Pitting binding against selection--electrophysiological measures of feature-based attention are attenuated by Gestalt object grouping.

Abstract

Humans have limited cognitive resources to process the nearly limitless information available in the environment. Endogenous, or 'top-down', selective attention to basic visual features such as color or motion is a common strategy for biasing resources in favor of the most relevant information sources in a given context. Opposing this top-down separation of features is a 'bottom-up' tendency to integrate, or bind, the various features that constitute objects. We pitted these two processes against each other in an electrophysiological experiment to test if top-down selective attention can overcome constitutive binding processes. Our results demonstrate that bottom-up binding processes can dominate top-down feature-based attention even when explicitly detrimental to task performance.

Full text

Page 1

Pitting binding against selection – electrophysiological
measures of feature-based attention are attenuated by
Gestalt object grouping
Adam C. Snyder,1,2Ian C. Fiebelkorn1and John J. Foxe1,2
1The Cognitive Neurophysiology Laboratory, Children’s Evaluation and Rehabilitation Center (CERC), Departments of Pediatrics
and Neuroscience, Albert Einstein College of Medicine, Van Etten Building – Wing 1C, 1225 Morris Park Avenue, Bronx, NY 10461,
USA
2The Cognitive Neurophysiology Laboratory, Program in Cognitive Neuroscience, Departments of Psychology & Biology, City
College of the City University of New York, New York, NY, USA
Keywords: attention, binding, ERP, feature-based selection, object-processing
Abstract
Humans have limited cognitive resources to process the nearly limitless information available in the environment. Endogenous, or
‘top-down’, selective attention to basic visual features such as color or motion is a common strategy for biasing resources in favor of
the most relevant information sources in a given context. Opposing this top-down separation of features is a ‘bottom-up’ tendency to
integrate, or bind, the various features that constitute objects. We pitted these two processes against each other in an
electrophysiological experiment to test if top-down selective attention can overcome constitutive binding processes. Our results
demonstrate that bottom-up binding processes can dominate top-down feature-based attention even when explicitly detrimental to
task performance.
Introduction
It has been estimated that approximately 107bits of data travel down
the optic nerve each waking second (Koch et al., 2006), and while our
perception is impressive and creates an illusion of continuity, we
consciously register but a fraction of that information. Every day we
must extract useful signals from among the information in our
environment if we are to perceive, learn, remember and plan actions
useful to our survival. This filtering of information from the
environment is achieved through a combination of endogenous
(‘top-down’) and exogenous (‘bottom-up’) biasing that resolves the
competition for limited processing resources. The top-down form of
this biasing, known as ‘selective attention’, involves allocating
processing resources such that sources or classes of information
expected to be relevant in a given context are given priority at the
expense of irrelevant or distracting information (Broadbent, 1958;
Kahneman, 1973; Neely, 1977; Schneider & Shiffrin, 1977; Shiffrin &
Schneider, 1977; Schneider & Fisk, 1982). In this article we focus on
selective attention to basic visual features, such as color and motion
(Snyder & Foxe, 2010). Feature-based selective attention is an
efficient way to guide visual search, as when meeting a friend at a
crowded train station who has mentioned in advance, ‘I’ll be wearing
red’. Experimentally, people informed in this way about the likely
color of an upcoming task-relevant stimulus are faster to respond if the
stimulus does in fact appear with the expected color than if it appears
with another color (Most & Astur, 2007; Egner et al., 2008; see also
Hopf et al., 2004).
Seemingly in opposition to this intentional boost in processing of
separable features is an inherent bias to integrate (or ‘bind’) features
that co-occur in time and space. Previous literature has demonstrated
that enhanced processing spreads from an object’s task-relevant
features to its irrelevant features (O’Craven et al., 1999; Schoenfeld
et al., 2003; Wylie et al., 2004; Molholm et al., 2007; Katzner et al.,
2009; Fiebelkorn et al., 2010a,b). What’s more, this spreading of
attention across the features of an object seems to be highly automatic
(i.e. ‘bottom-up’; Fiebelkorn et al., 2010a).
We reasoned that such a spreading of attention across the features of
an object might undermine attempts at feature-based selective
attention. Selective attention entails a segregation of features whereas
binding entails an integration of features, so the two processes have
opposing aims. Our prediction was that binding, as an exogenous
process, would dominate the more endogenous feature-based selective
attention.
We tested this prediction in a prior behavioral study (Snyder &
Foxe, 2011), which used response time (RT) facilitation to measure
attentional biasing effects. Visual cues directed attention to one of two
basic visual features (color or motion), and a non-informative neutral
cue was used as a reference condition. Following the cue, participants
tried to detect a subtle variation in either the color or motion of a field
of moving dots. The cue correctly indicated which feature contained
the target 80% of the time (i.e. valid cues), but participants had to
Correspondence: Prof. J. J. Foxe,1Departments of Pediatrics and Neuroscience, as above.
E-mail: john.foxe@einsten.yu.edu
Received 16 November 2011, revised 21 December 2011, accepted 23 December 2011
European Journal of Neuroscience, pp. 1–8, 2012
doi:10.1111/j.1460-9568.2012.08016.x
ª 2012 The Authors. European Journal of Neuroscience ª 2012 Federation of European Neuroscience Societies and Blackwell Publishing Ltd
EJN 8016
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62

Page 2

respond to all targets, including those that were invalidly cued.
Moreover, we pinned performance below ceiling levels, such that
processing resources were limited, and any attentional enhancement of
one feature would necessarily come at the expense of some other
process. On top of this biased attention task, we used well-established
Gestalt grouping principles to manipulate the degree to which the
imperative stimulus was perceived as a single coherent object. In the
‘STRONG’ binding condition, all dots moved at the same speed and in
the same direction, which leads to the percept that the dots are
‘painted’ on a single transparent surface that is itself sliding past an
aperture. In the ‘WEAK’ binding condition dots moved in the same
direction but each at a unique rate, which unlike the STRONG
condition does not lead to the percept of a single object. We found that
for WEAK stimuli, valid cue information could be utilized to speed
responses relative to neutral cues, but no such RT benefit was found
for STRONG stimuli (Fig. 1). In other words, increasing the tendency
to perceive the stimulus as a single coherent object eliminated the
behavioral signature of biased attention.
Using a similar design, the current experiment follows up on our
earlier behavioral results by using event-related potential (ERP)
methods to examine the neurophysiological underpinnings of the
observed interference between selective attention and binding pro-
cesses. Use of ERP methods affords several advantages that comple-
ment and expand the behavioral approach just described. Most
obviously, ERPs provide information regarding the spatiotemporal
dynamics of the brain activity underlying the observed behavior.
Additionally, ERPs enable the use of an experimental design in which
the uncued feature can be completely ignored (rather than merely ‘less
attended’). This is in contrast to behavioral approaches, which require
overt responses to all targets, including those that were not indicated
by the cue. Because of this, the uncued feature is always task-relevant,
and should be attended to some degree, if only less so than the cued
feature. ERPs, which provide measures of brain activity that are not
dependent on overt behavior, do not have this drawback. Use of a
design in which the unattended feature is completely irrelevant creates
the greatest advantage for strongly biasing attention, and creates the
most powerful context for testing our predictions regarding the
automaticity of binding.
Previous research into feature-based selective attention has reported
ERP differences emerging about 200 ms post-stimulus onset that
localize to brain regions implicated in processing the attended feature
(Anllo-Vento & Hillyard, 1996; Hopf et al., 2004; Schoenfeld et al.,
2007). While the current experiment differs in several key respects
from these preceding examples, a similar sort of effect might be
expected for this experiment when feature-based selective attention is
effective (i.e. for WEAK stimuli). The key prediction of the current
experiment, however, is that attention effects in the ERP should be
diminished, delayed or entirely absent for STRONG stimuli, as
enhanced processing is reallocated to support binding across all the
features of the object, including the task-irrelevant features. Because
processing capacity is limited in this experiment, such a reallocation of
processing resources should involve a withdrawal of resources from
the attended feature, leading to a less-biased processing state.
Materials and methods
Participants
Sixteen adults (11 male, three left-handed) aged 20–47 years
(mean ± SD – 27.19 ± 6.75 years) participated in the experiment.
Three participants were excluded due to data quality concerns. The
resultant sample contained 10 males and three females aged 20–
47 years (27.92 ± 7.32 years), three of whom were left-handed.
Participants were sourced from the undergraduate and graduate
student populations of The City College of New York, and from the
local community. None of the participants had any history of brain
injury or disease, per self-report. Participants had normal or corrected-
to-normal visual acuity and normal color vision per self-report. All
participants provided informed consent prior to the experiment. All
materials and procedures were approved by the institutional review
board of The City College of New York in accordance with the United
States Public Health Service Act (US 45 CFR 46) and the Declaration
of Helsinki.
Stimuli
The experiment was administered in a light- and sound-attenuated
chamber using Presentation software version 14.4 (Neurobehavioral
Systems, Albany, CA, USA). All stimuli were presented on a
34.5 · 55.0 cm LCD monitor with a 60-Hz refresh rate (ViewSonic,
model VP2655wb). Trials began with a warning cue consisting of a
white fixation dot on a black background for 1 s, followed by a cue
word (‘COLOR’, ‘HUE’, ‘MOTION’ or ‘DIRECTION’) in white
block capitals for 1 s. The words ‘color’ and ‘hue’ both directed
attention to the color of the upcoming imperative stimulus (the second
stimulus, or ‘S2’). Likewise, ‘motion’ and ‘direction’ both directed
attention to the motion of the stimulus. The use of multiple cue words
for each feature was to reduce the automatization of the task through
implicit learning, with an aim to maintain the engagement of
endogenous orienting mechanisms throughout the session. After an
interval of 1.7–2.3 s (random, and evenly distributed) during which
only a black screen was displayed, the S2 was presented for 0.2 s.
Each S2 was followed by a 1-s response interval. Each subsequent trial
began immediately following the response interval (Fig. 2A).
The S2 consisted of an array of 1000 dots, each subtending 0.05? of
visual angle, constrained to a square aperture subtending 5? of visual
angle. In one condition, designed to engage Gestalt grouping
processes, all dots moved in the same direction at a speed of 21? of
visual angle per second. We refer to this condition as the STRONG
condition, indicating that Gestalt processes are strongly engaged. A
common percept of this type of stimulus is that the dots are ‘painted’
on a single transparent surface that is itself sliding past the aperture. In
Fig. 1. Response time (RT) results from Snyder & Foxe (2011). For the
depicted experiment, participants responded to all targets, including those that
were invalidly cued. Benefits for valid cue information were seen for WEAK
stimuli, but not for STRONG stimuli. That task-irrelevant binding processes led
to behavioral costs suggests that binding is automatic and interferes with
successful feature-based selective attention.
COLOR
2 A. C. Snyder et al.
ª 2012 The Authors. European Journal of Neuroscience ª 2012 Federation of European Neuroscience Societies and Blackwell Publishing Ltd
European Journal of Neuroscience, 1–8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62

Page 3

another condition, designed to minimize the engagement of Gestalt
grouping processes, each dot moved at a unique speed between 14 and
28? of visual angle per second, and all dots moved in the same
direction. For this type of stimulus the dots are not perceived as parts
of a single object as is the case when the dots all move at the same
speed. We refer to this condition as the WEAK condition, indicating
that Gestalt processes are weakly engaged (relative to the STRONG
condition). Dots ‘wrapped around’ the edges of the square aperture, so
that the total amount of illumination was held constant. STRONG and
WEAK S2s each occurred on 50% of trials in an unpredictable order.
It should be emphasized that the STRONG⁄WEAK manipulation was
irrelevant to the task. Participants were not informed by the
experimenter of this manipulation until debriefing, at which point
several participants expressed that they were unaware of the
STRONG⁄WEAK distinction during the experiment.
Dots were colored with hues from an isoluminant plane of DKL
color-space (Derrington et al., 1984). This color-space uses the
response properties of neurons in macaque lateral geniculate nucleus
to create a subjective luminance axis, planes orthogonal to which are
approximately isoluminant. The use of this color-space enables the
continuous variation of hue needed to derive hue discrimination
thresholds while controlling for subjective luminance.
Task
The task was adapted from previously reported experiments (Snyder &
Foxe, 2010, 2011). Half of the trials consisted of uniformly-colored
S2s (color standard), whereas for the other half of trials 20% of dots
had a slightly different color from the majority (color target).
Orthogonally to this color manipulation, half of trials contained S2s
for which all dots moved on a common linear trajectory (motion
standard) and the other half contained S2s for which the dots moved
on a curved trajectory (motion target). Schematized examples of
targets are illustrated in Fig. 2B. The degree of difference for each of
the relevant targets was titrated on a per-subject basis to 80% detection
rate prior to beginning each experiment using an up-down transformed
response modified staircase procedure (Wetherill & Levitt, 1965). No
particular value of any feature indicated a target – subjects had to
detect a particular feature ‘variation’ in the stimulus. This strategy was
used to reduce competition within a feature processing area (if subjects
were attending to red and suppressing green, for example). The goal,
rather, was to have subjects attend to ‘color’ and suppress ‘motion’, or
vice versa as the cue indicated.
Participants were instructed to attend to the cued feature only,
ignoring the uncued feature entirely, and to respond by pressing a
computer mouse button with the right index finger upon detection of a
target in the cued feature. The requirement to ignore the uncued
feature was intended to make any target properties appearing in the
uncued feature distracting, so that object-related spreading of attention
could provide no task-related benefits. Participants completed 10 10-
min blocks, which included breaks every 12 trials to reduce fatigue
and maintain a high level of alertness throughout the session.
Electroencephalogram (EEG) recording
Continuous EEG was acquired through the ActiveTwo BioSemi
(Amsterdam, the Netherlands) electrode system from 168 scalp
electrodes,digitizedat512 Hzwitha31.25 nVquantizationresolution.
Data were band-pass filtered during acquisition from 0.1 to 100 Hz.
With the BioSemi system, every electrode or combination of electrodes
can be assigned as the reference, which is done purely in software after
acquisition. BioSemi replaces the ground electrodes that are used in
conventional systems with two separate electrodes: Common Mode
Sense active electrode and Driven Right Leg passive electrode. These
two electrodes forma feedbackloop,whichdrives theaverage potential
of the subject as close as possible to the reference voltage of the analog-
to-digital converter, thus rendering them references. One electrode
placed 1 cm posterior to each orbital canthus and one electrode placed
on the nasion were used to monitor eye movements.
Data processing
EEG data were processed using the FieldTrip toolbox (Donders
Institute for Brain, Cognition and Behaviour, Radboud University
Nijmegen, the Netherlands) for MATLAB (The MathWorks, Natick,
MA, USA). Only trials with correct behavior (responses to targets and
withheld responses to non-targets) were included in the analysis. Data
A
B
Fig. 2. Task. (A) Schematic of task procedures. For each trial, subjects first viewed a fixation dot for 1 s, followed by a cue word in block capitals for 1 s. The cue
(S1) was followed by an interval of 1.7–2.3 s with no stimulation. After the cue–target interval, the random dot stimulus (S2) was shown for 0.2 s, followed by a 1-s
response period. The next trial began immediately following the response interval. The arrow in the S2 represents the motion of the dots and was not actually present
in the stimulus. Details have been enhanced for clarity of illustration. Timeline is not to scale. (B) Schematized illustration of target examples. Arrows represent
motion and were not actually present in the stimulus. The length of each arrow represents the speed of the adjacent dot. Color targets were defined by the presence of
two colors of dots. Motion targets were defined by the presence of curved motion (i.e. sequential presentation of different motion directions). Note that no one
particular color or motion direction was indicative of target presence, and that the examples given here are not exhaustive. Both color and motion targets could be
present in a single stimulus. Stimuli are shown here on a white background for illustration; stimuli were on a black background in the experiment. Color and motion
differences are enhanced for clarity. Dots in the actual stimuli were smaller and more numerous.
COLOR
Feature attention and objects3
ª 2012 The Authors. European Journal of Neuroscience ª 2012 Federation of European Neuroscience Societies and Blackwell Publishing Ltd
European Journal of Neuroscience, 1–8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62

Page 4

were epoched from 100 ms pre-S2 onset to 500 ms post-S2 onset.
Each trial was first visually inspected, and channels showing transient,
large-amplitude electronic artifacts were linearly interpolated on a per-
trial basis. Trials with more than five interpolated channels were
excluded from further analysis. Then, the data were decomposed into
independent components that were used for correction of muscular
and ocular artifacts.
To perform the independent components analysis (ICA), we first
concatenated all the trials for each individual participant. This resulted
in very large datasets, and so we next downsampled the data by a
factor of four for efficiency of the ICA algorithm. The concatenated,
downsampled data were then used to derive unmixing matrices using
the FastICA algorithm with default settings (Hyva ¨rinen, 1999). The
downsampled data were then discarded, and the unmixing matrices
were used to unmix the original, undownsampled data for each
participant into independent components. Components reflecting
artifacts were identified using an automated approach described by
Whittingstall et al. (2010). Artifacts were identified based on temporal
and spatial characteristics. For the temporal criterion, component time
courses were rectified and then converted to z-scores, and components
containing z-scores exceeding a value of 7 were discarded. This
criterion identifies infrequently-occurring, large-amplitude, transient
artifacts. For the spatial criterion, each component topography (i.e.
each column of the unmixing matrix) was converted to z-scores, and
components for which the largest z-score exceeded a value of 9, or for
which the sum of the two largest z-scores exceeded 15, were
discarded. This criterion identifies artifacts that are limited to one or
two electrodes. The choice of z-score threshold values was adopted
from the report by Whittingstall et al. (2010). In a preliminary stage,
we tested this method with several z-score thresholds and found that
the particular set of components rejected was consistent across a
moderate range of threshold values. After rejecting components
reflecting artifacts, the remaining components were remixed into
sensor-space and re-divided into individual epochs for further
processing and analysis.
Artifact-corrected data were re-referenced to a frontopolar electrode
equivalent to Fpz in the 10–20 system. This reference was chosen
because it was a clean recording site in all participants far from
occipital cortices, which were hypothesized to be the regions likely to
show maximal effects in this visual paradigm. Data were then low-
pass filtered below 30 Hz and then high-pass filtered above 0.3 Hz. In
each case a 6th-order zero-phase digital Butterworth filter was used.
Expanded epochs (2 s pre-S2 to 2 s post-S2) were used for filtering to
avoid contamination by artifacts introduced by discontinuities at the
edge of the epochs. Subsequent to filtering, the epochs were trimmed
back to 100 ms pre-S2 onset to 500 ms post-S2 onset and averaged
across trials. A mean of 161 trials (± 25.4 SD) was included in the
ERP for each condition for each participant.
Analysis
Behavioral analysis
We analysed both performance accuracy and RTs. In each case we
used a repeated-measures anova with factors of ‘feature’ (two levels:
color and motion) and ‘S2 type’ (two levels: WEAK and STRONG).
Our dependent measure for performance accuracy was the discrim-
inability index (d¢), a measure of performance that takes into account
response bias. For RT, we tested the median RTs across trials for each
participant for each cell of the anova design matrix. We chose the
median as the measure of central tendency because it is robust to
outliers, which were likely in this case because participants were not
instructed to respond rapidly. Behavioral statistics were analysed using
pasw version 18.0.0 (SPSS, Somers, NY, USA). We should mention
that unlike our prior study (Snyder & Foxe, 2010) in which behavioral
effects of biased attention were explicitly measured, no behavioral
differences were predicted for the current study. The reasons for this
are twofold: (1) responses were only made to the cued feature, so there
were no ‘valid’ or ‘invalid’ cues between which RT would differ; and
(2) we explicitly titrated performance to equate task difficulty across
all conditions. Therefore, our aim in analysing behavioral outcomes
for this study was simply to confirm that task difficulty was equated
across conditions.
Statistical hypothesis testing of ERP differences
We had a directed hypothesis that the ERP indices of attentional
biasing would be greater for WEAK S2s than for STRONG S2s. To
identify the ERP indices of our feature-based attention manipulation,
we performed a running dependent samples t-test between conditions
of attention-to-motion and attention-to-color for all channels and time
points. We used a clustering approach to control for inflation of type I
error due to multiple comparisons (cf. Guthrie & Buchwald, 1991).
The rationale for this method is that type I errors are unlikely to occur
simultaneously at adjacent electrodes and unlikely to endure for
several consecutive time points. Because the EEG signal does not
change arbitrarily fast, however, there is some dependence between
consecutive time points. We examined the autocorrelation of the noise
in the baseline interval of the ERP to determine the minimum time
delay at which time points in the EEG are not significantly correlated,
which we found to be 15 samples (29 ms). We thus required two-
tailed P-values below 0.05 to be simultaneously present at two or more
adjacent electrodes and to persist for at least 15 samples to consider
the effects significant. This statistical analysis was applied separately
to the STRONG and WEAK conditions. The results of this analysis
indicated whether a significant attention effect was present for either
STRONG or WEAK stimuli. We also performed an orthogonal
comparison of feature-based attentional deployment across binding
conditions. To do this, we first derived difference waves for each
subject by subtracting the ERPs when motion was attended from the
ERPs when color was attended, separately for each binding condition.
We then tested for significant differences between these difference
waves using a non-parametric repeated-measures randomization
procedure as follows: (1) we first measured the mean difference for
each channel and time point between the observed WEAK and
STRONG difference waves; (2) we then selected a random subset of
our sample and swapped the categorization of the WEAK and
STRONG difference waves for those participants (this swapping
reflects the null hypothesis that the WEAK and STRONG conditions
are equivalent); and (3) we noted the difference between the two
randomly repartitioned groups. This random repartitioning procedure
was repeated 1000 times to generate a distribution of such differences.
The probability (P-value) that the observed difference between the
WEAK and STRONG attention-related difference waves is due to
chance (under the null hypothesis) is equal to the proportion of the
distribution that exceeds that observed difference. We also performed
a running dependent samples t-test comparing all STRONG with
WEAK responses, regardless of which feature was attended, to
characterize main effects of the Gestalt grouping manipulation.
Source estimation
To visualize the cortical regions most likely to be contributing to
effects observed at the scalp, we employed a minimum norm current
source estimation (MNE) method using Brain Electrical Source
4 A. C. Snyder et al.
ª 2012 The Authors. European Journal of Neuroscience ª 2012 Federation of European Neuroscience Societies and Blackwell Publishing Ltd
European Journal of Neuroscience, 1–8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62

Page 5

Analysis software (BESA GmbH, ver. 5.1.8), using the following
settings: depth weighting and spatiotemporal weighting using
subspace correlation (dimension = 7); and noise estimation using
the 15% lowest values, weighted by averaging over channels. Source
localization was applied to subtraction waveforms representing
effects of interest. Subtractions were made as follows: (1) attend-
colorWEAK minusattend-motion
STRONG minus attend-motion STRONG; and (3) STRONG minus
WEAK.
WEAK;(2)attend-color
Results
Behavior
Performance accuracy
Discriminability
(mean ± SD – 2.44 ± 0.76), indicating that all participants performed
above chance (i.e. d¢ > 0) and below ceiling (i.e. d¢ < 4.65, which
reflects a 99% hit rate and 1% false alarm rate). There were no main
effects or interactions for the factors of ‘feature’ and ‘S2 type’ on d¢
scores (all P > 0.12). The lack of significant differences was
consistent with our aim to equate task difficulty across conditions
(Fig. 3A).
scoresrangedfrom
d¢ = 0.88to
d¢ = 4.22
RTs
Median RTs ranged from 511.7 ms to 957.0 ms (698.8 ± 113.9 ms).
There was a main effect of ‘feature’ (F1,12= 6.593, P = 0.025). A
post hoc paired-samples t-test showed that responses to motion targets
were slower than responses to color targets (P = 0.025). Overall,
participants responded 42.2 ms (SD – 59.3 ms) slower to motion than
color, and 10 out of 13 participants showed this general pattern.
Slower RTs for motion judgments compared with color judgments are
not uncommon in feature-based attention studies (Anllo-Vento &
Hillyard, 1996; Schoenfeld et al., 2007), and likely reflect the
integrative nature of motion processing (Fig. 3B).
No main effect was found for the factor of ‘S2 type’ (F1,12= 1.241,
P = 0.287). The ‘feature’ by ‘S2 type’ interaction was not significant
(F1,12= 0.207, P = 0.657).
Statistical testing of ERP differences
For WEAK stimuli, ERP differences between attend-color and attend-
motion were found at 11 parieto-occipital electrodes from 180 ms to
221 ms post-stimulus onset (Figs 4 and 5A). The ERP was more
negative for attend-color than for attend-motion at these locations for
this time period. In contrast, no attention effects for STRONG stimuli
were found (Fig. 4).
We compared attention effects between WEAK and STRONG
stimuli using a non-parametric randomization test. Difference waves
reflecting feature-based attention were compared across binding
conditions. Significant differences were found from 174 to 219 ms,
which closely overlaps the period of significant attention effects
within the WEAK condition. Figure 6 illustrates the spatial extent of
the effect during the overlapping time period. We found that attention-
related differences for WEAK stimuli were greater than those for
STRONG stimuli at six of the 11 electrodes that showed an attention
effect within the WEAK condition (P < 0.05; Fig. 6). There was a
trend towards significance at another nine nearby electrodes
(P < 0.1).
A comparison between the ERPs for STRONG and WEAK stimuli
revealed a robust effect at parieto-occipital electrodes from 273 to
404 ms (Fig. 7A and B). The ERP was more positive for STRONG
stimuli than for WEAK stimuli at these times and locations.
It is reasonable to suppose that the moving dot stimuli used for this
experiment might induce eye movements that could potentially
confound the electrophysiological results reported here. We examined
the bipolar-referenced electrooculogram channels prior to artifact
correction, and found that eye movements were not made prior to
400 ms and, even thereafter, any eye movements that were seen did
not differ across conditions. Thus, it seems unlikely that eye
movements are confounding the observed effects.
Source estimation
The MNE analysis localized the early (180–221 ms) attention effect
for WEAK stimuli to occipital pole regions and posterior ventral
occipito-temporal cortex bilaterally (Fig. 5B). The general Gestalt
binding effect (i.e. STRONG vs. WEAK) localized to right lateral
occipital cortex and left dorsolateral prefrontal cortex (Fig. 7C).
Discussion
Here we tested whether the electrophysiological effects of feature-
based selective attention persist despite environmental cues (i.e.
A
B
Fig. 3. Summary of behavioral results for the current experiment. (A)
Performance accuracy in d¢ scores. (B) Response times (RTs) in ms.
*P < 0.05 (repeated-measures anova). Error bars indicate two standard errors
of the mean.
COLOR
Feature attention and objects5
ª 2012 The Authors. European Journal of Neuroscience ª 2012 Federation of European Neuroscience Societies and Blackwell Publishing Ltd
European Journal of Neuroscience, 1–8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62

Page 6

common fate) that typically lead to object-based integration through
the bottom-up spreading of attention (O’Craven et al., 1999; Schoen-
feld et al., 2003; Busse et al., 2005;
words, we examined whether the well-established ability to bias
processing toward a specific feature (color or motion) would be
counteracted by the well-established bias to form coherent objects. To
test the automaticity of the spreading of attention, which is thought to
reflect feature binding, we designed a task where object-based
integration would be detrimental to performance.
On each trial, participants were cued to detect either color or motion
targets, while overall performance was held near 80%. Pinning
performance below 100% accuracy served to saturate attentional
resources, such that any boost in processing the irrelevant feature
would necessarily detract from processing the relevant one. If Gestalt
principles in the STRONG condition led to a spreading of attention
from the relevant feature to the irrelevant feature, we would predict
that the two features would be processed with greater parity. In other
words, any differences in the ERPs between attention to motion and
attention to color should be smaller for STRONG than for WEAK
stimuli. Evidence of decreased feature-based selectivity in the
presence of environmental cues for binding would suggest that the
inherent bias to process objects as wholes supersedes task-driven
attempts to bias processing toward a particular feature in order to
optimize performance.
1
Fiebelkorn et al., 2010a). In other
Our electrophysiological measure of feature-selective processing
was the difference between ERPs when color was attended compared
with when motion was attended. For the WEAK condition, robust
evidence of biased processing for the cued feature was evident at
about 180 ms; whereas, for the STRONG condition, evidence of
biased processing for the cued feature was not observed. We therefore
conclude that feature-based selective attention is substantially atten-
uated by object-based binding processes. These data emphasize the
strength of the tendency to integrate features that share a common fate,
even when doing so will interfere with the task at hand, as was
demonstrated by our earlier behavioral report (Snyder & Foxe, 2011;
Fig. 1).
In addition to providing confirmatory support for the previously
reported behavioral effects, the results of the current experiment
provide important additional information. Firstly, the use of an
instructional cue in the current experiment speaks to the robustness
of the effect. In our prior experiment, participants were required to
respond to all targets, including the infrequent cases when the target
appeared in the uncued feature. Thus, the optimal strategy in that
case would not have been to completely ignore the uncued feature,
but merely to attend to it less. Thus, biasing might not have been
particularly strong to begin with, increasing susceptibility of feature-
based attention to interference by binding. In the current design,
however, the uncued feature never warranted a response, and could
in fact be a source of distracting information. Thus, the optimal
strategy would be to actively ignore the uncued feature. In this way,
the strongest anticipatory biasing is encouraged. That disruptions of
feature-based attention by binding were still observed in this
extreme case of attentive biasing indicate the robustness of the
effect.
Moreover, the ERP method provides information regarding the
spatiotemporal dynamics of the brain activity underlying the effect.
The significant differences that we observed between attention to color
and attention to motion for the WEAK condition are similar in terms
of latency and topography to previously observed feature-specific
attention effects (Anllo-Vento & Hillyard, 1996; Hopf et al., 2004;
Schoenfeld et al., 2007). Source localizations implicated ventral
temporal lobes bilaterally, with additional contributions from occipital
pole regions. Because color processing has been previously localized
Fig. 4. Feature-based attention effects for WEAK and STRONG stimuli. ERPs in response to S2s are plotted for a representative electrode (electrode location
highlighted in insets) above the P-values for a running t-test of the comparison ‘color vs. motion’. The green dashed line represents P = 0.05. The light green
rectangle highlights the time period of significant difference.
AB
Fig. 5. Maps of feature-based attention effects. (A) Scalp topography of
attend-color ERP minus attend-motion ERP for WEAK stimuli for the time
period of significant difference. Black circles depict significant channels. (B)
MNE for time period with significant attention effects for WEAK stimuli (L,
left; P, posterior; R, right; V, ventral).
COLOR
COLOR
6A. C. Snyder et al.
ª 2012 The Authors. European Journal of Neuroscience ª 2012 Federation of European Neuroscience Societies and Blackwell Publishing Ltd
European Journal of Neuroscience, 1–8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62

Page 7

to the ventral occipito-temporal regions (Corbetta et al., 1991; Vaina,
1994), we interpret this pattern of results as enhanced processing in
color-related cortical areas when color is attended compared with
when motion is attended.
Because the evidence of feature-based selective attention was
present for WEAK stimuli at 180 ms, but absent for STRONG stimuli
at that time, we can conclude that object-based binding effects within
STRONG stimuli must have onset prior to 180 ms. An obvious
question is how the timing of our current effects accord with timing of
known binding processes. Evidence for the timing of object binding
comes from several electrophysiological investigations of the canon-
ical ‘illusory contour’ stimulus class. These studies have utilized
Kanisza illusory figures (cf. Kanizsa, 1976)
arrays of Pacman-shaped inducers oriented with their ‘mouths’
aligned toward a common point, such that the region interior to the
inducers is perceived as a geometric shape superimposed on the
background, although no such shape actually exists. If the inducers
are sufficiently misaligned the illusion is destroyed. By comparing the
ERP to the Kanisza figures with their non-illusion-inducing counter-
parts, an ‘illusory contour effect’ has been identified that onsets at
about 100 ms and peaks at about 170 ms (Herrmann et al., 1999;
Murray et al., 2002, 2004, 2006; Foxe et al., 2005; Shpaner et al.,
2009; Fiebelkorn et al., 2010b), the effect is taken to indicate binding
of the disparate inducers into a single Gestalt, as in the current
experiment.
While our inference that binding effects occurred prior to 180 ms
was indirect, we directly observed effects of our Gestalt grouping
manipulation during later time periods (273–404 ms). This effect was
characterized by greater positivity at posterior electrodes, and
implicated a robust current source in right lateral occipital cortex.
Lateral occipital cortex has previously been implicated in object-
related processing (Martı ´nez et al., 2006, 2007b; Lucan et al., 2010),
particularly with regards to integrating fragmented images (Doniger
et al., 2000, 2001; Sehatpour et al., 2008)
(Murray et al., 2002, 2004; Martı ´nez et al., 2007a; Shpaner et al.,
2009).
We found that participants were successfully able to bias processing
resources between the features of color and motion. This biased
attentional state was evident in differences in the ERPs between when
motion was attended and color was attended. These effects were
relatively early (180 ms) and attributed to feature-specific cortical
processing regions. However, these differences were only evident
when environmental cues for object-based binding were weak. When
environmental cues for binding were strong, evidence for biased
feature-specific processing was not observed. We therefore conclude
that bottom-up binding processes supercede top-down feature-based
selective attention, even when explicitly detrimental to task perfor-
mance.
Our results provide insight into how attention and binding processes
are used to guide interaction with the environment in a natural setting.
We propose the following model: first, feature-based attention, which
is not spatially specific, highlights objects with the attended feature for
subsequent spatial selection. Binding processes then take over, which
leads to a spread of attention between the attended object’s constituent
features, and an object-based attentional selection is formed. The
attended object can then be scrutinized with respect to the individual
features, but this scrutiny is likely deferred to post-perceptual
timeframes.
2
, which are composed of
3
and illusory contours
Acknowledgements
This work was primarily supported by a grant from the US National Science
Foundation to Professor Foxe (BCS0642584), with additional support coming
from a grant from the US National Institute of Mental Health (NIMH – RO1 –
MH085322). Dr Snyder was supported by a Ruth L. Kirschstein National
Research Service Award (NRSA) pre-doctoral fellowship from the National
Institute of Mental Health (NIMH – F31 – MH087077). The authors would like
to thank Ms Sara Schildkraut for help with data collection, and Mr Ted S.
Altschuler and Dr John Butler for insightful feedback on earlier versions of this
manuscript.
Fig. 6. Results of non-parametric statistical test for the comparison ‘WEAK
attention effect > STRONG attention effect’. Shading denotes P-values with a
threshold of P < 0.1. A cluster of 12 contiguous electrodes can be seen over
parietal occipital scalp where the effect of attention for WEAK stimuli is
significantly greater than the effect of attention for STRONG stimuli.
A
BC
Fig. 7. General Gestalt grouping effect (i.e. STRONG minus WEAK). (A)
ERPs for representative electrode (location depicted in inset). Shaded rectangle
indicates time period of significant difference. (B) Representative topography
of the difference ‘STRONG minus WEAK’ during significant period. Black
circles indicate significant electrodes. (C) MNE for time period with significant
effect.
COLOR
Feature attention and objects7
ª 2012 The Authors. European Journal of Neuroscience ª 2012 Federation of European Neuroscience Societies and Blackwell Publishing Ltd
European Journal of Neuroscience, 1–8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62

Page 8

Abbreviations
EEG, electroencephalogram; ERP, event-related potential; ICA, independent
components analysis; MNE, minimum norm current source estimation; RT,
response time.
References
Anllo-Vento, L. & Hillyard, S.A. (1996) Selective attention to the color and
direction of moving stimuli: electrophysiological correlates of hierarchical
feature selection. Percept. Psychophys., 58, 191–206.
Broadbent, D. (1958) Perception and Communication. Pergamon Press,
London, UK.
Busse, L., Roberts, K.C., Crist, R.E., Weissman, D.H. & Woldorff, M.G.
(2005) The spread of attention across modalities and space in a multisensory
object. Proc. Natl Acad. Sci. USA, 105, 16380–16385.
Corbetta, M., Miezin, F.M., Dobmeyer, S., Shulman, G.L. & Petersen, S.E.
(1991) Selective and divided attention during visual discriminations of
shape, color, and speed: functional anatomy by positron emission tomog-
raphy. J. Neurosci., 11, 2383–2402.
Derrington, A.M., Krauskopf, J. & Lennie, P. (1984) Chromatic mechanisms in
lateral geniculate nucleus of macaque. J. Physiol., 357, 241–265.
Doniger, G.M., Foxe, J.J., Murray, M.M., Higgins, B.A., Snodgrass, J.G.,
Schroeder, C.E. & Javitt, D. (2000) Activation timecourse of ventral visual
stream object-recognition areas: high density electrical mapping of percep-
tual closure processes. J. Cogn. Neurosci., 12, 615–621.
Doniger, G.M., Foxe, J.J., Schroeder, C.E., Murray, M.M., Higgins, B.A. &
Javitt, D.C. (2001) Visual perceptual learning in human object recognition
areas: a repetition priming study using high-density electrical mapping.
Neuroimage, 13, 305–313.
Egner, T., Monti, J.M.P., Trittschuh, E.H., Wieneke, C.A., Hirsch, J. &
Mesulam, M.-M. (2008) Neural integration of top-down spatial and feature-
based information in visual search. J. Neurosci., 28, 6141–6151.
Fiebelkorn, I.C., Foxe, J.J. & Molholm, S. (2010a) Dual mechanisms for the
cross-sensory spread of attention: how much do learned associations matter?
Cereb. Cortex, 20, 109–120.
Fiebelkorn, I.C., Foxe, J.J., Schwartz, T.H. & Molholm, S. (2010b) Staying
within the lines: the formation of visuospatial boundaries influences
multisensory feature integration. Eur. J. Neurosci., 31, 1737–1743.
Foxe, J.J., Murray, M.M. & Javitt, D.C. (2005) Filling-in in schizophrenia: a
high-density electrical mapping study and source-analysis investigation of
illusory contour processing. Cereb. Cortex, 15, 1914–1927.
Guthrie, D. & Buchwald, J.S. (1991) Significance testing of difference
potentials. Psychophysiology, 28, 240–244.
Herrmann, C.S., Mecklinger, A. & Pfeifer, E. (1999) Gamma responses and
ERPs in a visual classification task. Clin. Neurophysiol., 110, 636–642.
Hopf, J.-M., Boelmans, K., Schoenfeld, M.A., Luck, S.J. & Heinze, H.-J.
(2004) Attention to features precedes attention to locations in visual search:
evidence from electromagnetic brain responses in humans. J. Neurosci., 24,
1822–1832.
Hyva ¨rinen, A. (1999) Fast and robust fixed-point algorithms for independent
component analysis. IEEE Trans. Neural Netw., 10, 626–634.
Kahneman, D. (1973) Attention and Effort. Prentice-Hall, Englewood Cliffs,
NJ.
Kanizsa, G. (1976) Subjective Contours. Sci. Am., 234, 48–52.
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. Front. Syst. Neurosci., 3, 12.
Koch, K., McLean, J., Segev, R., Freed, M.A., Berry, M.J. II, Balasubrama-
nian, V. & Sterling, P. (2006) How much the eye tells the brain. Curr. Biol.,
16, 1428–1434.
Lucan, J.N., Foxe, J.J., Gomez-Ramirez, M., Satian, K. & Molholm, S. (2010)
Tactile shape discrimination recruits human lateral occipital complex during
early perceptual processing. Hum. Brain Mapp., 31, 1813–1821.
Martı ´nez, A., Teder-Sa ¨leja ¨rvi, W., Vazquez, M., Molholm, S., Foxe, J.J., Javitt,
D.C., Di Russo, F., Worden, M.S. & Hillyard, S.A. (2006) ???. J. Cogn.
Neurosci., 18, 298–310.
4
Martı ´nez, A., Teder-Sa ¨leja ¨rvi, W. & Hillyard, S.A. (2007a) Spatial attention
facilitates selection of illusory objects: evidence from event-related brain
potentials. Brain Res., 1139, 143–152.
Martı ´nez, A., Ramanathan, D.S., Foxe, J.J., Javitt, D.C. & Hillyard, S.A.
(2007b) The role of spatial attention in the selection of real and illusory
objects. J. Neurosci., 27, 7963–7973.
Martinez-Trujillo, J.C. & Treue, S. (2004) Feature-based attention increases the
sensitivity of population responses in primate visual cortex. Curr. Biol., 14,
744–751.
Molholm, S., Martinez, A., Shpaner, M. & Foxe, J.J. (2007) Object-based
attention is multisensory: co-activation of an object’s representations in
ignored sensory modalities. Eur. J. Neurosci., 26, 499–509.
Most, S.B. & Astur, R.S. (2007) Feature-based attentional set as a cause of
traffic accidents. Vis. Cogn., 75, 499–509.
Murray, M.M., Wylie, G.R., Higgins, B.A., Javitt, D.C., Schroeder, C.E. &
Foxe, J.J. (2002) The spatiotemporal dynamics of illusory contour process-
ing: combined high-density electrical mapping, source analysis, and
functional magnetic resonance imaging. J. Neurosci., 22, 5055–5073.
Murray, M.M., Foxe, D.M., Javitt, D.C. & Foxe, J.J. (2004) Setting boundaries:
brain dynamics of modal and amodal illusory shape completion in humans.
J. Neurosci., 24, 6898–6903.
Murray, M.M., Imber, M.L., Javitt, D.C. & Foxe, J.J. (2006) Boundary
completion is automatic and dissociable from shape discrimination.
J. Neurosci., 26, 12043–12054.
Neely, J.H. (1977) Semantic priming and retrieval from lexical memory: roles
of inhibitionless spreading activation and limited-capacity attention. J. Exp.
Psychol., 106, 226–254.
O’Craven, K.M., Downing, P.E. & Kanwisher, N. (1999) fMRI evidence for
objects as the units of attentional selection. Nature, 401, 584–587.
Schneider, W. & Fisk, A.D. (1982) Concurrent automatic and controlled visual
search: can processing occur without resource cost? J. Exp. Psychol. Learn.
Mem. Cogn., 8, 261–278.
Schneider, W. & Shiffrin, R.M. (1977) Controlled and automatic human
information processing: I. Detection, search, and attention. Psychol. Rev., 84,
1–66.
Schoenfeld, M.A., Tempelmann, C., Martinez, A., Hopf, J.-M., Sattler, C.,
Heinze, H.-J. & Hillyard, S.A. (2003) Dynamics of feature binding during
object-selective attention. Proc. Natl Acad. Sci. USA, 100, 11806–11811.
Schoenfeld, M.A., Hopf, M.-A., Martinez, A., Mai, H.M., Sattler, C., Gasde,
A., Heinze, H.-J. & Hillyard, S.A. (2007) Spatio-temporal analysis of
feature-based attention. Cereb. Cortex, 17, 2468–2477.
Sehatpour, P., Molholm, S., Schwartz, T.H., Mahoney, J.R., Mehta, A.D.,
Javitt, D.C., Stanton, P.K. & Foxe, J.J. (2008) A human intracranial study of
long-range oscillatory coherence across a frontal-occipital-hippocampal
brain network during visual object processing. Proc. Natl Acad. Sci. USA,
105, 4399–4404.
Shiffrin, R.M. & Schneider, W. (1977) Controlled and automatic human
information processing: II. Perceptual learning, automatic attending and a
general theory. Psychol. Rev., 84, 127–190.
Shpaner, M., Murray, M.M. & Foxe, J.J. (2009) Early processing in the human
lateral occipital complex is highly responsive to illusory contours but not to
salient regions. Eur. J. Neurosci., 30, 2018–2028.
Snyder, A.C. & Foxe, J.J. (2010) Anticipatory attentional suppression of visual
features indexed by oscillatory alpha-band power increases: a high-density
electrical mapping study. J. Neurosci., 30, 4024–4032.
Snyder, A.C. & Foxe, J.J. (2011) The countervailing forces of binding
and selection in vision. Cortex, ???, ???–???, doi: 10.1016/j.cortex.2011.
05.003.
Vaina, L.M. (1994) Functional segregation of color and motion processing in
the human visual cortex: clinical evidence. Cereb. Cortex, 4, 555–572.
Wetherill, G.B. & Levitt, H. (1965) Sequential estimation of points on a
psychometric function. Br. J. Math. Stat. Psychol., 18, 1–10.
Whittingstall, K., Bartels, A., Singh, V., Kwon, S. & Logothetis, N.K. (2010)
Integration of EEG source imaging and fMRI during continuous viewing of
natural movies. Magn. Reson. Imaging, 28, 1135–1142.
Wylie, G.R., Javitt, D.C. & Foxe, J.J. (2004) Don’t think of a white bear: an
fMRI investigation of the effects of sequential instructional sets on cortical
activity in a task-switching paradigm. Hum. Brain Mapp., 21, 279–297.
5
6
8A. C. Snyder et al.
ª 2012 The Authors. European Journal of Neuroscience ª 2012 Federation of European Neuroscience Societies and Blackwell Publishing Ltd
European Journal of Neuroscience, 1–8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62

Page 9

Author Query Form
Journal:EJN
Article:8016
Dear Author,
During the copy-editing of your paper, the following queries arose. Please respond to these by marking up your proofs with the
necessary changes/additions. Please write your answers on the query sheet if there is insucient space on the page proofs. Please
write clearly and follow the conventions shown on the attached corrections sheet. If returning the proof by fax do not write too
close to the paper’s edge. Please remember that illegible mark-ups may delay publication.
Many thanks for your assistance.
Query
reference
QueryRemarks
1AUTHOR: Busse et al., 2010 has been changed to Busse et al., 2005
so that this citation matches the Reference List. Please confirm that this
is correct.
2AUTHOR: Kanisza, 1976 has been changed to Kanizsa, 1976 so that
this citation matches the Reference List. Please confirm that this is
correct.
3 AUTHOR: Sehatpour, et al., 2006 has been changed to Sehatpour
et al., 2008 so that this citation matches the Reference List. Please
confirm that this is correct.
4AUTHOR: Please provide the document title for reference Martı ´nez et
al. (2006).
5 AUTHOR: Martinez-Trujillo & Treue (2004) has not been cited in the
text. Please indicate where it should be cited; or delete from the
Reference List.
6 AUTHOR: Please provide the volume number, page range for reference
Snyder & Foxe (2011).
EJN
8016B
-
Dispatch: 7.2.12Journal: EJN
CE: Nivedha
Journal Name Manuscript No.
Author Received:No. of pages: 8 PE: Indumathi

Page 10

USING e-ANNOTATION TOOLS FOR ELECTRONIC PROOF CORRECTION

Required software to e-Annotate PDFs: Adobe Acrobat Professional or Adobe Reader (version 8.0 or
above). (Note that this document uses screenshots from Adobe Reader X)
The latest version of Acrobat Reader can be downloaded for free at: http://get.adobe.com/reader/

Once you have Acrobat Reader open on your computer, click on the Comment tab at the right of the toolbar:




























































1. Replace (Ins) Tool Î for replacing text.

Strikes a line through text and opens up a text
box where replacement text can be entered.
How to use it
‚
‚
Highlight a word or sentence.
Click on the Replace (Ins) icon in the Annotations
section.
‚ Type the replacement text into the blue box that
appears.
This will open up a panel down the right side of the document. The majority of
tools you will use for annotating your proof will be in the Annotations section,
rkevwtgf"qrrqukvg0"YgÓxg"rkemgf"qwv"uqog"qh"vjgug"vqqnu"dgnqy<

2. Strikethrough (Del) Tool Î for deleting text.

Strikes a red line through text that is to be
deleted.
How to use it
‚
‚
Highlight a word or sentence.
Click on the Strikethrough (Del) icon in the
Annotations section.



3. Add note to text Tool Î for highlighting a section
to be changed to bold or italic.

Highlights text in yellow and opens up a text
box where comments can be entered.
How to use it
‚
‚
Highlight the relevant section of text.
Click on the Add note to text icon in the
Annotations section.
‚ Type instruction on what should be changed
regarding the text into the yellow box that
appears.
4. Add sticky note Tool Î for making notes at
specific points in the text.

Marks a point in the proof where a comment
needs to be highlighted.
How to use it
‚ Click on the Add sticky note icon in the
Annotations section.
‚ Click at the point in the proof where the comment
should be inserted.
‚ Type the comment into the yellow box that
appears.

Page 11

USING e-ANNOTATION TOOLS FOR ELECTRONIC PROOF CORRECTION

















































For further information on how to annotate proofs, click on the Help menu to reveal a list of further options:
5. Attach File Tool Î for inserting large amounts of
text or replacement figures.

Inserts an icon linking to the attached file in the
appropriate pace in the text.
How to use it
‚ Click on the Attach File icon in the Annotations
section.
‚
Enkem"qp"vjg"rtqqh"vq"yjgtg"{qwÓf"nkmg"vjg"cvvcejgf"
file to be linked.
‚ Select the file to be attached from your computer
or network.
‚ Select the colour and type of icon that will appear
in the proof. Click OK.
6. Add stamp Tool Î for approving a proof if no
corrections are required.

Inserts a selected stamp onto an appropriate
place in the proof.
How to use it
‚ Click on the Add stamp icon in the Annotations
section.
‚ Select the stamp you want to use. (The Approved
stamp is usually available directly in the menu that
appears).
‚
Enkem"qp"vjg"rtqqh"yjgtg"{qwÓf"nkmg"vjg"uvcor"vq"
appear. (Where a proof is to be approved as it is,
this would normally be on the first page).
7. Drawing Markups Tools Î for drawing shapes, lines and freeform
annotations on proofs and commenting on these marks.
Allows shapes, lines and freeform annotations to be drawn on proofs and for
comment to be made on these marks..

How to use it
‚ Click on one of the shapes in the Drawing
Markups section.
‚ Click on the proof at the relevant point and
draw the selected shape with the cursor.
‚ To add a comment to the drawn shape,
move the cursor over the shape until an
arrowhead appears.
‚ Double click on the shape and type any
text in the red box that appears.