mechanisms of attentional processing during visual search...
TRANSCRIPT
Mechanisms of attentional processing during
visual search: how distraction is handled by the
brain
by
Gaspar, John Manuel
M.A. (Psychology), Simon Fraser University, 2012
Thesis Submitted in Partial Fulfillment of the
Requirements for the Degree of
Doctor of Philosophy
in the
Department of Psychology
Faculty of Arts and Social Sciences
John Manuel Gaspar 2016
SIMON FRASER UNIVERSITY
Fall 2016
ii
Approval
Name:
Degree:
Title:
Examining Committee:
John Manuel Gaspar
Doctor of Philosophy
Mechanisms of attentional processing during visual search: how distraction is handled by the brain
Chair: Thomas Spalek Professor
John McDonaldSenior Supervisor Professor
Mario LiottiSupervisorProfessor
Urs RibarySupervisor Professor
Sam DoesburgInternal ExaminerAssociate ProfessorDepartment of Biomedical Physiology and Kinesiology
Joseph Hopfinger External ExaminerProfessor Department of Psychology and NeuroscienceUniversity of North Carolina at Chapel Hill
Date Defended/Approved: November 23, 2016
iii
Ethics Statement
iv
Abstract
In order to effectively search the visual environment, an observer must continually locate
objects of interest amid an abundance of irrelevant and distracting stimuli. These visual
distractors can sometimes inadvertently attract attention to their locations, even when an
observer is attempting to search for an entirely different object. To deal with visual
distractors, it has been well established that the visual system can implement a
suppression mechanism to filter out irrelevant stimuli. Within the past decade, event-
related potential (ERP) recordings have isolated an attentional component that is thought
to reflect this suppressive processing. This ERP component—termed the distractor
positivity (PD)—has been used to demonstrate that the sensory processing of irrelevant
information can be strongly modulated in line with the visual search goals of an observer.
Here, four electrophysiological studies of attention are presented which focus on yielding
insight into how the visual system deals with irrelevant information during visual search
and seeks to further our understanding of the PD component. Chapter 2 tests the stimulus
conditions necessary to elicit the distractor suppression indexed by the PD by examining
how differences in the salience of an irrelevant stimuli affect visual search. Chapter 3
explores how individual differences in target and distractor processing are associated with
variations in visual working memory (vWM) capacity. Chapter 4 asks how distractor
processing is altered during a disruption of attentional control by examining how visual
search is affected during the attentional blink (AB). Chapter 5 explores how high levels of
trait anxiety alter inhibitory control and the ability to ignore distracting information. In the
final chapter, future directions are discussed and a model for attentional processing is
proposed.
Keywords: suppression, attention, event-related potentials, distractor positivity, individual differences
v
Dedication
This work is dedicated to my mother and father,
for their love, patience, and sacrifice.
vi
Table of Contents
Approval .......................................................................................................................... ii Ethics Statement ............................................................................................................ iii Abstract .......................................................................................................................... iv Dedication ....................................................................................................................... v Table of Contents ........................................................................................................... vi List of Figures................................................................................................................. ix List of Acronyms ............................................................................................................ xiii
General Introduction ............................................................................... 1 1.1. How objects are prioritized for selection in the visual environment? ........................ 2 1.2. Top-down versus bottom-up processing: the debate .............................................. 4
1.2.1. Stimulus-driven capture ............................................................................. 5 1.2.2. Contingent-capture .................................................................................... 6
1.3. An alternative to the dichotomy: signal suppression ............................................... 7 1.3.1. The ERP technique and components associated with attention ................. 8 1.3.2. Evidence for the signal suppression hypothesis....................................... 10 1.3.3. The additional singleton task and signal suppression .............................. 11 1.3.4. Other evidence for signal suppression ..................................................... 13
1.4. Extending our knowledge of signal suppression ................................................... 13 1.4.1. Chapter 2: Eliciting signal suppression during visual search .................... 14 1.4.2. Chapter 3: Individual differences in working memory and visual
search ..................................................................................................... 15 1.4.3. Chapter 4: Signal suppression during a transient loss of attentional
control ..................................................................................................... 16 1.4.4. Chapter 5: Individuals with high levels of trait anxiety show
differences in selective attentional processing ......................................... 17
Eliciting signal suppression during visual search .............................. 19 2.1. Introduction ........................................................................................................... 19 2.2. Methods ............................................................................................................... 22
2.2.1. Materials and Methods ............................................................................ 22 2.2.2. Participants .............................................................................................. 23 2.2.3. Behavioural Pilot Stimuli and Apparatus .................................................. 23 2.2.4. Behavioural Pilot Procedure .................................................................... 23 2.2.5. Visual Search Task Stimuli and Apparatus .............................................. 24 2.2.6. Visual Search Task Procedure ................................................................ 25 2.2.7. Behavioural Analysis ............................................................................... 25 2.2.8. Electrophysiological Recording and Analysis ........................................... 26
2.3. Results ................................................................................................................. 28 2.3.1. Both high- and low-salience distractors produce behavioural
interference ............................................................................................. 28 2.3.2. High- but not low-salience distractors vary as a function of target-
distractor distance ................................................................................... 30 2.3.3. Interference without evidence of attentional capture ................................ 31
vii
2.3.4. Only salient distractors are suppressed during additional singleton search. .................................................................................................... 32
2.3.5. Distractor suppression can be indirectly observed in differences in N2pc amplitude. ....................................................................................... 33
2.4. Discussion ............................................................................................................ 36
Individual differences in working memory .......................................... 40 3.1. Introduction ........................................................................................................... 40 3.2. Materials and Methods ......................................................................................... 42
3.2.1. Participants .............................................................................................. 43 3.2.2. Working Memory Capacity Procedure ..................................................... 43 3.2.3. Visual Search Task Stimuli and Apparatus .............................................. 43 3.2.4. Visual Search Task Procedure ................................................................ 43 3.2.5. Behavioural Analysis ............................................................................... 44 3.2.6. Electrophysiological Recording and Analysis ........................................... 44
3.3. Results ................................................................................................................. 45 3.3.1. Behaviour in Change-Detection Task ...................................................... 45 3.3.2. Behavior in Visual Search Task ............................................................... 46 3.3.3. Neural activity associated with distractor suppression ............................. 46 3.3.4. Neural activity associated with distractor suppression predicts
individual differences in vWM .................................................................. 48 3.3.5. Neural activity associated with target processing ..................................... 51 3.3.6. Neural activity associated with target processing does not predict
individual differences in vWM .................................................................. 52 3.4. Discussion ............................................................................................................ 54
Signal suppression during a transient loss of attentional control .................................................................................................... 57
4.1. Introduction ........................................................................................................... 57 4.2. Methods ............................................................................................................... 59
4.2.1. Materials and Methods ............................................................................ 59 4.2.2. Participants .............................................................................................. 59 4.2.3. Attentional Blink Task Stimuli and Apparatus ........................................... 60 4.2.4. Attentional Blink Task Procedure ............................................................. 61 4.2.5. Behavioural Analysis ............................................................................... 62 4.2.6. Electrophysiological Recording and Analysis ........................................... 62
4.3. Results ................................................................................................................. 65 4.3.1. Visual search is delayed during the attentional blink ................................ 65 4.3.2. The N2pc is delayed within the attentional blink....................................... 66 4.3.3. The PPC is unaffected during the attentional blink ................................... 67 4.3.4. Individuals cannot recruit distractor suppression during the
attentional blink........................................................................................ 68 4.3.5. Behavioural evidence during the AB revisited .......................................... 69
4.4. Discussion ............................................................................................................ 70
viii
Individuals with high levels of trait anxiety show differences in selective attentional processing ........................................................... 74
5.1. Introduction ........................................................................................................... 74 5.2. Methods ............................................................................................................... 77
5.2.1. Materials and Methods ............................................................................ 77 5.2.2. STAI Prescreen ....................................................................................... 77 5.2.3. Participants .............................................................................................. 78 5.2.4. Visual Search Task Stimuli and Apparatus .............................................. 78 5.2.5. Visual Search Task Procedure ................................................................ 78 5.2.6. Behavioural Analysis ............................................................................... 79 5.2.7. Electrophysiological Recording and Analysis ........................................... 80
5.3. Results ................................................................................................................. 82 5.3.1. STAI scores ............................................................................................. 82 5.3.2. Search performance does not differ between individuals with high-
and low-anxiety individuals ...................................................................... 82 5.3.3. Suppression is preceded by an attentional shift to the distractor in
high-anxiety individuals ............................................................................ 84 5.3.4. Differences in target processing between high-anxiety and low-
anxiety individuals ................................................................................... 87 5.4. Discussion ............................................................................................................ 89
General Discussion ............................................................................... 94 6.1. Attentional capture revisited.................................................................................. 94 6.2. The PD is a measure of top-down signal suppression ........................................... 96 6.3. What is the clinical value of the PD as an index of attentional processing? ........... 98
6.3.1. ADHD ...................................................................................................... 98 6.3.2. Aging and attention .................................................................................. 98 6.3.3. Molecular biology, genetics, and selective attention ................................ 99
6.4. A proposed stream for visual processing ............................................................ 100
References .................................................................................................................103
ix
List of Figures
Figure 2.1. Trial Types. Example stimulus displays from the two experimental conditions. Subjects were instructed to attend to the yellow circle and to identify the orientation of the line inside of it. On 66% of trials, a distractor singleton was simultaneously presented within the display. ............................................................................................. 22
Figure 2.2. RTs associated with distractor interference. Mean response times (across participants; in milliseconds) for blue distractor, red distractor, and distractor absent (x) trials for Condition 1 and Condition 2 (left). Mean response times were then collapsed to create high-salience, low-salience, and distractor absent trials across the two experimental conditions (right). ....................................... 29
Figure 2.3. Target-distractor RT distance effects. Mean response times (across participants; in milliseconds) for five target-distractor distances (d1- d5) for both high- and low-salience distractor trials.......... 31
Figure 2.4. ERPs elicited by displays containing a midline target and a lateral distractor for each non-target condition. Time 0 reflects the onset of the search display, and negative voltage deflections are plotted above the x-axis, by convention. Waveforms were recorded over the lateral occipital scalp (electrodes PO7 and PO8). (A) ERPs recorded contralateral and ipsilateral to a high-salience distractor. (B) ERPs recorded contralateral and ipsilateral to a low-salience distractor. ............................................................................................... 33
Figure 2.5. ERPs elicited by displays containing a lateral target and a midline distractor for each non-target condition. ERPs are presented as contralateral-minus-ipsilateral difference waveforms for displays containing a lateral target and a distractor singleton, recorded over the lateral occipital scalp (electrodes PO7 and PO8). Difference waveforms are separated for trials where the high- and the low-salience distractor were presented in both (A) Condition 1 and (B) Condition 2. ............................................................................................ 34
Figure 3.1. ERPs elicited by displays containing a midline target and a lateral distractor. Time 0 reflects the onset of the search display, and negative voltage deflections are plotted above the x-axis, by convention. Waveforms were recorded over the lateral occipital scalp (electrodes PO7 and PO8). (A) ERPs recorded contralateral and ipsilateral to a high-salience distractor. (B) ERPs recorded contralateral and ipsilateral to a low-salience distractor. (C) Contralateral-minus-ipsilateral difference waveforms for both conditions. .............................................................................................. 47
x
Figure 3.2. Neural activity associated with salient distractor suppression predicts visual working memory capacity. (A) ERP waveforms recorded contralateral and ipsilateral to the salient distractor plotted separately for high-, medium-, and low-capacity groups. (B) Contralateral-minus-ipsilateral difference waveforms for high-, medium-, and low-capacity groups. ........................................................ 49
Figure 3.3. Neural activity associated with salient distractor suppression predicts visual working memory capacity. (a) Correlation between memory capacity (k) and the mean amplitude of the PD. (b) Correlation between memory capacity (k) and the “pure” PD area. The “pure” PD area reflects the area of the signed positive voltage under the curve between 200-350 ms minus the area of the signed positive voltage in the baseline between -150-0 ms prior to the onset of the search array. ................................................................. 51
Figure 3.4. Neural activity associated with target processing not predictive of visual working memory capacity. (A) Correlation between memory capacity (k) and pure N2pc area for lateral-target displays of interest. (B) Contralateral-minus-ipsilateral difference waveforms for high-, medium-, and low-capacity groups. ......................................... 53
Figure 4.1. Example stimulus display from the experiment. T1 was a number presented amongst letters in an RSVP stream. T2 was an additional singleton search display where participants were instructed to identify the orientation of the line inside the yellow colour singleton. Participants were instructed to give a speeded response to the search array first and then identify the number as either even or odd. ................................................................................. 61
Figure 4.2. Main behavioural results: (A) Accuracy rates for T1 and T2 on both lag 2 and lag 8 trials. (B) Mean response times (across participants; in milliseconds) for lag 2 and lag 8 trials. ............................ 66
Figure 4.3. ERPs elicited by trials with displays containing a lateral target and a midline distractor. Time 0 reflects the onset of the search display, and negative voltage deflections are plotted above the x-axis, by convention. Waveforms were recorded over the lateral occipital scalp (electrodes PO7 and PO8). (A) ERPs recorded contralateral and ipsilateral to a distractor for lag 8 and lag 2 trials. (B) Contralateral-minus-ipsilateral difference waveforms for lag 8 and lag 2 trials. ....................................................................................... 67
Figure 4.4. ERPs elicited by trials with displays containing a midline target and a lateral distractor. Time 0 reflects the onset of the search display, and negative voltage deflections are plotted above the x-axis, by convention. Waveforms were recorded over the lateral occipital scalp (electrodes PO7 and PO8). (A) ERPs recorded contralateral and ipsilateral to a distractor for lag 8 and lag 2 trials. (B) Contralateral-minus-ipsilateral difference waveforms for lag 8 and lag 2 trials. ....................................................................................... 69
xi
Figure 4.5. Target-distractor RT distance effects. Mean response times (across participants; in milliseconds) for lag 2 and lag 8 trials where the target and distractor appeared adjacent to one another and on trials where they appeared furthest from one another. ................ 70
Figure 5.1. Trial Types. Example stimulus displays from the two experimental conditions. Subjects were instructed to attend to the yellow circle and to identify the orientation of the line inside of it. On 50% of trials, a salient distractor singleton was simultaneously presented within the display. ................................................................................... 77
Figure 5.2 Target-distractor RT distance effects. Mean response times (across participants; in milliseconds) for five target-distractor distances (d1- d5) for both high- and low-anxiety individuals. ................. 83
Figure 5.3 PD ERPs elicited by trials with displays containing a midline target and a lateral distractor. Time 0 reflects the onset of the search display, and negative voltage deflections are plotted above the x-axis, by convention. Waveforms were recorded over the lateral occipital scalp (electrodes PO7 and PO8). (A) ERPs recorded contralateral and ipsilateral to a distractor for high- and low-anxiety individuals. (B) Contralateral-minus-ipsilateral difference waveforms for high- and low-anxiety individuals. .................................... 85
Figure 5.4 High-anxiety group ERPs for displays containing a midline target and a lateral distractor, separately for fast- and slow-response trials. (A) ERPs recorded contralateral and ipsilateral to a distractor for fastest and slowest trials. (B) Contralateral-minus-ipsilateral difference waveforms. ............................................................ 86
Figure 5.5 N2pc ERPs elicited by trials with displays containing a lateral target and no distractor. Time 0 reflects the onset of the search display, and negative voltage deflections are plotted above the x-axis, by convention. Waveforms were recorded over the lateral occipital scalp (electrodes PO7 and PO8). (A) ERPs recorded contralateral and ipsilateral to a target for high- and low-anxiety individuals. (B) Contralateral-minus-ipsilateral difference waveforms for high- and low-anxiety individuals. .................................... 88
Figure 5.6 ERPs for displays containing a midline target and a lateral distractor, separately for fast- and slow-response trials. (A) High-anxiety group ERPs recorded contralateral and ipsilateral to a distractor for fastest and slowest trials. (B) High anxiety group contralateral-minus-ipsilateral difference waveforms for high- and low-anxiety individuals............................................................................ 89
Figure 6.1. Adapted from Janatti et al., 2013, a proposed hypothetical processing stream thought to occur during the fixed-feature variant of the additional singleton search task. Listed below each stage is the ERP component associated with that particular level of processing. ....................................................................................... 100
xii
Figure 6.2. Hypothetical resolving of a visual search task based on the input image shown. The stars (top) represent the stimuli’s activation on the saliency map, with increased brightness denoting greater salience. The saliency map is then scanned by attention and suppression/enhancement are applied contingent on top-down attentional templates. ........................................................................... 101
xiii
List of Acronyms
ACT Attentional Control Theory
ADHD Attentional-deficit/hyperactivity disorder
ANOVA Analysis of variance
BOLD Blood oxygen level dependent
CIE Commission Internationale de l’Éclairage (International Commission on
Illumination)
EEG Electroencephalography
EOG Electrooculogram
ERP Event-related potential
ERPSS Event-related Potential Software System
Hz Hertz
K Estimate of vWM capacity
LAI Localized Attentional Interference
ms Millisecond
N2pc N2-posterior-contralateral component
PD Distractor positivity
PO Parieto-occipital
PPC Positivity, posterior contralateral
RSVP Rapid serial visual presentation
RT Reaction time
STAI State-trait anxiety inventory
T1 First target
T2 Second Target
xiv
µV Microvolts
vWM Visual working memory
1
General Introduction
Making sense of our visually complex world requires we be able to rapidly search
through the visual environment and appraise the information within it. From infinite
combinations of shapes, colours, and boundaries, the visual system works to ascribe
meaning to the objects that comprise our world. Yet, at any given moment, our brain lacks
the capacity to process all objects in a visual scene to a level of complete perceptual
awareness (Broadbent, 1957; Sperling, 1960; Treisman & Gelade, 1980; Tsotsos, 1990).
Rather than attempting to process all available information, the visual system has evolved
a strategy that allows for the parsing of a cluttered visual scene into computationally less
demanding chunks. The visual system acts to systematically shift resources to various
locations in space, preferentially processing pertinent information at the expense of the
rest. However, irrelevant objects can sometimes interfere with this process, leading to
distraction and inefficient behavioural performance. To operate efficiently, the visual
system must be able to contend with such potentially distracting information to accurately
select information essential to our present goals.
Attention is the general term used to describe the multitude of cognitive operations
that allow processing resources to be biased toward relevant information in the
environment. Understanding the role of attention as a selective mechanism dates at least
as far back as the late 19th century. In his famous treatise, The Principles of Psychology,
influential American philosopher William James (1890) describes attention as the
“withdrawal of some things in order to effectively deal with others” (p. 404), for to James,
“without selective interest, experience is an utter chaos” (p. 402). It was around this time
that psychological laboratories were beginning to crop up throughout the Western world;
however, attention would not become an essential topic for the fledgling science for
2
several years to come. The dominant influence of rigid behaviourism inspired little interest
in studying a mental process that has been poetically described to “lie at the crossroads
between perception and cognition” (Carrasco, 2011, p. 1485). However, toward the middle
of the twentieth century the fall of behaviourism would mark the rise of the cognitive
revolution and, with it, a resurged interest in the topic of attention. Important early
contributions from cognitive psychologists, including Donald Broadbent, Anne Triesman,
and Michael Posner would provide a theoretical foundation upon which the science of
selective attention would flourish. Since then, considerable progress has been made into
understanding how the selection and withdrawal of sensory information is incorporated
into conscious experience. Most notably in the last 40 years, with the advent of advanced
neurophysiological and neuroimaging technologies, scientists have made considerable
strides to understand exactly how it is we pay attention in our complex world. Although
attention interfaces with information arriving from multiple sensory modalities (e.g.,
audition, touch), the scope of this work will be limited to how selective attention alters early
visual processing and behaviour. The purpose of this chapter will be to review the key
concepts and theories of visual selection and present the experiments that will comprise
this thesis.
1.1. How objects are prioritized for selection in the visual
environment?
It is well understood that objects in the visual field are not selected at random but
rather based on their behavioural relevance and physical conspicuity. Many prominent
models of attention propose that for selection to occur, the visual system first
retinotopically encodes all objects in the visual field proportional to their sensory and
cognitive inputs (Bisley & Goldberg, 2010; Fecteau & Munoz, 2006; Itti & Koch, 2001;
Serences & Yantis, 2006; Thompson & Bichot, 2005). The object that elicits the greatest
retinotopic activation is selected and attentional processing is biased to the location of that
object in the visual field. By deploying attention to a location in space, sensory neurons
will alter their responses and enhance the encoding of attended information relative to all
other information in the visual field. Such attentional processing has been shown to have
an impact even at the earliest stages of visual processing, modulating neurons in the
3
lateral geniculate nucleus (O’Connor, Fukui, Pinsk & Kastner, 2002) and primary visual
cortex V1 (Kastner, Pinsk, De Weerd, Desimone & Ungerleider, 1999).
While attention can serve to amplify the sensory processing of a lone object
(Martínez-Trujillo & Treue, 2002; Reynolds, Pasternak & Desimone, 2000; Reynolds &
Desimone, 2003), its effects are greatest when multiple objects are presented
simultaneously at different locations in the visual field (Awh, Matsukura & Serences, 2003;
Awh & Pashler, 2000; Dosher & Lu, 2000; Kastner & Ungerleider, 2001; Moran &
Desimone, 1985; Reynolds, Chelazzi & Desimone, 1999). Specifically, when two objects
fall within the same visual receptive field (the area of visual space that a neuron codes
for), attention has been shown to have modulatory effects via the enhancement of relevant
objects and suppression of irrelevant ones (Chelazzi, Miller, Duncan & Desimone, 2001;
Miller, Gochin & Gross, 1993; Motter, 1993; Reynolds et al., 1999; Reynolds & Desimone,
1999). Findings such as these represent the basis for the biased competition theory, which
proposes that all objects in the visual field will contend for control a neuron's response
(Desimone & Duncan, 1995). Since neurons are constructed to optimally encode a single
object, when multiple objects fall within the same receptive field these objects will mutually
compete for representation.
With multiple objects competing for neural representation, it is often necessary to
selectively prioritize behaviourally relevant information and filter out other distracting
irrelevant information. Top-down attentional control describes the process of prioritizing
objects for selection in line with the goals, intentions, and expectations of an observer
(Bacon & Egeth, 1994; Folk & Remington, 1998; Krummenacher, Müller, Zehetleitner &
Geyer, 2009; Treisman & Gelade, 1980; Wolfe, Cave & Franzel, 1989). Here, feedback
from higher cortical areas—areas spanning superior frontal, inferior parietal and superior
temporal cortex—all act to modulate visual processing and selection is in line with the
behavioural goals an observer sets (Hopfinger, Buonocore & Mangun, 2000). Top-down
attentional control can be set based on knowing in advance where an object will appear
in space (spatial attention; e.g., Posner, 1980), the dimensional features of the object
(feature-based attention; e.g., Saenz, Buracas & Boynton, 2002; see also: dimensional
weighting; e.g., Found & Müller, 1996), or the specific object itself (object based attention;
e.g., O’Craven, Downing & Kanwisher, 1999).
4
Attentional priority can also be influenced by the bottom-up activations triggered
by the intrinsic properties of a stimulus. For example, objects possessing distinct colours
or unique motion paths will be more easily noticed than other objects in a visual scene
(Beck & Kastner, 2005). In such cases, the exogenous, sensory-driven aspects of a salient
object can be sufficient to grab attention automatically (e.g., Bruce & Tsotsos, 2009;
Theeuwes, 1991, 1992; 2010; Yantis & Johnson, 1990). These involuntary shifts of
attention have been shown to occur even when an object is known in advance to be
irrelevant and the observer is aware that attending to it will impair their behavioural
performance.
In addition to physical salience, several other cognitive factors can also
automatically prioritize the selection of an object. For example, the contents of working
memory (Olivers, Meijer & Theeuwes, 2006), the emotional valence of a stimulus
(Carretié, Mercado, Hinojosa, Martín-Loeches & Sotillo, 2004; Eimer & Kiss, 2007;
Pourtois, Grandjean, Sander & Vuilleumier, 2004), the anticipated reward contingency
associated with a stimulus (Anderson, Laurent & Yantis, 2011; Hickey, Chelazzi &
Theeuwes, 2010; Theeuwes & Belopolsky, 2012), or a previously shown stimulus (e.g.,
priming of pop-out, Maljkovic & Nakayama, 1994; inter-trial priming; Pinto, Olivers &
Theeuwes, 2005), can all lead to reflexive, automatic shifts of attention.
1.2. Top-down versus bottom-up processing: the debate
Top-down and bottom-up processes do not work in isolation but rather interact
dynamically to motivate attentional selection. Under certain conditions, however, conflict
can arise between the volitional goals of an observer and the strength of irrelevant sensory
inputs. In a typical cluttered visual scene where multiple stimuli are competing for
attention, can an observer bias processing toward a behaviourally relevant object or will
competing salient-but-irrelevant objects automatically capture attention? This precise
question has resulted in a longstanding controversy regarding the nature of attentional
biasing and the extent to which top-down processing can counteract the influence of
conspicuous competing stimuli.
5
1.2.1. Stimulus-driven capture
Decades of research has resulted in two predominant models of attention that fall
at opposing ends of the theoretical spectrum. At one end of the debate is the stimulus-
driven capture hypothesis (e.g., Theeuwes, 2010). According to this perspective,
attentional priority is driven entirely by the bottom-up features of a stimulus. This stimulus-
driven prioritization of processing occurs regardless of the intentions or expectations of
the observer and the most salient object in a visual scene will always automatically capture
attention. Top-down attentional settings can work to disengage and redeploy attention to
a behaviourally relevant stimulus but only subsequent to this involuntary initial shift.
Evidence for the stimulus-driven capture hypothesis comes primarily from visual
search studies using the additional singleton paradigm (Theeuwes, 1991, 1992, 1994). In
these tasks, participants search an array of non-targets for a target defined by a feature
unique to it alone (i.e., a singleton). For example, the non-targets might be green ellipses
and the target a green diamond. Typically, observers report the orientation of a line
segment contained within the target (e.g., horizontal or vertical). On a subset of trials, one
of the non-targets is also a singleton, and its salience can be greater than the target’s
(e.g., a red ellipse). The presence of this salient distractor results in longer response times
(RTs) to the target, and the magnitude of this behavioural interference cost varies with the
predictability of the stimuli. If the features of the target and distractor are randomly
swapped from trial to trial, the presence of a salient distractor singleton can delay RTs by
100-150 ms (Theeuwes, 1991). When the features of the target and distractor remain fixed
across trials, the presence of a salient distractor singleton delays RTs by 20-25 ms
(Theeuwes, 1992). This lattermost finding is notable because observers can configure
their attentional set perfectly for the target and the non-targets, and yet the behavioural
interference effect persists 1. Stimulus-driven capture accounts for this by holding that the
1 According to the stimulus-driven capture hypothesis, capture itself is associated with this 20-25
ms interference cost. The increased RT interference observed in the mixed-feature variant of the additional singleton cost is not associated with increased capture but rather with target uncertainty. Additional dwell time is necessary to determine whether the item selected is either a target or a distractor (Theeuwes, 2010).
6
initial deployment of attention is always made to the most salient item in view before being
redeployed to a target.
Opponents of this stimulus-driven capture hypothesis have proposed two
alternative explanations for the distractor interference observed in the additional singleton
task. The first explanation holds that the RT interference is not related to stimulus-driven
capture, but rather to the use of a singleton-detection strategy. By this account,
participants prime themselves to detect any unique singleton, as opposed to detecting a
singleton of a specific identity, even when they know the identity of the to-be-located
target. By altering the experimental parameters and by having participants search for a
target feature instead of a unique singleton, Bacon & Egeth (1994) showed that the
interference associated with a color singleton could be eliminated; RT inference was no
longer present even on trials where the displays were physically identical to those
presented in Theeuwes’ 1992 study. The authors concluded that the use of a singleton-
detection strategy leaves observers susceptible to interference from singletons defined by
task-irrelevant features. The second explanation attributes distractor interference to a non-
spatial filtering cost (Kahneman, Treisman & Burkell, 1983). By this account, search
displays with a target and a distractor singleton result in an increased amount of
competition between the two items. Additional filtering is necessary to suss out the target
singleton, which delays the deployment of attention to that item. Critically, no shift of
attention is made to the distractor. Evidence for this non-spatial filtering cost come from
experiments that show distractor singletons produce an additional RT cost in the absence
of any evidence for attentional capture (Folk & Remington, 2006; Folk & Remington, 1998;
Wykowska & Schubö, 2011).
1.2.2. Contingent-capture
At the opposite end of the debate is the contingent-capture hypothesis (e.g., Folk,
Remington & Johnston, 1992; Folk, Remington & Wright, 1994). This perspective
contends that volition serves to guide early visual processing so that selection is always
in line with the top-down attentional set adopted by the observer. According to contingent-
capture models of attentional selection, objects that are irrelevant to the observer’s
attentional set will not capture attention regardless of their salience.
7
Evidence for the contingent capture hypothesis has, in large part, come from a
modified spatial cueing paradigm developed by Folk and colleagues (1992, 1994). In their
original experiment, Folk and Remington (1992) had participants respond to a search
display containing either a single item (the target) or several items, with the target defined
in a unique color. A cue display was shown 150 ms prior to the target display, and the
color of the cue could either match or not match the color of the subsequent target. RTs
tended to be longer on invalid trials, when the location of the cue did not match the location
of the subsequent target, however the results varied with the features of the cue and
target. If the cue did not match the color of the target, it delayed search for an abruptly
flashed target but not for a unique-color target. Conversely, if the cue did match the color
of the target, it delayed search for a unique-color target but not for an abruptly flashed
target. Collectively. Folk et al. took this as evidence for the top-down prioritization for
selection: irrelevant stimuli that match an observer’s attentional set will capture attention
reflexively, whereas stimuli that do not match an attentional set can be ignored.
Opponents of the contingent capture hypothesis have argued that in the modified
spatial cueing paradigm, irrelevant cues do still capture attention but in a manner that is
much more short-lived and easily withdrawn. Theeuwes (2010) has argued that dissimilar
distractors produce a relatively short attentional dwell time; this dwell time for dissimilar
distractors have been estimated to last possibly as little as 100 ms (see: Theeuwes,
Atchley & Kramer, 2000). Thus, in the spatial cueing paradigm used by Folk and others,
the 150 ms delay between the presentation of the cue and the search display might have
been sufficient for an observer to disengage attention from the dissimilar cue prior to the
appearance of the target singleton. This rapid disengagement could serve to explain the
absence of an RT cost when the cue and target have the dissimilar defining characteristics
(e.g., a colour cue and an abrupt onset target).
1.3. An alternative to the dichotomy: signal suppression
To address this growing body of conflicting experimental results, researchers have
begun proposing more integrative, hybrid models of attentional selection. These models
incorporate aspects of both bottom-up and top-down selection biases, as well as highlight
8
other aspects of selection biases, such as dimensional weighting (e.g., Found & Müller,
1996) and selection history (e.g., Awh, Belopolsky & Theeuwes, 2012).
One model of attentional selection with growing empirical support is the signal-
suppression hypothesis (Sawaki & Luck, 2010). According to the signal-suppression
hypothesis, a suppression mechanism can act to mitigate the capture of salient-but-
irrelevant information in the visual field. This model utilizes the central tenets of both
salience-driven capture and contingent capture models to parsimoniously explain the
empirical findings of both. In line with salience-driven theories of selection, salient objects
will still automatically generate bottom-up stimulus driven signals. In line with contingent
capture theories of selection, these signals can be mitigated in concert with an observer’s
goals; salient-but-irrelevant objects can be actively suppressed in order to prevent them
from capturing attention. Over the past six years, mounting evidence has begun to
accumulate in favour of this hypothesis. This evidence has stemmed primarily from event-
related potential (ERP) studies focusing on electrophysiological markers of selective
attention.
1.3.1. The ERP technique and components associated with
attention
The ERP technique has proven to be an ideal method for testing the signal-
suppression hypothesis for a number of reasons. The first reason relates to the very nature
of how we shift attention: attentional shifts are quick and transient and can happen as
frequently as six per second (Woodman, 2010; Woodman & Luck, 2003). The temporal
resolution of the ERP technique offers the millisecond by millisecond precision necessary
for indexing these rapid changes in attentional processing (Luck, 2014). Second, ERPs
can be elicited in the absence of any overt behavioural response to a stimulus. This is
important because it allows inferences to be made about neural processing associated
with irrelevant, to-be-ignored objects in the visual field. Lastly, two attentional ERP
components—one associated with attentional capture and one associated with attentional
suppression—can be isolated and examined independently of one another. ERP research
testing the signal suppression hypothesis has primarily relied on these two well
documented electrophysiological components.
9
The first and more extensively studied of these components is the N2-posterior-
contralateral (N2pc) component. The N2pc is an enhanced negative potential observed
contralateral to attended objects in the visual field (Luck & Hillyard, 1994a, 1994b). It is
reported as the difference between ipsilateral versus contralateral voltages at posterior
scalp electrodes and has been shown to occur 150-300 ms after the presentation of a
stimulus array. The N2pc appears to be generated in the ventral visual stream and
presents maximally in area V4 and the lateral occipital complex (Hopf et al., 2000; Hopf,
Boelmans, Schoenfeld, Luck & Heinze, 2004). Several studies have demonstrated the
N2pc to be a reliable index of covert attentional selection (Hickey, McDonald & Theeuwes,
2006; Kiss, Jolicoeur, Dell’acqua & Eimer, 2008; Leblanc, Prime & Jolicoeur, 2008;
Woodman & Luck, 2003; Woodman & Luck, 1999) that reflects the enhancement of
processing at an attended location (Eimer, 1996; Hickey, Di Lollo & McDonald, 2009;
Mazza, Turatto & Caramazza, 2009; although see Tan & Wyble, 2015 for an alternative
explanation).
The other electrophysiological component is the distractor positivity (PD). Converse
to the N2pc, the PD is an enhanced positive potential observed contralateral, not to a
target, but rather to a task-irrelevant distractor singleton. It has been observed to occur in
the same time interval as the N2pc and has been shown to have a similar scalp
topography, with maximal differences in voltage observed at lateral parieto-occipital
electrode sites. The PD can be isolated by placing a target singleton on the vertical
meridian while simultaneously presenting a lateralized distractor singleton. The
presentation of a target on the midline nullifies any lateralized differences associated with
the target (Hickey et al., 2009; Woodman & Luck, 2003); therefore, any differences
between voltages at ipsilateral and contralateral electrode sites can be attributed entirely
to distractor processing. Several studies have now demonstrated the PD to be an index of
distractor suppression that can be observed in instances where the visual system must
resolve attentional competition between a target and a potentially distracting object
(Gaspar & McDonald, 2014; Jannati, Gaspar & McDonald, 2013; Kiss, Grubert, Petersen
& Eimer, 2012; McDonald, Green, Jannati & Di Lollo, 2013; Sawaki & Luck, 2010;
Wykowska & Schubö, 2010, 2011).
10
The existence of an electrophysiological index of distractor suppression was first
reported by Hickey, McDonald, and Di Lollo in 2009. In their study, participants viewed a
visual search array that contained either a green square or diamond (that was brighter
than the background) presented alongside either a short or long red line (that was
isoluminant to the background). Task instructions varied across blocks: on some trials,
participants were told to report the green shape (square versus diamond), whereas on
other trials, they were told to report the length of the red line (short versus long). When
the lateralized red line was the target and the midline green square was the distractor, the
red line elicited an N2pc. Critically, when the lateralized red line was the distractor and the
midline green square was the target, the red line elicited a contralateral positivity. This
contralateral positivity—which they sensibly named the distractor positivity—was
interpreted to reflect direct suppression of the cortical representation of the lateralized red
line. Hickey et al. observed a PD to the red line only when participants were required to
discriminate the feature of the target—it did not appear when observers merely had to
detect the presence of the target. This absence of the PD on detection trials is consistent
with the idea that the component reflects attentional processing that can be distinguished
from mere sensory processing.
1.3.2. Evidence for the signal suppression hypothesis
The discovery of PD component was paramount for the development of the signal
suppression hypothesis. Sawaki and Luck (2010) immediately used the newly discovered
component to investigate whether or not salient distractor singletons could be actively
suppressed. In a series of experiments, participants were instructed to search an array of
letter stimuli (four above and four below fixation) for a target letter (defined by its size and
identity, e.g., a large A). On a subset of the trials, a salient distractor singleton (a letter of
a different colour) replaced one of the non-target letters within the array. They anticipated
two potential outcomes: if the salient distractor singleton elicited an N2pc, this would be
taken as evidence that it captured attention; however, if the salient distractor singleton
elicited a PD, this would indicate that it was actively suppressed.
11
Sawaki and Luck found target letters to elicit an N2pc component; however, salient
distractors did not. Instead, in line with an active suppression account, salient distractor
singletons were found to elicited the PD component. From these findings, they concluded
that the salient-but-irrelevant singletons likely always generate signals associated with
attentional capture irrespective of an observer’s attentional control—what they colloquially
referred to as an attend-to-me signal. These signals could, however, be overridden by a
suppressive mechanism working in accordance with the top-down attentional set of an
observer. This suppressive mechanism would serve to mitigate the distractor singleton
from involuntarily capturing attention in instances where it would be competing with the
target for attentional selection.
Several studies have since used the PD to provide compelling evidence in favour
of the signal suppression hypothesis. These studies have shown that to-be-ignored
distractors can be actively suppressed under conditions where they compete for selection
with a target singleton. To date, such active suppression has been demonstrated for
distractors that are both physically salient (Burra & Kerzel, 2014; Gaspar & McDonald,
2014; Jannati et al., 2013; Kiss et al., 2012; McDonald et al., 2013) or that match the
contents of visual working memory (Sawaki & Luck, 2011).
1.3.3. The additional singleton task and signal suppression
To date, many of the ERP studies that have tested the signal suppression
hypothesis have employed additional singleton paradigms. In these paradigms,
behavioural interference associated with the presence of a salient distractor has
traditionally been interpreted as evidence for salience-driven capture (see: Section 1.2.1);
however, this interpretation is inconsistent with the findings of several of these ERP
studies. In fixed-feature variants of the additional singleton task—where the target and
distractor features remain consistent from trial to trial—distractor singletons have
repeatedly been shown to be actively suppressed (Burra & Kerzel, 2014; Feldmann-
Wüstefeld & Schubö, 2013; Gaspar & McDonald, 2014; Jannati et al., 2013). For example,
Jannati and colleagues (2013) had participants complete an additional singleton task
where the target singleton was defined by a fixed shape and the distractor singleton by a
fixed color. Distractor-present trials produced a behavioural cost relative to distractor-
12
absent trials, replicating the standard additional singleton behavioural effect (Theeuwes,
1991; 1992). The ERP analysis, however, showed that distractor singletons did not elicit
an N2pc, but rather a PD. In addition to this, the timing of target N2pc was unaffected by
the presence of the salient distractor. Based on these findings, Jannati and colleagues
concluded that distractor singletons in the additional singleton paradigm do not capture
attention, but can produce an increased competition. This competition can ultimately be
resolved by suppressing the location of the distractor.
In mixed-feature variants of the additional singleton task—where target and
distractor features are swapped randomly from trial to trial—evidence for distractor
suppression is not quite as clear cut. One of the first ERP studies to utilize a mixed-feature
additional singleton task reported that salient distractors captured attention (Hickey et al.,
2006); however, the likelihood of this capture has been called into question. In 2013,
McDonald and colleagues re-evaluated the findings of the original Hickey et al. (2006)
study, using the original data and collecting data from an additional 26 subjects. In their
follow-up analysis, they subdivided search trials into fast-response and slow-response
subsets, depending on whether a participant responded faster or slower than the median
RT. When ERPs were separately evaluated for fast and slow trials, capture by the
distractor singleton was only in evidence on the slowest subset of trials; on fast-response
trials, the distractor elicited only a PD.
Temporal factors, such as stimulus duration, also seems to play a critical role in
determining if distractors will be effectively suppressed in mixed-feature additional
singleton tasks. For example, a study by Kiss and colleagues (2012) varied the duration
for which the search display remained on the screen. When the search array was visible
until response, distractor singletons elicited an N2pc component that was followed by a
late PD component. This suggested to the authors that the distractor initially captured
attention but that it was subsequently suppressed. When search displays were brief
(presented for 200 ms), an N2pc to the distractor singleton was no longer observed;
instead they elicited only an earlier PD component in the N2pc time range, indicating the
rapid suppression of the distractor. Based on these findings, the authors concluded that
distractor singletons could be actively suppressed in instances where the demands of the
task necessitated more temporally efficient target processing.
13
1.3.4. Other evidence for signal suppression
Although evidence for the signal suppression hypothesis comes primarily from
ERP studies, the notion that salient distractors can be suppressed is not new. Evidence
that salient distractors are actively suppressed has been found using a wide range of
experimental approaches. Behavioural studies, for example, have shown participants to
be less accurate to report a probe presented at a location where a salient distractor
singleton was suppressed, relative to any other irrelevant location in a search array
(Gaspelin, Leonard & Luck, 2015). Similar spatial distractor suppression has been
reported even when the distractor is no longer on the screen. For example, Cepeda and
colleagues (1998) reported slower probe response RTs at locations where a distractor
had previously been presented. Using fMRI, Seidl, Peelen, and Kastner (2012) showed a
significant reduction in blood oxygen level dependent (BOLD) responses for stimuli
belonging to distractor categories relative to target and neutral categories, indicating that
the processing of distractor objects was suppressed. Such suppressive modulation of
BOLD signals can be observed even prior to the onset of the visual information. For
example, Serences and colleagues (2004) showed that anticipation of distractors on an
upcoming trial resulted in a suppression of baseline signals in visual cortex. In macaque
monkeys, salient distractors elicit reduced neural activity in the lateral intraparietal area—
a candidate location for where the attentional priority map is thought to reside—when a
distractor is ignored and attention is correctly deployed to a target (Ipata, Gee, Gottlieb,
Bisley & Goldberg, 2006). These, and a myriad of other findings, have led to top-down
distractor suppression mechanisms becoming a critical feature in a number of
computational models of attention (Navalpakkam & Itti, 2006; Reynolds & Heeger, 2009;
Usher & Niebur, 1996).
1.4. Extending our knowledge of signal suppression
A growing number of studies have used the PD component to demonstrate that the
sensory processing of irrelevant information can be strongly modulated by top-down
attentional control. The primary aim of this thesis was to further our understanding of the
PD component, as well as to yield further insight into how the visual system deals with
14
irrelevant information during visual search. The following subsections outline the
motivation for the four electroencephalography studies that comprise this thesis. Recurring
goals that will be readdressed throughout the thesis are, i) to better understand what the
processing indexed by the PD reflects, ii) to determine the stimulus conditions necessary
to elicit distractor suppression, and iii) to explore what factors might affect an individual's
ability to suppress.
1.4.1. Chapter 2: Eliciting signal suppression during visual search
Chapter 2 explores the stimulus conditions necessary to elicit the distractor
suppression indexed by the PD. While several studies have utilized the PD component to
provide compelling evidence that to-be-ignored distractor singletons can be actively
suppressed, the perceptual conditions necessary to elicit the component remain an issue
of debate. Some have argued that when distraction is highly likely, the PD could be
strategically recruited to suppress highly salient distractors that would otherwise be
prioritized for attentional selection (Gaspar & McDonald, 2014; Sawaki & Luck, 2010).
Alternatively, the PD could instead index a mechanism that is generally recruited to deal
with all unique distracting objects, regardless of salience; other studies have reported
instances where physically inconspicuous distractor singletons have also been shown to
elicit the PD (Hickey et al., 2009; Hilimire, Hickey & Corballis, 2012). In this chapter, I
examine how differences in distractor salience relate to signal suppression and the impact
these differences have on selective attention.
To investigate this, EEG was recorded from 40 participants while they performed
a unidimensional variant of the additional singleton search task (Gaspar & McDonald,
2014; Theeuwes, 1992). Participants were instructed to search a multi-item array for a
target singleton while attempting to ignore a task-irrelevant distractor singleton that could
simultaneously appear within the display. On 33% of trials, a lone yellow target singleton
was presented alongside nine uniformly coloured non-targets. On the remainder of trials,
one non-target item was replaced with either a red or a blue distractor singleton.
Participants were instructed to report the orientation of the line inside the yellow target
singleton while ignoring the presence of any distractor singleton. The color of the non-
targets was varied across two experimental conditions (green in one condition and orange
15
in the other) in order to reverse the salience of the red and blue distractor singletons.
Consequently, the red distractor was the highly salient when presented against green non-
targets, whereas it was no more salient than the target when presented against orange
non-targets. The reverse was true for distractors presented against orange non-targets.
In this chapter, I provide evidence that distractor suppression is implemented only when
a target competes with a more salient distractor. These findings will be discussed further
in the context of current models of attentional selection.
1.4.2. Chapter 3: Individual differences in working memory and
visual search
Chapter 3 explores how individual differences in target and distractor processing
contribute to variations in visual working memory (vWM) capacity. Studies have shown
that, when required to process multiple visual objects, low-capacity individuals have
difficulty filtering relevant from irrelevant information (McNab & Klingberg, 2008;
Shipstead, Zach, Lindsey, Marshall & Engle, 2014; Unsworth, Nash & Robison, 2014;
Vogel, McCollough & Machizawa, 2005). Together these findings suggest that an
individual’s working memory capacity might be determined by the degree to which relevant
information is remembered and irrelevant information is ignored; however, the neural basis
for such a filtering mechanism remains unknown. Theoretically, the inefficient filtering
observed in low-capacity individuals could be specifically linked to problems enhancing
target representations, ignoring distractor representations, or a combination of both. In this
chapter, I address these possibilities by examining how individual differences in vWM
relate to both target and distractor processing during visual search.
To investigate this, vWM capacities were derived for 48 normal young adults using
a delayed visual change detection task (Identical to that used by Luck & Vogel, 1997).
EEG was then recorded while subjects performed a competitive visual search task.
Identical to the visual search task used in Chapter 2, participants searched an array for a
pre-specified target singleton while attempting to ignore a simultaneously presented high-
or low-salience distractor singleton. ERP components associated with target selection
(N2pc) and distractor suppression (PD) were then separately assessed for individuals with
a high, medium, and low working memory capacity. This chapter will present evidence
16
indicating that differences in working memory capacity are specifically related to distractor-
suppression (but not target processing) activity in visual cortex. Specifically, high-capacity
individuals are able to actively suppress distractors, whereas low-capacity individuals
cannot suppress in time to prevent salient distractors from capturing attention.
1.4.3. Chapter 4: Signal suppression during a transient loss of
attentional control
The distractor suppression indexed by the PD is thought to reflect an active process
contingent on an observer’s top-down attentional set (Hickey et al., 2009; Hilimire, Hickey
& Corballis, 2012). However, others have proposed that the PD may instead characterize
a bottom-up process with stimulus characteristics, such as certain colours, given special
status in the context of visual search (Fortier-Gauthier, Dell'Acqua & Jolicœur, 2013).
Chapter 4 tests this assertion and asks whether the PD can be elicited during a disruption
of top-down attentional control. One manner of producing such a disruption would be to
exploit a behavioural phenomenon known as the attentional blink (AB). The attentional
blink, characterized as an impairment in identifying the second of two targets when they
are presented in close temporal succession, has been shown to affect attentional selection
by producing a transient loss of endogenous control over the visual system (Di Lollo,
Kawahara, Shahab Ghorashi & Enns, 2005). In this chapter, I examine how target
selection and distractor suppression processing operate within the window of the
attentional blink during visual search.
To investigate how visual search is affected during a disruption to the availability
of attentional control, EEG was recorded from 18 participants while they performed a
modified rapid serial visual presentation (RSVP)/visual search task. The first target (T1)
was a number within an RSVP stream of letters. The second target (T2) was a colour
singleton that appeared within a visual search array that also contained a salient distractor
singleton. Subjects were instructed to first make a speeded response to T2 (by identifying
the orientation of a line inside the target singleton) and were then subsequently probed to
respond to T1 (by identifying whether the number in the RSVP stream had been even or
odd). ERPs elicited by the T2 search array at lag 2 (within the attentional blink) and at lag
8 (outside the attentional blink) were then separately examined. In this chapter I will show
17
that during the attentional blink i) the processing of T2 is put on hold until after processing
of T1 is complete and ii) distractor suppression is not possible. These findings will be
presented in line with contemporary models of distractor suppression and the attentional
blink.
1.4.4. Chapter 5: Individuals with high levels of trait anxiety show differences in selective attentional processing
High levels of anxiety have been associated with deficits in attentional control. The
attentional control theory of anxiety (ACT; Derakshan et al., 2009; Eysenck, Derakshan,
Santos & Calvo, 2007) posits that these deficits could potentially relate to an inability of
high-anxiety individuals to suppress salient-but-irrelevant information. Behaviourally, this
has been shown in anti-saccade tasks where high-anxiety individuals are slower to initiate
eye-movements away from counter-predictive abrupt-onset cues presented in the
periphery (Derakshan et al., 2009). This observed increase in anti-saccade latency is
thought to reflect an inability to inhibit the automatic, reflexive shift of attention to the salient
counter-predictive cue (Eysenck & Derakshan, 2011). While trait anxiety appears to
disrupt the ability to ignore distracting information, the neural correlates of this effect are
not well understood. Differences in attentional biases between high- and low-anxiety
individuals may be due to an impaired ability to shift attention toward relevant stimuli or a
reduced ability to apply active suppression to irrelevant stimuli. Chapter 4 will examine
how high- versus low-trait anxiety relate to both target and distractor processing.
To investigate this, 219 young adults were initially screened using the State-Trait
Anxiety Inventory (STAI; Spielberger, Gorsuch, Lushene, Vagg & Jacobs, 1983), a 40-
item self-evaluation questionnaire pertaining to anxiety affect. To maximize power to
detect potential differences in brain responses, an extreme-groups design was used
(Yarkoni, Tal & Braver, 2010). Individuals whose trait anxiety scores were among the
highest and lowest were selected to participate in the ERP experiment (n = 36; 18 per
group). To measure the neural correlates of attention, EEG was recorded while subjects
performed an additional singleton search task identical to that previously used by Gaspar
and McDonald (2014). In this task participants searched for a color-singleton target and
on 50% of trials attempted to ignore a more salient color-singleton distractor. Participants
18
were instructed to indicate whether the orientation of the line inside the target was either
horizontal or vertical. The relationship between anxiety and i) target processing ERPs and
ii) distractor suppression processing ERPs were then separately assessed. This chapter
will relate these specific aspects of visual-search processing to the attentional control
theory of anxiety and show that distractor suppression is intact in high-anxiety individuals;
however, it is deployed reactively after the distractor is initially attended.
19
Eliciting signal suppression during visual search
2.1. Introduction
The natural visual environment typically contains an abundance of irrelevant and
distracting stimuli. These visual distractors can attract attention to their locations even
when an observer is attempting to search for an entirely different object (Burra & Kerzel,
2013; Hickey, Di Lollo & McDonald, 2009; Hickey, McDonald & Theeuwes, 2006; Leblanc,
Prime & Jolicoeur, 2008; Theeuwes, 1991, 1992). Moreover, this potential for distraction
is increased when irrelevant items are more physically salient than another competing
item of interest (Theeuwes, 1992; Theeuwes, 2010). Mechanisms for mitigating such
salience-driven distraction have been proposed (Bacon & Egeth, 1994; Belopolsky,
Zwaan, Theeuwes & Kramer, 2007; Folk et al., 1992; Gaspar & McDonald, 2014; Ipata et
al., 2006; Leber, 2010; Sawaki & Luck, 2010; Töllner, Zehetleitner, Gramann & Müller,
2010); however, these mechanisms remain contentious and poorly understood.
Although the means by which observers mitigate distraction remains open for
debate, accumulating evidence indicates that a suppression mechanism can be
implemented to filter out irrelevant stimuli (Awh et al., 2003; Caputo & Guerra, 1998;
Sawaki & Luck, 2010; Serences et al., 2004). A number of studies in humans have shown
this suppressive mechanism to facilitate the selection of relevant information by preventing
irrelevant objects from erroneously capturing attention (Gaspar & McDonald, 2014;
Gaspelin et al., 2015; Hickey et al., 2009). In monkeys, evidence for such a mechanism
can be found at a neural level: single-unit neurophysiological research has shown that
neurons in extrastriate cortex suppress their responses to unattended stimuli when they
fall within the same receptive field of an attended stimulus (Chelazzi et al., 1993;
Desimone & Duncan, 1995; Moran & Desimone, 1985; Reynolds et al.,1999). Such
modulated neural responses to irrelevant information have been shown to enable
20
monkeys to ignore distractor stimuli, even when they are more physically conspicuous
than a target (Ipata et al., 2006).
Recently, studies using event-related potential (ERP) recordings have isolated an
attentional selection component directly associated with the suppression of a distractor
singleton. These studies have generally found that laterally presented stimuli tend to elicit
contralateral negative or positive potentials depending on whether the item is attended or
ignored, respectively (Gaspar & McDonald, 2014; Hickey et al., 2009). Whereas the
negativity (called N2pc) is hypothesized to reflect attentional selection (Luck & Hillyard,
1994a, 1994b), the positivity (called distractor positivity, PD) is hypothesized to reflect
suppression that prevents the eliciting stimulus from inadvertently capturing attention
(Gaspar & McDonald, 2014; Hickey et al., 2009; Sawaki & Luck, 2010). Several sources
of evidence indicate that the PD component reflects a marker of the distractor suppression
recruited to aid target resolution. First, the PD component is only present for a distractor
stimulus that appears in competition with a target stimulus (Hilimire et al., 2012).
Additionally, this target stimulus must be perceptually resolved; the PD component is
absent for a distractor stimulus when a competing target need only be detected and not
discriminated (Hickey et al., 2009). Furthermore, the PD component has been shown to be
predictive of behavioural performance and the likelihood of distraction: on trials where
participants are fastest to identify a target stimulus, the PD is larger in amplitude and occurs
without a preceding shift of attention to the distractor (Gaspar & McDonald, 2014; Jannati
et al., 2013; Sawaki, Geng & Luck, 2012).
To date, several studies have utilized the PD component to provide compelling
evidence that to-be-ignored distractor singletons can be actively suppressed when
competing for selection with a target singleton. However, there remains some ambiguity
concerning the circumstances under which such a suppression mechanism would
manifest. Specifically, one outstanding question regarding the PD concerns whether or not
directing attention to a target singleton automatically entails the suppression of any unique
distractor singleton or exclusively those that are particularly salient. Some recent findings
support the notion that the visual system exclusively suppresses highly salient items that
would otherwise be prioritized for selection due to their dominant activations on an internal
saliency map (Gaspar & McDonald, 2014; Sawaki & Luck, 2010). However, physically
21
inconspicuous distractors have been found to elicit the PD in other studies (Hickey et al.,
2009; Hilimire et al., 2012), thereby obscuring the conditions under which active
suppression is required to resolve competition for attention.
To address this question, EEGs were recorded while participants performed a uni-
dimensional variant of the well-known additional singleton search paradigm (Theeuwes,
1992) that pitted a target against one of two potential distractors. Participants were
instructed to search a circular multi-item display for a pre-specified color singleton while
attempting to ignore a task-irrelevant color singleton that could potentially appear in the
same display. Distractor absent trials consisted of a yellow target singleton presented
alongside nine uniformly coloured non-targets. On distractor present trials, one non-target
item was replaced with one red or blue distractor singleton (Trials types can be observed
in Figure 2.1). The color of the non-targets was varied across experimental conditions (all
green or all orange) to disentangle distractor salience from distractor color. Specifically, in
Condition one the red distractor was the most salient singleton against green non-targets,
whereas in Condition two the blue distractor was the most salient singleton against orange
non-targets (which was confirmed in a behavioral pilot experiment). Target- and distractor-
related ERPs were then measured separately for low- and high-salience distractor trials
to determine whether or not attentional suppression occurred.
22
Figure 2.1. Trial Types. Example stimulus displays from the two experimental conditions. Subjects were instructed to attend to the yellow circle and to identify the orientation of the line inside of it. On 66% of trials, a distractor singleton was simultaneously presented within the display.
2.2. Methods
2.2.1. Materials and Methods
The Research Ethics Board at Simon Fraser University approved the research
protocol used in this study.
23
2.2.2. Participants
Fifty-five subjects from Simon Fraser University participated after giving informed
consent. Students received course credit for their participation as part of a departmental
research participation program. A grand total of 48 participants, 24 in Condition 1 (20
women, age 20.8 ± 2.2 y; 0 left-handed), and 24 in Condition 2 (13 women, age 20.0 ± 2.9
y; 1 left-handed) participated in the study. All subjects reported normal or corrected-to-
normal visual acuity and were tested for typical colour vision using Ishihara colour test
plates.
2.2.3. Behavioural Pilot Stimuli and Apparatus
To ensure only one distractor (per condition) was considerably more salient than
the target, an initial singleton-detection task was performed. The three singleton colors
were selected so that the salience of the target would be approximately equal to that of
one distractor and considerably less than that of the other distractor. Salience was
considered in two ways: as the local contrast between the (uniform) non-targets and each
color singleton, and as the rapidity with which each singleton could be detected. Local
contrast was measured as the distance in CIE [Commission Internationale de l’Éclairage
(International Commission on Illumination)] chromaticity space between a singleton and
the surrounding non-targets (herein called color distance). In each condition, the color
distance was considerably larger for one distractor (e.g., red distractor vs. green non-
targets) than for the target (e.g., yellow target vs. green non-targets). The method used to
measure the rapidity of item selection is described in the next section.
2.2.4. Behavioural Pilot Procedure
After candidate colors for all five items were selected, a behavioural pilot
experiment was conducted wherein participants (n = 8) were required to detect any
singleton (yellow, red, or blue) appearing with either the green or orange non-targets.
Stimuli were presented on a 23 inch, 120 Hz LCD monitor viewed from a distance of 57
cm. The six combinations of colors were presented in separate blocks, and on each trial,
24
there was an equal probability the target would be present or absent. Participants were
instructed to press a button when they detected the singleton. The mean RTs for the pilot
were analyzed in a repeated-measures ANOVA with factors for non-target color (green,
orange) and singleton color (yellow, blue, red). With green non-targets, RTs were shorter
for the red singleton (363 ms) than for the blue or yellow singleton (387 ms, 405 ms). With
orange non-targets, RTs were shorter for the blue singleton (364 ms) than for the red or
yellow singleton (411 ms, 402 ms). This pattern of results led to a significant interaction
between the non-target color and the singleton color [F(2,28) = 12.4; P < 0.001]. Planned
Bonferroni-corrected t tests confirmed the following: on green non-target trials, RTs to the
yellow singleton were longer than RTs to the red singleton [t(7) = 4.9; P = 0.003] but not
the blue singleton [t(7) = 2.1; P = 0.27]; on orange non-target trials, RTs to the yellow
singleton were significantly longer than RTs to the blue singleton [t(7) = 4.4; P = 0.008]
but not the red singleton [t(7) = 1.0; P = 0.37].
2.2.5. Visual Search Task Stimuli and Apparatus
Stimuli were presented on the same 23 inch, 120 Hz LCD monitor used in the
previous section. Viewing distance was 57 cm. Visual search arrays comprised 10 unfilled
circles presented equidistant (9.2°) from a central fixation point. Each circle was 3.4° in
diameter with a 0.3° thick outline. Eight or nine of the circles were uniformly coloured non-
targets, one was a target colour singleton, and one was a distractor color singleton (on
distractor-present trials). The target was dark yellow (x = 0.416, y = 0.519, 7.9 cd/m2), and
the distractor was either red (x = 0.640, y = 0.324, 7.0 cd/m2) or blue (x = 0.179, y = 0.199,
7.9 cd/m2). A randomly oriented vertical or horizontal gray line (x = 0.295, y = 0.361, 7.9
cd/m2) was contained within each of the circles. In Condition 1, the non-target circles were
green (x = 0.288, y = 0.636, 7.9 cd/m2), and in Condition 2, they were orange (x = 0.563,
y = 0.402, 7.9 cd/m2). All stimuli were presented on a uniform black background (0.5
cd/m2).
25
2.2.6. Visual Search Task Procedure
On each trial, a search display was presented after an 800-1,200 ms fixation
period, during which only the central fixation point was visible. Participants were instructed
to maintain fixation on the central point and to identify the orientation of the gray line inside
the target singleton by pressing one of two response buttons as quickly as possible. The
search array remained visible for 100 ms after a response was registered, at which point
the next trial began.
Displays contained a target singleton and one distractor singleton on 66% of trials
(distractor-present trials). On half of these trials, the distractor singleton was more salient
than the target (high-salience distractor trials), and on the other half of these trials, it was
no more salient than the target (low-salience distractor trials). On the remaining 33% of
trials, the target was the only singleton in the array (distractor-absent trials).
Target and distractor locations were varied to produce the following display
configurations: lateral target, no distractor (22.0%); midline target, no distractor (11.3%);
lateral target, midline distractor (14.7%); lateral target, ipsilateral distractor (14.7%); lateral
target, contralateral distractor (14.7%); midline target, lateral distractor (14.7%); midline
target, midline distractor (8.0%). There were equal numbers of both high- and low-salience
distractors that appeared within each block of trials. These display configurations were
randomly intermixed across trials. Each experimental block comprised 108 trials, with a 5-
second break after every 36 trials. At the end of the block, participants were given a
minimum 5-second rest period and could begin the subsequent block when ready. The
experiment contained 12 blocks, for a total of 1,296 trials per participant. At least 36
practice trials were given to each participant before commencing the experiment.
2.2.7. Behavioural Analysis
First, to ensure that distractor salience (and not distractor colour) modulated
search performance, RTs from distractor-present trials were analysed as a function of
distractor colour (red, blue) and non-target colour (orange, green). Upon confirmation,
high- and low-salience trials were combined and distractor interference was assessed by
26
comparing grand-averaged median reaction times (RTs) for each participant for high-
salience distractor, low-salience distractor, and distractor-absent trials.
Next, distractor proximity effects were assessed for both high- and low-salience
distractor present trials. This was accomplished by re-averaging distractor present trial
RTs to lateral-target displays according to the number of positions between target and
distractor singletons. This was done separately for high- and low-salience distractors. The
distance between the target and distractor singleton ranged from one (d1; adjacent) to five
(d5; four intervening non-targets) positions. The effect of the distractor proximity on search
performance was then analyzed in a 2 x 5 analysis of variance (ANOVA) with factors of
salience (high/low) and proximity (d1; d2; d3; d4; d5) tested. Lastly, RTs obtained with
target-distractor distance 5 (d5) was compared with the RT obtained on distractor-absent
trials for both high- and low-salience distractor trials. This analysis served to determine
whether the most distant distractors captured attention and produced a decrement in
behavioural search performance.
2.2.8. Electrophysiological Recording and Analysis
EEG and electrooculogram (EOG) were recorded from active sintered Ag-AgCl
electrodes (Biosemi Active Two system) from 125 standard sites and three nonstandard
subinion sites. Horizontal EOGs were recorded using two electrodes positioned 1 cm
lateral to the external canthi, and vertical EOGs were recorded using two electrodes
positioned above and below the right eye. All EEG and EOG signals were digitized at 512
Hz, referenced in real time to an active common-mode electrode, and low-pass filtered
using a fifth-order sinc filter with a -3 dB cutoff at 104 Hz. Electrode offsets were monitored
to ensure the quality of the data. After the data acquisition, EEG data for each channel
were high-pass filtered (-3 dB point at 0.05 Hz) and then converted from 24 bit to 12 bit
integers.
EEG processing and ERP averaging were performed using event-related potential
software system (ERPSS; University of California, San Diego). A semi-automated
procedure was used to discard epochs of EEG contaminated by blinks, eye movements,
or excessive noise (Green, Conder & McDonald, 2008). Any trial with an artifact within a
27
1-s interval (-200-800 ms post-stimulus) was rejected. Artifact-free epochs associated with
the various display configurations of interest were then averaged separately to create ERP
waveforms. The resulting ERPs were digitally low-pass filtered (-3 dB point at 32 Hz) and
digitally re-referenced to the average of the left and right mastoids. All ERP amplitudes
and baselines were computed using a 200 ms pre-stimulus window. The averaged event-
related horizontal EOGs did not exceed 2 μV for any individual participant, indicating their
gaze remained within 0.3° of the fixation point for most the trials (McDonald & Ward, 1999).
The primary analysis focused on ERPs elicited by the following display
configurations: (i) lateral target, no distractor; (ii) lateral target, midline high-salience
distractor; (iii) lateral target, midline low-salience distractor; (iv) midline target, lateral high-
salience distractor; (v) midline target, lateral low-salience distractor. ERPs to these search
displays were created by collapsing across left and right visual hemifields and left and
right electrodes (P07 and P08) to produce waveforms recorded ipsilateral and
contralateral to distractor stimuli. Negative voltages were plotted upward, so that the N2pc
and PD would appear in these difference waveforms as upward and downward deflections,
respectively. Lateralized ERP difference waveforms were then derived by subtracting the
ipsilateral waveform from the corresponding contralateral waveform. All ERP statistics
were computed using contralateral minus ipsilateral difference values.
PD and N2pc components were measured using a conventional mean-amplitude
approach. By convention, electrodes and time windows were selected a priori based on
existing studies that measured mean amplitudes (thereby avoiding problems of multiple
implicit comparisons; Luck, 2014). Both components were measured at lateral occipital
electrodes PO7/PO8, as in most previous papers. The N2pc window and the PD window
was measured 250-290 ms post stimulus onset and were chosen to replicate the
methodology used in Gaspar and McDonald’s (2014) study. Latency statistics were
computed 200-350 ms post stimulus onset using jackknifed averages. Latency onsets
were defined as the point at which the voltage of the component reached 50% of the peak
amplitude.
28
2.3. Results
2.3.1. Both high- and low-salience distractors produce behavioural
interference
For the main experiment, participants were instructed to search a circular multi-
item array for a yellow target singleton that appeared among green (Condition 1) and
orange (Condition 2) non-targets. On a subset of trials, a red or a blue distractor singleton
would replace a non-target item. Between the two conditions, only the colour of non-target
items were varied. This was done in order to disentangle the effects of distractor colour
from those of distractor salience. Colours were selected based on an initial pilot study to
gauge salience (See: 2.2.4. Behavioural Pilot Procedure). Among green non-targets, the
red distractor was selected to be more salient than the yellow target and the blue
distractor, while the blue distractor was selected to be nominally salient to the target.
Among orange non-targets, the blue distractor was selected to be more salient than the
yellow target and the red distractor, while the red distractor was selected to be nominally
salient to the target.
To determine the effectiveness of this manipulation in the visual search task,
median reaction times (RTs) were separately computed for each condition, for trial types
where the red or blue distractor appeared with the target. Inter-participant mean RTs were
then derived by averaging participant median RTs across these different trial types. RTs
from distractor-present trials were analysed as a function of distractor colour (red, blue)
and non-target colour (green, orange). As anticipated, the main effect was not significant,
[F(1,46) = .65; P = .425] but the distractor colour × non-target colour interaction was found
to be significant [F(1,46) = 55.9; P < .001]. Figure 2.2 illustrates the basis for this
interaction across the two conditions: RTs were slower when the display contained a high-
salience distractor than when the display contained a low-salience distractor, regardless
of the specific combination of distractor and non-target colours.
29
Figure 2.2. RTs associated with distractor interference. Mean response times (across participants; in milliseconds) for blue distractor, red distractor, and distractor absent (x) trials for Condition 1 and Condition 2 (left). Mean response times were then collapsed to create high-salience, low-salience, and distractor absent trials across the two experimental conditions (right).
This preliminary RT analysis confirmed that distractor salience rather than
distractor colour modulated performance in the additional singleton task employed here.
Accordingly, RT data from the two distractor colours were combined with RT data from the
two non-target colours to yield high- and low-salience distractor types. A distractor-absent
level was added to assess the overall effects of high- and low-salience distractors on
search performance. The results showed inter-participant mean RTs were shortest on
distractor-absent trials (606 ms), intermediate for low-salience distractor trials (613 ms),
and longest in the high-salience distractor trials (628 ms), leading to a significant main
effect of distractor type [F(2,92) = 75.5; P < .001]. Inter-participant mean RTs across the
three levels were all found to be statistically different from one another by pair-wise
comparison (Ps < .001). As can be seen in Figure 2.2, although both distractors delayed
search, the high-salience distractor caused a longer delay (22 ms) than did the low-
salience distractor (8 ms).
30
2.3.2. High- but not low-salience distractors vary as a function of
target-distractor distance
RT interference has been shown to vary not only as a function of distractor salience
(as shown here) but also as a function of target-distractor distance, with nearby distractors
causing more interference than more distant distractors (Gaspar & McDonald, 2014;
Hickey & Theeuwes, 2011; Jannati et al., 2013; Mounts, 2000). To examine the
dependency of interference on target-distractor distance for the different salience
conditions, RTs were submitted to an ANOVA with two within-subject factors—Target-
Distractor Distance (1, 2, 3, 4, 5; see Methods) and Distractor Salience (high, low)—and
a between-subject factor for Non-target Colour (green, orange). Overall, RTs were longer
for nearby distractors than distant distractors and for high-salience distractors than low-
salience distractors, resulting in significant main effects for Target-Distractor Distance;
F(4,184) = 37.09, P < .001, and Distractor Salience, F(4,46) = 55.07, P < .001,
respectively. Critically, a significant interaction between Target-Distractor Distance and
Distractor Salience was also found, F(4,184) = 38.30, P < .001. No between subjects
effects were observed for non-target colour [F(1,46) = 1.48, P = .23], further confirming
the success of the salience manipulation. Figure 2.3 illustrates the basis for this
interaction: RTs decreased in a monotonic fashion as the distance between target and
distractor increased for high-salience-distractor trials, but the interference effect changed
little over the five target-distractor distances for low-salience trials.
31
Figure 2.3. Target-distractor RT distance effects. Mean response times (across participants; in milliseconds) for five target-distractor distances (d1- d5) for both high- and low-salience distractor trials.
2.3.3. Interference without evidence of attentional capture
Target-distractor distance effects like the one observed here for high-salience
distractors have previously been attributed to distractor-driven attentional capture by some
(Hickey & Theeuwes, 2011; Mounts, 2000) and distractor suppression by others (Gaspar
& McDonald, 2014; Jannati et al., 2013). According to the capture account, attention is
deployed first to the location of the more salient distractor, and a concomitant zone of
inhibition around that location impairs selection of nearby targets. According to the
suppression account, inhibitory signals associated with the unattended distractor conflict
with excitatory signals associated with the to-be-attended target, leading to perceptual
ambiguity when the two singletons fall within the same receptive fields. While it is difficult
to disambiguate between these accounts on the basis of the behavioural evidence alone,
one aspect of the RT data favours a suppression account: the most distant salient
distractors (target-distractor distance d5) did not delay search relative to distractor-absent
trials [606 ms vs. 606 ms, respectively, t(47) = .04, P = .97]. That is to say, there was no
behavioural evidence to indicate that attention was inadvertently deployed to salient
distractors that appeared far away from the target.
32
2.3.4. Only salient distractors are suppressed during additional
singleton search.
The primary goal of this chapter was to determine whether all distractors or only
particularly salient ones are suppressed during visual search. To this end, ERPs were
examined separately for(i) displays containing a lateral, high-salience distractor and a
midline target and (ii) displays containing a lateral, low-salience distractor and a midline
target. Any lateralized ERP activity observed in response to such displays can be ascribed
to the distractor singleton because target singletons appearing above or below fixation are
incapable of eliciting such ERP lateralizations (Hickey et al., 2006, 2009; Luck, 2014;
Woodman & Luck, 2003).
The resultant PD ERPs confirmed the prediction that the visual system acts to
suppress the processing of high- but not low-salient items. As shown in Figure 2.4a, the
high-salience distractor elicited the PD in both non-target-colour conditions (green non-
targets, t(23) = 2.5; P = 0.02; orange non-targets, t(23) = 2.1; P = 0.046). Neither the onset
latency (274 ms vs. 293 ms; tc = -1.3; P = 0.19) nor the amplitude (0.53 vs. 0.44 μV; t[46]
= 0.3; P = 0.756) of the PD was found to differ significantly between the two conditions. As
shown in Figure 2.4b, no PD was in evidence on low-salience distractor trials (green non-
targets, t(23) = -1.0; P = 0.345; orange non-targets, t(23) = 0.7; P = 0.493.
To evaluate the differences statistically, PD amplitudes were analysed using an
ANOVA with a within-subject factors for Distractor Salience (high, low) and a between-
subject factor for Non-target Colour (green, orange). As anticipated, the PD was
significantly larger for high-salience distractors than for low-salience distractors, F(1,46) =
7.51, P = .009. Neither the Non-target Colour main effect nor the Distractor Salience x
Non-target Colour interaction (Fs < 1) were significant.
33
Figure 2.4. ERPs elicited by displays containing a midline target and a lateral distractor for each non-target condition. Time 0 reflects the onset of the search display, and negative voltage deflections are plotted above the x-axis, by convention. Waveforms were recorded over the lateral occipital scalp (electrodes PO7 and PO8). (A) ERPs recorded contralateral and ipsilateral to a high-salience distractor. (B) ERPs recorded contralateral and ipsilateral to a low-salience distractor.
2.3.5. Distractor suppression can be indirectly observed in
differences in N2pc amplitude.
Figured 2.5 shows grand averaged contralateral minus ipsilateral difference
waveforms for lateralized target singleton trials for the various display configurations. The
N2pc was measured as the difference in mean amplitude between contralateral and
ipsilateral activity at electrodes PO7/PO8 from 250 to 290 ms after the onset of the search
array. For Condition 1, a significant N2pc was observed in both high and low-salience
distractor conditions for the i) lateral target, contralateral distractor, ii) lateral target, midline
distractor, and iii) lateral target, ipsilateral distractor display configurations (t’s > 2.63, p <
34
.02). For Condition 2, a significant N2pc was observed for the same six display
configurations [t > 4.19, p < .001].
Figure 2.5. ERPs elicited by displays containing a lateral target and a midline distractor for each non-target condition. ERPs are presented as contralateral-minus-ipsilateral difference waveforms for displays containing a lateral target and a distractor singleton, recorded over the lateral occipital scalp (electrodes PO7 and PO8). Difference waveforms are separated for trials where the high- and the low-salience distractor were presented in both (A) Condition 1 and (B) Condition 2.
When attentional selection and distractor suppression coincide, the similar
temporal profiles of the N2pc and PD will result in the two components overlapping.
Specifically, on a search display trial that elicits target selection and distractor
35
suppression, the negative-going voltage of the N2pc will sum linearly with the positive-
going voltage of the PD (Hickey et al., 2009). The overlap of the two components results
in an observed difference in the amplitude of the more dominant N2pc component: the
N2pc is largest when the target and distractor appear on opposite sides of fixation,
intermediate when the distractor appears on the midline, and smallest when the target and
distractor appear on the same side of fixation (Gaspar & McDonald, 2014).
In line with the summation hypothesis outlined above, amplitude differences in the
N2pc can be interpreted as indirect evidence of the presence of the PD and distractor
suppression. Namely, if distractor suppression has occurred, the N2pc waves associated
with lateral-target displays will vary linearly due to their summation with the PD.
Alternatively, if distractor suppression has not occurred, the N2pc waves associated with
lateral-target displays will not change as a function of the relative distractor location (same
side, opposite side).
As shown in Figure 2.5, amplitude differences were observed for trials that
contained a high-salience distractor singleton. As would be predicted, for Condition 1 the
target N2pc was largest when the distractor was on the opposite side of fixation (mean
amplitude in 250-290 ms window: -1.60 µV), intermediate when the distractor was on the
vertical midline (-.91 µV), and smallest when the distractor was on the same side of fixation
(-.63 µV). The same pattern of results was observed for Condition 2 (-1.93 µV, -1.47 µV,
-1.25 µV respectively). These differences in N2pc amplitude were found to differ
significantly for all high-salience distractor trials across the three lateral-target
configurations [F(2,94) = 9.0, p < .001]. In contrast, low-salience distractor trials elicited
no differences in N2pc amplitudes across the same three display configurations [Condition
1: -1.42 µV, -1.37 µV, -1.35 µV; Condition 2: -1.36 µV, -1.20 µV, -1.47 µV; F (2,94) = 0.45;
P = .64]. Taken together, the resultant N2pc ERPs provide further confirmation that the PD
was present on high-salience distractor trials but not on low-salience distractor trials.
36
2.4. Discussion
The PD has proven to be a useful tool for understanding how the visual system
deals with distraction. While evidence continues to mount that the PD component reflects
distractor suppression, the sensory conditions necessary to elicit the PD component
remain an issue of some debate. Some have argued that when forthcoming distraction is
highly likely, the PD might be strategically utilized to prevent highly salient distractors from
capturing attention. On the other hand, the PD may reflect a preventive mechanism that
would automatically be assigned to deal with suppressing all unique and potentially
distracting objects. The main objective of the present study was to investigate how
differences in distractor salience affected the suppressive processing indexed by the PD
component during visual search. High-salience distractor stimuli were found to elicit a
contralateral positivity in the expected PD interval (250-290 ms), whereas low-salience
distractor stimuli did not. Differences in N2pc amplitude were also used to provide indirect
evidence for the presence of the PD component. On trials where the PD occurred, the
component was observed to summate with the N2pc (negativity) producing the same
pattern of amplitude effects previously reported by Gaspar and McDonald (2014). These
amplitude differences were observed to lateralized target stimuli solely on trials where the
high-salience distractor was also present in the search array.
The findings here support the notion that individuals can adopt distinct filtering
strategies depending on the given circumstances in order to deal with distraction and
facilitate target processing. In the present study, suppression was selectively engaged
depending on the physical properties of the stimulus: high- but not low-salience distractors
were actively suppressed during visual search. There are three potential reasons why
suppression may have been selectively implemented to deal with only salient distractors
in this paradigm. First, the recruitment of an active suppression mechanism likely requires
the allocation of top-down attentional resources (Engle et al., 1995). In instances where
the competition between a target and distractor stimulus is minimal, other low-level,
bottom-up mechanisms are often available to help resolve attention in favour of the target.
For example, the mere repetition of stimulus features across trials primes the visual
system for a previously selected stimulus (Christie, Livingstone & McDonald, 2015; Hickey
et al., 2011; Maljkovic & Nakayama, 1994). In the present study, such priming would have
37
biased selection in favor of the target over the equally salient distractor. Second, salience
computations are imprecise and most models of visual attention assume a level of noise
to impact the outcome of the coded saliency map (Li, 2002; Zehetleitner, Koch, Goschy &
Müller, 2013). The visual system would have difficulty locating a distractor versus a target
based on physical salience, as their output on the salience map would be roughly equal.
Accordingly, if the salience activation of a distractor does not meet a threshold on the
salience map, the visual system simply may not be able to pre-attentively index it for active
suppression to be applied at an attentional stage of processing. Third, suppressing a
distractor is not always an optimal search strategy and does not necessarily benefit target
resolution; actively suppressing a distractor stimulus will increase an observer’s response
time and error rate in a proximity-dependent manner (Gaspar & McDonald, 2014; Jannati
et al., 2013). In instances where stimulus competition is minimal and a target can be easily
resolved, not suppressing may prove an optimal strategy as suppression can actually
impair search on a large proportion of trials. As shown here, distractor suppression
benefits visual search performance when the target and distractor are spatially distant but
systematically hinders performance as they appear closer together in the visual field. High-
salience distractors produced 54 ms of interference (versus distractor absent trials) when
they appeared adjacent to the target singleton, whereas identically positioned low-
salience distractors produced only 6 ms of interference.
These proximity costs for high-salience distractors are worth further discussing, as
findings such as these have become commonly interpreted as a behavioral marker of
attentional capture (Caputo & Guerra, 1998, Eriksen & St. James, 1986, Hickey &
Theeuwes, 2011, Mounts, 2000, Mounts, 2005, Mounts & Gavett, 2004, Theeuwes,
Kramer & Kingstone, 2004; Zehetleitner, Proulx & Müller, 2009). According to the
Localized Attentional Interference hypothesis (LAI; Mounts, 2000, Mounts, 2005; Mounts
& Gavett, 2004), when attention is deployed to a distractor it produces a spatial inhibition
surrounding the object (colloquially referred to as a Mexican hat distribution). This
inhibitory ring makes it difficult to subsequently deploy attention to neighboring targets as
these targets now fall within the suppressed region of the previously attended distractor.
Inconsistent with this interpretation is the finding here that salient distractors produce no
behavioural interference when the target and distractor appear furthest from one another;
a salient distractor that initially captures attention would still invariably evoke some
38
interference cost, even if that salience-based interference was small. Furthermore, there
is no ERP evidence here to suggest that distractors captured attention: salient distractor
singletons were not found to elicit an N2pc under any of the display configurations. Given
the lack of evidence for attentional capture, it is argued here that actively suppressing the
salient distractor impairs search for nearby targets (see also Gaspar & McDonald, 2014;
Jannati et al., 2013). This spreading-inhibition account is similar to the LAI hypothesis,
except that the inhibition is not hypothesized to surround a central zone of attentional
excitation. Thus, when an attended target and a suppressed distractor are presented
within the same receptive field, facilitatory and inhibitory signals compete and make it
more difficult to resolve the identity of the target stimulus (Gaspar & McDonald, 2014;
Janatti, Gaspar & McDonald, 2013). The findings here are consistent with this spreading-
inhibition interpretation.
Thus far, the distractor interference reported in this chapter has been largely
discussed in the context of active suppression and the PD; however, it is important to note
that low-salience distractors also interfered with target discrimination. Although low-
salience distractors produced significantly less interference than high-salience distractors,
what is especially notable about this distractor interference is the fact that it was not
modulated by the spatial distance between the target and distractor. Low-salience
distractors were observed to produce a uniform pattern of interference irrespective of
target-distractor proximity. This finding is consistent with the non-spatial filtering costs that
have been attributed to the parallel filtering of irrelevant stimuli competing for attentional
priority (Folk & Remington, 1998, Folk & Remington, 2006 and Kahneman et al., 1983).
Thus, distractor singletons that do not capture attention can nonetheless produce a cost
during visual search that delay the allocation of attention to a target. However, an
interesting alternative explanation for this non-spatial interference is worth consideration:
perhaps low-salience distractors did capture attention—at least on a subset of trials.
According to the sequential sampling model of salience-based selection (Zehetleitner et
al., 2013), the salience map reflects a noisy estimate of the available sensory information
for the entire visual field and, due to this noise, a proportion of selection errors to a
distractor will invariably occur. The likelihood of making a selection error varies around a
probability distribution determined by the reliability of the salience map to index stimuli:
errors are low when the output of the salience map is effective at indexing stimuli that differ
39
significantly in physical salience; however, if two competing stimuli are approximately
equal in salience (as was the case here) distractor locations could erroneously be denoted
as most salient on a higher proportion of trials. Without an active suppression mechanism
to prevent their selection, distractor stimuli could have captured attention on a subset of
the low-salience distractor trials, resulting in the interference reported here. Although this
is an intriguing idea, the present experiment was not designed to directly test this
hypothesis and there is little evidence here to favour this explanation over one of non-
spatial filtering.
Despite the evidence here to suggest that only salient distractors are actively
suppressed, one outstanding question remains: why then have other researchers
observed non-conspicuous distractors to elicit the PD? Both the seminal PD study (Hickey
et al., 2009) and a more recent follow-up (Hilimire et al., 2012) reported a PD component
to an isoluminant distractor stimulus that was reasonably inconspicuous relative to the
target stimulus. One potential explanation for the discrepancy is that the perceptual load
was lower in those earlier studies that used two-item displays (a target and a distractor)
and it is possible that these low-load displays allowed an observer sufficient attentional
resources to process all items on the screen. This inference is consistent with perceptual
load theory (Lavie & Tsal, 1994), which states that under low perceptual loads any spare
capacity left over from the attentional processing of the target will be applied to the
processing of irrelevant distractors (e.g., Beck & Lavie, 2005, Lavie, 1995, Lavie & Cox,
1997; Lavie & Fox, 2000). In such instances, a more active form of attentional control can
act to minimize intrusion from all irrelevant stimuli (Lavie, Hirst, de Fockert, and Viding,
2004). However, under high load conditions the visual system is likely unable to allocate
this form of active processing to all irrelevant stimuli, selectively opting to suppress only
stimuli that are especially salient and distracting. Future research will be required to
elucidate the extent to which the PD component can be impacted by differences in
perceptual load.
40
Individual differences in working memory
This chapter is adapted from: Gaspar, J. M., Christie, G. J., Prime, D. J., Jolicœur, P. &
McDonald, J. J. (2016). Inability to suppress salient distractors predicts low visual working
memory capacity. Proceedings of the National Academy of Sciences, 113(13), 3693-3698.
3.1. Introduction
The ability to hold and manipulate visuospatial representations in memory is
fundamental to the way in which we navigate through our visual environment. The visual
working memory (vWM) system enables the temporary storage and processing of these
representations but has strict capacity limitations (Sperling, 1960; Irwin and Andrews,
1996; Luck and Vogel, 1997; Vogel et al., 2001). In order to effectively cope with these
limitations, attention can bias which objects will consume vWM's finite resources and
which objects will be ignored. In line with this, attention has been hypothesized to work as
a gatekeeper that controls the flow of information into working memory (McNab &
Klingberg, 2008; Awh & Vogel, 2008; Awh, Vogel & Oh, 2006). This hypothesis is
consistent with a growing literature on the functional relationship between attention and
vWM. Hemodynamic responses (Zanto, Rubens, Thangavel & Gazzaley, 2011), visually
evoked brain potentials (Zanto & Gazzaley, 2009), and single-cell recordings (Suzuki &
Gottlieb, 2013) all provide converging evidence that attention and WM are inextricably
linked, and may even share the same neural substrates (Awh & Jonides, 2001).
Research has shown that on average the capacity of vWM is approximately three
or four items; however, this limit has been found to vary significantly across individuals
(e.g., Luck & Vogel, 1997; Pashler, 1988; Sperling, 1960). According to cognitive-control
based theories of vWM, the inter-individual differences observed for vWM performance
are not related to memory capacity per se, but rather directly associated with variability in
41
attentional control (e.g., Engle & Kane, 2004; Hasher, Lustig, & Zack, 2007; Kane,
Conway, Hambrick & Engle, 2007; Lavie, Hirst, de Fockert & Viding, 2004). In line with
such theories, individuals are thought to differ primarily in their ability to encode
behaviourally relevant information in the presence of irrelevant, distracting information.
Consistent with this notion is a growing body of studies that have reported that the
presence of distractors in visual search tasks impact the ability to hold representations in
vWM (de Fockert, Rees, Frith & Lavie, 2001; Gazzaley & Nobre, 2012; Lavie & de Fockert,
2005; Luck & Vogel, 1013; McNab & Klingberg, 2008; Vogel, McCollough, Machizawa,
2005), particularly for individuals identified to have low vWM capacities (Fukuda and
Vogel, 2009; 2011). It has been hypothesized that this distractor cost among low-capacity
individuals is primarily associated with an inability to filter out irrelevant information.
Whereas high-capacity individuals will only store task-relevant information, low capacity
individuals will attempt to encode both task-relevant and task-irrelevant information. Thus,
when distractors are presented alongside to-be-remembered items, low-capacity
individuals overload of the vWM system by unnecessarily storing irrelevant distractors
(Vogel et al., 2005; Fukuda & Vogel, 2009; 2011).
Although the relationship between inefficient attentional filtering and vWM
performance has been extensively researched, the precise filtering mechanism(s) that are
affected remain largely unknown. Attentional filtering can modulate sensory processing in
two distinct manners: by enhancing the neural representation of to-be-remembered items
(Mazza, Turatto, & Caramazza, 2009; Treue & Trujillo, 1999) and by actively suppressing
to-be-ignored items (Duncan & Desimone, 1995; Luck et al., 1997; Serences et al., 2004;
Zanto & Gazzaley, 2008). What is unclear is whether the inefficient filtering observed for
low-capacity individuals reflects a general deficit in processing associated with the ability
to enhance, the ability to suppress, or both. While previous ERP studies have examined
the relationship between attentional filtering and vWM, these studies have typically
compared modulations of a single early sensory-evoked component (e.g., the amplitude
of the P1 or N1) relative to a baseline condition (Fukuda & Vogel, 2009; Rutman, Clapp,
Chadick & Gazzaley, 2010; Zanto & Gazzaley, 2009). The limitations to this approach are
twofold. First, the modulation of a component does not necessarily give any information
regarding temporal aspects of processing. Second, although comparing P1/ N1 amplitude
relative to a baseline can demonstrate that a difference exists, it is insensitive to whether
42
the difference reflects enhancement, suppression, or a change associated with processing
the baseline condition. What has been missing from the extant studies are dissociable
neurophysiological measures of attentional enhancement and of distractor suppression
that could resolve this question definitively.
In the present study, this limitation was addressed by isolating distinct filtering
operations associated with target enhancement and distractor suppression during visual
search. Recent electrophysiological studies of selective attention have found that laterally
presented stimuli tend to elicit unique ERP signatures contingent on whether they are
attended or ignored. These ERP components—termed the N2pc and the PD—are
observed over visual cortex, contralateral to the location in space where a subject is
actively processing a stimulus. Whereas the N2pc is thought to index the enhanced
processing of an attended stimulus (e.g., Mazza et al., 2009), the PD is thought to index
the suppressive processing of an irrelevant to-be-ignored stimulus (e.g., Gaspar &
McDonald, 2014; Hickey et al., 2009).
Using these two electrophysiological indices of selective attention, Chapter 3
investigates whether a relationship exists between attentional filtering and vWM
performance. Here, EEG was recorded while participants performed a visual search task
identical to the task used in Chapter 2. This task required the participant to search a visual
array of stimuli for a pre-specified color singleton while attempting to ignore another, task-
irrelevant color singleton. This task-irrelevant singleton could either be more salient or
equally salient to the target singleton (with salience having been confirmed by the pilot
experiment presented in Chapter 2). In addition to this, the vWM capacity of participants
was assessed using a change detection task (Phillips, 1974; Vogel & Luck, 1997). Utilizing
both measures, the relationship between vWM and attentional filtering was then assessed
by correlating vWM estimates (k) with electrophysiological indices of target enhancement
and distractor suppression.
3.2. Materials and Methods
The Research Ethics Board at Simon Fraser University approved the research
protocol used in this study.
43
3.2.1. Participants
Fifty-five subjects from Simon Fraser University participated after giving informed
consent. The subjects used here were sampled from the same data as those presented
in Chapter 2. Students received course credit for their participation as part of a
departmental research participation program. A grand total of 48 participants, 24 in
Condition 1 (20 women, age 20.8 ± 2.2 y; 0 left-handed), and 24 in Condition 2 (13 women,
age 20.0 ± 2.9 y; 1 left-handed) participated in the study. All subjects reported normal or
corrected-to-normal visual acuity and were tested for typical colour vision using Ishihara
colour test plates.
3.2.2. Working Memory Capacity Procedure
For the present experiment, participants first completed 120 trials of a change
detection task. Each trial consisted of an unmasked 150 ms memory array display with
set sizes consisting of two, four, six, or eight coloured squares. After a blank inter-stimulus
interval, a probe item was presented in one of the previously occupied locations.
Participants had to indicate whether the colour of the probe matched or did not match the
colour of the item from the memory array previously presented at that location. vWM
capacity (K) was computed for and averaged across the four set sizes (Cowan, 2001;
Pashler, 1998). All other methods were identical to those in Luck and Vogel (1997).
3.2.3. Visual Search Task Stimuli and Apparatus
All visual search task stimuli and apparatus were identical to those used in the
previous chapter.
3.2.4. Visual Search Task Procedure
All visual search task procedures were identical to those used in the previous
chapter.
44
3.2.5. Behavioural Analysis
The principle behavioural analysis sought to assess the relationship between vWM
capacity versus response variability and overall speed of processing. RTs were collapsed
across the three conditions (high-salience, low-salience, distractor absent) in order to
create an overall speed of processing (average median RT) and response variability (RT
standard deviation) estimate. Each individual’s measures were then correlated with their
respective vWM capacity score.
3.2.6. Electrophysiological Recording and Analysis
All EEG and electrooculogram (EOG) recording procedures were identical to those
in Chapter 2. In this chapter, the primary analysis focused on ERPs elicited by the following
display configurations: (i) lateral target, midline high-salience distractor; (ii) lateral target,
midline low-salience distractor; (iii) midline target, lateral high-salience distractor; (iv)
midline target, lateral low-salience distractor. For each participant, ipsilateral and
contralateral waveforms were constructed by combining trials where the stimulus of
interest appeared in the ipsilateral hemifield versus when the stimulus of interest appeared
in the contralateral hemifield relative to an electrode site. Lateralized ERP difference
waveforms were then computed for the display configurations of interest by subtracting
the ipsilateral waveform from the corresponding contralateral waveform at electrode sites
P07 and P08. As with all ERPs shown here, negative voltages were plotted upward so
that the N2pc would appear in these difference waveforms as an upward deflection and
the PD as a downward deflection.
The mean-amplitude of the PD and N2pc were first measured within windows that
were determined a priori based on previous published studies. The N2pc was measured
from 230-290 ms (based on the N2pc window used by Hickey et al., 2009) and the PD was
measured from 250-290 ms (based on the PD window used by Gaspar & McDonald, 2014).
The N2pc and PD were next measured using a novel signed-area technique. The
signed negative area of the N2pc and the signed positive area of the PD were computed
using each individual participant’s difference waveforms in a wide 200-350 ms post-
stimulus interval for the appropriate display configurations. As signed area measures are
45
biased to include only positive or negative values, they cannot be independently assessed
against zero using parametric statistics. In order to account for this, the signed area was
first subtracted against the signed area of an equal interval of baseline noise. This meant
the signed negative area of the N2pc was subtracted from the signed negative area of the
baseline and the signed positive area of the PD was subtracted from the signed positive
area of the baseline, both baselines spanning -150-0 ms. This signal-minus-noise
difference was then statistically assessed using standard parametric statistics.
Latency statistics were computed 200-350 ms post stimulus onset using jackknifed
averages. Relative to conventional approaches using single-participant average
waveforms, jackknife average waveforms provide accurate latency estimates with
increased statistical power (see Luck, 2014, for discussion regarding the benefits to the
jackknife technique). Here, latency onsets were defined as the point at which the voltage
of the component reached 50% of the peak amplitude. Statistical tests and degrees of
freedom were adjusted according to the standard proof described by Miller and colleagues
(e.g., Miller, Patterson, & Ulrich, 1998; Ulrich & Miller, 2001).
To visually assess the relationship between vWM capacity and ERP components
of interest, participants were evenly apportioned into a high-, medium-, and low-capacity
subgroup based on their vWM capacity estimates (n = 16 per group).
Lastly, the split-half reliability of the N2pc and PD components were computed by
randomly splitting the data into two halves and computing correlations of the half-data
mean amplitude averages for each component. All split-half correlations were corrected
for using the standard method (Anastasi & Urbina, 1997).
3.3. Results
3.3.1. Behaviour in Change-Detection Task
vWM capacity was measured for each participant using a change detection task
(Luck and Vogel, 1997). The average K estimate was 2.5, with scores ranging from 1.8 to
46
4.0. The K estimates were not found to differ significantly across the two non-target-colour
conditions [2.47 vs. 2.52; t(46) = 0.3; P = 0.778].
3.3.2. Behavior in Visual Search Task
Here, the relationship between vWM capacity and behavioural performance were
assessed for both response variability (RT standard deviation) and speed of processing
(median RT). Such measures have been linked to attentional control and shown to be
associated with differences in working memory capacity (Castellanos & Tannock, 2002,
Frye & Hale, 1996; Schachar & Logan, 1990). RTs were collapsed across all conditions in
order to create an overall participant speed and standard deviation estimate. Consistent
with previous findings, both median RT and RT standard deviation were found to
negatively correlate with vWM capacity in the present study (rs > -0.39; P < 0.007). This
indicates that vWM capacity was associated with faster and less variable responses.
3.3.3. Neural activity associated with distractor suppression
Having confirmed the success of the salience manipulation in Chapter 2, ERPs for
participants from both non-target conditions were combined to create three new main trial
types: high-salience distractor, low-salience distractor, and distractor absent trials.
Collapsed ERPs for lateral distractor present trials are shown in Figure 3.1a and 3.1b.
When collapsed, the mean amplitude of the PD remained significant for midline target,
lateral high-salience distractor trials [t(47) = 3.3; P = 0.002]. The data was then randomly
split into two equal halves of trials in order that the split-half reliability of the PD could be
computed. The PD constructed from the first half of data was found to moderately correlate
with the PD constructed from the second half of data (r = 0.57; P < 0.001).
47
Figure 3.1. ERPs elicited by displays containing a midline target and a lateral distractor. Time 0 reflects the onset of the search display, and negative voltage deflections are plotted above the x-axis, by convention. Waveforms were recorded over the lateral occipital scalp (electrodes PO7 and PO8). (A) ERPs recorded contralateral and ipsilateral to a high-salience distractor. (B) ERPs recorded contralateral and ipsilateral to a low-salience distractor. (C) Contralateral-minus-ipsilateral difference waveforms for both conditions.
In addition to the conventional mean-area approach, a novel signed area approach
was devised to measure the magnitude of the PD component. The signed area approach
used here has three primary benefits versus the conventional mean-area approach. First,
an unbiased and wide measurement window can be set without components of the
opposite polarity cancelling out their contribution. Second, by casting a wide measurement
window, subject variability is better represented within the measurement. The wide
window better detects the specific contribution of a participant’s component rather than
48
their contribution within a narrow window centered around the mean. This is particularly
important for the present experiment, as subject variability was central to the main
hypotheses. Third, and specific to the approach used here, baseline noise in the ERPs is
taken into consideration, reducing the likelihood of erroneously detecting a component
that does not exist. Here, the signed positive area of the PD was measured in an interval
from 200-350 ms post stimulus onset. This area was then subtracted from the signed
positive area in an equal interval within the baseline (-150-0 ms). The resultant magnitude
of the PD was found to be both statistically significant [t(47) = 4.1; P < 0.001] and internally
reliable (r = 0.61; P < 0.001).
3.3.4. Neural activity associated with distractor suppression predicts individual differences in vWM
In order to explore the basis for a relationship between distractor suppression and
vWM capacity, participants were equally divided into three groups contingent on their
behavioral K estimates to produce a high-, medium, and low-capacity group. ERP
waveforms recorded contralateral and ipsilateral to the salient distractor were then
averaged for the high, medium, and low K estimate groups (Figure 3.2a). vWM estimates
were 2.73 to 4.03 for the high-capacity group, 2.16 to 2.73 for the medium-capacity group,
and 1.60 to 2.13 for the low-capacity group. As can be seen in Figure 3.2b, the PD was
largest for the high-capacity group, reduced for the medium capacity group, and diminutive
for the low-capacity group. The differences between groups were significant for both the
mean amplitude and the signed positive area: F’s > 8.4 (P’s < 0.001).
49
Figure 3.2. Neural activity associated with salient distractor suppression predicts visual working memory capacity. (A) ERP waveforms recorded contralateral and ipsilateral to the salient distractor plotted separately for high-, medium-, and low-capacity groups. (B) Contralateral-minus-ipsilateral difference waveforms for high-, medium-, and low-capacity groups.
T-tests for both the mean amplitude and signed positive area confirmed the
presence of the PD in the high- and medium-capacity group (Ps < 0.006); however, the PD
was not found to be statistically significant in the low-capacity group (Ps > 0.18).
Additionally, in the low-capacity group a contralateral negativity was observed prior to the
non-significant contralateral positivity. Similar distractor elicited negativities (distractor
N2pcs) have been shown to reflect attentional capture by the distractor singleton (e.g.,
Burra & Kerzel, 2013; Hickey et al., 2006; Jannati et al., 2013). The mean amplitude of
the distractor elicited N2pc observed here was determined to be statistically significant
[t(15) = 2.3; P = 0.03]. Together, the results here indicate that while the majority of
participants were able to actively suppress the distractor singleton, individuals with the
50
lowest vWM capacities were not. As a result, the inability to suppress led to the attentional
capture of the salient-but-irrelevant singleton.
Next, correlations were used to investigate the relationship between the magnitude
of distractor suppression processing indexed by the PD component and vWM capacity. As
can be seen in Figure 3.3, the mean amplitude of the PD was observed to correlate
positively with vWM capacity (r = 0.55; P < 0.001), as was the signed positive area (r =
0.59; P < 0.001). In addition to differences in magnitude, the tertile split of the data also
revealed differences in the timing of the PD component. The PD appeared to begin earlier
for individuals with higher vWM capacity estimates relative to those with lower vWM
capacity estimates. To test this prediction, jackknifed estimates of PD onset were
correlated with each participant’s behavioral K estimate. Jackknifed estimates of PD onset
were found to also correlate positively with vWM capacity (r = -0.39; P < 0.006), confirming
a positive relationship between the onset of the PD and vWM capacity.
51
Figure 3.3. Neural activity associated with salient distractor suppression predicts visual working memory capacity. (a) Correlation between memory capacity (k) and the mean amplitude of the PD. (b) Correlation between memory capacity (k) and the “pure” PD area. The “pure” PD area reflects the area of the signed positive voltage under the curve between 200-350 ms minus the area of the signed positive voltage in the baseline between -150-0 ms prior to the onset of the search array.
3.3.5. Neural activity associated with target processing
In order to isolate the lateralized processing of the target, ERPs were first
constructed for lateral target, midline high-salience distractor trials. These trials allowed
for target processing to be assessed under the same stimulus load condition used to
assess distractor processing (Section 3.3.3). Prior to collapsing across the two non-target
52
conditions, an analysis was performed to ensure the target N2pc was present for both
experimental conditions and did not vary as a function of non-target colour. The lateral
target was confirmed to elicit an N2pc in each condition using both mean amplitude and
signed negative area measures [green non-targets: t(23) > 3.2 (P < 0.004); orange non-
targets: t(23) > 5.5 (P < 0.001)]. Neither N2pc mean amplitude nor signed negative area
was found to differ between the non-target conditions [t(23) > 1.6 (P < .15)], confirming
the conditions did not differ.
Distractor absent trials were additionally tested to assess the relationship between
target processing and vWM capacity. These trials offered a more isolated representation
of target enhancement in the absence of the concurrent suppression of the salient
distractor. Lateral target stimuli were again confirmed to elicit an N2pc in each condition
using both mean amplitude and signed negative area measures [green non-targets: t(23)
> 4.0 P < 0.001); orange non-targets: t(23) > 4.4; P < 0.001)]. Both N2pc mean amplitude
and negative signed area were found not to differ between the non-target conditions [t(23)
> .47 (P < .65)]. The split-half reliability estimates were observed to be significant for both
mean amplitude and signed negative area measures (rs > 0.67; Ps < 0.001).
3.3.6. Neural activity associated with target processing does not predict individual differences in vWM
The relationship between vWM capacity and target processing was assessed in
two ways. First, to visualize the relationship between vWM capacity and target processing,
participants were again apportioned into three subgroups contingent on their vWM
capacity (Figure 3.4a). In contrast to the differences observed for the PD component, one-
way ANOVAs revealed no significant differences for the N2pc across the K subgroups
(mean amplitude: Fs < 0.7, Ps > 0.512; negative area: Fs < 0.5, Ps > 0.62; onset latency:
Fs < 0.6; Ps > 0.552). Next, correlations were computed to assess the relationship
between target processing and vWM capacity. As shown in Figure 3.4a and 3.4b, neither
the mean amplitude nor the signed negative area of the N2pc was found to correlate with
vWM capacity in either of the display configurations (rs < 0.08; Ps > 0.59). In addition to
this, the jackknifed estimates of N2pc onset latency were also found to not correlate with
vWM capacity (rs < 0.13; Ps > 0.38). In contrast to the relationship observed for distractor
53
suppression, the results here reveal no discernable association between the
enhancement of the target singleton and vWM.
Figure 3.4. Neural activity associated with target processing not predictive of visual working memory capacity. (A) Correlation between memory capacity (k) and pure N2pc area for lateral-target displays of interest. (B) Contralateral-minus-ipsilateral difference waveforms for high-, medium-, and low-capacity groups.
54
3.4. Discussion
Cognitive-control based theories of vWM propose that individual differences in
performance are closely associated with variability in attentional control (e.g., Engle &
Kane, 2004; Kane, Conway, Hambrick, & Engle, 2007). Numerous studies have
repeatedly found high-capacity individuals to outperform their low-capacity counterparts
across a broad assortment of attention tasks (e.g., Kane et al., 2001; Bleckley et al., 2003;
Sobel et al., 2007). Based on such findings, it has been proposed that individual variability
may stem, not from a difference in memory capacity, but rather from differences in
attentional filtering (Cowan & Morey, 2006; Vogel et al., 2005). Specifically, low-capacity
individuals are thought to be inefficient filterers, which results in an inability to effectively
encode behaviourally relevant information in the presence of irrelevant, distracting
information (Fukuda & Vogel, 2009; Vogel et al., 2005). However, what has remained
unclear is whether inefficient filtering in low-capacity individuals reflects a deficit to
enhance relevant representations, suppress irrelevant representations, or both.
The present chapter sought to better understand the relationship between
attentional neural filtering and vWM capacity. Dissociable ERP measures of attentional
enhancement and suppression were isolated while participants performed a visual search
task. The results presented here revealed distractor suppression processing during visual
search to underlie optimal vWM performance. Specifically, an vWM capacity was found to
positively correlate with both the magnitude and onset latency of the PD component.
Additionally, whereas salient-but-irrelevant distractors were observed to evoke the PD for
high- and medium-capacity individuals, the PD was altogether absent for low-capacity
individuals. These results demonstrate that individual differences in vWM capacity are
associated with the magnitude and timing of a specific attentional control operation that is
necessary for suppressing the processing of salient-but-irrelevant visual objects.
Difficulty ignoring distracting information has been associated with general deficits
in cognitive performance and is symptomatic of several neurological disorders (e.g.,
Enright & Beech, 1993; Hahn et al., 2010; Melnick, Harrison, Park, Bennetto & Tadin,
2013; Wang et al, 2016). Electrophysiological studies of attention have found deficits in
inhibiting distractor representations to be predictive of lower vWM capacity (e.g.,
55
Gulbinaite et al., 2014; McNab & Klingberg, 2008; Rutman et al., 2010; Sreenivasan &
Jha, 2007; Zanto & Gazzaley, 2009). For example, individuals with low vWM capacity
more readily express deficits modulating the sensory processing of irrelevant information
at very early stages of visual processing (Sreenivasan, Katz, & Jha, 2007; Zanto &
Gazzaley, 2009). Furthermore, this impaired modulation in low-capacity individuals
appears to relate to a decrease in functional connectivity between the medial frontal and
dorsolateral prefrontal cortex, areas critical for attentional control (Gulbinaite et al., 2014).
The present study complements these findings by showing here that, not only were low-
capacity individuals unable to actively suppress salient-but-irrelevant distractors, but these
distractors captured their attention. This latter finding is consistent with a number of EEG
studies that have shown low-capacity individuals to have an increased propensity for
attentional capture by irrelevant distractors (Fukuda & Vogel, 2009). By attending to—
rather than suppressing—distractors, irrelevant items can inadvertently gain entry into
working memory and consume the finite resources of the system (Vogel & Machizawa,
2004; Vogel, McCollough, & Machizawa, 2005).
In this study, attentional enhancement was found to be unrelated to vWM capacity,
suggesting that vWM does not predict differences in processing a target processing during
visual search. The was confirmed twice: once using trials where the salient distractor was
present and once using trials where the distractor singleton was absent. The data
presented here is consistent with a recent study by Zanto and Gazzaley (2009), which
showed an individual’s vWM capacity to be independent of the attentional enhancement
of relevant information during the early stages of visual processing. The fact that target
processing was unrelated to vWM capacity does not infer that attentional enhancement is
unimportant for working memory—attentional enhancement is critical for the encoding of
information into the vWM system—but rather that the focusing of attention may represent
a more robust attentional mechanism, less susceptible to individual variability.
The finding here—that distractor suppression processing but not target
enhancement processing underlies individual differences in vWM—is consistent with
previous studies that have demonstrated a similar dissociable relationship in both typical
(Fukuda & Vogel, 2009; Zanto & Gazzaley, 2009) and aging populations (Gazzaley et al.,
2005; Gazzaley et al., 2008). Since the ability to focus on goal-relevant stimuli is
56
hypothesized to require working memory (Lavie, 2005), one possible explanation for this
finding is that low-capacity individuals may not have the available cognitive resources
necessary to sustain top-down control throughout the duration of a visual search task. As
a result, when a target and salient distractor compete, the initial selection may rely more
on bottom-up selection processes. This resource availability hypothesis is consistent with
studies that have shown that, as the availability of cognitive resources is depleted (by
increasing cognitive load), observers are more susceptible to attentional capture (Lavie &
de Fockert, 2005a, 2005b). If these resources are not available to begin with, low-capacity
individuals may not be able to recruit a distractor suppression mechanism, resulting in the
salient distractor capturing attention. Future research will be necessary to elucidate the
precise relationship between vWM, cognitive load, and the active suppression indexed by
the PD.
57
Signal suppression during a transient loss of attentional control
4.1. Introduction
The visual environment is replete with information; however, at any given moment
only a subset of this information is relevant to our behavioural goals. In order to effectively
navigate this environment, we must be able to locate the relevant information amid
cluttered and distracting visual conditions. It has been well established that selective
attention facilitates this process by directing our cognitive resources toward locations in
the environment contingent with our top-down volitional goals. By biasing processing
resources toward behaviourally relevant locations in the visual field, both the detectability
and discriminability of the information there is markedly improved (e.g., Liu, Pestilli &
Carrasco, 2005; Liu, Abrams & Carrasco, 2009; Luck, Hillyard, Mouloua & Hawkins, 1996;
Spitzer, Desimone & Moran, 1988; Yeshurun & Carrasco, 1998). Over the years, the
neural bases of the mechanisms controlling selective attention have been investigated
using electrophysiological recordings in humans (see Carrasco, 2011 and Luck, 2014 for
comprehensive reviews). Many of these electrophysiological studies have tracked
attention using the N2pc, an ERP component known to index spatially selective
processing during visual search (Luck & HIllyard, 1994a, 1994b; Hickey et al., 2009;
Mazza et al., 2009; Eimer, 1996). When a laterally presented item is attended, the N2pc
component typically presents as a greater negativity over posterior electrode sites
contralateral (versus ipsilateral) to the item. This processing is thought to reflect the
enhanced processing of an attended target when presented in competition among other
irrelevant items (Hickey et al., 2009; Luck, 2012, Mazza et al., 2009a, 2009b).
In addition to the enhanced processing of a relevant target item, the visual system
can also act to suppress the processing of irrelevant distractor items. In humans, studies
of visual attention have reported that task-irrelevant distractors can be inhibited in a top-
down manner when they are anticipated. This inhibitory mechanism serves to actively
58
suppress distractor representations and prevent these irrelevant objects from erroneously
capturing attention, even when they are especially salient (e.g., Gaspar & McDonald,
2014; Gaspelin et al., 2015; Hickey et al., 2009; Janatti, Gaspar & McDonald, 2013). The
electrophysiological correlate of this process is a contralateral positive-going voltage
recorded from electrodes over posterior-occipital scalp—an ERP component termed the
PD. The distractor suppression indexed by the PD is thought to reflect an active mechanism
that inhibits certain stimuli based on the parameters of an observer’s top-down attentional
set (e.g., Gaspar & McDonald, 2014; Hickey et al., 2009; Hilimire et al., 2012; Jannati et
al., 2013; Sawaki & Luck, 2010; Sawaki, Geng & Luck, 2011); however, direct evidence
for this active suppression is limited. Alternatively, it is possible that the PD may reflect
processing associated with the bottom-up physical properties of an object rather than the
object’s top-down status (Fortier-Gauthier et al., 2013).
If the PD is strongly related to endogenous attentional factors, it should be impaired
in situations where top-down control would be severely restricted. The primary purpose of
the present chapter was to investigate this possibility by disrupting attentional control
during visual search and examining how distractor processing is affected. One manner of
producing a disruption of attentional control is by presenting a critical target in rapid
succession of another. Although participants are easily able to identify the first target (T1),
their ability to identify the second target (T2) depends on the amount of time separating
the two items. Specifically, if the second target is presented within 200-500 ms of the first,
accuracy for the second target is heavily impaired. This impairment for processing the
second target is termed the attentional blink (AB; Broadbent & Broadbent, 1987;
Raymond, Shapiro, & Arnell, 1992). Although there is no consensus regarding the precise
nature of the mechanism(s) underlying the AB, most theories agree that the deficit in
processing T2 is associated with attentional resources being transiently disrupted by the
processing of T1 (Chun & Potter, 1995; Raymond, Shapiro, & Arnell, 1995; however, see
Olivers & Meeter, 2008; and Taatgen, Juvina, Schipper, Borst & Martens, 2009 for
alternative explanations).
In the present study, both the N2pc and PD, as well as behavioural performance,
were measured while participants performed a modified rapid serial visual presentation
(RSVP)/visual search task. The task here combined attentional-blink methodology with
59
those of the visual additional singleton search paradigm. The first target (T1) was a
number within an RSVP stream of letters and the second target (T2) was a colour singleton
that appeared within a visual search array that also contained a salient distractor singleton
(Figure 4.1). Subjects were instructed to first make a speeded response to T2 (by
identifying the orientation of a line inside the target singleton) and were then subsequently
probed to respond to T1 (by identifying whether the number in the RSVP stream had been
even or odd). Target and distractor processing ERPs to the T2 search array at short (within
the attentional blink) and at long (outside the attentional blink) separations from T1 were
then separately examined. Based on previous electrophysiological studies, it was
anticipated that target processing during visual search would be delayed during the blink
(Lagroix, Grubert, Spalek, Di Lollo & Eimer, 2015; Pomerleau, et al., 2014). Alternatively,
it was less clear was how distractor processing would be affected. If the distractor
suppression indexed by the PD reflects an active mechanism contingent on the observer
maintaining a high level of attention control, then it is expected that the AB would disrupt
this mechanism. However, if the PD reflects an exogenous process that is instead sensitive
to the physical properties of a stimulus, the component should remain unimpaired during
the AB.
4.2. Methods
4.2.1. Materials and Methods
The Research Ethics Board at Simon Fraser University approved the research
protocol used in this study.
4.2.2. Participants
Twenty students from Simon Fraser University participated after giving informed
consent. These students were given course credit for their participation as part of a
departmental research participation program. Eighteen subjects were included in the EEG
analysis (8 women, age 19.61, SD = 1.97; 3 left-handed), as two were excluded due to
excessive noise in the ocular channels. All subjects reported normal or corrected-to-
60
normal visual acuity and were tested for typical color vision, using Ishihara color test
plates.
4.2.3. Attentional Blink Task Stimuli and Apparatus
Stimuli were presented on a 23-inch, 120-Hz LCD monitor viewed from a distance
of 57 cm. RSVP streams were comprised of digits (x = 0.295, y = 0.361, 7.9 cd/ m2) and
uppercase letters (x = 0.295, y = 0.361, 7.9 cd/ m2) presented at a central fixation point.
Alphanumeric characters were approximately 1° in height and varied proportionally in
width. Visual search arrays were comprised of 10 unfilled circles presented equidistant
(9.2°) from a central fixation point. Each circle was 3.4° in diameter with a 0.3° thick outline.
Eight of the circles were uniformly colored non-targets, one was a target color singleton,
and one was a distractor color singleton. The target was dark yellow (x = 0.416, y = 0.519,
7.9 cd/ m2) and the distractor was red (x = 0.640, y = 0.324, 7.0 cd/ m2), and the non-target
circles were green (x = 0.288, y = 0.636, 7.9 cd/ m2). The red distractor singleton was the
most salient item in the search array (see: Gaspar & McDonald, 2014). A randomly
oriented vertical or horizontal gray line (x = 0.295, y = 0.361, 7.9 cd/ m2) was contained
within each of the circles. All stimuli were presented on a uniform black background (0.5
cd/ m2).
61
Figure 4.1. Example stimulus display from the experiment. T1 was a number presented amongst letters in an RSVP stream. T2 was an additional singleton search display where participants were instructed to identify the orientation of the line inside the yellow colour singleton. Participants were instructed to give a speeded response to the search array first and then identify the number as either even or odd.
4.2.4. Attentional Blink Task Procedure
For each trial, the display sequence was preceded by an 800-1,200 ms fixation
period. During this time, only the central fixation point was visible (Figure 4.1). In
preparation for the presentation of the display sequence, participants were instructed to
maintain fixation on the central point. The display sequence consisted of an initial 14-item
RSVP stream comprised of 13 letters and a single digit. Letter stimuli (A, B, C, D, E, F, G,
H, J, K, L, M, N, P, Q, R, S, T, U, V, W, X, Y, Z) were selected at random with the constraint
that the same letter could not appear twice in the stream. A digit stimulus (1, 2, 3, 4, 5, 6,
7, 8) was selected at random with the constraint that an equal number of even and odd
numbers would be presented within each block. The stimulus onset asynchrony (SOA)
between successive items in the RSVP stream was 100 ms. Upon the completion of the
RSVP stream, a search display was presented for 200 ms. The search array was
subsequently masked for 200 ms.
The first target (T1) was the digit inserted into the RSVP stream that appeared at
either the seventh (lag 8) or thirteenth (lag 2) position. The second target (T2) was the
target singleton in the search array. Participants were instructed to first make a speeded
response to T2 by identifying the orientation of the gray line inside the target singleton by
pressing one of two response buttons as quickly as possible. Participants were then
62
probed to indicate whether T1 had been an even or an odd number. After a response was
made to the probe, the next trial began.
The search display (T2) contained one target singleton and one highly salient
distractor singleton. Target and distractor locations were varied to produce the following
display configurations: lateral target, midline distractor (50%); midline target, lateral
distractor (50%). The order of the display configurations was pseudo-randomly intermixed
within each block of trials. Each experimental block was comprised of 36 trials. At the end
of the block, participants were given a minimum 5 second rest period and were permitted
to begin the next block whenever they decided. The experiment contained 24 blocks, for
a total of 864 trials per participant. At least 36 practice trials were given to each participant
prior to the start of the experiment.
4.2.5. Behavioural Analysis
Trials on which the participant responded incorrectly to either T1 or T2 were
automatically excluded from the analysis. Trials with anticipatory RTs and excessively
slow responses were excluded from analysis (less than 1% of all correct trials). Median
reaction times to T2 were derived for search displays for each participant. The means of
these median reaction times were then computed for both lag 2 and lag 8 trials.
Differences were statistically assessed using a paired t-tests. A repeated-measures
ANOVA with two factors (target-distractor distance and lag) was then used to assess
search performance as a function of target-distractor proximity. Next, the RT obtained on
target-distractor distance 1 trials was compared with target-distractor distance 4 trials for
both lag 2 and lag 8. This analysis sought to determine whether target-distractor proximity
differentially affected search performance within versus outside of the attentional blink.
4.2.6. Electrophysiological Recording and Analysis
EEG and electrooculogram (EOG) were recorded from active sintered Ag-AgCl
electrodes (Biosemi Active Two system) from 32 electrodes, using a modified montage
that included electrode sites FP1, FP2, AF3, AFZ, AF4, F7, F3, FZ, F4, F8, FC5, FCZ,
FC6, T7, C3, CZ, C4, T8, CP5, CPZ, CP6, P7, P3, PZ, P4, P8, PO7, POZ, PO8, O1, OZ,
63
O2. Horizontal EOGs were recorded using two electrodes positioned 1 cm lateral to the
external canthi, and vertical EOGs were recorded using two electrodes positioned above
and below the right eye. All EEG and EOG signals were digitized at 512 Hz, referenced in
real time to an active common-mode electrode, and low-pass filtered using a fifth-order
sinc filter with a -3 dB cutoff at 104 Hz. Electrode offsets were monitored to ensure the
quality of the data. After the data acquisition, EEG data for each channel were high-pass
filtered (-3 dB point at 0.05 Hz) and then converted from 24-bit to 12-bit integers.
EEG processing and ERP averaging were performed using event-related potential
software system (ERPSS) (University of California, San Diego). A semi-automated
procedure was used to discard epochs of EEG contaminated by blinks, eye movements,
or excessive noise (Green et al., 2008). Any trial with an artifact within a 1-s interval (-200-
800 ms post-stimulus) was rejected. Artifact-free epochs associated with the various
display configurations of interest were then averaged separately to create ERP
waveforms. The resulting ERPs were digitally low-pass filtered (-3 dB point at 32 Hz) and
digitally re-referenced to the average of the left and right mastoids. All ERP amplitudes
and baselines were computed using a 200 ms pre-stimulus window. The averaged event-
related horizontal EOGs did not exceed 2 μV for any individual participant, indicating their
gaze remained within 0.3° of the fixation point for a majority of the trials (McDonald &
Ward, 1999).
The primary analysis focused on ERPs elicited by search displays with following
display configurations: (1) lag 2, lateral target, midline distractor; (2) lag 2, midline target,
lateral distractor, (3) lag 8, lateral target, midline distractor; (4) lag 8, midline target, lateral
distractor. On each trial, a search display contained a lateral singleton and a midline
singleton. Biasing a singleton to the midline on each trial allowed for an equal number of
trials where the N2pc and PD could be isolated.
For each participant, ERPs to the various search displays were collapsed across
left and right visual hemi-fields, as well as left and right electrodes, to produce waveforms
indexing the processing of a lateralized singleton. Lateralized ERP difference waveforms
were then computed for the display configurations of interest by subtracting the ipsilateral
waveform from the corresponding contralateral waveform at electrode sites P07 and P08.
64
All statistics were performed on lateralized ERP difference waveforms. For all ERPs
shown here, negative voltages were plotted upward so that the N2pc would appear in
these difference waveforms as an upward deflection and the PD as a downward deflection.
The mean-amplitude of the N2pc was first measured in a window 230-290 ms post
stimulus onset. This window was determined a priori based on the findings of Hickey and
colleagues (2009) and was identical to the N2pc window set in Chapter 3. As it was
anticipated that the N2pc would be delayed during the AB, the window for lag 2 trials was
shifted by 30 ms to match the ~30 ms latency shift observed by Lagroix et al., 2015.
The PD was computed in a window 290-330 ms post stimulus onset, approximately
centered around the peak of the observed component. The 40 ms duration of the window
used to measure the component was selected to mirror the size of the statistical windows
used in the previous chapters but shifted later in time to account for later onset observed
here. Because the component appeared later than it had in the previous chapters, an
unbiased signed area measurement was also used to confirm the presence of the
component. Identical to what was done in Chapter 3, the signed positive area of the PD
was computed using each individual participant’s difference waveforms in a wide 200-350
ms post-stimulus interval. The signed positive area of the PD was subtracted from the
signed positive area of the baseline from -150-0 ms. This signal-minus-noise difference
was then statistically assessed using standard parametric statistics.
On midline target, lateral distractor trials, the PPC was computed in a 50 ms
window from 120-170 ms. This window is consistent with previous studies that have
typically reported the PPC to occur between 120-190 ms (Fortier-Gauthier et al., 2012;
Jannati et al., 2014; Leblanc, Prime, & Jolicœur, 2008).
Latency onsets were defined as the 50%-of-peak-amplitude voltage in the 0-200
ms interval for the PPC and 200-400 ms interval for the N2pc and PD. Statistical tests were
performed on jackknife-averaged ERPs and statistical thresholds were adjusted
accordingly (Ulrich and Miller, 2001). All statistics were computed relative to a 100 ms pre-
stimulus interval.
65
4.3. Results
4.3.1. Visual search is delayed during the attentional blink
Response accuracy data is shown in Figure 4.2. The mean proportion of correct
responses to the first target (T1) was 91.2%. A t-test performed on the T1 data yielded no
significant difference for Lag 2 versus Lag 8 trials [t(17) = 1.23, P = .23]. As is common
procedure, the accuracy for the second target (T2) was computed using only those trials
where the T1 had been identified correctly. The mean proportion of correct responses to
T2 was 94.7%. A t-test performed on the T2 data also yielded no significant effect of lag
[t(17) = .50, P = .62].
The absence of an effect of lag on either T1 or T2 accuracy was not unexpected
given the experimental design used here. The present study sought to maximize response
accuracy in order that a maximal number of trials might be retained for the ERP analysis.
In order to deal with the response ceiling for accuracy, RTs were used as the dependent
measure of the attentional blink. The idea that response speed can be used to index the
AB was initially proposed by Ruthruff and Pashler (2001) and has been since supported
by studies that have used RT as a dependent measure (e.g., Ghorashi, Smilek & Di Lollo,
2007; LaGroix et al., 2015). Estimates of the RT were again based exclusively on trials
where correct responses were made to both T1 and T2. Median RTs to the onset of the
T2 search array were calculated for each observer at both lag 2 and lag 8. As shown in
Figure 4.2b, responses to the T2 search array were slower on lag 2 trials (890 ms) than
on lag 8 trials (816 ms). This 64 ms RT difference was found to be significant [t(17) =
11.05, p < .001], confirming that a substantial AB deficit had occurred on lag 2 trials.
66
Figure 4.2. Main behavioural results: (A) Accuracy rates for T1 and T2 on both lag 2 and lag 8 trials. (B) Mean response times (across participants; in milliseconds) for lag 2 and lag 8 trials.
4.3.2. The N2pc is delayed within the attentional blink
Figured 4.3a shows grand averaged ERP waveforms contralateral and ipsilateral
to the lateral T2 target singleton for both lag 2 and lag 8 trials. For lag 8 trials, the N2pc
was measured as the difference in mean amplitude between contralateral and ipsilateral
activity at electrodes PO7/PO8 from 230 to 290 ms after the onset of T2. For lag 2 trials,
the window was shifted by 30 ms from 260 to 320 ms after the onset of T2. The mean
N2pc amplitudes for the lateral-target display configurations were found to differ
significantly from zero for both lag 2 [t(17) = 2.29, P = .03] and lag 8 [t(17) = 2.38, P = .03]
trials. A follow-up t-test further revealed that there was no difference in amplitude across
the two lag conditions [-.58 µV vs. -.56 µV; t(17) = .07, P = .94].
As shown in Figure 4.3b, although the N2pc did not differ in amplitude between
Lag 2 and Lag 8 trials, clear differences in the onset latency of the N2pc were evident,
with the N2pc occurring earlier on lag 8 trials than on lag 2 trials. A paired samples t tests
revealed that the N2pc to the T2 search array did in fact emerge 36 ms sooner at lag 8
than at lag 2 [288 ms vs. 252 ms; tc = 2.5, P = 0.02], confirming an AB in N2pc onset
latency. These findings corroborate the hypothesis that visual search is postponed during
67
the period of the AB (Ghorashi et al., 2007) and are consistent with other ERP studies of
the AB that have shown the N2pc to be delayed when the lag between targets was shorter
(Lagroix et al., 2015; Pomerleau et al., 2014).
Figure 4.3. ERPs elicited by trials with displays containing a lateral target and a midline distractor. Time 0 reflects the onset of the search display, and negative voltage deflections are plotted above the x-axis, by convention. Waveforms were recorded over the lateral occipital scalp (electrodes PO7 and PO8). (A) ERPs recorded contralateral and ipsilateral to a distractor for lag 8 and lag 2 trials. (B) Contralateral-minus-ipsilateral difference waveforms for lag 8 and lag 2 trials.
4.3.3. The PPC is unaffected during the attentional blink
A PPC (positivity posterior contralateral) component was observed to occur on
midline target, lateral distractor trials. Thought to reflect pre-attentive sensory processing,
the PPC presents as an early positivity in the time interval of the N1 over parieto-occipital
scalp (Fortier-Gauthier et al., 2012; Jannati et al., 2013; Pomerleau et al., 2014). In the
68
present study, the PPC component can be clearly seen in the difference waveforms on
trials where the salient distractor was the lateralized singleton (Figure 4.4b). To confirm
the presence of the PPC components, a pairwise t test compared the difference in mean
amplitude at electrodes PO7/PO8 from 120 to 170 ms post the onset of T2. The PPC was
determined to be significant on both Lag 2 [t(17) = 3.01, P = .008] and Lag 8 [t(17) = 2.94,
P = .009] trials. The PPC did not differ in either amplitude [t(17) = .64, P = .53] or latency
[tc = .22, P = .73] for either of the two lag conditions. Collectively, these data are consistent
with a pre-attentive role for the PPC: the proportional amplitudes and latencies for lag 2
and lag 8 trials support the idea that, unlike the PD, the PPC reflects a process unaffected
by a disruption of attentional control.
4.3.4. Individuals cannot recruit distractor suppression during the attentional blink
One of the principle questions asked in this chapter was whether the processing
indexed by the PD reflects an active mechanism contingent on top-down control. To test
this possibility, distractor processing was compared under conditions where the availability
of top-down control was restricted (during the AB) and unrestricted (outside of the AB).
Figured 4.4a shows grand averaged ERP waveforms contralateral and ipsilateral to the
lateralized T2 distractor singleton for both lag 2 and lag 8 trials. The PD was measured as
the difference in mean amplitude between contralateral and ipsilateral activity at
electrodes PO7/PO8 from 290 to 330 ms after the onset of T2. A one-way ANOVA on PD
mean amplitude yielded a main effect of lag [F(1,34) = 5.8, P = .02]. The presence of the
component was further confirmed using a paired samples t-test. The PD differed
significantly from zero on lag 8 [t(17) = 3.90, P = .001] but not on lag 2 [t(17) = 1.22, P =
.24] trials, which indicates that the distractor suppression mechanism was not recruited
during the AB. This finding was corroborated when the signed positive area was compared
to baseline noise [lag 8: t(17) = 3.31, P = .004; lag 2: t(17) = 1.19, P = .25]. The absence
of an observable PD on lag 2 trials indicates that the distractor suppression mechanism
indexed by the PD was not recruited during the AB.
69
Figure 4.4. ERPs elicited by trials with displays containing a midline target and a lateral distractor. Time 0 reflects the onset of the search display, and negative voltage deflections are plotted above the x-axis, by convention. Waveforms were recorded over the lateral occipital scalp (electrodes PO7 and PO8). (A) ERPs recorded contralateral and ipsilateral to a distractor for lag 8 and lag 2 trials. (B) Contralateral-minus-ipsilateral difference waveforms for lag 8 and lag 2 trials.
4.3.5. Behavioural evidence during the AB revisited
In instances where the distractor is more salient than the target, behavioural
interference increases as target-distractor separation decreases. This pattern of
interference has been used to argue the misallocation of attention to a distractor (Caputo
& Guerra, 1998; Hickey & Theeuwes, 2011; McCarley & Mounts, 2007, 2008; Mounts,
2000a, 2000b, 2005; Mounts & Gavett, 2004), while others have attributed it to distractor
inhibition spreading to the location of a nearby target (Gaspar & McDonald, 2014; Jannati
70
et al., 2013). To test these competing interpretations here, a proximity analysis was
conducted on RT data for both lag 2 and lag 8. RTs were separated for the nearest-
distractor and farthest-distractor conditions and a repeated measures ANOVA with the
factors lag (lag 2 vs. lag 8) and proximity (Nearest-distractor vs. Farthest-distractor) was
computed to statistically assess the data. This test revealed no main effect for proximity
[F(1,17) = .903, P = .355] but a main effect for lag [F(1,17) = 66.10, p < .001]. The
interaction between proximity and lag was also found to be significant [F(1,34) = 13.42, P
= .001]. As illustrated in Figure 4.5, these results indicate two distinct patterns of
interference during visual search. For Lag 8 trials participants were slower (22 ms) to
respond to the target when it appeared close to the distractor versus when it appeared
furthest. For Lag 2 trials participants were instead faster (14 ms) to respond to the target
when it appeared close to the distractor versus when it appeared furthest.
Figure 4.5. Target-distractor RT distance effects. Mean response times (across participants; in milliseconds) for lag 2 and lag 8 trials where the target and distractor appeared adjacent to one another and on trials where they appeared furthest from one another.
4.4. Discussion
The purpose of the present research was to examine how mechanisms of object
selection are impacted by a transient disruption to attentional control. To this end,
electrophysiological activity was recorded while participants completed a task comprised
71
of two well-known attention paradigms—the additional singleton paradigm, to measure
selective object processing, and the RSVP paradigm, to elicit the attentional blink. The AB
is a well-known behavioural consequence of the limitations of attention that can be
observed in dual-task paradigms where two target stimuli are presented in rapid
succession. The AB manifests as a deficit in reporting the second target stimulus (T2)
compared to the first target stimulus (T1). Although there is a vast literature dedicated to
debating precisely what causes the AB, it has been shown to be an effective means of
disrupting attention control (Akyürek et al., 2010; Brisson & Jolicœur, 2007; Dell'acqua et
al., 2006; Jolicoeur et al., 2006a; 2006b; Zhang et al., 2009). The present experiment
required participants perform the visual search task both within (lag 2) and outside of (lag
8) the AB to study how target and distractor processing were independently affected by
the availability of attentional control.
To date, a handful of electrophysiological studies have explored how selective
attentional processing during visual search is affected by the AB (e.g., Pomerleau et al.,
2011; Lagroix et al., 2015). These studies have revealed that the AB produces a transient
disruption in the ability to process the T2 visual search targets—that is, the AB results in
an impairment in guiding attention towards a relevant object in a timely manner (e.g.
Ghorashi and colleagues, 2007). Both the behavioural and ERP findings in the present
study are consistent with this hypothesis. Using visual search RT performance as the
dependent measure, a deficit in the processing of T2 was observed during the AB.
Relative to lag 8 trials, on lag 2 trials RTs were slower and the onset of the target N2pc
was delayed. These findings are consistent with those of Pomerleau et al. (2014), which
showed similar behavioural and electrophysiological delays during the AB period.
By comparison, only a single study has explored how distractor processing is
affected by the AB during visual search (Pomerleau et al., 2011). A primary motivation for
the current study was to examine whether the suppression of salient (but irrelevant) stimuli
is accomplished during the AB, if at all. In doing so, the current study also resolves an
important, outstanding question regarding the cognitive mechanism indexed by the PD.
The PD is thought to reflect a voluntary and flexible suppressive mechanism that can be
modulated by task instructions (Hickey et al., 2009; Hilimire et al., 2012). Such
endogenous processing would likely be disrupted during the AB interval, and would be
72
evidenced by a delay and/or attenuation of the mean amplitude of the PD. Several studies
have offered evidence for this view and have shown the PD to reflect a top-down process
that allows observers to actively suppress salient, task irrelevant items and prevent
capture (e.g., Gaspar and McDonald, 2014; Jannati et al., 2013; Sawaki & Luck, 2010).
However, an alternate account of the PD is that it reflects an automatic process that
indexes the location—but does not necessarily the suppression of—expected, salient
distractors. According to this view, the PD is activated by the bottom-up properties of an
object and not based on top-down control (Fortier-Gauthier, Dell'Acqua & Jolicœur, 2013).
If this were the case, the PD should not be affected by the AB and should therefore be
approximately equal in both timing and amplitude at both lags. The present study
disconfirmed this latter hypothesis, as the PD elicited by the T2 display varied considerably
between lag 2 and lag 8. Whereas the PD was elicited to the lateralized distractor singleton
on lag 8 trials (outside of the AB), it was altogether absent on lag 2 trials (within the AB).
This result is consistent with the interpretation that there was a transient disruption in the
ability to suppress salient distractors during the AB while selective processing
mechanisms remain engaged processing T1.
Previous studies have found that observers are slower to respond to objects
presented in close proximity of a salient object, and that this behavioral search penalty is
reduced as the distance between the target and distractor increases (Awh, Matsukura &
Serences, 2003; Bahcall & Kowler, 1999; Gaspar & McDonald, 2014; Hopfinger,
Buonocore & Mangun, 2000; Jannati et al., 2013; Muller, Mollenhauer, Rosler &
Kleinschmidt, 2005). This interference has been explained in terms of the automatic
inhibitory processing surrounding an attended area. In instances where the distractor is
more salient than the target, this proximity interference has been used to argue for the
misallocation of attention to a distractor (Caputo & Guerra, 1998; Hickey & Theeuwes,
2011; McCarley & Mounts, 2007, 2008; Mounts, 2000a, 2000b, 2005; Mounts & Gavett,
2004). Alternatively, it has been proposed that this increased interference for resolving a
target is not related to an erroneous shift of attention but rather to the suppression of the
distractor. When a distractor is actively inhibited, the suppressive processing can spread
to neighboring objects in the visual field (Gaspar & McDonald, 2014). According to this
account, neural ambiguity can occur when an attended target and a suppressed distractor
fall within the same cellular receptive fields. As a result, facilitative signals associated with
73
the target and suppressive signals associated with the distractor conflict, increasing the
difficulty to resolve the target stimulus (Luck, Girelli, McDermott & Ford, 1997). The
findings reported here seem to support this latter hypothesis. On Lag 8 trials, in which the
PD was present, RT interference was shown to increase as the distance between the two
singletons decreased (cf., Gaspar and McDonald, 2014). By comparison, on Lag 2 trials,
in which the PD was absent, no such RT interference was observed. Rather, subjects
responded significantly faster when the target and the distractor were adjacent to each
other, rather than when they were furthest away. This could suggest that, in the absence
of active suppression, the distractor singleton initially captured attention. The time
necessary to disengage and redeploy attention to a distant target may have been longer
than the time necessary to disengage and redeploy attention to an adjacent one, resulting
in the reversed proximity effect observed for lag 2 trials. However, since the lateral
distractor was not observed to elicit an N2pc on these trials, a more parsimonious
explanation is that the proximate grouping of the target and distractor singleton likely led
to less spatial uncertainty, and a more rapid engagement of attention to the target.
One final notable result of the present study was the observation that lateralized
distractor singletons triggered the PPC (positivity posterior contralateral), a component
thought to reflect a pre-attentive salience signal (Fortier-Gauthier, Moffat, Dell’Acqua,
McDonald & Jolicœur, 2012; Leblanc, Prime & Jolicœur, 2008). The Ppc has been shown
to be elicited by both target and distractor singletons in visual search tasks, and is not
related to the efficiency of search or singleton processing (Jannati et al., 2013). Rather,
the PPC seems to be tied to low-level sensory processes associated with the pre-attentive
processing of the most salient object in the visual field (Corriveau, et al., 2011; Fortier-
Gauthier et al., 2012). Based on such findings, the PPC has been hypothesized to reflect
an initial index of activation on a pre-attentive salience map that could serve to assist
attentional selection. In the current study, the PPC was observed to the most salient
object—the distractor—both within and outside of the AB. This seems to suggest that,
whereas later attentional processes are fragile and susceptible to impairment by
disruptions to attentional control, early, pre-attentive processes are robust and unaffected.
and lend further support to the idea that the subsequent PD reflects an active process
generated by an observer’s attentional goals.
74
Individuals with high levels of trait anxiety show differences in selective attentional processing
5.1. Introduction
High levels of trait anxiety have long been associated with the preferential biasing
of attention toward threat-related information, even when this information is known to be
irrelevant to the task at hand (for a meta-analysis of 217 studies, see Bar-Haim, Lamy,
Pergamin, Bakermans-Kranenburg & van Ijzendoorn, 2007). This negative attentional bias
has been linked to an impairment in the ability to filter out emotionally salient information.
As a result, high-anxiety individuals are more likely to be distracted by a threatening
stimulus—be it a fearful face or a negative word—leading to such stimuli inadvertently
capturing attention (Eimer & Kiss, 2007; Fox, Russo & Georgiou, 2005; McTeague
Shumen, Wieser, Lang & Keil, 2011; Moser, Becker & Moran, 2012). Furthermore, it has
been proposed that this filtering deficit may play a causal role in the etiology and
maintenance of clinical anxiety disorders. Because anxious individuals are more sensitive
to—and unnecessarily process more—emotionally salient information, this additional
processing may serve to promote the intrusive thoughts, heightened rumination, and other
anxiety-related behaviours that are typically associated with affective pathologies
(Wadlinger & Isaacowitz, 2010).
Although trait anxiety has been generally linked to the impaired filtering of
emotionally salient stimuli, researchers have recently reported that this impairment can
also be observed to stimuli that have no affective significance (e.g., Ansari & Derakshan,
2011a, 2011b; Berggren & Derakshan, 2013; Bishop, 2009; Pacheco-Unguetti, Acosta,
Callejas & Lupiáñez, 2010). Behaviourally, this has been shown in antisaccade tasks
where individuals are instructed to make an eye movement away from an emotionally
neutral abrupt-onset stimulus presented in the periphery. High anxiety individuals are
slower to initiate a saccade away from this stimulus (Derakshan et al., 2009; Wieser, Pauli
& Mühlberger, 2009), which suggests that they have a reduced ability to volitionally
75
override orienting toward a stimulus and commensurately reduced top-down attentional
control necessary to counteract the reflexive processes (Hutton & Ettinger, 2006).
Furthermore, high-anxiety individuals perform worse still on these antisaccade tasks under
higher levels of cognitive load (Berggren, Hutton, Derakshan, 2011; Berggren et al, 2013).
However, one shortcoming of the antisaccade task is that the salient stimulus is always
task-relevant—participants must first direct their attention to the stimulus onset in order to
make an eye movement away from it. This makes it difficult to determine whether the
slower saccades in high-anxiety individuals are the result of a deficit in the ability to inhibit
the stimulus or are the result of a deficit in executing a shift of attention away from the
stimulus.
Recently, it has been argued that an optimal way to assess salience- and goal-
driven attentional selection in high-anxiety individuals is with the additional singleton
paradigm (Moser et al., 2012; Moran & Moser, 2015; Moser, Moran & Leber, 2015). In the
additional singleton paradigm, observers locate a salient target defined by a unique
feature (i.e. a singleton, often a unique form) while simultaneously ignoring a more salient
distractor (often an item of a unique color). Using the additional singleton task, studies
have shown distractor interference to be exaggerated for individuals with high levels of
trait anxiety (Moran & Moser, 2015; Moser et al., 2012). These findings have been
interpreted as reflecting enhanced susceptibility in anxious individuals for distraction by
physically salient but irrelevant information. Using additional singleton paradigms, similar
findings have been reported for patients with other anxiety-related affective disorders
including depression (Bredemeier, Berenbaum, Brockmole, Boot, Simons & Most, 2011)
and posttraumatic stress disorder (PTSD; Esterman, Rosenberg & Noonan, 2013).
Collectively, these findings have provided strong support for an influential model
of anxiety known as the Attentional Control Theory (ACT; Derakshan & Eysenck, 2009
and Eysenck et al., 2007). According to ACT, individuals with high trait anxiety suffer from
a general deficit in top-down attentional control. This deficit is thought to primarily impair
inhibitory processing, increasing the influence of bottom-up attention and leading to a
greater sensitivity for irrelevant and distracting information. To compensate for this deficit,
high-anxiety individuals are hypothesized to invest additional processing resources,
76
thereby allowing them to reach a standard level of behavioural performance under certain
circumstances (Eysenck et al., 2007).
Recent models of anxiety have further proposed that the impaired inhibitory control
exhibited by high-anxiety individuals may be associated with the impoverished ability to
engage proactive attentional control (Braver et al., 2007; Braver, 2012; Aron, 2011).
Proactive attentional control is necessary for the active maintenance of goal-related
information and serves to facilitate preparatory attentional biasing. Such proactive control
can anticipate distractor representations and prevent them from interfering with the
processing of task-relevant information (Geng, 2014). Alternatively, high-anxiety
individuals are thought to engage attention in a reactive manner by exerting processing
resources as needed, typically after a salient-but-irrelevant stimulus is encountered. The
default use of reactive attention is thought to reflect a deficit in maintaining an active and
persistent top-down attentional set, leading to an increased distractibility in anxious
individuals.
While trait anxiety appears to disrupt the ability to ignore distracting information,
the neural correlates of this effect are not well understood. Differences in attentional
biases between high- and low-anxiety individuals may be due to a reduced ability to apply
active suppression to irrelevant stimuli, a greater reliance on reactive shifts of attention,
or both. To investigate this, individuals were initially screened using the State-Trait Anxiety
Inventory (STAI; Spielberger et al., 1983), a 40-item self-evaluation questionnaire
pertaining to anxiety affect. Those whose trait anxiety scores were among the highest and
lowest were selected to participate in the ERP experiment. To measure the neural
correlates of attention, EEG was recorded while subjects performed an additional
singleton search task identical to that previously used by Gaspar and McDonald (2014).
In this task participants searched for a color-singleton target and on 50% of trials
attempted to ignore a more salient color-singleton distractor (Figure 5.1). Participants were
instructed to indicate whether the orientation of the line inside the target was either
horizontal or vertical. The relationship between anxiety and i) target processing ERPs and
ii) distractor suppression processing ERPs were then separately assessed.
77
Figure 5.1. Trial Types. Example stimulus displays from the two experimental conditions. Subjects were instructed to attend to the yellow circle and to identify the orientation of the line inside of it. On 50% of trials, a salient distractor singleton was simultaneously presented within the display.
5.2. Methods
5.2.1. Materials and Methods
The Research Ethics Board at Simon Fraser University approved the research
protocol used in this study.
5.2.2. STAI Prescreen
In total, 218 students from Simon Fraser University volunteered to be prescreened
for potential inclusion into an EEG experiment. Students were prescreened using the
State-Trait Anxiety Inventory (STAI; Spielberger et al., 1983), a 40-item self-evaluation
questionnaire pertaining to anxiety affect. Subjects were contacted and invited to
participate in the full EEG experiment if their trait-anxiety score was above 50 (N = 20;
high-anxiety group) or below 35 (N = 20; low-anxiety group).
78
5.2.3. Participants
Forty students from Simon Fraser University participated after giving informed
consent. These students were given course credit for their participation as part of a
departmental research participation program. Prior to the EEG collection, subjects were
again asked to complete the STAI to ensure that they still fulfilled the predetermined
criteria for high- or low-anxiety. Of the 40 subjects, one was excluded due to excessive
noise in the ocular channels and another was excluded for failing to answer all questions
on the STAI. Of the remaining 38 participants, 19 (16 women, age 20.26, SD = 1.97; 1
left-handed) were characterized as high-anxiety and 19 (14 women, age 20.94, SD = 5.60;
4 left-handed) were characterized as low-anxiety. All subjects reported normal or
corrected-to-normal visual acuity and were tested for typical color vision, using Ishihara
color test plates.
5.2.4. Visual Search Task Stimuli and Apparatus
Stimuli were presented on a 23-inch, 120-Hz LCD monitor viewed from a distance
of 57 cm. Visual search arrays were comprised of 10 unfilled circles presented equidistant
(9.2°) from a central fixation point. Each circle was 3.4° in diameter with a 0.3° thick outline.
Eight or nine of the circles were uniformly colored non-targets, one was a target color
singleton, and one was a distractor color singleton (on distractor-present trials). The target
was dark yellow (x = 0.416, y = 0.519, 7.9 cd/m2) and the distractor was red (x = 0.640, y
= 0.324, 6.95 cd/m2), and the non-target circles were green (x = 0.288, y = 0.636, 7.85
cd/m2). A randomly oriented vertical or horizontal gray line (x = 0.295, y = 0.361, 7.89
cd/m2) was contained within each of the circles. All stimuli were presented on a uniform
black background (0.5 cd/m2).
5.2.5. Visual Search Task Procedure
On each trial, a search display was preceded by an 800-1,200 ms fixation period.
During this time only the central fixation point was visible. Upon the presentation of the
search display, participants were instructed to maintain fixation on the central point and to
identify the orientation of the gray line inside the target singleton by pressing one of two
79
response buttons as quickly as possible. The search array remained visible for 100 ms
after a response was registered, at which point the next trial began.
Displays contained a target singleton and one distractor singleton on 50% of trials
(distractor-present trials). On the remaining 50% of trials, the target was the only singleton
in the array (distractor-absent trials). Target and distractor locations were varied to
produce the following display configurations: lateral target, no distractor (22.0%); midline
target, no distractor (11.3%); lateral target, midline distractor (14.7%); lateral target,
ipsilateral distractor (14.7%); lateral target, contralateral distractor (14.7%); midline target,
lateral distractor (14.7%); midline target, midline distractor (8.0%). The order of the display
configurations was pseudo-randomly intermixed within each block of trials. Each
experimental block comprised 36 trials. At the end of the block, participants were given a
minimum 5-s rest period and were permitted to begin the next block whenever they
decided. The experiment contained 35 blocks, for a total of 1260 trials per participant. At
least 36 practice trials were given to each participant prior to the start of the experiment.
5.2.6. Behavioural Analysis
Median reaction times were derived for distractor-present and distractor-absent
trials for each participant. Trials on which the participant responded incorrectly, too quickly
(RT < 200 ms) or too slowly (RT > 1,500 ms) were excluded from the analysis. The means
of these median reaction times were then computed for both high- and low-anxiety groups.
Distractor interference was assessed by comparing the overall reaction times (RTs)
obtained on distractor present versus distractor absent trials. Group differences were
statistically assessed using a paired t-test.
Next, distractor proximity effects were assessed for both high- and low-anxiety
participants. This was accomplished by re-averaging distractor present trial RTs to lateral-
target displays according to the number of positions between target and distractor
singletons. The distance between the target and distractor singleton ranged from one (d1;
adjacent) to five (d5; four intervening non-targets) positions. The effect of the distractor
proximity on search performance was then analyzed in a 2 x 5 analysis of variance
(ANOVA) with factors of anxiety (high/low) and proximity (d1; d2; d3; d4; d5) tested.
80
5.2.7. Electrophysiological Recording and Analysis
EEG and electrooculogram (EOG) were recorded from active sintered Ag-AgCl
electrodes (Biosemi Active Two system) from 32 electrodes, using a modified montage
that included electrode sites FP1, FP2, AF3, AFZ, AF4, F7, F3, FZ, F4, F8, FC5, FCZ,
FC6, T7, C3, CZ, C4, T8, CP5, CPZ, CP6, P7, P3, PZ, P4, P8, PO7, POZ, PO8, O1, OZ,
O2. Horizontal EOGs were recorded using two electrodes positioned 1 cm lateral to the
external canthi, and vertical EOGs were recorded using two electrodes positioned above
and below the right eye. All EEG and EOG signals were digitized at 512 Hz, referenced in
real time to an active common-mode electrode, and low-pass filtered using a fifth-order
sinc filter with a -3 dB cutoff at 104 Hz. Electrode offsets were monitored to ensure the
quality of the data. After the data acquisition, EEG data for each channel were high-pass
filtered (-3 dB point at 0.05 Hz) and then converted from 24-bit to 12-bit integers.
EEG processing and ERP averaging were performed using event-related potential
software system (ERPSS) (University of California, San Diego). A semi-automated
procedure was used to discard epochs of EEG contaminated by blinks, eye movements,
or excessive noise (Green et al., 2008). Any trial with an artifact within a 1-s interval (-200-
800 ms post-stimulus) was rejected. Artifact-free epochs associated with the various
display configurations of interest were then averaged separately to create ERP
waveforms. The resulting ERPs were digitally low-pass filtered (-3 dB point at 32 Hz) and
digitally re-referenced to the average of the left and right mastoids. All ERP amplitudes
and baselines were computed using a 200 ms pre-stimulus window. The averaged event-
related horizontal EOGs did not exceed 2 μV for any individual participant, indicating their
gaze remained within 0.3° of the fixation point for a majority of the trials (McDonald &
Ward, 1999).
The primary analysis focused on ERPs elicited by search displays with following
display configurations: (1) lateral target, contralateral distractor; (2) lateral target, midline
distractor; (3) lateral target, ipsilateral distractor; (4) midline target, lateral distractor; (5)
lateral target, no distractor. As previously mentioned, displays containing a lateral
singleton and a midline singleton enable isolation of lateralized the N2pc and PD
components because these midline singletons do not trigger ERP lateralizations
(Woodman & Luck, 2003; Hickey et al., 2009).
81
For each participant, ERPs to the various search displays were collapsed across
left and right visual hemifields, as well as left and right electrodes, to produce waveforms
indexing the processing of a lateralized singleton. Lateralized ERP difference waveforms
were then computed for the display configurations of interest by subtracting the ipsilateral
waveform from the corresponding contralateral waveform at electrode sites P07 and P08.
As with all ERPs shown here, negative voltages were plotted upward so that the N2pc
would appear in these difference waveforms as an upward deflection and the PD as a
downward deflection. The mean amplitude of the PD was computed in a window 270-310
ms post stimulus onset, approximately centered around the peak of the component for
both groups. The 40 ms duration of the window was selected to mirror the size of the
statistical windows used in previous chapters but shifted later in time to account for later
onset observed here. Because the component appeared later than had been predicted a
priori, an unbiased signed area measurement was also used to confirm the presence of
the component. The signed positive area was measured within a 200-350 ms window and
subtracted from a baseline of equal duration.
On midline target, lateral distractor trials, mean amplitudes for an early N2pc was
computed in a 50 ms window from 170-220 ms. The early N2pc window used here was
identical to that used by Eimer and Kiss (2007) for testing early N2pcs to emotionally
salient stimuli. On lateral target, midline distractor trials, mean amplitudes for the N2pc
were computed in the same 230-290 ms post stimulus onset window used in the previous
chapters. All mean amplitudes were computed relative to a 100 ms pre-stimulus interval.
Finally, in addition to the mean amplitude measures, signed negative area was measured
within a 200-400 ms window from each individual participant’s contralateral-minus-
ipsilateral difference waveform. The signed negative area was then subtracted from a
baseline of equal duration, from -200-0 ms. The procedures used to compute signed area
in this chapter are identical to those described in Chapter 3 (Section 3.3.6.).
ERPs were also assessed separately for fast-response and slow-response trials.
Individual trials with RTs falling below or above the median RT for the display configuration
of interest were defined as fast-response and slow-response trials, respectively.
82
Split-half reliability of the N2pc and PD components were conducted by randomly
splitting the data into two halves and computing correlations of the half-data mean
amplitude averages for each component. All split-half correlations were corrected for using
the standard method (Anastasi & Urbina, 1997).
5.3. Results
5.3.1. STAI scores
Subjects were recruited for the EEG experiment subsequent to an initial screening
that determined their STAI trait-anxiety score. To maximize the power to detect potential
differences in brain responses, an extreme-groups design was used (Yarkoni et al., 2010).
Subjects were chosen for these groups contingent on their prescreen score on the STAI
trait-anxiety scale: high-anxiety individuals were defined as those scoring 50 or more and
low-anxiety individuals were defined as those scoring 35 and below. These thresholds
were chosen based on similar thresholds from other recent ERP studies of anxiety (Fox,
Derakshan & Leor, 2008). Prior to their participation in the EEG experiment, subjects were
again asked to complete the STAI. Mean trait anxiety was 62.37 (SD = 6.0) for the high-
anxiety individuals (N = 20) and 26.79 (SD = 3.5) for the low-anxiety individuals (N = 20).
5.3.2. Search performance does not differ between individuals with high- and low-anxiety individuals
Differences in behavioral performance between high- and low-anxiety groups was
tested using a repeated measure analysis of variance (ANOVA) with ‘trial type’ (distractor
present and distractor absent) and ‘group’ (high and low) as factors. The ANOVA revealed
that responses were faster for distractor absent (671 ms) relative to distractor present
trials [693 ms; F(1,36) = 114.10, p < .001]. Although low-anxiety individuals were
marginally faster than high-anxiety individuals on both distractor absent (664 vs. 678) and
distractor present (685 vs. 702) trials, this difference was not found to be statistically
significant [F(1,36) = .28, P = .65]. The magnitude of the interference effect (measured as
the RT difference between distractor present and distractor absent trials) was also not
83
found to differ between groups [t(18) = .46, P = .65]; RT interference was nearly identical
for both the high- and low-anxiety group (23 ms and 21 ms, respectively).
Search performance was also assessed as a function of target-distractor distance
for both high- and low-anxiety participants. Similar to what was reported in Chapter 2, RTs
were observed to mostly decrease as the target and distractor appeared farther from one
another (Figure 5.2). To examine the dependency of interference on target-distractor
distance for the different anxiety groups, RTs were submitted to an ANOVA with a within-
subject factors of Target-Distractor Distance (1, 2, 3, 4, 5; see Methods) and a between-
subject factor for anxiety (high, low). Overall, RTs were longer for nearby distractors than
distant distractors, resulting in significant main effects for Target-Distractor Distance
[F(4,36) = 47.43, P < .001]; however, no between subjects effect was observed for anxiety
[F(1,36) = .28, P = .60]. This suggests that although RTs differed as a function of the target
and distractor proximity, this RT effect did not differ between the high- and low-anxiety
group.
Figure 5.2 Target-distractor RT distance effects. Mean response times (across participants; in milliseconds) for five target-distractor distances (d1- d5) for both high- and low-anxiety individuals.
Lastly, RT standard deviations were computed to determine if response speed was
more variable among either group. RT standard deviation was not found to differ between
high- and low-anxiety participants for either distractor present or distractor absent trials (ts
< 0.66, ps > .52).
84
5.3.3. Suppression is preceded by an attentional shift to the distractor in high-anxiety individuals
Figured 5.3a shows grand averaged ERP waveforms for midline target, lateral
distractor trials for both high- and low-anxiety individuals. The PD was measured as the
difference in mean amplitude between contralateral and ipsilateral activity at electrodes
PO7/PO8 from 270 to 310 ms after the presentation of the search array. The mean PD
amplitudes for the lateral-target display configurations were found to differ significantly
from zero for both high- [t(18) = 2.25, P = .04] and low-anxiety [t(18) = 2.49, P = .02]
individuals. A split half reliability test found the PD to have a low but significant reliability
across the two halves of data (r = 0.39; P = 0.02).
The presence of the PD component was additionally confirmed by computing the
signed positive area within a 150 ms time window and subtracting it from an equally wide
pre-stimulus baseline (noise) interval (See methods). The PD was significant for both high-
[t(18) = 2.32, P = .03] and low-anxiety [t(18) = 2.65, P = .02] individuals. A follow-up t-test
further revealed that the PD did not differ across the high- and low-anxiety groups for either
mean amplitude [.47 µV vs. .55 µV; t(18) = .26, P = .80] or latency [278 ms vs. 273 ms; tc
= .35, P = .73].
For high-anxiety individuals, an N2pc was observed prior to the onset of the PD
component, with its early phase overlapping with the N1 component. Although beginning
quite early—at approximately 170 ms—this enhanced negativity is consistent with
previously reported early N2pc components (Eimer & Kiss, 2007; Gaspar & McDonald,
2014). The early N2pc was measured as the difference in mean amplitude between
contralateral and ipsilateral activity at electrodes PO7/PO8 from 170 to 220 ms after the
presentation of the search array. The mean N2pc amplitude for the lateral-target display
configuration was found to differ significantly from zero for high-anxiety [t(18) = 3.00, P =
.008] but not for low-anxiety [t(18) = .66, P = .52] individuals.
85
Figure 5.3 PD ERPs elicited by trials with displays containing a midline target and a lateral distractor. Time 0 reflects the onset of the search display, and negative voltage deflections are plotted above the x-axis, by convention. Waveforms were recorded over the lateral occipital scalp (electrodes PO7 and PO8). (A) ERPs recorded contralateral and ipsilateral to a distractor for high- and low-anxiety individuals. (B) Contralateral-minus-ipsilateral difference waveforms for high- and low-anxiety individuals.
In the high-anxiety group, the presence of an N2pc preceding the PD may indicate
that—after an initial shift of attention to the distractor singleton—a corrective mechanism
was invoked to suppress the distractor and reorient attention toward the target (see Geng,
2014). This may reflect a distinct search strategy among high-anxiety individuals, whereby
reactive, rather than proactive, mechanisms of attentional control are more readily invoked
during visual search (Braver, Gray & Burgess, 2007; Fales et al., 2008). However, an
alternative explanation is that high-anxiety individuals exhibit greater variability in their
capacity to maintain top-down attentional control which could lead to a different sequence
of processing on different trials. In line with this notion, it is plausible that the distractor
86
captured attention only on the most inefficient subset of trials, whereas on trials where
performance was optimal, this initial capture of attention did not occur and distractor
processing would closely resemble that of the low-anxiety group. To test these
possibilities, distractor processing in high-anxiety individuals was separately assessed for
both fast and slow-response trials. Figure 5.5 illustrates the difference in distractor
processing ERPs for high-anxiety individuals on fast- and slow-response trials. Both the
early N2pc and PD component were present on both the fastest and slowest half of trials.
Statistically, neither the N2pc nor the PD were observed to differ in amplitude when tested
in the same windows used above (ts < 1.21, P > .24). A similar analysis also found no
difference in PD amplitude for the low-anxiety group [t(18) = 1.09, P = .29].
Figure 5.4 High-anxiety group ERPs for displays containing a midline target and a lateral distractor, separately for fast- and slow-response trials. (A) ERPs recorded contralateral and ipsilateral to a distractor for fastest and slowest trials. (B) Contralateral-minus-ipsilateral difference waveforms.
87
5.3.4. Differences in target processing between high-anxiety and low-anxiety individuals
To assess the relationship between selective target processing and anxiety, target
N2pc waves were isolated for lateral target, distractor absent display configurations. Trials
where the distractor was absent were used to assess target processing here, as the N2pc
elicited on these trials would in no way be confounded by any attentional processing
associated with the salient distractor (see Chapter 2 for an explanation of how N2pc
amplitude can be modulated by distractor processing). Figured 5.5a shows grand
averaged ERP waveforms for lateral target, distractor absent trials for both high- and low-
anxiety individuals. The N2pc was measured as the difference in mean amplitude between
contralateral and ipsilateral activity at electrodes PO7/PO8 from 230 to 290 ms after the
presentation of the search array. The mean N2pc amplitudes for these lateral-target
display configurations were found to differ significantly from zero for both high- [t(18) =
5.57, p < .001] and low-anxiety [t(18) = 2.87, P = .01] individuals. An internal consistency
test found this N2pc to be highly reliable (r = 0.69; P < 0.001).
Figure 5.5b illustrates the difference in target processing ERPs for high- and low-
anxiety individuals. In addition to the conventional mean amplitude, the N2pc was isolated
from the waveform by computing the signed negative area within a 200 ms time window
and subtracted from an equally wide pre-stimulus baseline (noise) interval. Despite
appearing to differ in Figure 5.5b, neither the mean of the resultant N2pc area nor mean
amplitude were not found to reach statistical significance [t(18) = 1.702; P < .11; t(18) =
1.49; P = .15]2. The onset latency was also not found to differ between high- and low-
anxiety individuals [244 ms vs. 250 ms; tc = .95, P = .36].
2 It should be noted that the resultant mean amplitudes did not significantly differ due to a single
outlier in the high-anxiety group that had a large positivity in the N2pc time range. When the subject with the most extreme positive amplitude was removed from each group, this difference was found to reach significance [t(17) = 2.195, P = .04].
88
Figure 5.5 N2pc ERPs elicited by trials with displays containing a lateral target and no distractor. Time 0 reflects the onset of the search display, and negative voltage deflections are plotted above the x-axis, by convention. Waveforms were recorded over the lateral occipital scalp (electrodes PO7 and PO8). (A) ERPs recorded contralateral and ipsilateral to a target for high- and low-anxiety individuals. (B) Contralateral-minus-ipsilateral difference waveforms for high- and low-anxiety individuals.
To determine if response efficiency was associated with a unique sequence of
target processing, differences in target selection ERPs were examined separately for fast-
and slow-response trials. As can be seen in Figure 5.6, the N2pc component did not differ
for the high-anxiety group on fast- versus slow-response trials [t(18) = 0.05, P = .96]. In
contrast, the N2pc was observed to be markedly attenuated for the low-anxiety group on
slow-response trials [-0.24 vs. -0.87 μV; t(18) = 3.38, P = .003]. A reduction in the
amplitude of the N2pc component on slow response trials has been previously reported
by Jannati and colleagues (2013). Considered together, the results here indicate that
89
inefficient search is associated with a reduction in target processing for low-anxiety
individuals; however, no such relationship exists for the high-anxiety group.
Figure 5.6 ERPs for displays containing a midline target and a lateral distractor, separately for fast- and slow-response trials. (A) High-anxiety group ERPs recorded contralateral and ipsilateral to a distractor for fastest and slowest trials. (B) High anxiety group contralateral-minus-ipsilateral difference waveforms for high- and low-anxiety individuals.
5.4. Discussion
High levels of trait anxiety are associated with an increased sensitivity to threat-
related information, even when that information is known to be behaviourally
inconsequential (Bar-Haim et al., 2007). This negative attentional bias has been linked to
an impairment in the ability to filter out emotionally salient information (e.g., Ansari &
90
Derakshan, 2011a; 2011b; Berggren & Derakshan, 2013; Bishop, 2009; Pacheco-
Unguetti et al., 2010). Several recent studies have reported that this filtering impairment
in individuals with high-trait anxiety can also be observed to stimuli that have no affective
significance (e.g., Ansari & Derakshan, 2011a, 2011b; Berggren & Derakshan, 2013;
Bishop, 2009). For example, studies have found trait anxiety to effectively predict deficits
in filtering out irrelevant emotionally neutral stimuli, which led to their unnecessary storage
in visual working memory (Moriya & Sugiura, 2013; Qi et al., 2014; Stout, Shackman &
Larson, 2013). These findings seem to suggest that individuals with high trait anxiety may
have a more general deficit in attentional control. To date, however, the precise filtering
mechanism(s) used, and how their operation differs between high- and low-anxiety
individuals, remains unclear.
The main objective of the present study was to investigate whether the filtering
inefficiency in highly anxious individuals is related to the ability to actively suppress salient-
but-irrelevant distractor stimuli during a competitive visual search task. Specifically, the
question asked here was: does the suppressive processing indexed by the PD differ
between high- and low-anxiety individuals? To investigate this, distractor processing was
isolated in the ERPs by segregating trials where the distractor was the only lateralized
singleton in the search array (see previous chapters for further details regarding this
methodology). On these trials, distractor suppression was observed for all individuals: the
lateralized high-salience distractor stimulus was found to elicit a contralateral distractor
positivity in both low- and high-trait anxiety groups. Furthermore, the observed PD was not
found to differ in either latency or amplitude across the two groups. However, among
individuals with high trait anxiety, the PD was preceded by an initial early deflection in the
ERPs of the opposite polarity. This negativity—which began at approximately 170 ms after
the onset of the search display and continued until the onset of the PD—likely reflects an
early N2pc component, indicating that the distractor singleton initially captured attention.
Furthermore, the presence of the subsequent PD suggests that this initial capture of
attention was rapidly followed by the active suppression of the distractor location. Notably,
similar patterns of suppression (in which an N2pc is immediately followed by a PD
component) have been previously reported in several other studies (Jannati et al., 2013;
Sawaki et al., 2012; Sawaki & Luck, 2010). This sequence of events is thought to reflect
a reactive suppression mechanism. By this account, in situations in which a distracting
91
stimulus erroneously captures attention, a corrective mechanism is invoked to suppress
the location of the distractor, which in turn facilitates the selection of the target at another
location (see Geng, 2014).
Recently, models of anxiety have begun to distinguish between proactive and
reactive mechanisms of attentional control (Braver et al., 2007; Braver, 2012; Aron, 2011).
Whereas low-anxiety individuals are thought to engage top-down attentional control in a
sustained and proactive manner, high-anxiety individuals have been shown to rely more
on a reactive recruitment of attentional control (Braver, Gray & Burgess, 2007; Fales et
al., 2008). This notion is consistent with the findings reported here. Among the high-
anxiety group, a preceding shift of attention was observed to the distractor prior to it being
suppressed. This suggests that the inefficient filtering in high-anxiety individuals may not
stem from an inability to suppress distractor representations but rather from an inability to
proactively ignore them.
On first blush, the finding that the PD did not differ across groups appears to conflict
with ACT’s prediction that inhibitory processing is impaired in high-anxiety individuals; this
link between anxiety and inhibition is central to ACT and has been reported in a number
of studies (e.g., Derakshan et al., 2009; Eysenck and Byrne, 1992; Fox, 1993a; 1993b;
Wieser et al., 2009; Wood, Mathews, Dalgleish, 2001). However, one intriguing
explanation for this contradiction may relate to the different functional networks that govern
varying aspects of inhibitory processing and attentional control. Whereas the ventral
attention network is involved in stimulus-driven attentional orienting, the fronto-parietal
network is responsible for implementing increased levels of top-down cognitive control.
While anxiety is associated with various patterns of deficits for both networks (see
Sylvester et al., 2012 for a review), it may be the case that either i) the deficit is more
pronounced for the ventral network or ii) the compensatory effort observed in high-anxiety
individuals may primarily activate the fronto-parietal network. This suggests that activation
of fronto-parietal network regions may be necessary to implement the top-down active
suppression associated with the PD component. This would be consistent with the finding
that distractor suppression is closely correlated with activation of the dorsolateral
prefrontal cortex, a key brain region in the fronto-parietal network (Suzuki & Gottlieb, 2013)
92
that has been previously implicated in proactive attentional control (Braver & Barch 2006;
Braver et al. 2007).
In several the behavioral studies examining trait anxiety, reaction time measures
have been regarded as a principal index of processing efficiency (Eysenck et al., 2007).
However, others have criticized the validity of such measures, arguing reaction time to be
an indirect measure of the outcome of processing rather than a measure of the processing
itself (Basten, Stelzel & Fiebach, 2012; Eysenck & Derakshan, 2011). In the present study,
individuals with high trait anxiety were characterized by their inefficient processing of the
distractor at a neural level; however, this difference did not result in any decrement in
behavioural performance. Although reaction time and distractor interference costs were
observed, no differences were found for the high-anxiety group relative to the low-anxiety
group. This dissociation between neural and behavioural efficiency is not entirely
unexpected, as several previous studies have shown no effects of trait anxiety on
behaviour, yet significant effects of trait anxiety on neural processing as measured by EEG
and fMRI (e.g., Ansari & Derakshan, 2011a; Basten et al., 2012; Eysenck & Derakshan,
2011; Fales et al., 2008 Osinsky, Alexander, Gebhardt & Hennig, 2010). ACT accounts
for these findings, arguing that under some circumstances high-anxiety individuals may
show no behavioural evidence of disrupted attentional control; high-anxiety individuals
may be able to compensate for behavioural deficiencies by recruiting additional resources
and investing greater effort, allowing them to maintain a level of task performance on par
with their low-anxiety counterparts. Support for this idea comes from a study by Hayes
and colleagues (2009) that found anxiety to affect performance on a learning task when
learning is effortless; however, this behavioural deficit was eliminated when learning was
effortful and required higher motivation. In the additional singleton paradigm employed
here, where the target and distractor singleton remained fixed throughout, the cognitive
demands required to perform the search task would be considerably low. Thus, highly
anxious individuals may have been able to directly compensate for their potential
inefficiency in processing by recruiting additional cognitive resources to resolve the target.
Neuroimaging studies have substantiated the prediction that reduced efficiency in
high-anxiety individuals is associated with increased levels of processing (e.g., Basten,
Stelzel & Fiebach, 2011; Fales et al., 2008). For example, Santos, Wall, and Eysenck
93
(2006) found anxiety to be associated with greater activation in the right lateral prefrontal
cortex, an area of the brain implicated in shifting attention. To test the idea that high-
anxiety individuals recruit more attentional resources processing than low-anxiety
individuals, differences in the target N2pc amplitude were assessed here. Recently, it has
been proposed that the amplitude of the N2pc can be modulated by the allocation of top-
down attentional resources: during target selection, if more attentional resources are
directed towards the target, then a higher N2pc amplitude is observed (Liu et al., 2016).
The prediction here was that high-anxiety individuals would have an increased levels of
target processing relative to low-anxiety individuals. The mean area of the N2pc differed
across the two groups in the direction expected—high-anxiety individuals had a larger and
more sustained N2pc than did low-anxiety individuals—however, this difference did not
reach statistical significance (but likely would have with a larger sample; see Section
5.3.4). In addition, there was no reduction in N2pc amplitude on fast- relative to slow-
response trials for high-anxiety individuals, indicating that high-anxiety individuals devote
considerable attentional resources irrespective of their response efficiency. This may
suggest that among high-anxiety individuals, relative to their low-anxiety counterparts,
response efficiency is more readily associated with a deficit in post-selection processing
than that of selection itself. Further research on the link between anxiety and processing
efficiency is necessary given the results of previous studies are inconsistent and
controversial (e.g., Bishop, 2009; Wood et al., 2001).
94
General Discussion
Attention is a complex cognitive phenomenon that fills the divide between
perception and conscious experience. It is thought to be what “turns looking into seeing”
(pg. 1484, Carrasco, 2011). An ongoing goal of experimental psychology and
neuroscience has been to explore the many processes that comprise attention to gain a
better understanding of the role they play in our behaviour and cognition. The central
theme of this thesis has focused, perhaps counter-intuitively, on the processes that
mediate our ability to not pay attention. The ability to ignore irrelevant—but highly salient—
information is essential in allowing our limited cognitive resources to process only the
information that is relevant to our current goals. This thesis has focused primarily on an
event related potential component, the PD, which is thought to index this ability to ignore
distracting information. In the four chapters of experimental data presented herein, I have
attempted to expand our nascent understanding of the functional significance of the PD
and its role in attentional processing. In this final chapter I will briefly recapitulate these
findings in the context of the broader theoretical questions they address and propose
future directions for subsequent research.
6.1. Attentional capture revisited
The question of whether or not salient-but-irrelevant stimuli automatically capture
attention remains a contentious issue among vision scientists and has resulted in a
longstanding debate about the nature of attentional biasing. The more prominent models
of attentional control agree on a rigid dichotomy between top-down and bottom-up control,
but make diametrically different predictions regarding the extent to which volition can
counteract the influence of salience. On the one hand, the salience-driven selection
perspective maintains that attentional priority is driven entirely by the bottom-up features
of a stimulus, regardless of the intentions or expectations of the observer. On the other
hand, the contingent-capture perspective maintains that volition serves to guide early
95
visual processing so that selection is always in line with the top-down attentional set
adopted by the observer.
The series of experiments presented in this thesis were constructed to be able to
test whether or not salient distractor singletons capture attention. Using search displays
containing a lateral distractor and midline target, salient distractor singletons would be
expected to elicit either an N2pc (in favour of bottom-up perspectives) or a PD (in favour
of top-down perspectives). Based on previous studies that have employed fixed-feature
additional-singleton search paradigms, salient-but-irrelevant distractors have been
repeatedly found to not capture attention (e.g., Burra & Kerzel, 2014; Gaspar & McDonald,
2014; Jannati et al., 2013). Although this finding is initially replicated in Chapter 2, it is then
immediately challenged in subsequent chapters where subject variability is considered.
For example, in Chapter 3, vWM capacity was observed to be a predictive factor for
whether or not an individual would be able to avoid distractor-driven capture. While those
identified to have high- and medium-levels of working memory performance were able to
actively suppress the salient distractor singleton, those who performed at a low level were
not. Salient distractor singletons were observed to elicit an N2pc for the individuals with
the poorest vWM performance, a group which comprised one-third of the total sample
collected.
Evidence for distractor-driven attentional capture was again reported in Chapter 5,
where individual differences in trait anxiety were linked to two distinct patterns of cognitive
control. Whereas individuals with low trait anxiety performed as anticipated, high-anxiety
individuals invoked a reactive search strategy when dealing with distractors. The ERP
data in Chapter 5 showed that high-anxiety individuals were initially captured by the
distractor singleton but could rapidly apply suppression in a corrective manner to facilitate
the selection of the target singleton. This signature of stimulus suppression has been
previously reported for additional singleton tasks where task difficulty is increased by
swapping target and distractor features across trials (Eimer et al., 2012; Jannati et al.,
2013; McDonald et al., 2013). This finding demonstrates that individuals with high anxiety
may exhibit a general difficulty initially filtering out salient information, forcing them to rely
on a more effortful reactive search strategy.
96
A central point of the findings presented throughout this thesis illustrate that
attentional filtering and top-down control are acutely susceptible to individual differences.
Findings such as these should be considered when conducting neuroimaging studies that
require subject averages be combined for statistical purposes. This is epitomized by the
different conclusions about suppression that are drawn in Chapter 2 versus Chapter 3.
The conclusion from Chapter 2 is that individuals can suppress salient distractors during
visual search; however, when many of these same subjects were grouped per their vWM
capacity, the conclusion changes to tell something different: low capacity individuals are
unable to suppress salient distractors during visual search.
Despite contentious debate for several decades, it is becoming increasing
apparent that purely top-down and purely bottom-up models of attentional selection are
insufficient to fully explain the complexity of selection influences. The results presented
here serve to underscore the limitation of a strict theoretical dichotomy. While an
individual's current behavioural goals and physical salience do certainly serve to bias an
object’s priority for selection, more integrative frameworks are necessary to fully
incorporate the varying selection influences and individual differences that can alter the
guidance of attention. With a growing number of studies that cannot be easily explained
by this traditional dichotomy, considering these more integrative models of attention will
undoubtedly serve to extend our knowledge of selection phenomena.
6.2. The PD is a measure of top-down signal suppression
One such integrated framework of selection with growing empirical support is
signal-suppression hypothesis of controlled attention capture (Sawaki & Luck, 2010).
According to this hypothesis, salient singletons are automatically indexed by the visual
system and will trigger an attend-to-me signal irrespective of an individual's behavioural
goals. If the singleton is consistent with an individual’s target template, attention will be
deployed to its location. If the singleton does not match the template, a top-down
suppression mechanism may be elicited to prevent the object from erroneously capturing
attention. The PD is believed to index this top-down suppression and has been an essential
tool for the development of the signal suppression hypothesis.
97
One overarching motivation of the experiments conducted for this thesis was to
further validate the role of the PD component as the top-down attentional processing that
indexes this suppression mechanism. In Chapter 2, I sought to address this question by
altering the salience of the distractor singleton to be highly- versus equally-salient relative
to the target singleton. If the signal suppression hypothesis was correct and the PD is a
measure of top-down suppression triggered by an attend-to-me signal, it was expected
that the PD would be strategically recruited to suppress only the highly salient distractors.
Alternatively, the PD could have reflected an automatic mechanism recruited to deal with
any unique distracting object, regardless of that object's attentional priority. In this case,
both high- and low-salience distractors would likely elicit the ERP component. Consistent
with the signal suppression hypothesis, I found the visual system worked to selectively
suppress high- but not low-salient distractor singletons. Thus, the findings here seemed
to suggest that although salient signals are prioritized for selection based on their bottom-
up activation, they can be actively suppressed in a top-down manner.
This question of whether the PD reflects top-down active suppression was
addressed again in Chapter 4. Here a visual search array was placed both inside and
outside the attentional blink and the question was asked: can the suppression indexed by
the PD occur during a disruption of attentional control? Similar to the predictions made in
Chapter 2, it was hypothesized that if the PD reflected a form of top-down active
suppression, it should be severely impaired when attentional control was disrupted.
Alternatively, if instead the PD reflected a bottom-up form of suppression (similar to the
lateral inhibition surrounding an attended object), it should not be affected by the
attentional blink. The results were again in favour of a top-down account, as the PD to be
observed only on trials where the search array appeared outside of the attentional blink.
Together the findings in Chapter 2 and Chapter 4 support the notion that the
attentional suppression indexed by the PD requires a substantive amount of top-down
control. One outstanding curiosity, however, is that, despite the PD not being elicited on
trials that appeared within the attentional blink, there was no ERP evidence that attention
was captured by the salient singleton: the distractor singleton did not elicit an N2pc on lag
2 trials. Although individuals were markedly faster to respond to search arrays where the
PD was present, it would be premature to link this behavioural difference to distractor
98
suppression. As there were no “distractor absent” trials, it is difficult to ascertain if the
attentional blink produced an increase in distractor interference or simply a difference in
terms of speed of processing. Future study will be necessary to better appreciate the
relationship between the behavioural findings and the ERP findings in Chapter 4.
6.3. What is the clinical value of the PD as an index of attentional processing?
6.3.1. ADHD
Researchers have already begun to use the PD to study how individuals with
attention-deficit/hyperactivity disorders (ADHD) direct their attention. For example, a
recent study by Wang and colleagues (2016) used the PD component to examine
differences in attentional suppression among children with ADHD. They reported children
with ADHD to elicit a smaller PD component overall and to be more susceptible to capture
by salient-but-irrelevant distractors. Furthermore, this decreased PD amplitude was found
to predict a higher level of severity for inattentive symptoms and poorer behavioural
performance in children with ADHD. Preliminary data from our lab further suggests that
similar deficits may also be present in subclinical populations that self-report higher
distractibility. As the primary symptoms of ADHD can manifest quite differently between
individuals, future studies should seek to elucidate the specific factors that may predict
impaired suppression, as well as selective target processing. A nuanced appreciation of
electrophysiological differences across the subtypes of ADHD may offer a more targeted
clinical approach for understanding the disorder and constructing effective individualized
therapies.
6.3.2. Aging and attention
Current research on the developmental trajectory for visual search performance
finds that search proficiency increases from childhood until young adulthood and then
ultimately declines late in life (e.g., Plude, Enns & Brodeur, 1994; Trick & Enns, 1998).
Although it has been well established that numerous aspects of attention tend to
deteriorate with age, the mechanisms that are primarily affected by this decline remain
99
elusive. Some studies have proposed that these age-related declines in performance may
be specifically associated with impaired top-down suppression (Gazzaley, Cooney,
Rissman & Esposito, 2005; Gazzaley, Clapp, Kelley, McEvoy, Knight & Esposito, 2007).
Future studies should seek to use the PD to explore whether this impaired suppression of
distracting information may serve as a useful index for differentiating the variable impact
of the aging process. The PD may prove a functional marker for identifying differences in
cognitive processing that could predict healthy aging and preserved top-down modulation
in older adults.
6.3.3. Molecular biology, genetics, and selective attention
Although the clear majority of research on selective attention has sought to
understand the neural mechanisms associated with processing task-relevant and task-
irrelevant stimuli, separate from this line of inquiry are other studies that have sought to
assess the role particular neurotransmitters have on attentional control. For example,
neurotransmitters, such as dopamine, norepinephrine, and acetylcholine are often
implicated in the control of attention. Neurons that produce these neurotransmitters—
found largely throughout the brainstem and midbrain nuclei—send dense projections to
the prefrontal cortex, where attentional control signals are generated, as well as to
posterior sensory regions, where ERP correlates of selective attention manifest. Although
the clinical significance of these neurotransmitter systems and attentional control has been
extensively researched, not much extant evidence has sought to directly relate differences
in neurotransmitter production to differences in neural signals. Future research might seek
to assess these relationships by examining how functional gene variants that affect
neurotransmission might also affect attentional neural processing. It would be interesting
to see if there exists a relationship between an individual's genotype, distraction, and the
PD component. Such a study would be the first of its kind, to link molecular genetics with
an attentional ERP component.
100
6.4. A proposed stream for visual processing
Figure 6.1. Adapted from Janatti et al., 2013, a proposed hypothetical processing stream thought to occur during the fixed-feature variant of the additional singleton search task. Listed below each stage is the ERP component associated with that level of processing.
Adapted from Figure 7 in Jannati et al. (2013), Figure 6.1 illustrates an updated
proposal for a hypothetical attentional processing stream during visual search. At the pre-
attentive stage, the visual system first processes an entire scene in mass parallel,
encoding all objects on a topographical salience map proportional to their sensory inputs.
These sensory inputs can over time be altered by cognitive inputs that reflect dimensional
weighting, training, selection history, etc... Next, at the attentive stage, the saliency map
is scanned and the objects that elicit the greatest activations generate “attend-to-me”
signals and are selected. Over time, attention has built up a top-down module/heuristic
that contains a functional template of what the target and the distractor are. The location
of singleton that matches the target template is selected for enhanced processing while
the location of the singleton that matches the distractor template is selected for
suppression. Only information located at the enhanced location can be subsequently
identified and consolidated in memory.
The proposed model relies on several theoretical assumptions on how attention
may work during a competitive visual search task. First, the model necessitates that
individuals be able to simultaneously select more than one spatial location at a given
moment. Evidence for the parallel allocation of attention to multiple spatial locations come
101
from a series of recent papers by Grubert and Eimer (2015; 2016), who report that—when
an observer is instructed to report multiple feature-specific targets—the attentional set can
be flexibly configured to select the items in parallel. Since the top-down attentional
templates necessary for visual search are assumed to be stored in visual working memory,
the upper limit of the number of objects that can be selected in parallel would likely
correspond to an individual’s working memory capacity. This notion is consistent with the
finding reported in Chapter 3, where the group of individuals with lowest vWM scores
(ranging from 1.60 to 2.13), failed to elicit the PD—that is, they failed to maintain a template
for two items (the target and the distractor).
Figure 6.2. Hypothetical resolving of a visual search task based on the input image shown. The stars (top) represent the stimuli’s activation on the saliency map, with increased brightness denoting greater salience. The saliency map is then scanned by attention and suppression/enhancement are applied contingent on top-down attentional templates.
Another assumption of this proposed model is that top-down control is applied both
at pre-attentive and attentive stages of processing, albeit in functionally distinct ways. At
a pre-attentive stage, top-down control can act to increase or decrease the sensitivity for
sensory processing based on low-level features (e.g., colours, orientations, intensities).
This form of top-down control would serve to up-weight and down-weight the feature
dimensions of behaviourally relevant and irrelevant objects, altering their representation
on the salience map. This is consistent with several other models of selection which
propose top-down control to have an impact prior to the construction of a master saliency
map (Aziz & Mertsching, 2008; Found & Müller, 1996; Hu, Xie, Ma, Chia & Rajan, 2004;
102
Itti & Koch, 2001). This form of top-down control would be more reflexive, more easily
reset, and more sensitive to aspects such as selection history. In contrast, top-down
control at the attentive stage would reflect a more specialized process that would serve to
determine the flow processing instructions. This form of top-down modulation would more
closely resemble a Labergian module of attentional control (LaBerge, 2002). Here, this
control module requires iterative feedback to instantiate the higher-order attentional
instruction, be it to enhance or suppress processing. After a number of trials, the task
becomes routine, the instructions are consolidated and the attention module activated.
Although a distinction between forms of top-down attentional control is consistent with
some limited empirical findings (e.g., Ganis & Kosslyn, 2007), future research will be
necessary to establish the scope and the influence for such mechanisms to bias
attentional selection.
103
References
Akyürek, E. G., Leszczyński, M. & Schubö, A. (2010). The temporal locus of the interaction between working memory consolidation and the attentional blink. Psychophysiology, 47(6), 1134-1141.
Anastasi, A. & Urbina, S. (1997). Psychological Testing (7th ed.). Upper Saddle River, N.J.: Prentice Hall.
Ansari, T. L. & Derakshan, N. (2011a). The neural correlates of cognitive effort in anxiety: Effects on processing efficiency. Biological Psychology,86(3), 337-348.
Ansari, T. L. & Derakshan, N. (2011b). The neural correlates of impaired inhibitory control in anxiety. Neuropsychologia, 49(5), 1146-1153.
Anderson, B. A., Laurent, P. A. & Yantis, S. (2011). Value-driven attentional capture. Proceedings of the National Academy of Sciences of the United States of America, 108(25), 10367-10371.
Aron, A. R. (2011). From reactive to proactive and selective control: developing a richer model for stopping inappropriate responses. Biological psychiatry, 69(12), e55-e68.
Awh, E., Belopolsky, A. V. & Theeuwes, J. (2012). Top-down versus bottom-up attentional control: a failed theoretical dichotomy. Trends in Cognitive Sciences, 16(8), 437-443.
Awh, E. & Jonides, J. (2001). Overlapping mechanisms of attention and spatial working memory. Trends in cognitive sciences, 5(3), 119-126.
Awh, E., Matsukura, M. & Serences, J. T. (2003). Top-down control over biased competition during covert spatial orienting. Journal of Experimental Psychology. Human Perception and Performance, 29(1), 52-63.
Awh, E. & Pashler, H. (2000). Evidence for split attentional foci. Journal of Experimental Psychology. Human Perception and Performance, 26(2), 834-846.
Awh, E., Vogel, E. K. & Oh, S. H. (2006). Interactions between attention and working memory. Neuroscience, 139(1), 201-208.
Awh, E. & Vogel, E. K. (2008). The bouncer in the brain. Nature neuroscience, 11(1), 5-6.
104
Aziz, M. Z. & Mertsching, B. (2008). Fast and robust generation of feature maps for region-based visual attention. IEEE transactions on image processing, 17(5), 633-644.
Bacon, W. F. & Egeth, H. E. (1994). Overriding stimulus-driven attentional capture. Perception & Psychophysics, 55(5), 485-496.
Bahcall, D. O. & Kowler, E. (1999). Attentional interference at small spatial separations. Vision research, 39(1), 71-86.
Bar-Haim, Y., Lamy, D., Pergamin, L., Bakermans-Kranenburg, M. J. & Van Ijzendoorn, M. H. (2007). Threat-related attentional bias in anxious and nonanxious individuals: a meta-analytic study. Psychological bulletin, 133(1), 1.
Basten, U., Stelzel, C. & Fiebach, C. J. (2011). Trait anxiety modulates the neural efficiency of inhibitory control. Journal of Cognitive Neuroscience,23(10), 3132-3145.
Basten, U., Stelzel, C. & Fiebach, C. J. (2012). Trait anxiety and the neural efficiency of manipulation in working memory. Cognitive, Affective & Behavioral Neuroscience, 12(3), 571-588.
Beck, D. M. & Kastner, S. (2005). Stimulus context modulates competition in human extrastriate cortex. Nature Neuroscience, 8(8), 1110-1116.
Beck, D. M. & Lavie, N. (2005). Look here but ignore what you see: effects of distractors at fixation. Journal of Experimental Psychology: Human Perception and Performance, 31(3), 592.
Belopolsky, A. V., Zwaan, L., Theeuwes, J. & Kramer, A. F. (2007). The size of an attentional window modulates attentional capture by color singletons. Psychonomic Bulletin & Review, 14(5), 934-938.
Berggren, N. & Derakshan, N. (2013). Attentional control deficits in trait anxiety: why you see them and why you don’t. Biological Psychology, 92(3), 440-446.
Berggren, N., Hutton, S. B. & Derakshan, N. (2011). The effects of self-report cognitive failures and cognitive load on antisaccade performance.Cognitive and affective control, 19.
Bishop, S. J. (2009). Trait anxiety and impoverished prefrontal control of attention. Nature neuroscience, 12(1), 92-98.
Bisley, J. W. & Goldberg, M. E. (2010). Attention, intention, and priority in the parietal lobe. Annual Review of Neuroscience, 33, 1-21.
105
Braver, T. S. (2012). The variable nature of cognitive control: a dual mechanisms framework. Trends in cognitive sciences, 16(2), 106-113.
Braver, T. S. & Barch, D. M. (2006). Extracting core components of cognitive control. Trends in cognitive sciences, 10(12), 529-532.
Braver, T. S., Gray, J. R. & Burgess, G. C. (2007). Explaining the many varieties of working memory variation: Dual mechanisms of cognitive control.Variation in working memory, 76-106.
Bredemeier, K., Berenbaum, H., Brockmole, J. R., Boot, W. R., Simons, D. J. & Most, S. B. (2012). A load on my mind: Evidence that anhedonic depression is like multi-tasking. Acta psychologica, 139(1), 137-145.
Brisson, B. & Jolicœur, P. (2007). Electrophysiological evidence of central interference in the control of visuospatial attention. Psychonomic Bulletin & Review, 14(1), 126-132.
Broadbent, D. E. (1957). A mechanical model for human attention and immediate memory. Psychological Review, 64(3), 205-215.
Bruce, N. D. B. & Tsotsos, J. K. (2009). Saliency, attention, and visual search: an information theoretic approach. Journal of Vision, 9(3), 5.1-24.
Burra, N. & Kerzel, D. (2014). The distractor positivity (Pd) signals lowering of attentional priority: evidence from event-related potentials and individual differences. Psychophysiology, 51(7), 685-696.
Caputo, G. & Guerra, S. (1998). Attentional selection by distractor suppression. Vision Research, 38(5), 669-689.
Carretié, L., Mercado, F., Hinojosa, J. A., Martín-Loeches, M. & Sotillo, M. (2004). Valence-related vigilance biases in anxiety studied through event-related potentials. Journal of Affective Disorders, 78(2), 119-130.
Cepeda, N. J., Cave, K. R., Bichot, N. P. & Kim, M. S. (1998). Spatial selection via feature-driven inhibition of distractor locations. Perception & Psychophysics, 60(5), 727-746.
Chelazzi, L., Miller, E. K., Duncan, J. & Desimone, R. (2001). Responses of neurons in macaque area V4 during memory-guided visual search. Cerebral Cortex , 11(8), 761-772.
Christie, G. J., Livingstone, A. C. & McDonald, J. J. (2014). Searching for inefficiency in visual search. Journal of cognitive neuroscience.
106
Corriveau, I., Fortier-Gauthier, U., Pomerleau, V. J., McDonald, J., Dell'Acqua, R. & Jolicoeur, P. (2012). Electrophysiological evidence of multitasking impairment of attentional deployment reflects target-specific processing, not distractor inhibition. International Journal of Psychophysiology, 86(2), 152-159.
de Fockert, J. W., Rees, G., Frith, C. D. & Lavie, N. (2001). The role of working memory in visual selective attention. Science, 291(5509), 1803-1806.
Dell'Acqua, R., Sessa, P., Jolicœur, P. & Robitaille, N. (2006). Spatial attention freezes during the attention blink. Psychophysiology, 43(4), 394-400.
Derakshan, N. & Eysenck, M. W. (2009). Anxiety, processing efficiency, and cognitive performance: New developments from attentional control theory. European Psychologist, 14(2), 168-176.
Derakshan, N., Nazanin, D., Ansari, T. L., Miles, H., Leor, S. & Eysenck, M. W. (2009). Anxiety, Inhibition, Efficiency, and Effectiveness. Experimental Psychology, 56(1), 48-55.
Desimone, R. & Duncan, J. (1995). Neural mechanisms of selective visual attention. Annual Review of Neuroscience, 18, 193-222.
Di Lollo, V., Kawahara, J.-I., Shahab Ghorashi, S. M. & Enns, J. T. (2005). The attentional blink: resource depletion or temporary loss of control? Psychological Research, 69(3), 191-200.
Dosher, B. A. & Lu, Z. L. (2000). Mechanisms of perceptual attention in precuing of location. Vision Research, 40(10-12), 1269-1292.
Eimer, M. (1996). The N2pc component as an indicator of attentional selectivity. Electroencephalography and Clinical Neurophysiology, 99(3), 225-234.
Eimer, M. & Kiss, M. (2007). Attentional capture by task-irrelevant fearful faces is revealed by the N2pc component. Biological Psychology, 74(1), 108-112.
Engle, R. W., Conway, A. R., Tuholski, S. W. & Shisler, R. J. (1995). A resource account of inhibition. Psychological Science, 6(2), 122-125.
Engle, R. W., & Kane, M. J. (2004). Executive attention, working memory capacity, and a two-factor theory of cognitive control. Psychology of learning and motivation, 44, 145-199.
Esterman, M., Rosenberg, M. D. & Noonan, S. K. (2014). Intrinsic fluctuations in sustained attention and distractor processing. The Journal of Neuroscience, 34(5), 1724-1730.
107
Eysenck, M. W. & Byrne, A. (1992). Anxiety and susceptibility to distraction. Personality and Individual Differences, 13(7), 793-798.
Eysenck, M. W. & Derakshan, N. (2011). New perspectives in attentional control theory. Personality and Individual Differences, 50(7), 955-960.
Eysenck, M. W., Derakshan, N., Santos, R. & Calvo, M. G. (2007). Anxiety and cognitive performance: attentional control theory. Emotion , 7(2), 336-353.
Fales, C. L., Barch, D. M., Rundle, M. M., Mintun, M. A., Snyder, A. Z., Cohen, J. D., ... & Sheline, Y. I. (2008). Altered emotional interference processing in affective and cognitive-control brain circuitry in major depression. Biological psychiatry, 63(4), 377-384.
Fecteau, J. H. & Munoz, D. P. (2006). Salience, relevance, and firing: a priority map for target selection. Trends in Cognitive Sciences, 10(8), 382-390.
Feldmann-Wüstefeld, T. & Schubö, A. (2013). Context homogeneity facilitates both distractor inhibition and target enhancement. Journal of Vision, 13(3).
Folk, C. L. & Remington, R. (1998). Selectivity in distraction by irrelevant featural singletons: evidence for two forms of attentional capture. Journal of Experimental Psychology. Human Perception and Performance, 24(3), 847-858.
Folk, C. L. & Remington, R. (2006). Top-down modulation of preattentive processing: Testing the recovery account of contingent capture. Visual Cognition, 14(4-8), 445-465.
Folk, C. L., Remington, R. W. & Johnston, J. C. (1992). Involuntary covert orienting is contingent on attentional control settings. Journal of Experimental Psychology. Human Perception and Performance, 18(4), 1030-1044.
Folk, C. L., Remington, R. W. & Wright, J. H. (1994). The structure of attentional control: contingent attentional capture by apparent motion, abrupt onset, and color. Journal of Experimental Psychology. Human Perception and Performance, 20(2), 317-329.
Fortier‐Gauthier, U., Dell'Acqua, R. & Jolicœur, P. (2013). The “red‐alert” effect in visual search: Evidence from human electrophysiology.Psychophysiology, 50(7), 671-679.
Fortier-Gauthier, U., Moffat, N., Dell'Acqua, R., McDonald, J. J. & Jolicœur, P. (2012). Contralateral cortical organisation of information in visual short-term memory: Evidence from lateralized brain activity during retrieval.Neuropsychologia, 50(8), 1748-1758.
108
Found, A. & Müller, H. J. (1996). Searching for unknown feature targets on more than one dimension: Investigating a “dimension-weighting” account. Perception & Psychophysics, 58(1), 88-101.
Fox, E. (1993a). Allocation of visual attention and anxiety. Cognition and Emotion, 7(2), 207-215.
Fox, E. (1993b). Attentional bias in anxiety: selective or not?. Behaviour research and therapy, 31(5), 487-493.
Fox, E., Derakshan, N. & Shoker, L. (2008). Trait anxiety modulates the electrophysiological indices of rapid spatial orienting towards angry faces.Neuroreport, 19(3), 259-263.
Fox, E., Russo, R. & Georgiou, G. A. (2005). Anxiety modulates the degree of attentive resources required to process emotional faces. Cognitive, Affective & Behavioral Neuroscience, 5(4), 396-404.
Fukuda, K. & Vogel, E. K. (2009). Human variation in overriding attentional capture. The Journal of Neuroscience, 29(27), 8726-8733.
Fukuda, K. & Vogel, E. K. (2011). Individual differences in recovery time from attentional capture. Psychological science, 22(3), 361-368.
Gaspar, J. M. & McDonald, J. J. (2014). Suppression of salient objects prevents distraction in visual search. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 34(16), 5658-5666.
Gaspelin, N., Leonard, C. J. & Luck, S. J. (2015). Direct Evidence for Active Suppression of Salient-but-Irrelevant Sensory Inputs. Psychological Science, 26(11), 1740-1750.
Gazzaley, A., Clapp, W., Kelley, J., McEvoy, K., Knight, R. T. & D'Esposito, M. (2008). Age-related top-down suppression deficit in the early stages of cortical visual memory processing. Proceedings of the National Academy of Sciences, 105(35), 13122-13126.
Gazzaley, A., Cooney, J. W., Rissman, J. & D'Esposito, M. (2005). Top-down suppression deficit underlies working memory impairment in normal aging. Nature neuroscience, 8(10), 1298-1300.
Geng, J. J. (2014). Attentional mechanisms of distractor suppression.Current Directions in Psychological Science, 23(2), 147-153.
Ghorashi, S. M., Smilek, D. & Di Lollo, V. (2007). Visual search is postponed during the attentional blink until the system is suitably reconfigured. Journal of Experimental psychology: Human perception and performance, 33(1), 124.
109
Green, J. J., Conder, J. A. & McDonald, J. J. (2008). Lateralized frontal activity elicited
by attention‐directing visual and auditory cues.Psychophysiology, 45(4), 579-587.
Grubert, A. & Eimer, M. (2015). Rapid parallel attentional target selection in single-color and multiple-color visual search. Journal of Experimental Psychology: Human Perception and Performance, 41(1), 86.
Grubert, A. & Eimer, M. (2016). All Set, Indeed! N2pc Components Reveal Simultaneous Attentional Control Settings for Multiple Target Colors.
Gulbinaite, R., Johnson, A., de Jong, R., Morey, C. C., & van Rijn, H. (2014). Dissociable mechanisms underlying individual differences in visual working memory capacity. Neuroimage, 99, 197-206.
Hardin, M. G., Mandell, D., Mueller, S. C., Dahl, R. E., Pine, D. S. & Ernst, M. (2009). Inhibitory control in anxious and healthy adolescents is modulated by incentive and incidental affective stimuli. Journal of Child Psychology and Psychiatry, 50(12), 1550-1558.
Hasher, L., Lustig, C., & Zacks, R. T. (2007). Inhibitory mechanisms and the control of attention. Variation in working memory, 19, 227-249.
Hickey, C., Chelazzi, L. & Theeuwes, J. (2010). Reward changes salience in human vision via the anterior cingulate. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 30(33), 11096-11103.
Hickey, C., Di Lollo, V. & McDonald, J. J. (2009). Electrophysiological indices of target and distractor processing in visual search. Journal of Cognitive Neuroscience, 21(4), 760-775.
Hickey, C., McDonald, J. J. & Theeuwes, J. (2006). Electrophysiological evidence of the capture of visual attention. Journal of Cognitive Neuroscience, 18(4), 604-613.
Hickey, C., Olivers, C., Meeter, M. & Theeuwes, J. (2011). Feature priming and the capture of visual attention: Linking two ambiguity resolution hypotheses. Brain research, 1370, 175-184.
Hickey, C. & Theeuwes, J. (2011). Context and competition in the capture of visual attention. Attention, Perception & Psychophysics, 73(7), 2053-2064.
Hilimire, M. R., Hickey, C. & Corballis, P. M. (2012). Target resolution in visual search involves the direct suppression of distractors: evidence from electrophysiology. Psychophysiology, 49(4), 504-509.
Hopfinger, J. B., Buonocore, M. H. & Mangun, G. R. (2000). The neural mechanisms of top-down attentional control. Nature Neuroscience, 3(3), 284-291.
110
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. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 24(8), 1822-1832.
Hopf, J. M., Luck, S. J., Girelli, M., Hagner, T., Mangun, G. R., Scheich, H. & Heinze, H. J. (2000). Neural sources of focused attention in visual search. Cerebral Cortex , 10(12), 1233-1241.
Hu, Y., Xie, X., Ma, W. Y., Chia, L. T. & Rajan, D. (2004, November). Salient region detection using weighted feature maps based on the human visual attention model. In Pacific-Rim Conference on Multimedia (pp. 993-1000). Springer Berlin Heidelberg.
Hutton, S. B. & Ettinger, U. (2006). The antisaccade task as a research tool in psychopathology: a critical review. Psychophysiology, 43(3), 302-313.
Ipata, A. E., Gee, A. L., Gottlieb, J., Bisley, J. W. & Goldberg, M. E. (2006). LIP responses to a popout stimulus are reduced if it is overtly ignored. Nature Neuroscience, 9(8), 1071-1076.
Itti, L. & Koch, C. (2001). Computational modelling of visual attention. Nature Reviews. Neuroscience, 2(3), 194-203.
Jannati, A., Gaspar, J. M. & McDonald, J. J. (2013). Tracking target and distractor processing in fixed-feature visual search: Evidence from human electrophysiology. Journal of Experimental Psychology. Human Perception and Performance, 39(6), 1713-1730.
Jannati, A., McDonald, J. J. & Di Lollo, V. (2015). Individual differences in rate of encoding predict estimates of visual short-term memory capacity (K).Canadian Journal of Experimental Psychology/Revue canadienne de psychologie expérimentale, 69(2), 213.
Jensen, A. R. (1992). The importance of intraindividual variation in reaction time. Personality and individual Differences, 13(8), 869-881.
Jolicoeur, P. (1999). Concurrent response-selection demands modulate the attentional blink. Journal of Experimental psychology: Human perception and performance, 25(4), 1097.
Jolicœur, P., Sessa, P., Dell’Acqua, R. & Robitaille, N. (2006a). On the control of visual spatial attention: Evidence from human electrophysiology.Psychological research, 70(6), 414-424.
111
Jolicœur, P., Sessa, P., Dell'Acqua, R. & Robitaille, N. (2006b). Attentional control and capture in the attentional blink paradigm: Evidence from human electrophysiology. European Journal of Cognitive Psychology, 18(4), 560-578.
Kahneman, D., Treisman, A. & Burkell, J. (1983). The cost of visual filtering. Journal of Experimental Psychology. Human Perception and Performance, 9(4), 510-522.
Kane, M. J., Conway, A. R., Hambrick, D. Z., & Engle, R. W. (2007). Variation in working memory capacity as variation in executive attention and control. Variation in working memory, 1, 21-48.
Kastner, S., Pinsk, M. A., De Weerd, P., Desimone, R. & Ungerleider, L. G. (1999). Increased activity in human visual cortex during directed attention in the absence of visual stimulation. Neuron, 22(4), 751-761.
Kastner, S. & Ungerleider, L. G. (2001). The neural basis of biased competition in human visual cortex. Neuropsychologia, 39(12), 1263-1276.
Keppel, G. & Wickens, T. D. (2004). Design and analysis: A researcher’s handbook (4th ed.). Upper Saddle River, NJ: Pearson-Prentice Hall.
Kiesel, A., Miller, J., Jolicœur, P. & Brisson, B. (2008). Measurement of ERP latency differences: A comparison of single‐participant and jackknife‐based scoring methods. Psychophysiology, 45(2), 250-274.
Kiss, M., Grubert, A., Petersen, A. & Eimer, M. (2012). Attentional capture by salient distractors during visual search is determined by temporal task demands. Journal of Cognitive Neuroscience, 24(3), 749-759.
Kiss, M., Jolicoeur, P., Dell’acqua, R. & Eimer, M. (2008). Attentional capture by visual singletons is mediated by top-down task set: new evidence from the N2pc component. Psychophysiology, 45(6), 1013-1024.
Krummenacher, J., Müller, H. J., Zehetleitner, M. & Geyer, T. (2009). Dimension- and space-based intertrial effects in visual pop-out search: modulation by task demands for focal-attentional processing. Psychological Research, 73(2), 186-197.
LaBerge, D. (2002). Attentional control: brief and prolonged. Psychological research, 66(4), 220-233.
Lagroix, H. E., Grubert, A., Spalek, T. M., Di Lollo, V. & Eimer, M. (2015). Visual search is postponed during the period of the AB: An event‐related potential study. Psychophysiology, 52(8), 1031-1038.
112
Lavie, N. (1995). Perceptual load as a necessary condition for selective attention. Journal of Experimental Psychology: Human perception and performance, 21(3), 451.
Lavie, N. (2000). Selective attention and cognitive control: Dissociating attentional functions through different types of load. Attention and performance XVIII, 175-194.
Lavie, N., Hirst, A., De Fockert, J. W. & Viding, E. (2004). Load theory of selective attention and cognitive control. Journal of Experimental Psychology: General, 133(3), 339.
Lavie, N. & Cox, S. (1997). On the efficiency of visual selective attention: Efficient visual search leads to inefficient distractor rejection. Psychological Science, 8(5), 395-396.
Lavie, N. & De Fockert, J. (2005). The role of working memory in attentional capture. Psychonomic bulletin & review, 12(4), 669-674.
Lavie, N. & Fox, E. (2000). The role of perceptual load in negative priming.Journal of Experimental Psychology: Human Perception and Performance,26(3), 1038.
Lavie, N. & Tsal, Y. (1994). Perceptual load as a major determinant of the locus of selection in visual attention. Perception & Psychophysics, 56(2), 183-197.
Leblanc, E., Prime, D. J. & Jolicoeur, P. (2008). Tracking the location of visuospatial attention in a contingent capture paradigm. Journal of Cognitive Neuroscience, 20(4), 657-671.
Li, Z. (2002). A saliency map in primary visual cortex. Trends in cognitive sciences, 6(1), 9-16.
Liu, T., Abrams, J. & Carrasco, M. (2009). Voluntary attention enhances contrast appearance. Psychological Science, 20(3), 354-362.
Liu, Q., Li, H., Campos, J. L., Wang, Q., Zhang, Y., Qiu, J., ... & Sun, H. J. (2009). The N2pc component in ERP and the lateralization effect of language on color perception. Neuroscience letters, 454(1), 58-61.
Liu, T., Pestilli, F. & Carrasco, M. (2005). Transient attention enhances perceptual performance and fMRI response in human visual cortex. Neuron,45(3), 469-477.
Luck, S. J. (2012). Electrophysiological correlates of the focusing of attention within complex visual scenes: N2pc and related ERP components. The Oxford handbook of event-related potential components, 329-360.
Luck, S. J. (2014). An Introduction to the Event-Related Potential Technique. MIT Press.
113
Luck, S. J., Girelli, M., McDermott, M. T. & Ford, M. A. (1997). Bridging the gap between monkey neurophysiology and human perception: An ambiguity resolution theory of visual selective attention. Cognitive psychology, 33(1), 64-87.
Luck, S. J. & Hillyard, S. A. (1994a). Electrophysiological correlates of feature analysis during visual search. Psychophysiology, 31(3), 291-308.
Luck, S. J. & Hillyard, S. A. (1994b). Spatial filtering during visual search: evidence from human electrophysiology. Journal of Experimental Psychology. Human Perception and Performance, 20(5), 1000-1014.
Luck, S. J., Hillyard, S. A., Mouloua, M. & Hawkins, H. L. (1996). Mechanisms of visual-spatial attention: Resource allocation or uncertainty reduction?. Journal of Experimental Psychology: Human Perception and Performance, 22(3), 725.
Luck, S. J. & Vogel, E. K. (1997). The capacity of visual working memory for features and conjunctions. Nature, 390(6657), 279-281.
Maljkovic, V. & Nakayama, K. (1994). Priming of pop-out: I. Role of features. Memory & Cognition, 22(6), 657-672.
Markett, S. A., Montag, C. & Reuter, M. (2010). The association between dopamine DRD2 polymorphisms and working memory capacity is modulated by a functional polymorphism on the nicotinic receptor gene CHRNA4.Journal of cognitive neuroscience, 22(9), 1944-1954.
Martínez-Trujillo, J. & Treue, S. (2002). Attentional modulation strength in cortical area MT depends on stimulus contrast. Neuron, 35(2), 365-370.
Mazza, V., Turatto, M. & Caramazza, A. (2009). Attention selection, distractor suppression and N2pc. Cortex; a Journal Devoted to the Study of the Nervous System and Behavior, 45(7), 879-890.
McDonald, J. J., Green, J. J., Jannati, A. & Di Lollo, V. (2013). On the electrophysiological evidence for the capture of visual attention. Journal of Experimental Psychology. Human Perception and Performance, 39(3), 849-860.
McNab, F. & Klingberg, T. (2008). Prefrontal cortex and basal ganglia control access to working memory. Nature Neuroscience, 11(1), 103-107.
McTeague, L. M., Shumen, J. R., Wieser, M. J., Lang, P. J. & Keil, A. (2011). Social vision: sustained perceptual enhancement of affective facial cues in social anxiety. Neuroimage, 54(2), 1615-1624.
Melnick, M. D., Harrison, B. R., Park, S., Bennetto, L., & Tadin, D. (2013). A strong interactive link between sensory discriminations and intelligence.Current Biology, 23(11), 1013-1017.
114
Miller, E. K., Gochin, P. M. & Gross, C. G. (1993). Suppression of visual responses of neurons in inferior temporal cortex of the awake macaque by addition of a second stimulus. Brain Research, 616(1-2), 25-29.
Moran, J. & Desimone, R. (1985). Selective attention gates visual processing in the extrastriate cortex. Science, 229(4715), 782-784.
Moran, T. P. & Moser, J. S. (2015). The color of anxiety: Neurobehavioral evidence for distraction by perceptually salient stimuli in anxiety. Cognitive, Affective & Behavioral Neuroscience, 15(1), 169-179.
Moriya, J. & Sugiura, Y. (2013). Socially anxious individuals with low working memory capacity could not inhibit the goal-irrelevant information.Frontiers in human neuroscience, 7, 840.
Moser, J. S., Becker, M. W. & Moran, T. P. (2012). Enhanced attentional capture in trait anxiety. Emotion, 12(2), 213.
Moser, J. S., Moran, T. P. & Leber, A. B. (2015). Manipulating attention to nonemotional distractors influences state anxiety: a proof-of-concept study in low-and high-anxious college students. Behavior therapy, 46(6), 834-843.
Motter, B. C. (1993). Focal attention produces spatially selective processing in visual cortical areas V1, V2, and V4 in the presence of competing stimuli. Journal of Neurophysiology, 70(3), 909-919.
Mounts, J. R. (2000). Attentional capture by abrupt onsets and feature singletons produces inhibitory surrounds. Perception & Psychophysics,62(7), 1485-1493.
Mounts, J. R. & Gavett, B. E. (2004). The role of salience in localized attentional interference. Vision research, 44(13), 1575-1588.
Mounts, J. R. (2005). Attentional selection: A salience-based competition for representation. Perception & psychophysics, 67(7), 1190-1198.
Müller, N. G., Mollenhauer, M., Rösler, A. & Kleinschmidt, A. (2005). The attentional field has a Mexican hat distribution. Vision research, 45(9), 1129-1137.
Navalpakkam, V. & Itti, L. (2006). Top-down attention selection is fine grained. Journal of Vision, 6(11), 1180-1193.
O’Connor, D. H., Fukui, M. M., Pinsk, M. A. & Kastner, S. (2002). Attention modulates responses in the human lateral geniculate nucleus. Nature Neuroscience, 5(11), 1203-1209.
O’Craven, K. M., Downing, P. E. & Kanwisher, N. (1999). fMRI evidence for objects as the units of attentional selection. Nature, 401(6753), 584-587.
115
Olivers, C. N. & Meeter, M. (2008). A boost and bounce theory of temporal attention. Psychological review, 115(4), 836.
Olivers, C. N. L., Meijer, F. & Theeuwes, J. (2006). Feature-based memory-driven attentional capture: visual working memory content affects visual attention. Journal of Experimental Psychology. Human Perception and Performance, 32(5), 1243-1265.
Osinsky, R., Alexander, N., Gebhardt, H. & Hennig, J. (2010). Trait anxiety and dynamic adjustments in conflict processing. Cognitive, Affective & Behavioral Neuroscience, 10(3), 372-381.
Pacheco-Unguetti, A. P., Acosta, A., Callejas, A. & Lupiáñez, J. (2010). Attention and anxiety different attentional functioning under state and trait anxiety. Psychological science, 21(2), 298-304.
Pashler, H. E. & Sutherland, S. (1998). The psychology of attention (Vol. 15). Cambridge, MA: MIT press.
Phillips, W.A. (1974). On the distinction between sensory storage and short-term visual memory. Perception & Psychophysics, 16(2), 283–290.
Pinto, Y., Olivers, C. N. L. & Theeuwes, J. (2005). Target uncertainty does not lead to more distraction by singletons: intertrial priming does. Perception & Psychophysics, 67(8), 1354-1361.
Pomerleau, V. J., Fortier-Gauthier, U., Corriveau, I., McDonald, J. J., Dell’Acqua, R. & Jolicœur, P. (2014). The attentional blink freezes spatial attention allocation to targets, not distractors: Evidence from human electrophysiology. Brain Research, 1559, 33-45.
Posner, M. I. (1980). Orienting of attention. The Quarterly Journal of Experimental Psychology, 32(1), 3-25.
Pourtois, G., Grandjean, D., Sander, D. & Vuilleumier, P. (2004). Electrophysiological correlates of rapid spatial orienting towards fearful faces. Cerebral Cortex, 14(6), 619-633.
Qi, S., Chen, J., Hitchman, G., Zeng, Q., Ding, C., Li, H. & Hu, W. (2014). Reduced representations capacity in visual working memory in trait anxiety.Biological psychology, 103, 92-99.
Raymond, J. E., Shapiro, K. L. & Arnell, K. M. (1992). Temporary suppression of visual processing in an RSVP task: An attentional blink?.Journal of experimental psychology: Human perception and performance,18(3), 849.
116
Reynolds, J. H., Chelazzi, L. & Desimone, R. (1999). Competitive mechanisms subserve attention in macaque areas V2 and V4. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 19(5), 1736-1753.
Reynolds, J. H. & Desimone, R. (1999). The role of neural mechanisms of attention in solving the binding problem. Neuron, 24(1), 19-29, 111-25.
Reynolds, J. H. & Desimone, R. (2003). Interacting roles of attention and visual salience in V4. Neuron, 37(5), 853-863.
Reynolds, J. H. & Heeger, D. J. (2009). The normalization model of attention. Neuron, 61(2), 168-185.
Reynolds, J. H., Pasternak, T. & Desimone, R. (2000). Attention increases sensitivity of V4 neurons. Neuron, 26(3), 703-714.
Ruthruff, E. & Pashler, H. (2001). Perceptual and central interference in dual-task performance. The limits of attention: Temporal constraints in human information processing, 100-123.
Rutman, A. M., Clapp, W. C., Chadick, J. Z., & Gazzaley, A. (2010). Early top–down control of visual processing predicts working memory performance. Journal of Cognitive Neuroscience, 22(6), 1224-1234.
Saenz, M., Buracas, G. T. & Boynton, G. M. (2002). Global effects of feature-based attention in human visual cortex. Nature Neuroscience, 5(7), 631-632.
Sawaki, R. & Luck, S. J. (2010). Capture versus suppression of attention by salient singletons: Electrophysiological evidence for an automatic attend-to-me signal. Attention, Perception & Psychophysics, 72(6), 1455-1470.
Sawaki, R. & Luck, S. J. (2011). Active suppression of distractors that match the contents of visual working memory. Visual Cognition, 19(7), 956-972.
Sawaki, R., Geng, J. J. & Luck, S. J. (2012). A common neural mechanism for preventing and terminating the allocation of attention. The Journal of Neuroscience, 32(31), 10725-10736.
Seidl, K. N., Peelen, M. V. & Kastner, S. (2012). Neural evidence for distracter suppression during visual search in real-world scenes. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 32(34), 11812-11819.
Serences, J. T. & Yantis, S. (2006). Selective visual attention and perceptual coherence. Trends in Cognitive Sciences, 10(1), 38-45.
117
Serences, J. T., Yantis, S., Culberson, A. & Awh, E. (2004). Preparatory activity in visual cortex indexes distractor suppression during covert spatial orienting. Journal of Neurophysiology, 92(6), 3538-3545.
Shapiro, K. L., Raymond, J. E. & Arnell, K. M. (1997). The attentional blink.Trends in cognitive sciences, 1(8), 291-296.
Sperling, G. (1960). The information available in brief visual presentations. Psychological Monographs: General and Applied, 74(11), 1.
Spielberger, C. D., Gorsuch, R. L., Lushene, R., Vagg, P. R. & Jacobs, G. A. (1983). Consulting Psychologists Press; Palo Alto, CA: 1983. Manual for the state-trait anxiety inventory.
Spitzer, H., Desimone, R. & Moran, J. (1988). Increased attention enhances both behavioral and neuronal performance. Science, 240(4850), 338-340.
Sreenivasan, K. K., & Jha, A. P. (2007). Selective attention supports working memory maintenance by modulating perceptual processing of distractors. Journal of cognitive neuroscience, 19(1), 32-41.
Sreenivasan, K. K., Katz, J., & Jha, A. P. (2007). Temporal characteristics of top-down modulations during working memory maintenance: an event-related potential study of the N170 component. Journal of Cognitive Neuroscience, 19(11), 1836-1844.
Stahl, J. & Gibbons, H. (2004). The application of jackknife‐based onset detection of lateralized readiness potential in correlative approaches.Psychophysiology, 41(6), 845-860.
Stout, D. M., Shackman, A. J. & Larson, C. L. (2013). Failure to filter: anxious individuals show inefficient gating of threat from working memory.
Suzuki, M. & Gottlieb, J. (2013). Distinct neural mechanisms of distractor suppression in the frontal and parietal lobe. Nature neuroscience, 16(1), 98-104.
Sylvester, C. M., Corbetta, M., Raichle, M. E., Rodebaugh, T. L., Schlaggar, B. L., Sheline, Y. I., ... & Lenze, E. J. (2012). Functional network dysfunction in anxiety and anxiety disorders. Trends in neurosciences, 35(9), 527-535.
Taatgen, N. A., Juvina, I., Schipper, M., Borst, J. P. & Martens, S. (2009). Too much control can hurt: A threaded cognition model of the attentional blink. Cognitive psychology, 59(1), 1-29.
Tan, M. & Wyble, B. (2015). Understanding how visual attention locks on to a location: Toward a computational model of the N2pc component. Psychophysiology, 52(2), 199-213.
118
Theeuwes, J. (1991). Cross-dimensional perceptual selectivity. Perception & Psychophysics, 50(2), 184-193.
Theeuwes, J. (1992). Perceptual selectivity for color and form. Perception & Psychophysics, 51(6), 599-606.
Theeuwes, J. (1994). Endogenous and exogenous control of visual selection. Perception, 23(4), 429-440.
Theeuwes, J. (2010). Top-down and bottom-up control of visual selection. Acta Psychologica, 135(2), 77-99.
Theeuwes, J., Atchley, P. & Kramer, A. F. (2000). On the time course of top-down and bottom-up control of visual attention. Control of cognitive processes: Attention and performance XVIII, 105-124.
Theeuwes, J., Kramer, A. F. & Kingstone, A. (2004). Attentional capture modulates perceptual sensitivity. Psychonomic Bulletin & Review, 11(3), 551-554.
Theeuwes, J. & Belopolsky, A. V. (2012). Reward grabs the eye: Oculomotor capture by rewarding stimuli. Vision Research, 74, 80-85.
Thompson, K. G. & Bichot, N. P. (2005). A visual salience map in the primate frontal eye field. Progress in Brain Research, 147, 251-262.
Töllner, T., Zehetleitner, M., Gramann, K. & Müller, H. J. (2010). Top-down weighting of visual dimensions: behavioral and electrophysiological evidence. Vision Research, 50(14), 1372-1381.
Treisman, A. M. & Gelade, G. (1980). A feature-integration theory of attention. Cognitive Psychology, 12(1), 97-136.
Tsotsos, J. K. (1990). Analyzing vision at the complexity level. The Behavioral and Brain Sciences, 13(03), 423-445.
Ulrich, R. & Miller, J. (2001). Using the jackknife-based scoring method for measuring LRP onset effects in factorial designs. Psychophysiology, 38(05), 816-827.
Usher, M. & Niebur, E. (1996). Modeling the Temporal Dynamics of IT Neurons in Visual Search: A Mechanism for Top-Down Selective Attention. Journal of Cognitive Neuroscience, 8(4), 311-327.
Vogel, E. K., McCollough, A. W. & Machizawa, M. G. (2005). Neural measures reveal individual differences in controlling access to working memory. Nature, 438(7067), 500-503.
119
Wadlinger, H. A. & Isaacowitz, D. M. (2010). Fixing our focus: Training attention to regulate emotion. Personality and Social Psychology Review.
Wang, E., Sun, L., Sun, M., Huang, J., Tao, Y., Zhao, X., ... & Wang, Y. (2016). Attentional Selection and Suppression in Children With Attention-Deficit/Hyperactivity Disorder. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging.
Wieser, M. J., Pauli, P. & Mühlberger, A. (2009). Probing the attentional control theory in social anxiety: An emotional saccade task. Cognitive, Affective & Behavioral Neuroscience, 9(3), 314-322.
Wolfe, J. M. (1994). Guided search 2.0 a revised model of visual search.Psychonomic bulletin & review, 1(2), 202-238.
Wolfe, J. M., Cave, K. R. & Franzel, S. L. (1989). Guided search: an alternative to the feature integration model for visual search. Journal of Experimental Psychology. Human Perception and Performance, 15(3), 419-433.
Wood, J., Mathews, A. & Dalgleish, T. (2001). Anxiety and cognitive inhibition. Emotion, 1(2), 166.
Woodman, G. F. (2010). A brief introduction to the use of event-related potentials in studies of perception and attention. Attention, Perception & Psychophysics, 72(8), 2031-2046.
Woodman, G. F. & Luck, S. J. (1999). Electrophysiological measurement of rapid shifts of attention during visual search. Nature, 400(6747), 867-869.
Woodman, G. F. & Luck, S. J. (2003). Serial deployment of attention during visual search. Journal of Experimental Psychology. Human Perception and Performance, 29(1), 121-138.
Wykowska, A. & Schubö, A. (2010). On the temporal relation of top-down and bottom-up mechanisms during guidance of attention. Journal of Cognitive Neuroscience, 22(4), 640-654.
Wykowska, A. & Schubö, A. (2011). Irrelevant singletons in visual search do not capture attention but can produce nonspatial filtering costs. Journal of Cognitive Neuroscience, 23(3), 645-660.
Yantis, S. & Johnson, D. N. (1990). Mechanisms of attentional priority. Journal of Experimental Psychology. Human Perception and Performance, 16(4), 812-825.
120
Yarkoni, T., Tal, Y. & Braver, T. S. (2010). Cognitive Neuroscience Approaches to Individual Differences in Working Memory and Executive Control: Conceptual and Methodological Issues. In The Springer Series on Human Exceptionality (pp. 87-107).
Yeshurun, Y. & Carrasco, M. (1998). Attention improves or impairs visual performance by enhancing spatial resolution. Nature, 396(6706), 72-75.
Zanto, T. P. & Gazzaley, A. (2009). Neural suppression of irrelevant information underlies optimal working memory performance. The Journal of Neuroscience, 29(10), 3059-3066.
Zanto, T. P., Rubens, M. T., Thangavel, A. & Gazzaley, A. (2011). Causal role of the prefrontal cortex in top-down modulation of visual processing and working memory. Nature neuroscience, 14(5), 656-661.
Zehetleitner, M., Proulx, M. J. & Müller, H. J. (2009). Additional-singleton interference in efficient visual search: A common salience route for detection and compound tasks. Attention, Perception & Psychophysics, 71(8), 1760-1770.
Zehetleitner, M., Koch, A. I., Goschy, H. & Müller, H. J. (2013). Salience-based selection: Attentional capture by distractors less salient than the target. PloS one, 8(1), e52595.
Zhang, D., Zhou, X. & Martens, S. (2009). The impact of negative attentional set upon target processing in RSVP: An ERP study. Neuropsychologia, 47(12), 2604-2614.