color red, or horizontal orientation. However, attending to the specific direction of a moving stimulus might be different from attending to motion in general, as in the sports example from earlier. Likewise, to generally facilitate color processing, attention to a specific hue is apparently different from attending to color in general. In their seminal work, Corbetta, Miezin, Dobmeyer, Shulman, and Petersen (1990) reported distinct cortical activation patterns when subjects were discriminating objects according to either their shape, color, or velocity. The results of this study were suggestive of an FBA-dependent modulation of neuronal activity particularly in those extrastriate visual areas specialized for processing the selected feature. Another event-related functional MRI study showed higher V5 baseline activity when participants were expecting a motion target, and higher V4 baseline activity when they were expecting a color target, even in the absence of visual stimulation (Chawla, Rees, & Friston, 1999b). In these studies, subjects were attending not to a specific feature attribute but rather to a class of features. Yet the fact that the authors still found nonspatial, feature-specific attentional modulation suggests that instead of being exclusively specific for the selected feature attribute, FBA may generally increase neuronal activity in cortical modules processing the relevant feature class. These data confirmed results from previous visual search experiments predicting a selective benefit for feature dimensions that were known in advance. When tested for singletons that were presented with stimuli either from the same feature dimension or from a different dimension, search was parallel in both cases—i.e., with no set-size effects—but for the within-dimension condition, reaction times (RTs) were faster than for the across-dimension condition (Found & Müller, 1996; Müller, Heller, & Ziegler, 1995). Based on these data, the dimensional-weighting account was proposed, according to which neuronal responses are biased towards the feature dimension containing the target attribute; subsequent electroencephalogram and imaging work has provided further support for this model (Gramann, Toellner, Krummenacher, Eimer, & Müller, 2007; Gramann, Töllner, & Müller, 2010; Pollmann, Weidner, Müller, & von Cramon, 2000; Weidner, Pollmann, Müller, & von Cramon, 2002). 
Yet a general dimension-based weighting of neuronal responses predicts increased activity of neurons even if they are stimulated with a task-irrelevant feature attribute. In other words, if attention is directed to, e.g., upward motion, a dimension-based approach would predict increased responses to motion even if visual stimulation consists of downward motion. Conversely, if attention is directed to green, color processing should be generally increased but motion processing attenuated. These predictions have been tested in recent single-cell studies, but derived different results. For example, motion-selective MT neurons with a tuning for the attended motion direction are not modulated—or can even become inhibited—if stimulated with motion opposite to the attended direction (Martinez-Trujillo & Treue, 2004; Treue & Martínez Trujillo, 1999). Moreover, attentional effects in area MT seem not to depend on perceptual demands such as attending to a certain feature dimension: If the behavioral task required discrimination of a motion signal, effects of feature-based attention were found to be almost identical to those in a task requiring the discrimination of the color signal of the same stimulus (Katzner, Busse, & Treue, 2009) or partially showed even decreasing responses for the attend-motion as compared to the attend-color condition (Chen, Hoffmann, Albright, & Thiele, 2012). Based on these results, it remains unclear whether FBA modulates visual responses depending on the attended feature dimension or whether attribute- and dimension-specific processes exist concurrently. 
We investigated this issue by recording event-related potentials (ERPs) during a delayed-match-to-sample (DMS) task, using a visual stimulation protocol that allowed for the simultaneous analysis of attribute- and dimension-specific FBA. The logic behind the experimental protocol was to directly compare neural modulations in response to a spatially nonattended stimulus if it shared the attribute of the target (i.e., if it was identical to it) with those to stimuli defined in the same feature dimension versus stimuli from another feature dimension. We show that FBA consists of two spatiotemporally distinct processes, one of which is found for all stimuli sharing the relevant feature dimension and the other of which is specific to attribute-matching stimuli. 
Material and methods
Participants
Experiments 1 and 2 were conducted with 11 (seven women, mean age: 26.0 years, range: 22–31 years) and seven (five women, mean age: 25.3 years, range 22–28 years) naïve subjects, respectively, who were paid for participation. All were right-handed and were tested for normal or corrected-to-normal vision (Bach, 1996, 2007), and for normal color and stereo vision (Ishihara, 1917; Kanazawa & Kanatani, 1997). Participants received detailed task instructions but remained naïve to the actual purpose of the experiment. Prior to electrophysiological recordings, all participants performed a training block. The study conformed to the Code of Ethics of the World Medical Association (Declaration of Helsinki) and was approved by the local authorities. 
Visual stimulation and task
Stimuli were displayed on a Samsung SyncMaster 1100 MB monitor (NVIDIA GeForce 8500GT; 1280 × 1024 pixels; 100 Hz) in a darkened room. A constant viewing distance of 45 cm was provided by a head-chin rest. Maintenance of fixation within 1.5° was monitored online using a custom-made remote video-oculography system (WAT-902H2-Supreme; Minolta Objective Lens MD 50 mm 1:1.7). 
Experiment 1 consisted of a DMS task requiring detection of a target stimulus that possessed either a precued color hue or a precued motion direction. Stimuli consisted of pairs of random-dot patterns (RDPs) presented on a dark background. Each RDP (diameter: 6.4°; eccentricity: 7.0°) contained 160 elements (each 0.1° in diameter), with occasionally some frames having a few less due to rounding during computation of feature coherence. Test RDPs were defined either by color or by motion direction (Figure 1A). Color-defined test RDPs (red, green, blue, and yellow) were static and displayed with a color coherence of 60%. The remaining dots were composed of seven other uniformly distributed colors. Motion-defined test RDPs were achromatic and drifting at 2.5°/s in one of four directions (0°, 90°, 180°, and 270°). Motion coherence was 90% due to higher detection difficulty. The remaining 10% of dots were moving in seven different directions separated by 45°. Color hues and the shade of gray of all dots were kept isoluminant (25.2 cd/m2). Nontest RDPs either contained both a main color and a main motion direction or contained none of them (e.g., red-90°, gray-static; Figure 1B). 
Figure 1
 
Experimental conditions and stimulus pools. Example (A) test and (B) nontest RDPs. (C) Three stimuli instructing the same attentional selection of a 90° target motion direction at the precued location (spatial focus). For visual-presentation purposes, RDPs are shown as simplified discs representing the main color hue or motion direction. Outside the spatial focus of attention, RDPs may be defined by having the other feature dimension (nonattended), by having the same dimension but a different feature attribute (dimension-attended), or by sharing the target's feature attribute (attribute-attended). (D) Test stimulus combined with different cues, defining three different attentional conditions. (E) Subset of stimuli of the three stimulus pools, utilizing the same RDPs and cues but in different combination. Each test stimulus was presented in an attend-left and attend-right version (only attend-right stimuli are shown). ERPs were averaged over all templates of each pool. See Material and methods for details.
Figure 1
 
Experimental conditions and stimulus pools. Example (A) test and (B) nontest RDPs. (C) Three stimuli instructing the same attentional selection of a 90° target motion direction at the precued location (spatial focus). For visual-presentation purposes, RDPs are shown as simplified discs representing the main color hue or motion direction. Outside the spatial focus of attention, RDPs may be defined by having the other feature dimension (nonattended), by having the same dimension but a different feature attribute (dimension-attended), or by sharing the target's feature attribute (attribute-attended). (D) Test stimulus combined with different cues, defining three different attentional conditions. (E) Subset of stimuli of the three stimulus pools, utilizing the same RDPs and cues but in different combination. Each test stimulus was presented in an attend-left and attend-right version (only attend-right stimuli are shown). ERPs were averaged over all templates of each pool. See Material and methods for details.
Test stimuli always consisted of a chromatic-static and an achromatic-moving test RDP (e.g., green-static together with gray-270°). Each test stimulus was also shown in a mirrored version, in which the positions of the two RDPs were exchanged and attention was directed to the other side, thus avoiding a spatial bias in stimulation. Nontest stimuli contained at least one nontest RDP, combined with another nontest or test RDP (e.g., green-180° with red-90° or gray-static with gray-270°). Only test stimuli were used for analysis, while nontest stimuli were used only to increase the overall attentional demand of the task and to avoid participants' categorizing both types of RDP by just one feature, i.e., either color or motion being present or absent. Test and nontest stimuli were presented in a ratio of 3:1. 
Test stimuli were used to define three experimental conditions regarding the stimulus outside the spatial focus of attention: nonattended, dimension-attended, and attribute-attended (Figure 1C, D). For example, if the cue indicated an achromatic target stimulus moving upwards, the RDP opposite to the spatial focus of attention was considered nonattended if it was colored and static (i.e., if neither the feature attribute nor the feature dimension matched the target). Alternatively, it was considered dimension-attended if it was achromatic and moving but with a motion direction different from the precued target motion direction (i.e., if only the feature dimension and not the feature attribute matched the target), and it was considered attribute-attended if it was identical to the precued target RDP (i.e., if both feature dimension and feature attribute matched the target; Figure 1C). By this scheme, different cues allowed for the presentation of the same test stimulus in each of the three attention conditions: For example, a static green RDP outside the spatial focus of attention was considered nonattended if the cue indicated an achromatic RDP moving upwards, it was considered dimension-attended if the cue indicated a static red target, and it was considered attribute-attended if the cue indicated a static green target (Figure 1D). Since this logic applies to any of the cues and pairs of test RDPs, using the entire combination range of cues and RDP pairs allowed for the creation of stimulus pools containing cues and RDP pairs in identical type and number, differing only in their specific combination (Figure 1E). 
The DMS task is outlined in Figure 2A. Target location (left/right) was kept constant in blocks of 64 trials each, and the target's feature dimension (color/motion) was kept constant in sub-blocks of 16 trials. Each session consisted of eight blocks, resulting in a total of 512 trials. Target feature attributes were cued at the beginning of each trial and varied from trial to trial. About 25% of the trials ended without the appearance of a target, to prevent expectation effects at late trial intervals. Trials with behavioral errors (false alarms, misses, eye movements deviating more than 1.5° from the fixation point) were repeated later in the same sub-block. 
Figure 2
 
Task design and behavioral performance of Experiment 1. (A) Example trial of the DMS task. Target location was kept constant for a block of 64 trials, and target dimension was kept constant for sub-blocks of 16 trials each. For visualization purposes, RDPs are shown as simplified discs. A foveally presented cue indicated the target attribute (specific color hue or a gray arrow pointing towards a specific motion direction) at trial start. Prior to target appearance, up to three test stimuli were presented, separated by a jittered interstimulus interval. RDPs at the spatially nonattended location shared either the target's attribute, the target's dimension, or none of these (attribute-attended, dimension-attended, and nonattended conditions). (B) Mean percentages of hits and errors, and cumulative RT distributions separately for color and motion trials, and (C) for attend-left and attend-right trials. Error bars indicate standard deviation.
Figure 2
 
Task design and behavioral performance of Experiment 1. (A) Example trial of the DMS task. Target location was kept constant for a block of 64 trials, and target dimension was kept constant for sub-blocks of 16 trials each. For visualization purposes, RDPs are shown as simplified discs. A foveally presented cue indicated the target attribute (specific color hue or a gray arrow pointing towards a specific motion direction) at trial start. Prior to target appearance, up to three test stimuli were presented, separated by a jittered interstimulus interval. RDPs at the spatially nonattended location shared either the target's attribute, the target's dimension, or none of these (attribute-attended, dimension-attended, and nonattended conditions). (B) Mean percentages of hits and errors, and cumulative RT distributions separately for color and motion trials, and (C) for attend-left and attend-right trials. Error bars indicate standard deviation.
Each trial was initiated by central fixation (white spot, 0.4° square) and a button press. The attribute cue (colored bar or gray arrow of identical size and luminance; Figure 1C, D) appeared with a delay of 250 ms at the center of the screen (with the fixation point superimposed on it) and remained visible for the entire trial. Following the cue period of 1000 ms, sequences of one to four RDP pairs were presented. Each RDP pair was shown for 300 ms and was separated from the subsequent RDP pair by a jittered interstimulus interval of 1000 ± 100 ms (10-ms steps). Trials of different lengths (i.e., consisting of a sequence of one, two, three, or four RDP pairs) were presented in equal quantity. During pretarget intervals, stimuli differed in the degree of correspondence between the feature attribute of the cued target and the feature attribute of the spatially uncued RDP. They belonged to either the nonattended, dimension-attended, or attribute-attended condition if they were part of a test stimulus. Alternatively, they belonged to none of these conditions if they were part of a nontest stimulus. 
Participants were required to respond to the appearance of the target RDP by releasing the button pressed at the beginning of the trial as fast as possible. Very fast RTs were indicated by a particularly pleasant auditory feedback, primarily to keep the participants' engagement high during the entire course of the experiment. Trials were separated by a blank intertrial interval of 1500 ms. 
Experiment 2 was designed as a control condition for the results of Experiment 1. Here, the task was irrelevant for data analysis and was only used to keep participants engaged. Participants were asked to detect the appearance of a dot stimulus displayed randomly at one of six possible locations, while keeping their gaze at fixation. Prior to the onset of the dot, two different combinations of test RDPs and cues were shown. RDPs and cues were irrelevant to the task and were only passively observed by the participants. The rationale of Experiment 2 was to test whether differences in RDP/cue pairing are capable of inducing amplitude differences in ERPs. The experiment was performed with a subset of stimuli taken from each of the stimulus pools of Experiment 1. As in Experiment 1, these reduced stimulus pools consisted of RDP pairs and cues identical in type and number but differing in their specific combination (compare Figure 1E). All stimuli were also shown in a mirrored version, to provide the same visual stimulation for both hemispheres. The timing of Experiment 2 was identical to that of Experiment 1, with the exception that the second interval was followed by the appearance of the dot stimulus, randomly delayed between 300 and 800 ms. 
Electrophysiological recordings
The electroencephalogram was recorded continuously (500-Hz sampling rate, 0.5–120 Hz bandpass filtered, Neurofax EEG-1100, Nihon Kohden, Germany) from 25 silver/silver chloride electrodes (Fp1, Fp2, F7, F3, Fz, F4, F8, T7, C3, Cz, C4, T8, P7, P3, Pz, P4, P8, PO7, PO3, POz, PO4, PO8, O1, Oz, O2), corresponding to the 10-20 system (American Electroencephalographic Society, 1994). Signal traces were referenced to Cz and re-referenced off-line to the average of both mastoids. Electrode impedances were kept below 10 kΩ. In addition to video oculography, eye movements were recorded with two differential electrodes above and below one eye. 
Data analysis
The data were corrected for outliers by assigning color and motion trials with RTs slower than the mean RT plus two standard deviations of all color and motion trials, respectively, as misses. Performance was calculated by relating the sum of correct trials (including correct rejections of trials ending without target appearance) to the sum of all trials, disregarding fixation errors. 
For correct trials, pretarget test ERPs were averaged time-locked to stimulus onset (−100 to 500 ms) and baseline-corrected to the 100-ms prestimulus interval using the EEGLAB v10.2.5.6 toolbox (Delorme & Makeig, 2004) in MATLAB 7.13 (The MathWorks, Natick, MA). Epochs showing excessive muscle activity or amplitudes exceeding ±70 μV were rejected. For visual-presentation purposes, the data were smoothed with a weighted local regression after averaging (LOESS filter, second-order polynomial, span = 0.25). In both experiments, one subject was excluded from analysis due to large electroencephalogram artifacts. 
Data analysis was restricted to electrode sites contralateral to spatially unattended stimuli ([F3, F4], [C3, C4], [P3, P4], [PO3, PO4], [O1, O2], for attention to the left and right hemisphere, respectively). Dimension-specific FBA effects (DimAtt) were calculated by the amplitude difference of ERPs from the nonattended and the dimension-attended condition:    
Attribute-specific FBA effects (AttrAtt) were calculated by subtracting DimAtt from the ERP amplitude difference of the nonattended and the attribute-attended conditions:    
This logic is based on the fact that each stimulus matching the targets' feature attribute also matched its feature dimension. Therefore, proper isolation of attribute-specific ERP modulation needs consideration of those effects already showing up in response to stimuli not matching the attended feature attribute but belonging to the attended feature dimension. 
Statistical procedures were conducted with the Statistics toolbox in MATLAB and with custom-written scripts available from MATLAB Central (http://www.mathworks.de/matlabcentral/). Statistics consisted of one-way repeated-measures analyses of variance (RM-ANOVAs) and paired t tests, unless otherwise noted. Statistical results are reported as significant at p < 0.05. All statistics were derived using the means of the data of individual participants. 
Results
Behavioral data
We studied the neuronal processes underlying FBA using a DMS task that required detection of a random-dot pattern with a precued color hue or a precued motion direction (Experiment 1, Figure 2A). Participants performed the task with a mean accuracy of 83.9% ± 5.6% but were significantly better during color-hue detection (91.4% ± 1.5%) than during motion-direction detection (76.5% ± 9.7%), two-tailed paired t test, p < 0.001, n = 10. However, RTs for correctly performed trials were almost identical (color: 379 ± 34 ms, motion: 381 ± 40 ms), illustrated also by largely overlapping RT distributions (Figure 2B). To obtain a spatially unbiased database, all test RDP pairs were presented in one of two mirrored versions, thus providing the same stimulation for the left and right hemisphere. Subjects performed equally well at both sides (attend left: 82.7% ± 3.8%, attend right: 84.9% ± 6.9%), with essentially identical RTs (left: 376 ± 33 ms; right: 377 ± 34 ms) and almost perfectly overlapping RT distributions (Figure 2C). Approximately 7.5% of all trials were terminated due to imprecise fixation or eye blinks. False alarms and misses accounted for 8.6% and 7.5% of all trials, respectively. In Experiment 2, serving as a control, mean accuracy was 92.7% ± 2.8% and mean RT was 251 ± 34 ms. 
Event-related potentials depending on behavioral condition
During the DMS task, up to three pretarget intervals potentially contained test stimuli, each made up of two RDPs. The RDP at the spatially uncued, nonattended location was used to define the attentional condition. For example, if subjects were cued to an achromatic target stimulus moving upwards, the RDP opposite to the spatial focus of attention was considered nonattended if it was colored and static (i.e., if neither the feature attribute nor the feature dimension matched the target). Alternatively, it was considered dimension-attended if it was achromatic and moving but with a different motion direction (i.e., if only the feature dimension, and not the feature attribute, matched the target), and it was considered attribute-attended if it was identical to the precued target RDP (Figure 1C). Importantly, the same test stimulus could be presented in each of the three attention conditions depending on the cue (Figure 1D). Hence, using the entire combination range allowed the creation of stimulus pools containing cues and RDPs of identical type and number, only differing in their specific combination (Figure 1E). Averaging responses over all stimuli for each of the three pools thus allowed investigation of attribute- and dimension-specific effects of FBA as a function of correspondence between cue-instructed attentional deployment and representation of a spatially unattended pretarget RDP. FBA-specific effects were dissociated from motor responses by restricting the data analysis to test stimuli presented in pretarget intervals. 
At occipital (O1, O2), parieto-occipital (PO3, PO4), parietal (P3, P4), central (C3, C4), and frontal (F3, F4) electrode sites, the grand average ERP waveforms (n = 10) contralateral to the unattended location revealed a clear influence of FBA, as represented by amplitude differences of ERP waveforms recorded in the nonattended and the attribute-attended conditions (blue and red lines in Figure 3A). FBA effects appeared first as a negative deflection of the ERP (indicated by the black arrow in the upper left plot of Figure 3A), followed by a later positive deflection at predominantly parietal, central, and frontal electrodes. For the remainder of this article, we concentrate on the early FBA component, widely referred to as the selection negativity (SN) period, representing an event-related long-latency measure of feature selection (Anllo-Vento & Hillyard, 1996; Anllo-Vento, Luck, & Hillyard, 1998; Harter & Aine, 1984; Keil & Müller, 2010; Torriente, Valdes-Sosa, Ramirez, & Bobes, 1999). 
Figure 3
 
Grand average ERPs and differences during SN period demonstrating widespread effects of FBA. (A) Event-related potentials as averaged over the mean of all subjects in the three attentional conditions. ERP waveforms represent the response contralateral to the unattended location. Individual ERPs are arranged according to the electrode layout shown in the center of the figure. The black arrow in upper left plot highlights the SN period. (B) Mean amplitude differences between the nonattended condition and the dimension- and attribute-attended conditions, respectively, as calculated from the mean ERP amplitudes of the two electrodes at each anteriority. Shading represents standard error of the mean. Dashed lines indicate beginning of the SN period. (C) Graphical illustration of the results of one-way RM-ANOVA (df: 2, 9) comparing the means of each subject's ERP amplitudes for the three attention conditions, applied to bins of 10 ms for electrode pairs at different anteriorities. Color scaling indicates the p-value on a logarithmic scale. *p < 0.05; **p < 0.01; ***p < 0.001.
Figure 3
 
Grand average ERPs and differences during SN period demonstrating widespread effects of FBA. (A) Event-related potentials as averaged over the mean of all subjects in the three attentional conditions. ERP waveforms represent the response contralateral to the unattended location. Individual ERPs are arranged according to the electrode layout shown in the center of the figure. The black arrow in upper left plot highlights the SN period. (B) Mean amplitude differences between the nonattended condition and the dimension- and attribute-attended conditions, respectively, as calculated from the mean ERP amplitudes of the two electrodes at each anteriority. Shading represents standard error of the mean. Dashed lines indicate beginning of the SN period. (C) Graphical illustration of the results of one-way RM-ANOVA (df: 2, 9) comparing the means of each subject's ERP amplitudes for the three attention conditions, applied to bins of 10 ms for electrode pairs at different anteriorities. Color scaling indicates the p-value on a logarithmic scale. *p < 0.05; **p < 0.01; ***p < 0.001.
During the SN period, ERP amplitudes at occipital to frontal sites were most prominent if the spatially unattended location was stimulated with an RDP not matching the target's feature dimension (nonattended condition), but much smaller if stimulated with an RDP identical to the target (attribute-attended condition). This effect was evident at all electrode sites and did not show any differences depending on laterality. Yet for the purpose of this study, the main question of interest is the representation of the dimension-attended condition: For this condition, following a strict attribute-specific FBA approach, ERPs are expected to be similar to those in the nonattended condition, because they do not share the attended feature attribute. In the alternative, following a pure dimension-specific FBA approach, ERPs should be similar to those in the attribute-attended condition, since in both cases they are defined within the same, attended feature dimension. However, the data clearly show that neither of these exclusive alternatives is true. Instead, during the SN period, ERPs of the dimension-attended condition are situated in between those of the two other conditions at many electrode sites, indicating that visual stimuli of all three attentional conditions were differently processed depending on their relation to the precued target stimulus, i.e., the currently attended feature attribute (Figure 3A). 
For statistical testing, we averaged our data into bins of 10 ms for each subject and experimental condition and collapsed the ERPs from the two electrodes at each anteriority ([F3, F4], [C3, C4], [P3, P4], [PO3, PO4], [O3, O4]). For each of the 10-ms bins, we then performed an RM-ANOVA with the single factor Attentional Condition and the three corresponding groups nonattended, dimension-attended, and attribute-attended (df: 2, 9). ANOVAs confirmed a widespread influence of FBA on neural activity, with significant differences between the three attentional conditions at all electrodes of interest (Figure 3C). At frontal and central electrode sites, differences between the nonattended condition and at least one of the two FBA conditions started to be significantly different around 160 ms postonset and lasted for about 90 ms. At parietal, parieto-occipital, and occipital sites, differences became significant at around 190 ms and lasted for 90 ms and more. We did not observe any earlier differences between ERP traces, even though some studies have also reported FBA-dependent modulations of the P1 component (Gramann et al., 2010; Schoenfeld et al., 2007), and several experiments have shown modulations of the P1 and N1 components during spatial attention (for a review, see Herrmann & Knight, 2001; Luck, Woodman, & Vogel, 2000). 
To illustrate the effect of FBA separately for the two feature-attention conditions, we subtracted ERPs of the dimension- and attribute-attended conditions from the nonattended condition (Figure 3B). At frontal and central electrodes, the negative deflection of the ERP was approximately equally strong in both conditions, while at occipital to parietal electrodes it was of larger amplitude in the attribute-attended condition when compared to the dimension-attended condition. Thus, we obtained three main results: First, at frontal and central electrode sites, FBA-specific differences arise significantly earlier than at more posterior electrodes. Second, at all electrode sites, spatially unattended stimuli matching the target's feature dimension are processed differently from those of a different feature dimension. Third, only between parietal and occipital electrodes, stimuli matching the target's feature attribute elicit some additional processing modulation not evident for other stimuli of the attended feature dimension. 
Spatiotemporal profile of dimension- and attribute-specific FBA
To investigate these FBA characteristics in more detail, we isolated pure attribute-specific effects by subtracting the difference of the nonattended and dimension-attended conditions from the difference of the nonattended and attribute-attended conditions. With this procedure we accounted for the fact that any stimulus sharing the target's feature attribute necessarily also shares its feature dimension (see Material and methods). Scalp plots of the averaged ERP voltage for each of the 10-ms bins between 140 and 300 ms following stimulus onset illustrate the distinct spatiotemporal profile of both FBA processes. Dimension-specific FBA first developed over frontocentral electrodes and then moved over to more posterior electrode sites, with maximal effect size between 200 and 240 ms after stimulus onset (Figure 4A). Attribute-specific FBA, in contrast, was absent at frontal electrodes and developed most prominently between 220 and 280 ms postonset, predominantly at occipital, parieto-occipital, and parietal electrodes (Figure 4B). Paired t tests (n = 10) for each of the 10-ms bins corroborated the modulation pattern seen in the scalp plots. Dimension-specific FBA first became significant at frontal and central electrodes, with a stimulus-onset delay of about 170 to 180 ms, and was statistically evident at posterior electrode sites starting around 200 ms postonset. Attribute-specific FBA developed with some delay on top of dimension-specific FBA around 210 to 220 ms postonset, and was most pronounced at parietal, parieto-occipital, and occipital electrodes (Figure 4C). Both dimension- and attribute-specific FBA were statistically evident for a sustained period of time lasting about 50 ms and more. 
Figure 4
 
Dimension- and attribute-specific FBA. Scalp plots of the difference (A) between the nonattended and dimension-attended conditions (dimension-specific FBA) and (B) between the nonattended and attribute-attended conditions, minus dimension-specific FBA (attribute-specific FBA). Data were collapsed to represent ERP responses contralateral (left side) and ipsilateral (right side) to the unattended location. Color scaling is the same for all plots. (C) Graphical illustration of p-values derived from paired t tests for dimension- and attribute-specific FBA at electrode pairs contralateral to the unattended location, for each of the 10-ms bins. Color scaling is the same for both plots. *p < 0.05; **p < 0.01; ***p < 0.001.
Figure 4
 
Dimension- and attribute-specific FBA. Scalp plots of the difference (A) between the nonattended and dimension-attended conditions (dimension-specific FBA) and (B) between the nonattended and attribute-attended conditions, minus dimension-specific FBA (attribute-specific FBA). Data were collapsed to represent ERP responses contralateral (left side) and ipsilateral (right side) to the unattended location. Color scaling is the same for all plots. (C) Graphical illustration of p-values derived from paired t tests for dimension- and attribute-specific FBA at electrode pairs contralateral to the unattended location, for each of the 10-ms bins. Color scaling is the same for both plots. *p < 0.05; **p < 0.01; ***p < 0.001.
Control for attention-independent factors
As noted before, stimulus pools for the three experimental conditions consisted of cues and test stimuli identical in type and number but combined in different ways. In Experiment 2, we asked whether this difference in stimulus statistics is able to explain the difference in ERP waveforms. We used subsets of stimuli from each of the stimulus pools used in the DMS task, again identical in type and number regarding both cues and RDP pairs, but different in combination (compare Figure 1E). All stimuli were also presented in a mirrored version, providing the same stimulation to both hemispheres. Participants only passively viewed these stimuli, but were required to detect the appearance of a dot at a pseudorandom location and time after the disappearance of the test stimuli to keep them engaged. Differences in ERP amplitudes showed some fluctuations at all anteriorities, but no systematic deviance between pools (Figure 5A). RM-ANOVAs with the single factor Stimulus Pool and the three corresponding groups (df: 2, 5), run for each anteriority over bins of 10 ms, revealed two bins with a significant difference (PO3/4 between 260 and 280 ms) but no differences at other electrodes and time bins (Figure 5B). Thus, amplitude variations during the SN period as seen for the different attentional conditions in Experiment 1 are unlikely to be induced by differences in the combination of cues and test stimuli. The distinct spatiotemporal ERP profile found in Experiment 1 indicates a systematic influence of FBA and suggests two concurrently existing, spatially global processes, one being specific for the dimension and the other being specific for the attribute of an attended feature. 
Figure 5
 
ERP differences and statistical results of Experiment 2. (A) ERP amplitude differences between responses obtained with pool 1 and pool 2, and pool 1 and pool 3, respectively, analogous to the ERP differences investigated in Experiment 1. Shading indicates standard error of the mean. (B) Results of one-way RM-ANOVA comparing the means of each subject's ERP amplitudes for the three stimulus pools, applied to bins of 10 ms for electrode pairs at different anteriorities. Color scaling is the same as in Figure 3B.
Figure 5
 
ERP differences and statistical results of Experiment 2. (A) ERP amplitude differences between responses obtained with pool 1 and pool 2, and pool 1 and pool 3, respectively, analogous to the ERP differences investigated in Experiment 1. Shading indicates standard error of the mean. (B) Results of one-way RM-ANOVA comparing the means of each subject's ERP amplitudes for the three stimulus pools, applied to bins of 10 ms for electrode pairs at different anteriorities. Color scaling is the same as in Figure 3B.
Discussion
Previous results from electrophysiological and neuroimaging studies suggest that feature-based attention may select an entire feature dimension, such as color or motion (Chawla et al., 1999b; Found & Müller, 1996; Gramann et al., 2007; Pollmann et al., 2000; Pollmann, Weidner, Müller, Maertens, & von Cramon, 2006; Weidner et al., 2002), or an individual feature attribute within a single dimension, such as red versus green or upwards motion versus downwards motion (Liu et al., 2007; M. M. Müller et al., 2006; O'Craven, Rosen, Kwong, Treisman, & Savoy, 1997; Saenz et al., 2002). Yet it is an open question whether dimension- and attribute-specific processes constitute two competitive forms of FBA, with the one or the other process being activated depending on task requirements, or whether both processes coexist, acting together to support selection of the relevant feature information. In the current work, we investigated both aspects of FBA in a single experiment and under essentially identical visual stimulation. Our results suggest that FBA consists of both dimension- and attribute-specific processes, each with a distinct spatiotemporal profile. Since all experimental conditions were defined according to the stimulus outside the spatial focus of attention, and data analysis was restricted to ERPs in response to behaviorally irrelevant stimuli, the underlying neuronal mechanisms are effective in a global, spatially independent manner. The results suggest that FBA is characterized by an initial boost of neuronal responses to stimuli matching the attended dimension, emerging over frontal electrodes before moving to more posterior sites. Attribute-specific effects develop on top of this but are mostly restricted to electrodes at occipital to parietal sites. The coexistence of these two facets of FBA also underlines the importance of a unique terminology in order to better organize the results from the various studies investigating the influence of feature-based attention on both neuronal activity and behavioral performance (compare Carrasco, 2011). 
Alternative interpretations of the ERP results
Before comparing our results to those other studies of attention-dependent selection and processing, alternative interpretations should be considered. First, we obtained our data by utilizing a large database of visual stimuli to optimally balance stimulation across the three experimental conditions. However, even though the three stimulus databases were identical according to the type and number of cues and RDPs, pairings of cue and stimulus differed in individual stimuli, due to defining each of the experimental conditions. Even though the control experiment revealed two significantly different time bins at electrodes PO3/4, the overall pattern of ERPs does not support the interpretation that subtle differences in visual stimulation account for the ERP differences associated with the three attentional conditions. Especially for the time period between 200 and 260 ms postonset, during which we found the strongest and most widespread signal differences in the main experiment, the control experiment revealed essentially identical ERP traces. 
Second, we attempted to reach corresponding behavioral performance for motion and color trials, but still subjects made significantly more errors in trials with motion targets. It might be argued that motion and color trials are likely to have differently influenced the ERP results. However, we only analyzed successfully performed trials. For these, we obtained essentially identical RT distributions, indicating very similar behavioral performance within the subset of successfully completed trials. Thus, differences in performance are unlikely to have biased the ERP results. 
Third, we obtained our results by a behavioral task requiring delayed matching to sample. DMS tasks have been widely used to investigate visual attention but are also common in memory research. During the delay period between sample and match presentation, single-unit studies have found memory-related increases in neuronal activity in numerous cortical areas, including prefrontal, parietal, and extrastriate cortex (Chafee & Goldman-Rakic, 1998; Chelazzi, Duncan, Miller, & Desimone, 1998; Colombo & Gross, 1994; Constantinidis & Steinmetz, 1996; Fuster & Jervey, 1981; Mendoza-Halliday, Torres, & Martinez-Trujillo, 2014; Miller, Li, & Desimone, 1993; Salazar, Dotson, Bressler, & Gray, 2012). Thus, the cortical distribution of memory-related signals fits well with the distribution of ERP effects found in the present study. Two particular memory processes might be considered: working memory as the process to maintain a template of the task-relevant information, and recognition memory as a process associated with familiarity between a memorized and a currently displayed item. Yet two arguments make both processes unlikely candidates to explain our results. 
First, motion and color trials were shown in blocks, and (different from the classical DMS task) the cue indicating the task-relevant feature attribute was displayed throughout the trial. Thus, memory load was kept to a minimum. Second, the known ERP time courses of both working memory and recognition memory do not fit with the time course of the main effects found in the present study. Recognition memory has been most commonly associated with the FN400 and with a parietal ERP component, starting approximately 300 and 400 ms postonset, respectively, and represented by positive-going amplitude changes at most of the affected electrode sites (Addante, Ranganath, & Yonelinas, 2012; Curran, 2000; Rugg & Curran, 2007). Other studies have also found earlier effects at frontal sites, but again positive-going for familiar stimuli (Tsivilis, Otten, & Rugg, 2001), and a parietal component with more negative amplitudes for known items around the P1 component, between 100 and 175 ms (Speer & Curran, 2007). Thus, familiarity-associated effects are most commonly reported outside the time window of the effects we describe here, and are usually represented by amplitude modulations opposite to those of the present study. Working memory, as the second candidate memory process, is associated with a sustained negativity (SPCN component). Yet just like familiarity-related ERP correlates, the SPCN comes with longer latency, starting around 300 ms postonset and lasting for several hundred milliseconds (Eimer & Kiss, 2010; Vogel & Machizawa, 2004). Thus, although we cannot exclude an influence of memory-related processes in our main experiment, their known characteristics do not support an interpretation of the sustained negativity between 160 and 270 ms postonset as a memory-related ERP correlate. 
Comparison to other studies of attention
Attention and memory are both fundamental endogenous processes critical for processing behaviorally relevant information, and they closely interact (Awh, Vogel, & Oh, 2006). While memory processes are critical for the temporal maintenance of relevant information, attention is fundamental to its initial encoding. In accordance with this, several studies investigating ERP correlates of working memory have reported an earlier, attention-dependent negativity preceding the SPCN (Eimer, Kiss, & Nicholas, 2011; Mazza, Turatto, Umiltà, & Eimer, 2007; Vogel & Machizawa, 2004). This negativity can be distinguished from the SPCN by differences in terms of both scalp distribution (McCollough, Machizawa, & Vogel, 2007) and amplitude modulation following memory-load increases (Jolicoeur, Brisson, & Robitaille, 2008), and is suggested to represent processes of attentional selection. It is consistently reported in a time period around 200 ms postonset, representing the N2pc and/or the selection negativity period (Hillyard & Anllo-Vento, 1998; Luck et al., 2000; Vecera & Luck, 2002), and matches well with the scalp distribution observed in the present study. 
In our data, the onset of this negativity started at around 170 ms. Attention-dependent modulations of the ERP that occur even earlier, such as during the P1 and N1 waves, have been reported following spatial attention (Heinze et al., 1994; Mangun, Hopfinger, Kussmaul, Fletcher, & Heinze, 1997; Martinez et al., 2001). Nonspatial factors such as task demands, arousal, and discriminability may also influence the latency and/or amplitude of these components (Hopf, Vogel, Woodman, Heinze, & Luck, 2002; Taylor, 2002; Vogel & Luck, 2000), but early feature-specific attentional modulation has usually been designated to later components such as N2pc and SN (Hillyard & Anllo-Vento, 1998; McGinnis & Keil, 2011), starting around 160 ms postonset. While this is in close correspondence with our results, recent ERP evidence suggests that attention-related feature selection may even boost neural activity as early as 100 ms postonset (Gramann et al., 2010; Schoenfeld et al., 2007). However, as in the present study, the earliest effects occurred for stimuli sharing the attended feature dimension, indicating that selection of an entire feature dimension is associated with a more rapid neural facilitation than selection of a specific attribute within a single dimension (Schoenfeld et al., 2007). 
Spatiotemporal profile of feature-based attention
The spatiotemporal profile of dimension- and attribute-specific ERP signatures as found in the present study has two important implications: First, FBA as measured during the SN is globally effective. While this is in general accordance with numerous electrophysiological (Bichot, Rossi, & Desimone, 2005; David, Hayden, Mazer, & Gallant, 2008; Treue & Martínez Trujillo, 1999) and imaging studies (Saenz et al., 2002; Serences & Boynton, 2007; Sohn, Chong, Papathomas, & Vidnyánzky, 2005), it is in contrast with the finding that feature-specific modulation of the SN is negligible outside the spatial focus of attention (Anllo-Vento & Hillyard, 1996; Hillyard & Münte, 1984). However, attention-related changes in neural activity adjust with attentional load and the degree of stimulus competition (Bahrami, Lavie, & Rees, 2007; Rees, Frith, & Lavie, 1997; Schwartz et al., 2005), which was small in these early experiments (Anllo-Vento & Hillyard, 1996; Hillyard & Münte, 1984) but higher in the present study. 
Second, the emergence of the initial, dimension-specific modulation at anterior electrodes suggests that FBA starts with prefrontal areas sending facilitatory signals to feature-processing modules in the visual cortex. Note, however, that we did not investigate the actual cortical source of these potentials, which may differ from their localization on the scalp (Slotnick, 2004). This interpretation lines up with the involvement of the prefrontal cortex (PFC) in creating the attentional set and with its role in the voluntary deployment of attention (Corbetta & Shulman, 2002; Kastner & Ungerleider, 2000; Knudsen, 2007; Squire, Noudoost, Schafer, & Moore, 2013). The PFC conveys transient modulatory signals following feature-specific attention shifts (Greenberg, Esterman, Wilson, Serences, & Yantis, 2010; Liu, Slotnick, Serences, & Yantis, 2003) and sustained modulatory signals to maintain the attentional focus (Liu et al., 2003). Many neurons in the PFC show distinct responses to color and motion, depending on task requirements (Lauwereyns et al., 2001), and PFC lesions impair attention-dependent visual discrimination performance, particularly if task demands are changing frequently (Rossi, Bichot, Desimone, & Ungerleider, 2007). One of the prefrontal areas involved in the control of attention is the frontal eye field (FEF). Voluntary firing-rate control of the FEF is associated with spatially selective visual attention on both a behavioral and a neurophysiological level (Schafer & Moore, 2011), and electrical microstimulation of FEF neurons mimics visual attention effects in extrastriate area V4 (Armstrong & Moore, 2007; Moore & Armstrong, 2003). Importantly, feature-based attentional modulations in the FEF come with shorter latencies than in V4, suggesting that the FEF provides a biasing signal for modulation of visual processing in early visual areas (Gregoriou, Gotts, Zhou, & Desimone, 2009; Zhou & Desimone, 2011). Thus, the time course we found for the development of FBA over frontal to occipital electrode sites is very well in agreement with the data from monkey neurophysiology and with the role of the PFC in controlling and maintaining visual attention. The novel aspect, however, is that the initial FBA modulation is specific to the feature dimension of the target, not its attribute, at least under the task requirements of our study. 
Attribute-specific FBA was mostly restricted to electrode sites over occipital to parietal cortex. We investigated these by subtracting the ERP amplitude difference of the nonattended and dimension-attended conditions from the overall effect found for the attribute-attended condition. This procedure ensured isolation of attention-dependent modulation occurring solely in response to stimuli matching the attended feature attribute, not in response to other stimuli defined in the same feature dimension. Our results match previous neurophysiological findings of attribute-specific modulation in extrastriate visual cortex (Martinez-Trujillo & Treue, 2004; McAdams & Maunsell, 2000; Treue & Martínez Trujillo, 1999). Interestingly, using functional MRI, Saenz et al. (2002) showed attribute-specific modulation of neuronal activity in several early visual areas, including those regions apparently homologous to monkey areas V4 and MT, but they did not report such effects for regions outside extrastriate cortex. Shulman et al. (1999) investigated FBA using dynamic noise patterns with short periods of coherent motion. Subjects were required to detect the coherent motion and were provided with a directional cue on some trials. The authors compared neuronal activity in trials with a directional cue with activity in other trials using a neutral, noninformative cue. During the cue period, when subjects could focus attention on the cued direction (i.e., on the relevant feature attribute), they found attentional modulation in many regions involved in motion processing but no such modulations in prefrontal regions. 
Thus, the spatiotemporal profile of the two FBA components we describe here is consistent with many previous reports on feature-selective attention. A dimension- but not attribute-specific signal that emerges over frontal electrodes and then moves to more posterior sites is also in accordance with the finding that even in the absence of visual stimulation, baseline activity within those cortical modules that process the attended, behaviorally relevant feature dimension might be increased. Such baseline effects have been described following spatial allocation of attention (Hopfinger, Buonocore, & Mangun, 2000; Kastner, Pinsk, DeWeerd, Desimone, & Ungerleider, 1999; Luck, Chelazzi, Hillyard, & Desimone, 1997; Reynolds, Pasternak, & Desimone, 2000; Woldorff et al., 2004), but also in nonspatial attention tasks, suggesting that they represent a general mechanism of attention. For example, Chawla, Lumer, and Friston (1999a) reported enhanced functional MRI baseline activity in area V4 when subjects performed color judgments and in MT when they performed motion judgments. Based on computational work, the authors argued that such baseline increments improve functional connectivity and dynamic integration between two populations of neurons (Chawla et al., 1999b; Chawla et al., 2000), thereby increasing response sensitivity and providing a competitive advantage for the attended feature (Pessoa, Kastner, & Ungerleider, 2003). Similar to baseline shifts in the absence of visual stimulation, a general increase in neuronal responses for stimuli of the same, attended dimension may allow for enhanced stimulus discriminability within the task-relevant cortical feature domain, resulting in improved representation of the attended, task-relevant feature attribute. 
It should be noted, however, that even though the basic finding of a baseline shift following attention to either color or motion has been confirmed by others, this shift might be visible in several extrastriate areas at the same time, representing a more global, dimension-independent increase of neuronal responses (Fannon, Saron, & Mangun, 2008; McMains, Fehd, Emmanouil, & Kastner, 2007). A likely reason for this diversity of findings is a difference in the interaction of object- and feature-based attention mechanisms. This issue has been addressed recently by the three-step model of visual attention, stating that task demands and stimulus characteristics determine the degree of object-based coselection of nonattended features (Wegener, Galashan, Aurich, & Kreiter, 2014). The model assumes a task-dependent modulatory top-down signal to distinct feature channels, and subsequent attentional modulation within each channel. An additional processing benefit for the attended feature attribute arises on top of this dimension-specific modulation and presumably emerges only within the respective cortical processing module. 
Acknowledgments
DW was supported by DFG grants KR 1844/1-2 and WE 5469/2-1. DG was supported by a PhD grant from the University of Bremen. The authors thank Carina Volk and Fingal Orlando Galashan for support in data acquisition and data analysis, Axel Grzymisch for comments on an earlier draft, and all volunteers for participation. The authors declare no competing financial interest or conflict of interest. DW, DG, and MF designed the research; DG conducted the research; DG, DW, and CG analyzed the data; and DW and DG wrote the article. 
Commercial relationships: none. 
Corresponding author: Detlef Wegener. 
Email: wegener@brain.uni-bremen.de. 
Address: Brain Research Institute, Center for Cognitive Science, University of Bremen, Bremen, Germany. 
References
Addante R. J., Ranganath C., Yonelinas A. P. (2012). Examining ERP correlates of recognition memory: Evidence of accurate source recognition without recollection. NeuroImage, 62, 439–450. [PubMed]
American Electroencephalographic Society. (1994). Guideline thirteen: Guidelines for standard electrode position nomenclature. Journal of Clinical Neurophysiology, 11, 111–113. [PubMed]
Anllo-Vento L., Hillyard S. A. (1996). Selective attention to the color and direction of moving stimuli: Electrophysiological correlates of hierarchical feature selection. Perception & Psychophysics, 58, 191–206. [PubMed]
Anllo-Vento L., Luck S. J., Hillyard S. A. (1998). Spatio-temporal dynamics of attention to color: Evidence from human electrophysiology. Human Brain Mapping, 6, 216–238. [PubMed]
Armstrong K. M., Moore T. (2007). Rapid enhancement of visual cortical response discriminability by microstimulation of the frontal eye field. Proceedings of the National Academy of Sciences, USA, 104, 9499–9504. [PubMed]
Awh E., Vogel E. K., Oh S. H. (2006). Interactions between attention and working memory. Neuroscience, 139, 201–208. [PubMed]
Bach M. (1996). The “Freiburg Visual Acuity Test”—Automatic measurement of visual acuity. Optometry and Vision Science, 73, 49–53. [PubMed]
Bach M. (2007). The Freiburg Visual Acuity Test—Variability unchanged by post-hoc re-analysis. Graefe's Archive for Clinical and Experimental Ophthalmology, 245, 965–971. [PubMed]
Bahrami B., Lavie N., Rees G. (2007). Attentional load modulates responses of human primary visual cortex to invisible stimuli. Current Biology, 17, 509–513. [PubMed]
Bichot N. P., Rossi A. F., Desimone R. (2005, April 22). Parallel and serial neural mechanisms for visual search in macaque area V4. Science, 308, 529–534. [PubMed]
Carrasco M. (2011). Visual attention: The past 25 years. Vision Research, 51, 1484–1525. [PubMed]
Chafee M. V., Goldman-Rakic P. S. (1998). Matching patterns of activity in primate prefrontal area 8a and parietal area 7ip neurons during a spatial working memory task. Journal of Neurophysiology, 79, 2919–2940. [PubMed]
Chawla D., Lumer E. D., Friston K. J. (1999 a). The relationship between synchronization among neuronal populations and their mean activity levels. Neural Computation, 11, 1389–1411. [PubMed]
Chawla D., Lumer E. D., Friston K. J. (2000). Relating macroscopic measures of brain activity to fast, dynamic neuronal interactions. Neural Computation, 12, 2805–2821. [PubMed]
Chawla D., Rees G., Friston K. J. (1999 b). The physiological basis of attentional modulation in extrastriate visual areas. Nature Neuroscience, 2, 671–676. [PubMed]
Chelazzi L., Duncan J., Miller E. K., Desimone R. (1998). Responses of neurons in inferior temporal cortex during memory-guided visual search. Journal of Neurophysiology, 80, 2918–2940. [PubMed]
Chen X., Hoffmann K. P., Albright T. D., Thiele A. (2012). Effect of feature-selective attention on neuronal responses in macaque area MT. Journal of Neurophysiology, 107, 1530–1543. [PubMed]
Colombo M., Gross C. G. (1994). Responses of inferior temporal cortex and hippocampal neurons during delayed matching to sample in monkey (Macaca fascicularis). Behavioral Neuroscience, 108, 443–455. [PubMed]
Constantinidis C., Steinmetz M. A. (1996). Neuronal activity in posterior parietal area 7a during the delay periods of a spatial memory task. Journal of Neurophysiology, 76, 1352–1355. [PubMed]
Corbetta M., Miezin F. M., Dobmeyer S., Shulman G. L., Petersen S. E. (1990, June 22). Attentional modulation of neural processing of shape, color, and velocity in humans. Science, 248, 1556–1559. [PubMed]
Corbetta M., Shulman G. L. (2002). Control of goal-directed and stimulus-driven attention in the brain. Nature Reviews Neuroscience, 3, 215–229. [PubMed]
Curran T. (2000). Brain potentials of recollection and familiarity. Memory and Cognition, 28, 923–938. [PubMed]
David S. V., Hayden B. Y., Mazer J. A., Gallant J. L. (2008). Attention to stimulus features shifts spectral tuning of V4 neurons during natural vision. Neuron, 59, 509–521. [PubMed]
Delorme A., Makeig S. (2004). EEGLAB: An open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. Journal of Neuroscience Methods, 134, 9–21. [PubMed]
Eimer M., Kiss M. (2010). An electrophysiological measure of access to representations in visual working memory. Psychophysiology, 47, 197–200. [PubMed]
Eimer M., Kiss M., Nicholas S. (2011). What top-down task sets do for us: An ERP study on the benefits of advance preparation in visual search. Journal of Experimental Psychology: Human Perception and Performance, 37, 1758–1766. [PubMed]
Fannon S. P., Saron C. D., Mangun G. R. (2008). Baseline shifts do not predict attentional modulation of target processing during feature-based visual attention. Frontiers in Human Neuroscience, 1, 7, doi:10.3389/neuro.09.007.2007. [PubMed]
Found A., Müller H. J. (1996). Searching for unknown feature targets on more than one dimension: Investigating a “dimension-weighting” account. Perception & Psychophysics, 58, 88–101. [PubMed]
Fuster J. M., Jervey J. P. (1981, May 22). Inferotemporal neurons distinguish and retain behaviorally relevant features of visual stimuli. Science, 212, 952–955. [PubMed]
Gramann K., Toellner T., Krummenacher J., Eimer M., Müller H. J. (2007). Brain electrical correlates of dimensional weighting: An ERP study. Psychophysiology, 44, 277–292. [PubMed]
Gramann K., Töllner T., Müller H. J. (2010). Dimension-based attention modulates early visual processing. Psychophysiology, 47, 968–978. [PubMed]
Greenberg A. S., Esterman M., Wilson D., Serences J. T., Yantis S. (2010). Control of spatial and feature-based attention in frontoparietal cortex. The Journal of Neuroscience, 30, 14330–14339. [PubMed]
Gregoriou G. G., Gotts S. J., Zhou H., Desimone R. (2009, May 29). High-frequency, long-range coupling between prefrontal and visual cortex during attention. Science, 324, 1207–12010. [PubMed]
Harter M. R., Aine C. J. (1984). Brain mechanisms of visual selective attention. In Parasuraman R. Driver J. (Eds.) Varieties of attention (pp. 293–321). New York: Academic Press.
Heinze, H. J., Mangun G. R., Burchert W., Hinrichs H., Scholz M., Münte T. F., Hillyard S. A. (1994). Combined spatial and temporal imaging of brain activity during visual selective attention in humans. Nature, 372, 543–546. [PubMed]
Herrmann C. S., Knight R. T. (2001). Mechanisms of human attention: Event-related potentials and oscillations. Neuroscience & Biobehavioral Reviews, 25, 465–476. [PubMed]
Hillyard S. A., Anllo-Vento L. (1998). Event-related brain potentials in the study of visual selective attention. Proceedings of the National Academy of Sciences, USA, 95, 781–787. [PubMed]
Hillyard S. A., Münte T. F. (1984). Selective attention to color and location: An analysis with event-related brain potentials. Perception & Psychophysics, 36, 185–198. [PubMed]
Hopf J. M., Vogel E., Woodman G., Heinze H. J., Luck S. J. (2002). Localizing visual discrimination processes in time and space. Journal of Neurophysiology, 88, 2088–2095. [PubMed]
Hopfinger J. B., Buonocore M. H., Mangun G. R. (2000). The neural mechanisms of top-down attentional control. Nature Neuroscience, 3, 284–291. [PubMed]
Ishihara S. (1917). Tests for colour-blindness. Tokyo, Japan: Hongo Harukicho.
Jolicoeur P., Brisson B., Robitaille N. (2008). Dissociation of the N2pc and sustained posterior contralateral negativity in a choice response task. Brain Research, 1215, 160–172. [PubMed]
Kanazawa Y., Kanatani K. (1997). Infinity and planarity test for stereo vision. IEICE Transactions on Information and Systems, E80-D, 774–779.
Kastner S., Pinsk M. A., DeWeerd P., Desimone R., Ungerleider L. G. (1999). Increased activity in human visual cortex during directed attention in the absence of visual stimulation. Neuron, 22, 751–761. [PubMed]
Kastner S., Ungerleider L. G. (2000). Mechanisms of visual attention in the human cortex. Annual Review of Neuroscience, 23, 315–341. [PubMed]
Katzner S., Busse L., Treue S. (2009). Attention to the color of a moving stimulus modulates motion-signal processing in macaque area MT: Evidence for a unified attentional system. Frontiers in Systems Neuroscience, 3, 12, doi:10.3389/neuro.06.012.2009. [PubMed]
Keil A., Müller M. M. (2010). Feature selection in the human brain: Electrophysiological correlates of sensory enhancement and feature integration. Brain Research, 1313, 172–184. [PubMed]
Knudsen E. I. (2007). Fundamental components of attention. Annual Review of Neuroscience, 30, 57–78. [PubMed]
Lauwereyns J., Sakagami M., Tsutsui K. I., Kobayashi S., Koizumi M., Hikosaka O. (2001). Responses to task-irrelevant visual features by primate prefrontal neurons. Journal of Neurophysiology, 86, 2001–2010. [PubMed]
Liu T., Larsson J., Carrasco M. (2007). Feature-based attention modulates orientation-selective responses in human visual cortex. Neuron, 55, 313–323. [PubMed]
Liu T., Mance I. (2011). Constant spread of attention across the visual field. Vision Research, 51, 26–33. [PubMed]
Liu T., Slotnick S. D., Serences J. T., Yantis S. (2003). Cortical mechanisms of feature-based attentional control. Cerebral Cortex, 13, 1334–1343. [PubMed]
Luck S. J., Chelazzi L., Hillyard S. A., Desimone R. (1997). Neural mechanisms of spatial selective attention in areas V1, V2, and V4 of macaque visual cortex. Journal of Neurophysiology, 77, 24–42. [PubMed]
Luck S. J., Woodman G. F., Vogel E. K. (2000). Event-related potential studies of attention. Trends in Cognitive Sciences, 4, 432–440. [PubMed]
Mangun G. R., Hopfinger J. B., Kussmaul C. L., Fletcher E. M., Heinze H. J. (1997). Covariations in ERP and PET measures of spatial selective attention in human extrastriate visual cortex. Human Brain Mapping, 5, 273–279. [PubMed]
Martinez A., DiRusso F., Anllo-Vento L., Sereno M. I., Buxton R. B., Hillyard S. A. (2001). Putting spatial attention on the map: Timing and localization of stimulus selection processes in striate and extrastriate visual areas. Vision Research, 41, 1437–1457. [PubMed]
Martinez-Trujillo J. C., Treue S. (2004). Feature-based attention increases the selectivity of population responses in primate visual cortex. Current Biology, 14, 744–751. [PubMed]
Maunsell J. H. R., Treue S. (2006). Feature-based attention in visual cortex. Trends in Neurosciences, 29, 317–322. [PubMed]
Mazza V., Turatto M., Umiltà C., Eimer M. (2007). Attentional selection and identification of visual objects are reflected by distinct electrophysiological responses. Experimental Brain Research, 181, 531–536. [PubMed]
McAdams C. J., Maunsell J. H. (2000). Attention to both space and feature modulates neuronal responses in macaque area V4. Journal of Neurophysiology, 83, 1751–1755. [PubMed]
McCollough A. W., Machizawa M. G., Vogel E. K. (2007). Electrophysiological measures of maintaining representations in visual working memory. Cortex, 43, 77–94. [PubMed]
McGinnis E. M., Keil A. (2011). Selective processing of multiple features in the human brain: Effects of feature type and salience. PLoS ONE, 6, e16824. [PubMed]
McMains S. A. F., Fehd H. M., Emmanouil T.-A., Kastner S. (2007). Mechanisms of feature- and space-based attention: Response modulation and baseline increases. Journal of Neurophysiology, 98, 2110–2121. [PubMed]
Mendoza-Halliday D., Torres S., Martinez-Trujillo J. C. (2014). Sharp emergence of feature-selective sustained activity along the dorsal visual pathway. Nature Neuroscience, 17, 1255–1262. [PubMed]
Miller E. K., Li L., Desimone R. (1993). Activity of neurons in anterior inferior temporal cortex during a short-term memory task. The Journal of Neuroscience, 13, 1460–1478. [PubMed]
Moore T., Armstrong K. M. (2003). Selective gating of visual signals by microstimulation of frontal cortex. Nature, 421, 370–373. [PubMed]
Müller H. J., Heller D., Ziegler J. (1995). Visual search for singleton feature targets within and across feature dimensions. Perception & Psychophysics, 57, 1–17. [PubMed]
Müller M. M., Andersen S., Trujillo N. J., Valdés-Sosa P., Malinowski P., Hillyard S. A. (2006). Feature-selective attention enhances color signals in early visual areas of the human brain. Proceedings of the National Academy of Sciences, USA, 103, 14250–14254. [PubMed]
O'Craven K., Rosen B. R., Kwong K. K., Treisman A., Savoy R. L. (1997). Voluntary attention modulates fMRI activity in human MT-MST. Neuron, 18, 591–598. [PubMed]
Pessoa L., Kastner S., Ungerleider L. G. (2003). Neuroimaging studies of attention: From modulation of sensory processing to top-down control. The Journal of Neuroscience, 23, 3990–3998. [PubMed]
Pollmann S., Weidner R., Müller H. J., Maertens M., von Cramon D. Y. (2006). Selective and interactive neural correlates of visual dimension changes and response changes. NeuroImage, 30, 254–265. [PubMed]
Pollmann S., Weidner R., Müller H. J., von Cramon D. Y. (2000). A fronto-posterior network involved in visual dimension changes. Journal of Cognitive Neuroscience, 12, 480–494. [PubMed]
Rees G., Frith C. D., Lavie N. (1997, Nov 28). Modulating irrelevant motion perception by varying attentional load in an unrelated task. Science, 278, 1616–1619. [PubMed]
Reynolds J. H., Pasternak T., Desimone R. (2000). Attention increases sensitivity of V4 neurons. Neuron, 26, 703–714. [PubMed]
Rossi A. F., Bichot N. P., Desimone R., Ungerleider L. G. (2007). Top-down attentional deficits in macaques with lesions of lateral prefrontal cortex. The Journal of Neuroscience, 27, 11306–11314. [PubMed]
Rugg M. D., Curran T. (2007). Event-related potentials and recognition memory. Trends in Cognitive Sciences, 11, 251–257. [PubMed]
Saenz M., Buracas G. T., Boynton G. M. (2002). Global effects of feature-based attention in human visual cortex. Nature Neuroscience, 5, 631–632. [PubMed]
Saenz M., Buracas G. T., Boynton G. M. (2003). Global feature-based attention for motion and color. Vision Research, 43, 629–637. [PubMed]
Salazar R. F., Dotson N. M., Bressler S. L., Gray C. M. (2012, Nov 23). Content-specific fronto-parietal synchronization during visual working memory. Science, 338, 1097–1100. [PubMed]
Schafer R. J., Moore T. (2011, June 24). Selective attention from voluntary control of neurons in prefrontal cortex. Science, 332, 1568–1571. [PubMed]
Schoenfeld M. A., Hopf J. M., Martinez A., Mai H. M., Sattler C., Gasde A., Hillyard S. A. (2007). Spatio-temporal analysis of feature-based attention. Cerebral Cortex, 17, 2468–2477. [PubMed]
Schwartz S., Vuilleumier P., Hutton C., Maravita A., Dolan R. J., Driver J. (2005). Attentional load and sensory competition in human vision: Modulation of fMRI responses by load at fixation during task-irrelevant stimulation in the peripheral visual field. Cerebral Cortex, 15, 770–786. [PubMed]
Serences J. T., Boynton G. M. (2007). Feature-based attentional modulations in the absence of direct visual stimulation. Neuron, 55, 301–312. [PubMed]
Shulman G. L., Ollinger J. M., Akbudak E., Conturo T. E., Snyder A. Z., Petersen S. E., Corbetta M. (1999). Areas involved in encoding and applying directional expectations to moving objects. The Journal of Neuroscience, 19, 9480–9496. [PubMed]
Slotnick S. D. (2004). Source localization of ERP generators. In Handy T. C. (Ed.) Event-related potentials: A methods handbook (pp. 149–166). Cambridge, MA: MIT Press.
Sohn, W., Chong S. C., Papathomas T., Vidnyánzky Z. (2005). Cross-feature spread of global attentional modulation in human area MT+. NeuroReport, 16, 1389–1393. [PubMed]
Speer N. K., Curran T. (2007). ERP correlates of familiarity and recollection processes in visual associative recognition. Brain Research, 1174, 97–109. [PubMed]
Squire R. F., Noudoost B., Schafer R. J., Moore T. (2013). Prefrontal contributions to visual selective attention. Annual Review of Neuroscience, 36, 451–466. [PubMed]
Taylor M. J. (2002). Non-spatial attentional effects on P1. Clinical Neurophysiology, 113, 1903–1908. [PubMed]
Torriente I., Valdes-Sosa M., Ramirez D., Bobes M. A. (1999). Visual evoked potentials related to motion-onset are modulated by attention. Vision Research, 39, 4122–4139. [PubMed]
Treue S., Martínez Trujillo J. C. (1999). Feature-based attention influences motion processing gain in macaque visual cortex. Nature, 399, 575–579. [PubMed]
Tsivilis D., Otten L. J., Rugg M. D. (2001). Context effects on the neural correlates of recognition memory: An electrophysiological study. Neuron, 31, 497–505. [PubMed]
Vecera S. P., Luck S. J. (2002). Attention. In Ramachandran V. S. (Ed.) Encyclopedia of the human brain (pp. 269–284). San Diego, CA: Academic Press.
Vogel, E. K., Luck S. J. (2000). The visual N1 component as an index of a discrimination process. Psychophysiology, 37, 190–203. [PubMed]
Vogel E. K., Machizawa M. G. (2004). Neural activity predicts individual differences in visual working memory capacity. Nature, 428, 748–751. [PubMed]
Wegener D., Ehn F., Aurich M. K., Galashan F. O., Kreiter A. K. (2008). Feature-based attention and the suppression of non-relevant object features. Vision Research, 48, 2696–2707. [PubMed]
Wegener D., Galashan F. O., Aurich M. K., Kreiter A. K. (2014). Attentional spreading to task-irrelevant object features: Experimental support and a 3-step model of attention for object-based selection and feature-based processing modulation. Frontiers in Human Neuroscience, 8, 414, doi:10.3389/fnhum.2014.00414. [PubMed]
Weidner R., Pollmann S., Müller H. J., von Cramon D. Y. (2002). Top-down controlled visual dimension weighting: An event-related fMRI study. Cerebral Cortex, 12, 318–328. [PubMed]
White A. L., Carrasco M. (2011). Feature-based attention involuntarily and simultaneously improves visual performance across locations. Journal of Vision, 11 (6): 15, 1–10, doi:10.1167/11.6.15. [PubMed] [Article]
Woldorff M. G., Hazlett C. J., Fichtenholtz H. M., Weissman D. H., Dale A. M., Song A. W. (2004). Functional parcellation of attentional control regions of the brain. Journal of Cognitive Neuroscience, 16, 149–165. [PubMed]
Zhou H., Desimone R. (2011). Feature-based attention in the frontal eye field and area v4 during visual search. Neuron, 70, 1205–1217. [PubMed]
Figure 1
 
Experimental conditions and stimulus pools. Example (A) test and (B) nontest RDPs. (C) Three stimuli instructing the same attentional selection of a 90° target motion direction at the precued location (spatial focus). For visual-presentation purposes, RDPs are shown as simplified discs representing the main color hue or motion direction. Outside the spatial focus of attention, RDPs may be defined by having the other feature dimension (nonattended), by having the same dimension but a different feature attribute (dimension-attended), or by sharing the target's feature attribute (attribute-attended). (D) Test stimulus combined with different cues, defining three different attentional conditions. (E) Subset of stimuli of the three stimulus pools, utilizing the same RDPs and cues but in different combination. Each test stimulus was presented in an attend-left and attend-right version (only attend-right stimuli are shown). ERPs were averaged over all templates of each pool. See Material and methods for details.
Figure 1
 
Experimental conditions and stimulus pools. Example (A) test and (B) nontest RDPs. (C) Three stimuli instructing the same attentional selection of a 90° target motion direction at the precued location (spatial focus). For visual-presentation purposes, RDPs are shown as simplified discs representing the main color hue or motion direction. Outside the spatial focus of attention, RDPs may be defined by having the other feature dimension (nonattended), by having the same dimension but a different feature attribute (dimension-attended), or by sharing the target's feature attribute (attribute-attended). (D) Test stimulus combined with different cues, defining three different attentional conditions. (E) Subset of stimuli of the three stimulus pools, utilizing the same RDPs and cues but in different combination. Each test stimulus was presented in an attend-left and attend-right version (only attend-right stimuli are shown). ERPs were averaged over all templates of each pool. See Material and methods for details.
Figure 2
 
Task design and behavioral performance of Experiment 1. (A) Example trial of the DMS task. Target location was kept constant for a block of 64 trials, and target dimension was kept constant for sub-blocks of 16 trials each. For visualization purposes, RDPs are shown as simplified discs. A foveally presented cue indicated the target attribute (specific color hue or a gray arrow pointing towards a specific motion direction) at trial start. Prior to target appearance, up to three test stimuli were presented, separated by a jittered interstimulus interval. RDPs at the spatially nonattended location shared either the target's attribute, the target's dimension, or none of these (attribute-attended, dimension-attended, and nonattended conditions). (B) Mean percentages of hits and errors, and cumulative RT distributions separately for color and motion trials, and (C) for attend-left and attend-right trials. Error bars indicate standard deviation.
Figure 2
 
Task design and behavioral performance of Experiment 1. (A) Example trial of the DMS task. Target location was kept constant for a block of 64 trials, and target dimension was kept constant for sub-blocks of 16 trials each. For visualization purposes, RDPs are shown as simplified discs. A foveally presented cue indicated the target attribute (specific color hue or a gray arrow pointing towards a specific motion direction) at trial start. Prior to target appearance, up to three test stimuli were presented, separated by a jittered interstimulus interval. RDPs at the spatially nonattended location shared either the target's attribute, the target's dimension, or none of these (attribute-attended, dimension-attended, and nonattended conditions). (B) Mean percentages of hits and errors, and cumulative RT distributions separately for color and motion trials, and (C) for attend-left and attend-right trials. Error bars indicate standard deviation.
Figure 3
 
Grand average ERPs and differences during SN period demonstrating widespread effects of FBA. (A) Event-related potentials as averaged over the mean of all subjects in the three attentional conditions. ERP waveforms represent the response contralateral to the unattended location. Individual ERPs are arranged according to the electrode layout shown in the center of the figure. The black arrow in upper left plot highlights the SN period. (B) Mean amplitude differences between the nonattended condition and the dimension- and attribute-attended conditions, respectively, as calculated from the mean ERP amplitudes of the two electrodes at each anteriority. Shading represents standard error of the mean. Dashed lines indicate beginning of the SN period. (C) Graphical illustration of the results of one-way RM-ANOVA (df: 2, 9) comparing the means of each subject's ERP amplitudes for the three attention conditions, applied to bins of 10 ms for electrode pairs at different anteriorities. Color scaling indicates the p-value on a logarithmic scale. *p < 0.05; **p < 0.01; ***p < 0.001.
Figure 3
 
Grand average ERPs and differences during SN period demonstrating widespread effects of FBA. (A) Event-related potentials as averaged over the mean of all subjects in the three attentional conditions. ERP waveforms represent the response contralateral to the unattended location. Individual ERPs are arranged according to the electrode layout shown in the center of the figure. The black arrow in upper left plot highlights the SN period. (B) Mean amplitude differences between the nonattended condition and the dimension- and attribute-attended conditions, respectively, as calculated from the mean ERP amplitudes of the two electrodes at each anteriority. Shading represents standard error of the mean. Dashed lines indicate beginning of the SN period. (C) Graphical illustration of the results of one-way RM-ANOVA (df: 2, 9) comparing the means of each subject's ERP amplitudes for the three attention conditions, applied to bins of 10 ms for electrode pairs at different anteriorities. Color scaling indicates the p-value on a logarithmic scale. *p < 0.05; **p < 0.01; ***p < 0.001.
Figure 4
 
Dimension- and attribute-specific FBA. Scalp plots of the difference (A) between the nonattended and dimension-attended conditions (dimension-specific FBA) and (B) between the nonattended and attribute-attended conditions, minus dimension-specific FBA (attribute-specific FBA). Data were collapsed to represent ERP responses contralateral (left side) and ipsilateral (right side) to the unattended location. Color scaling is the same for all plots. (C) Graphical illustration of p-values derived from paired t tests for dimension- and attribute-specific FBA at electrode pairs contralateral to the unattended location, for each of the 10-ms bins. Color scaling is the same for both plots. *p < 0.05; **p < 0.01; ***p < 0.001.
Figure 4
 
Dimension- and attribute-specific FBA. Scalp plots of the difference (A) between the nonattended and dimension-attended conditions (dimension-specific FBA) and (B) between the nonattended and attribute-attended conditions, minus dimension-specific FBA (attribute-specific FBA). Data were collapsed to represent ERP responses contralateral (left side) and ipsilateral (right side) to the unattended location. Color scaling is the same for all plots. (C) Graphical illustration of p-values derived from paired t tests for dimension- and attribute-specific FBA at electrode pairs contralateral to the unattended location, for each of the 10-ms bins. Color scaling is the same for both plots. *p < 0.05; **p < 0.01; ***p < 0.001.
Figure 5
 
ERP differences and statistical results of Experiment 2. (A) ERP amplitude differences between responses obtained with pool 1 and pool 2, and pool 1 and pool 3, respectively, analogous to the ERP differences investigated in Experiment 1. Shading indicates standard error of the mean. (B) Results of one-way RM-ANOVA comparing the means of each subject's ERP amplitudes for the three stimulus pools, applied to bins of 10 ms for electrode pairs at different anteriorities. Color scaling is the same as in Figure 3B.
Figure 5
 
ERP differences and statistical results of Experiment 2. (A) ERP amplitude differences between responses obtained with pool 1 and pool 2, and pool 1 and pool 3, respectively, analogous to the ERP differences investigated in Experiment 1. Shading indicates standard error of the mean. (B) Results of one-way RM-ANOVA comparing the means of each subject's ERP amplitudes for the three stimulus pools, applied to bins of 10 ms for electrode pairs at different anteriorities. Color scaling is the same as in Figure 3B.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×