Modulating emotional bias in attentional competition
Briana L. Kennedy
Faculty of Science School of Psychology
2016
Thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy
Supervisor
Dr. Steven B. Most
Declaration | i
THE UNIVERSITY OF NEW SOUTH WALES Thesis/Dissertation Sheet
Surname or Family name: Kennedy
First name: Briana Other name/s: Lee
Abbreviation for degree as given in the University calendar: PhD
School: Psychology Faculty: Science
Title: Modulating emotional bias in attentional competition
Abstract 350 words maximum:
Emotionally powerful stimuli can impair the awareness for other items presented in their temporal wake – a phenomenon known as emotion-induced blindness (EIB). EIB is believed to occur because of a spatiotemporally- driven competition, in which emotionally powerful distractors gain a competitive edge over task-relevant targets. The present study had two aims: (1) to verify a competition account of EIB, and (2) to investigate if and how the competitive edge of emotional distractors can be reduced, thereby benefitting target perception. The experiments in Chapter 1 tested whether eye-gaze might present an alternative explanation for the spatially localized nature of EIB. In an EIB task, targets that appear in the same spatial location as a preceding emotional distractor are poorly reported, but targets that appear in a different location are not. This spatially localized pattern of impairment supports a competition account for EIB, but could instead be explained simply by where participants look during the task. The experiments in Chapter 1 employed gaze-contingent eye-tracking and ruled out eye-gaze, supporting a spatiotemporal competition account for EIB. Chapter 2 provided evidence that the bias toward emotional distractors is not dependent on the context in which distractors appear. In Chapter 2, EIB was observed regardless of the distractors’ categorical similarity to other items in the stream. However, in Chapter 3, the bias toward emotional distractors did change based on the way participants regarded the task-relevance of distractors: participants exhibited greater EIB when they considered the distractor to be task-relevant rather than task-irrelevant. In Chapter 4, the competition was also modulated by proactive control: when participants were told the kind of emotional distractor to expect in a given trial, they were better able to overcome emotion-induced blindness. Taken together, the experiments in this thesis supported a spatiotemporal competition account of emotion- induced blindness and demonstrated that the competition between emotional distractors and task-relevant targets can be modulated in an emotion-induced blindness paradigm. This work provides insight into the mechanisms involved in emotion-induced blindness and identifies ways that emotional distraction may be overcome.
Declaration relating to disposition of project thesis/dissertation
I hereby grant to the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or in part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all property rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.
I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstracts International (this is applicable to doctoral theses only).
………………………………………… ………………………………………… ………………………………………… Signature Witness Date
The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years must be made in writing. Requests for a longer period of restriction may be considered in exceptional circumstances and require the approval of the Dean of Graduate Research.
FOR OFFICE USE ONLY Date of completion of requirements for Award:
ii | Declaration
ORIGINALITY STATEMENT
‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.’
Signed ……………………………………………......
Date ……………………………………………......
Declaration | iii
COPYRIGHT STATEMENT
‘I hereby grant the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all proprietary rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation. I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstract International (this is applicable to doctoral theses only). I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; where permission has not been granted I have applied/will apply for a partial restriction of the digital copy of my thesis or dissertation.'
Signed ……………………………………………......
Date ……………………………………………......
AUTHENTICITY STATEMENT
‘I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.’
Signed ……………………………………………......
Date ……………………………………………......
Acknowledgements | v
Acknowledgements
Before starting my degree, I did not realize how important a PhD supervisor would be, which makes me all the luckier that my supervisor is Steve Most. Thank you for being such an inspiring leader, and for your unfaltering and generous dedication as a mentor. I am forever grateful to you for giving me two thrills of a lifetime: for taking a chance on me as your student and for letting me tail you all the way to Australia. Thank you for our meetings filled with laughter, for letting me sometimes challenge your brilliance, for allowing me to be independent but always having my back, and for sharing your unfiltered wisdom, creativity, and fresh perspective in science. I hope that as I move on in my career, I can share some of the perspective, patience, and wealth of knowledge that you have imparted on me.
Thank you to my co-supervisors, both official and unofficial. Brett Hayes, for your always generous and thoughtful advice and for being such a great ally. Tom
Carlson, for your honest insights, great friendship, and teaching me about quality. Jim
Hoffman, for teaching me the basics of vision science and your methodical way of asking questions. Tom Beesley, for asking tough questions and helping me answer them. Patrick Goodbourn, for sharing your wisdom from a few years ahead of me. And
Nathan Holmes, for your always inspiring conversation.
I am grateful to the many members of the Motivated Attention and Perception
Lab. Especially Lingling Wang, Mandy Skoranski, Jenna Zhao, and Vera Newman, for your sisterhood in research and in life. And Kyle Dobson, Sandy Onie, David Sutton,
Lucy Hudson, and Alger Petras, for your dependably kind spirits and contagious enthusiasm, and for helping me along the way. vi | Acknowledgements
I offer my sincerest gratitude to Daniel Pearson and Tim Schofield, for sharing your knowledge and skills with me whenever I asked. And to Jonathan Solomon, for helping me immediately every time I called.
I particularly wish to thank Dominic Tran, Ashley Luckman, Ly Huynh, Elizabeth
Barrett-Cheetham, Alice Towler, Johanna Bergmann, Peter Baldwin, Melissa Black,
Kelvin Wong, Galang Lufityanto, and Matthew Patten for never letting me feel lonely in the process, and for being such supportive colleagues and friends. My very special thanks to Stefanie LoSavio for making Delaware a fun place to live, to Luke Vu for sharing your passion for science and friendship, to Kate Allen for your loving support, and to Daniel de Zilva for taking me under your wing to make Sydney and UNSW my home. And to David Morrow, for your tremendous encouragement and support, for making tough days easier and great days even better, and for joining me on the adventure.
I also wish to acknowledge the dedicated participants in these experiments. It took the generosity of many people to sit through long experiments and knowingly expose themselves to disgusting images. I have genuinely been inspired by these strangers who were devoted to their task, asked enthusiastic questions during debriefings, and offered the occasional word of encouragement to melt my heart.
Most of all, I wish to thank Mary Jo and Michelle Kennedy. For being my support throughout this entire journey, mostly from the other side of the world. You have taught me to follow my dreams, to get up every morning with a smile on my face, and to believe in myself. You have contributed to every word in this thesis, and I could not have done it without you.
Abstract | vii
Abstract
Emotionally powerful stimuli can impair the awareness for other items presented in their temporal wake – a phenomenon known as emotion-induced blindness (EIB). EIB is believed to occur because of a spatiotemporally-driven competition, in which emotionally powerful distractors gain a competitive edge over task-relevant targets. The present study had two aims: (1) to verify a competition account of EIB, and (2) to investigate if and how the competitive edge of emotional distractors can be reduced, thereby benefitting target perception.
The experiments in Chapter 1 tested whether eye-gaze might present an alternative explanation for the spatially localized nature of EIB. In an EIB task, targets that appear in the same spatial location as a preceding emotional distractor are poorly reported, but targets that appear in a different location are not. This spatially localized pattern of impairment supports a competition account for EIB, but could instead be explained simply by where participants look during the task. The experiments in Chapter 1 employed gaze-contingent eye-tracking and ruled out eye-gaze, supporting a spatiotemporal competition account for EIB.
Chapter 2 provided evidence that the bias toward emotional distractors is not dependent on the context in which distractors appear. In Chapter 2, EIB was observed regardless of the distractors’ categorical similarity to other items in the stream. However, in Chapter 3, the bias toward emotional distractors did change based on the way participants regarded the task-relevance of distractors: participants exhibited greater EIB when they considered the distractor to be task-relevant rather than task-irrelevant. In Chapter 4, the competition was also viii | Abstract
modulated by proactive control: when participants were told the kind of emotional distractor to expect in a given trial, they were better able to overcome emotion-induced blindness.
Taken together, the experiments in this thesis supported a spatiotemporal competition account of emotion-induced blindness and demonstrated that the competition between emotional distractors and task-relevant targets can be modulated in an emotion-induced blindness paradigm. This work provides insight into the mechanisms involved in emotion-induced blindness and identifies ways that emotional distraction may be overcome.
Publications | ix
Publications
Kennedy, B.L., Pearson, D., Sutton, D.J., Beesley, T., & Most, S.B. (in
preparation). Affective penetration of vision: Behavioral and eye-tracking
evidence that emotion helps shape perception.
[Chapter 1]
Kennedy, B.L., & Most, S.B. (2015). Affective stimuli capture attention
regardless of categorical distinctiveness: An emotion-induced blindness
study. Visual Cognition, 23(1-2), 105-117.
[Chapter 2]
Kennedy, B.L. & Most, S.B. (in preparation). Emotionally-infused stimuli
disrupt visual awareness regardless of task goals.
[Chapter 3]
Kennedy, B.L. & Most, S.B. (in preparation). Proactive deprioritization of
emotional distractors enhances target perception.
[Chapter 4]
I certify that these publications were a direct result of my research towards this
PhD, and that reproduction in this thesis does not breach copyright regulations.
Briana L. Kennedy Date x | Contents
Table of contents
Declaration i Acknowledgements v Abstract vii Publications ix
Introduction and Overview 1
Chapter 1. Eye-tracking evidence of localized competition 25 in emotion-induced blindness Experiment 1 31 Experiment 2 42
Chapter 2. Affective stimuli capture attention regardless of 63 categorical distinctiveness: An emotion-induced blindness study Experiment 3 69
Chapter 3. Emotionally-infused stimuli disrupt visual awareness 89 regardless of task goals Experiment 4 96
Chapter 4. Proactive deprioritization of emotional distractors 115 enhances target perception Experiment 5 121 Experiment 6 129 Experiment 7 133
General Discussion 145
Contents | xi
List of figures
Introduction
Figure 1. Schematic of a partial trial in a typical emotion-induced blindness task
Figure 2. Results from Most et al., 2005
Figure 3. Schematic of a partial trial in the two-stream, emotion-induced blindness task used in Most & Wang, 2011
Figure 4. Results from Most & Wang, 2011
Chapter 1
Figure 5. Alternative account predictions
Figure 6. Schematic of a partial trial in Experiment 1
Figure 7. Experiment 1 target accuracy
Figure 8. Experiment 1 post-distractor eye-gaze results
Figure 9. Experiment 2 target accuracy
Figure 10. Experiment 2 distractor memory performance
Figure 11. Experiment 2 post-distractor eye-gaze results
Chapter 2
Figure 12. Example of part of an experimental trial in Experiment 3
Figure 13. Experiment 3 results
Figure 14. Results from Experiment 3 replication
Chapter 3
Figure 15. Schematic of a partial trial sequence in Experiment 4
Figure 16. Experiment 4 performance accuracy for T2
Figure 17. Performance accuracy for T2 given correct T1 identification
Figure 18. Performance accuracy for T1
Chapter 4
Figure 19. Schematic of a partial trial sequence in Experiment 5
Figure 20. Experiment 5 results
Figure 21. Experiment 6 results
Figure 22. Experiment 7 results
Introduction and overview | 1
Introduction and Overview
In our visual world, we tend to prioritize stimuli that are emotionally powerful
(Anderson & Phelps, 2001; Anderson, 2005; Öhman, Flykt, & Esteves, 2001) – even when they are irrelevant to our current goals (Dolcos & McCarthy, 2006; Kensinger,
Garoff-Eaton, & Schacter, 2006; Mather & Sutherland, 2011; Mogg & Bradley, 1998;
Most, Chun, Widders, & Zald, 2005; Williams, Mathews, & MacLeod, 1996). For example, emotional distractors can disrupt awareness of subsequent targets that appear in their temporal wake, an effect known as emotion-induced blindness (EIB; Arnell,
Killman, & Fijavz, 2007; Kennedy, Rawding, Most, & Hoffman, 2014; Most, Chun,
Johnson, & Kiehl, 2006; Most et al., 2005; Most & Jungé, 2008; Smith, Most,
Newsome, & Zald, 2006; Wang, Kennedy, & Most, 2012).
Whether driving past a graphic billboard or working in an emergency room, EIB carries important implications for individuals encountering emotionally powerful stimuli in the world. The motivation for this series of studies was two-fold: (1) to inform the understanding of mechanisms underlying emotion-attention interactions, and
(2) to understand strategies that may help individuals better overcome emotional distraction.
Emotion-induced blindness as a way to examine emotional distraction 2 | Introduction and overview
Figure 1. Schematic of a partial trial in a typical emotion-induced blindness task. Participants search for one rotated picture in a rapidly presented stream of upright images. See text for details.
In a standard EIB task, participants view a rapid serial visual presentation (RSVP) of upright landscape and architectural pictures presented at a rate of 100 milliseconds
(ms) per image (Most et al., 2005). Participants are instructed to look for one picture rotated 90 degrees clockwise or counterclockwise and then report the direction in which it was rotated (Figure 1).
Participants are generally accurate in reporting the direction of the target’s rotation despite the fast presentation rate. However, when an emotional distractor appears soon before the target, target accuracy is impaired. Lag, which refers to the temporal distance between the distractor and target, varies based on the trial: for example, lag-2 means that target was the second item after the distractor, lag-8 means that the target was the eighth item after the distractor. As demonstrated by results from
Most and colleagues (2005), EIB is typically observed at lag-2, but the system is usually fully recovered by lag-8 (see Figure 2). Additional studies demonstrate that at a rate of
100ms per item, EIB is usually observed through lag-4 (400ms; Ciesielski, Armstrong,
Zald, & Olatunji, 2010; Kennedy & Most, 2015). Introduction and overview | 3
Figure 2. Results from Most et al., 2005. Participants were impaired in their ability to detect targets when they appeared soon after negative distractors, compared to trials with neutral or scrambled distractors. When the distractors and targets were separated by a greater amount of time, such impairment was no longer observed. Figure adapted from Most et al., 2005.
The emotion-induced impairment can be elicited by negatively (Kennedy & Most,
2012; Most et al., 2006, 2005) or positively (Arnell et al., 2007; Most, Smith, Cooter,
Levy, & Zald, 2007) arousing stimuli (see the Appendix for valence and arousal ratings of the emotional distractors used throughout this thesis). Negative images typically include depictions of medical injuries, threatening animals, and violence. EIB has also been found using highly arousing positive images such as erotica, suggesting that EIB stems from the emotional intensity of a distractor image rather than its negative vs. positive valence (Most et al., 2007). Critically, neutral distractors (which are not emotionally powerful but also depict animals and people) do not impair target detection as much as emotional distractors, suggesting that the impairment from emotionally powerful stimuli stems from factors above and beyond their “oddball” depiction of people or animals in a stream otherwise comprised of landscape images. The emotion- induced impairment is also greater than that caused by scrambled versions of the emotional images, suggesting that the effect cannot be attributed to differences in low- level properties like color and luminance. Moreover, the impairment from emotional 4 | Introduction and overview
stimuli does not emerge only from stimuli that depict inherently emotional visual information. EIB has been demonstrated using words (Arnell et al., 2007; Ihssen &
Keil, 2009), as well as neutral stimuli that have been conditioned to be emotionally powerful (Smith et al., 2006).
Underlying mechanisms in emotion-induced blindness
Placed within the literature of emotion-attention interactions, EIB appears to reflect a relatively unique impact of emotion on attention. Understanding of the underlying mechanisms in EIB can be informed by considering the processing steps involved in the task. One might speculate that targets in emotion-induced blindness tasks are seen, but simply not remembered for report at the end of the stream. However, in one study, emotion-induced impairment was observed even when participants were instructed to make their response as soon as they saw the target (Kennedy & Most,
2012). These results suggests EIB occurs at a perceptual stage, rather than in the active maintenance of target information in working memory after participants see the target.
In order to further understand the mechanisms that drive EIB, it is useful to consider the substantial literature on the attentional blink (AB), a phenomenon that seems to be at least superficially similar. In the attentional blink, participants tend to have difficulty reporting the second of two targets (T2) when it appears too soon after the first target (T1; e.g., Chun & Potter, 1995; Raymond, Shapiro, & Arnell, 1992). In both phenomena, attention to a first stimulus impairs the ability to report an item after it, suggesting the attentional blink and EIB may share underlying mechanisms.
Attentional blink effects are also subject to modulation by emotionally powerful targets.
An emotional T1 produces a greater AB for T2 (Schwabe & Wolf, 2010; Schwabe et al., 2011), while an emotional T2 attenuates the blink (Anderson & Phelps, 2001;
Anderson, 2005; Ihssen & Keil, 2009; Schwabe & Wolf, 2010; Schwabe et al., 2011). Introduction and overview | 5
Traditional theories of the attentional blink have attributed it to a relatively central processing bottleneck. For example, some have suggested that it reflects competition for consolidation into visual working memory (e.g., Chun & Potter, 1995), from the disruption of a filter that distinguishes targets from non-targets (Di Lollo,
Kawahara, Ghorashi, & Enns, 2005), or from retrieving target information from memory (Shapiro, Raymond, & Arnell, 1994). Chun and Potter (1995) suggested that the attentional blink occurs in a two-stage process, such that both targets are perceived in the first stage of processing (with high capacity), but that the second stage – consolidating targets into working memory – is capacity limited, and requires consolidating one target at a time. In this framework, the ability to report the second target is impaired because the perceiver’s representation of it fades or is overwritten during the time it takes to consolidate T1. In another explanation (Di Lollo et al.,
2005), an attentional template is established in order to filter targets from non-targets.
When a target is detected, it temporarily disrupts control of the filter to allow the target to access further processing. Any items with target qualities are let through, however in the state of lost control, the filter must be reset in order to again filter out non-targets.
In this case, the second target is missed because it occurs during this resetting period.
However, EIB and the attentional blink may be more dissociable than their surface level properties suggest. “Lag-1 sparing” is a pattern commonly observed in the attentional blink, such that performance accuracy for reporting the second target remains intact when both targets are presented sequentially (Chun & Potter, 1995;
Dell’Acqua, Sessa, Jolicœur, & Robitaille, 2006; Di Lollo et al., 2005; Olivers, van der
Stigchel, & Hulleman, 2007). EIB does not exhibit the typical lag-1 sparing pattern, such that targets presented immediately after distractors are impaired just as much as when presented at lag-2 (Kennedy & Most, 2015). An EIB study using event-related 6 | Introduction and overview
potentials also revealed that while there is a great similarity in the neural patterns of EIB
Figure 3. Schematic of a partial trial in the two stream, emotion-induced blindness task used in Most & Wang, 2011. Participants searched for the one rotated picture within two simultaneous RSVP streams. The distractor appeared in the same stream or in the opposite stream from the target. Figure adapted from Most & Wang, 2011. See text for details.
and attentional blink, there may also be differences (Kennedy et al., 2014). Trade-offs in the amplitude of N2 and P3 components between the distractor and target in EIB mimicked previous findings in the attentional blink (Sergent, Baillet, & Dehaene, 2005;
Shapiro, Schmitz, Martens, Hommel, & Schnitzler, 2006), but there was also a posterior positivity that predicted behavioural performance in EIB which has not been cited in previous attentional blink studies. While neither of these findings necessarily indicate that EIB is separable from the attentional blink, they do suggest nuances to consider when comparing their mechanisms.
One finding that does seem to differentiate mechanisms underlying EIB and traditional attentional blink findings comes from an EIB study that employed two simultaneously presented streams of stimuli (Figure 3; Most & Wang, 2011). In this task, participants were instructed to report the direction of one rotated image that could appear in either of two streams. The distractor (emotional or neutral) could appear Introduction and overview | 7
either in the same stream as the target or in the opposite stream from the target. EIB was found only when emotional distractor appeared in the same stream as the target (see
Figure 4). When they appeared in different streams, emotional distractors no longer impaired target performance more than neutral distractors.
The spatially specific impairment in EIB contrasts with the patterns observed in the attentional blink and the predictions of traditional attentional blink theories. Indeed, attentional blink studies usually demonstrate impairment regardless of the spatial
Figure 4. Results from Most & Wang, 2011. Emotion-induced blindness was localized to trials in which the distractor and target appeared in the same stream. Figure adapted from Most & Wang, 2011.
relationship between the two targets (Lunau & Olivers, 2010; Shih, 2000; but see
Kristjansson & Nakayama, 2002). If EIB were to occur because of a central interference, there should be no reason for the impairment to be specific to the spatial location of the distractor – as the theories propose the impairment to occur at a time 8 | Introduction and overview
after perceiving the target and at a central bottleneck of target processing. As such, in a central interference framework both the distractor and target should be perceptually encoded, but both cannot be consolidated into memory. The spatially specific impairment in EIB thus makes it difficult to explain with traditional AB theories.
Instead, these results suggest that EIB impairs target perception at a relatively early stage in perceptual processing. One possibility is that EIB reflects a biased competition between the emotional distractor and target (Wang et al., 2012). In this framework, emotional distractors and the subsequent targets appear in the same space and so close in time that their representations compete with one another, with the prioritized emotional distractor biased to win neural representation and access to further processing at the cost of the target representation.
The spatiotemporal competition account for EIB builds from several theories of biased competition in the attentional literature (Desimone & Duncan, 1995; Keysers &
Perrett, 2002), which suggest that items that share receptive fields compete for neural representation. For example, a stimulus that is presented alone tends to elicit a greater neural response than when it is presented with other stimuli in the visual field, suggesting that the stimulus has to compete with other nearby items for representation
(Beck & Kastner, 2005; Scalf, Torralbo, Tapia, & Beck, 2013). Keysers and Perrett suggest that items in an RSVP can compete in a similar way, such that their neural responses overlap because of their shared spatial location and close temporal relationship (Keysers & Perrett, 2002).
A biased, spatiotemporal competition account for EIB suggests that emotional distractors compete for representation with the subsequent target, and dominate the representation because of their emotional power. This account is compatible with results that emotional distractors can impair the detection for targets that appear one Introduction and overview | 9
item before the distractor (Most & Jungé, 2008). That is, emotional distractors and targets compete for representation no matter their order of presentation. The spatiotemporal competition account for EIB also makes predictions about the ability to modulate the power of emotional distractors. That is, if the competition is somehow weighted away from the distractor, and/or toward the target, this theory would predict less EIB.
Emotion-induced blindness in the context of other emotion-attention tasks
Of course, EIB is one of many different tasks used to explore emotion-attention interactions. For example, visual search tasks often demonstrate the prioritization of emotional stimuli at the expense of other task-relevant stimuli: emotionally arousing stimuli (e.g., snakes) are detected faster in a visual search than are non-emotional stimuli (e.g., mushrooms; Öhman et al., 2001). The popular “dot-probe task” also demonstrates the attention-grabbing power of emotional distractors (MacLeod,
Mathews, & Tata, 1986; Macleod & Mathews, 2012; Mogg & Bradley, 1998). In this task, participants are instructed to respond to a target as fast as they can. Just before the target is presented, two task-irrelevant images flash on the computer screen, and the target appears in the same spatial location as one of those stimuli. Participants tend to respond to the target fastest when it appears in the location of an emotional stimulus, and slowest when it appears in the location of a neutral stimulus that is presented at the same time as an emotional stimulus. These results suggest that attention is drawn toward the emotional stimuli, allowing a quick response to targets that appear at the same location. However, if the target appears in the opposite location, attention must disengage from the spatial location of the emotional stimulus, ultimately delaying the target response. 10 | Introduction and overview
At first blush, the findings in the dot-probe task (MacLeod et al., 1986; Mogg &
Bradley, 1998) seem contrary to those of EIB. That is, in EIB, participants have a difficult time reporting targets that appear in the same location as an emotional distractor, whereas in the dot-probe, participants respond faster to cues that appear in the same location. The key difference between these two paradigms may come down to the difference between spatial attention and biased competition, a difference that may be revealed by the different use of masking in the two procedures. When targets are left unmasked in EIB (that is, when no filler items follow the target presentation in EIB), participants tend to respond better to targets that appear in the same spatial location as emotional distractors (Most & Wang, 2011). Removing the masks in EIB likely alleviates the temporal ambiguity of the distractor and target and allows the target to resonate in iconic memory. When EIB is in a traditional RSVP, with targets and distractors both embedded within the stream (and thereby masked by succeeding items), the temporal ambiguity of their positions increases. In turn, the overlapping representations of the distractor and target likely increase the temporal competition between them, and ultimately bias processing toward the salient emotional distractors at expense of the target. When the distractor and target representations become more separable (due to their explicit temporal order and the resonance of target representation in iconic memory), spatiotemporal competition should be reduced, and spatial attention effects become the mechanism observed.
Together, these results highlight the importance of understanding the different processing levels of attention-emotion interactions. EIB seems to reflect a relatively early interaction in attentional resources between the emotional distractor and subsequent target processing. There are many ways to study emotional distraction – visual search, the dot-probe, and EIB being only a few (see Bar-Haim, Lamy, Pergamin, Introduction and overview | 11
& Ijzendoorn, 2007; Yiend, 2010 for reviews). Different paradigms often demonstrate different consequences for stimuli presented nearby an emotionally powerful stimulus
(sometimes even showing benefits; Anderson, Wais, & Gabrieli, 2006; Bocanegra &
Zeelenberg, 2009; Mather, Clewett, Sakaki, & Harley, in press; Mather & Sutherland,
2011), likely reflecting different mechanisms that are tapped by the different tasks, or different task-goals or priorities of the stimuli at the time of emotional arousal (see
Mather & Sutherland, 2011).
The emotion-attention field seems particularly focused on spatial-attention tasks like the dot-probe, but understanding of EIB can provide insights into emotion-attention interactions that are otherwise being ignored. EIB seems to occur as a result of representational competition from emotional distractors, which is a seemingly different level of processing than those being observed in other tasks. At least one theoretical model has described such different levels of attentional processing, identifying at least three stages of attentional interference from two competing stimuli: 1) competitive interference, 2) attentional mapping and localized inhibition, and 3) working memory consolidation (Wyble & Swan, 2015). EIB likely represents a competitive interference for representation, which is distinct from later interference from working memory that are largely explored in other emotion-attention tasks. By understanding the mechanisms of EIB, we are able to ask questions that have not yet been addressed in the emotion-attention literature, while simultaneously asking questions for the attention field more broadly that are not being discussed.
Emotion-induced blindness as a way to examine emotion’s influence on perception
Attempts to make claims about emotion’s effect on perception have a long and storied history. One of the most notable movements was the “New Look” movement in 12 | Introduction and overview
the 1940s, which asserted that perception is influenced by top-down factors such as motivation, knowledge, and emotion (Bruner & Goodman, 1947). At one level, EIB could be seen as a piece of this larger literature concerned with the penetrability of top- down effects on perception from emotion.
However, questions about whether such top-down factors influence perception, per se, have been the subject of vociferous debate. Pylyshyn (1999), for example, has argued that the very early stages of perception (“early vision”) are modular and unaffected by such factors. In his view, motivation and emotion might impact the selection of information for input into a visual computation apparatus and might influence post-perceptual judgments that shape conscious experience, but the early visual computations themselves remain unmoved. More recently, a more extreme position has been adopted by Firestone and Scholl, who argue that effects of motivation and emotion do not reflect impacts on perception even beyond the bounds of early vision (Firestone & Scholl, in press). Instead, they suggest that any current evidence claiming for such penetrability has fallen for one of six “pitfalls”. These pitfalls include: (1) Predictions that cannot be disconfirmed, such that a measure is unable to reject an effect when the theory would predict none; (2) conclusions confusing post- perceptual judgment for perception; (3) task demands and response bias that make the participant respond in a way they believe they should; (4) low-level differences in the experimental stimuli that drive differences otherwise interpreted; (5) peripheral attentional effects, such that selected attention effects are confused as perceptual effects; and (6) measuring memory instead of perception, but interpreting the results as a perceptual effect. Notably, the patterns of EIB do not seem to fall prey for the pitfalls outlined by Firestone and Scholl (in press) – a point I will return to in the discussion.
The current study Introduction and overview | 13
This thesis tackles two overarching aims: (1) to develop a greater understanding of the mechanisms involved in EIB, particularly to verify a competition between emotional distractors and subsequent targets, and, (2) to identify and understand factors that modulate this competition.
Although the spatially localized pattern in EIB has been interpreted as mechanistically informative, and suggestive of spatiotemporal competition, an alternative explanation may be that such a pattern arises simply because of where participants look. For example, if a participant happens to be looking at the location containing the target, an emotional distractor appearing in the opposite stream might have less impact than one appearing in the same stream because participants do not process it. Experiment 1 considers both explanations, using gaze-contingent eye- tracking to rule out eye-gaze in the spatially localized effect. Experiment 2 further investigates this question by using task-relevant distractors, as used in the attentional blink, to better compare the spatial patterns of emotion-induced blindness and the attentional blink.
Subsequent experiments further examine conditions that may modulate the impact of emotional distractors. Experiment 3 considers the extent to which categorical distinctiveness of the emotional distractor from other items in its context can modulate its impact. Experiment 4 expands the investigation of task-relevancy in EIB, directly comparing performance when distractors are task-irrelevant and task-relevant. Finally,
Experiments 5, 6, and 7 explore how participants may be able to utilize proactive control of attention to overcome the distraction from emotional stimuli. 14 | Introduction and overview
References
Anderson, A. K. (2005). Affective influences on the attentional dynamics supporting
awareness. Journal of Experimental Psychology: General, 134(2), 258–281.
http://doi.org/10.1037/0096-3445.134.2.258
Anderson, A. K., & Phelps, E. A. (2001). Lesions of the human amygdala impair
enhanced perception of emotionally salient events. Nature, 411(6835), 305–9.
http://doi.org/10.1038/35077083
Anderson, A. K., Wais, P. E., & Gabrieli, J. D. E. (2006). Emotion enhances
remembrance of neutral events past. PNAS, 103(5), 1599–604.
http://doi.org/10.1073/pnas.0506308103
Arnell, K. M., Killman, K. V, & Fijavz, D. (2007). Blinded by emotion: target misses
follow attention capture by arousing distractors in RSVP. Emotion, 7(3), 465–477.
http://doi.org/10.1167/4.8.359
Bar-Haim, Y., Lamy, D., Pergamin, L., & van Ijzendoorn, M. H. (2007). Threat-related
attentional bias in anxious and nonanxious individuals: A meta-analytic study.
Psychological Bulletin, 133(1), 1–24. http://doi.org/10.1037/0033-2909.133.1.1
Beck, D. M., & Kastner, S. (2005). Stimulus context modulates competition in human
extrastriate cortex. Nature Neuroscience, 8(8), 1110–1116.
http://doi.org/10.1038/nn1501
Bocanegra, B. R., & Zeelenberg, R. (2009). Emotion improves and impairs early vision.
Psychological Science, 20(6), 707–713. http://doi.org/10.1111/j.1467-
9280.2009.02354.x
Bruner, J. S., & Goodman, C. C. (1947). Value and need as organizing factors in
perception. The Journal of Abnormal and Social Psychology, 42(1), 33–44.
http://doi.org/10.1037/h0058484 Introduction and overview | 15
Chun, M. M., & Potter, M. C. (1995). A two-stage model for multiple target detection in
rapid serial visual presentation. Journal of Experimental Psychology: Human
Perception and Performance, 21(1), 109–27. http://doi.org/10.1037/0096-
1523.21.1.109
Ciesielski, B. G., Armstrong, T., Zald, D. H., & Olatunji, B. O. (2010). Emotion
modulation of visual attention: Categorical and temporal characteristics. Plos
One, 5(11). http://doi.org/10.1371/journal.pone.0013860
Dell’Acqua, R., Sessa, P., Jolicœur, P., & Robitaille, N. (2006). Spatial attention freezes
during the attention blink. Psychophysiology, 43(4), 394–400.
http://doi.org/10.1111/j.1469-8986.2006.00411.x
Desimone, R., & Duncan, J. (1995). Neural mechanisms of selective visual attention.
Annual Review of Neuroscience, 18, 193–222.
http://doi.org/10.1146/annurev.ne.18.030195.001205
Di Lollo, V., Kawahara, J., Ghorashi, S. M. S., & Enns, J. T. (2005). The attentional
blink: resource depletion or temporary loss of control? Psychological Research,
69(3), 191–200. http://doi.org/10.1007/s00426-004-0173-x
Dolcos, F., & McCarthy, G. (2006). Brain systems mediating cognitive interference by
emotional distraction. The Journal of Neuroscience, 26(7), 2072–9.
http://doi.org/10.1523/JNEUROSCI.5042-05.2006
Firestone, C., & Scholl, B. J. (in press). Cognition does not affect perception:
Evaluating the evidence for “top-down” effects. Behavioral and Brain Sciences.
http://doi.org/10.1017/S0140525X15000965
Ihssen, N., & Keil, A. (2009). The costs and benefits of processing emotional stimuli
during rapid serial visual presentation. Cognition and Emotion, 23, 37–41.
http://doi.org/10.1080/02699930801987504 16 | Introduction and overview
Kennedy, B. L., & Most, S. B. (2012). Perceptual, not memorial, disruption underlies
emotion-induced blindness. Emotion, 12(2), 199–202.
http://doi.org/10.1037/a0026380
Kennedy, B. L., & Most, S. B. (2015). The Rapid Perceptual Impact of Emotional
Distractors. Plos One, 10(6), e0129320.
http://doi.org/10.1371/journal.pone.0129320
Kennedy, B. L., Rawding, J., Most, S. B., & Hoffman, J. E. (2014). Emotion-induced
blindness reflects competition at early and late processing stages: An ERP study.
Cognitive, Affective & Behavioral Neuroscience, 14(4), 1485–1498.
http://doi.org/10.3758/s13415-014-0303-x
Kensinger, E. A., Garoff-Eaton, R. J., & Schacter, D. L. (2006). Memory for specific
visual details can be enhanced by negative arousing content. Journal of Memory
and Language, 54(1), 99–112. http://doi.org/10.1016/j.jml.2005.05.005
Keysers, C., & Perrett, D. I. (2002). Visual masking and RSVP reveal neural
competition. Trends in Cognitive Sciences, 6(3), 120–125.
http://doi.org/10.1016/S1364-6613(00)01852-0
Kristjánsson, A., & Nakayama, K. (2002). The attentional blink in space and time.
Vision Research, 42, 2039–2050. http://doi.org/10.1016/S0042-6989(02)00129-3
Lang, P.J., Bradley, M. M., & Cuthbert, B.N. (2001). International Affective Picture
System (IAPS): Instruction manual and affective ratings (Tech. Rep. No. A-5).
Gainesville: University of Florida, Center for Research in Psychophysiology.
Lunau, R., & Olivers, C. N. L. (2010). The attentional blink and lag 1 sparing are
nonspatial. Attention Perception Psychophysics, 72(2), 317–325.
http://doi.org/10.3758/APP.72.2.317 Introduction and overview | 17
MacLeod, C., & Mathews, A. (2012). Cognitive bias modification approaches to
anxiety. Annual Review of Clinical Psychology, 8, 189–217.
http://doi.org/10.1146/annurev-clinpsy-032511-143052
MacLeod, C., Mathews, A., & Tata, P. (1986). Attentional bias in emotional disorders.
Journal of Abnormal Psychology, 95(1), 15–20. http://doi.org/10.1037/0021-
843X.95.1.15
Mather, M., Clewett, D., Sakaki, M., & Harley, C. W. (in press). Norepinephrine ignites
local hot spots of neuronal excitation: How arousal amplifies selectivity in
perception and memory. Behavioral and Brain Sciences.
http://doi.org/10.1017/S0140525X15000667
Mather, M., & Sutherland, M. R. (2011). Arousal-Biased Competition in perception and
memory. Perspectives on Psychological Science, 6(2), 114–133.
http://doi.org/10.1177/1745691611400234
Mogg, K., & Bradley, B. P. (1998). A cognitive-motivational analysis of anxiety.
Behavior Research and Therapy, 36, 809–848. http://doi.org/10.1016/S0005-
7967(98)00063-1
Most, S. B., Chun, M. M., Johnson, M. R., & Kiehl, K. a. (2006). Attentional
modulation of the amygdala varies with personality. NeuroImage, 31(2), 934–44.
http://doi.org/10.1016/j.neuroimage.2005.12.031
Most, S. B., Chun, M. M., Widders, D. M., & Zald, D. H. (2005). Attentional
rubbernecking: cognitive control and personality in emotion-induced blindness.
Psychonomic Bulletin & Review, 12(4), 654–61.
http://doi.org/10.3758/BF03196754 18 | Introduction and overview
Most, S. B., & Jungé, J. A. (2008). Don’t look back: Retroactive, dynamic costs and
benefits of emotional capture. Visual Cognition, 16(2), 262–278.
http://doi.org/10.1080/13506280701490062
Most, S. B., Smith, S. D., Cooter, A. B., Levy, B. N., & Zald, D. H. (2007). The naked
truth: Positive, arousing distractors impair rapid target perception. Cognition &
Emotion, 21(5), 964–981. http://doi.org/10.1080/02699930600959340
Most, S. B., & Wang, L. (2011). Dissociating spatial attention and awareness in
emotion-induced blindness. Psychological Science, 22(3), 300–5.
http://doi.org/10.1177/0956797610397665
Öhman, A., Flykt, A., & Esteves, F. (2001). Emotion drives attention : Detecting the
snake in the grass. Emotion, 130(3), 466–478. http://doi.org/10.1037/AXJ96-
3445.130.3.466
Olivers, C. N. L., van der Stigchel, S., & Hulleman, J. (2007). Spreading the sparing:
against a limited-capacity account of the attentional blink. Psychological
Research, 71(2), 126–39. http://doi.org/10.1007/s00426-005-0029-z
Pylyshyn, Z. W. (1999). Is vision continuous with cognition? The case for cognitive
impenetrability of visual perception. Behavioral and Brain Sciences, 22, 341-365.
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), 849–60.
http://doi.org/10.1037/0096-1523.18.3.849
Scalf, P. E., Torralbo, A., Tapia, E., & Beck, D. M. (2013). Competition explains
limited attention and perceptual resources : implications for perceptual load and
dilution theories. Frontiers in Psychology, 4, 1–9.
http://doi.org/10.3389/fpsyg.2013.00243 Introduction and overview | 19
Schwabe, L., Merz, C. J., Walter, B., Vaitl, D., Wolf, O. T., & Stark, R. (2011).
Emotional modulation of the attentional blink: The neural structures involved in
capturing and holding attention. Neuropsychologia, 49(3), 416–425.
http://doi.org/10.1016/j.neuropsychologia.2010.12.037
Schwabe, L., & Wolf, O. T. (2010). Emotional modulation of the attentional blink: is
there an effect of stress? Emotion, 10(2), 283–288.
http://doi.org/10.1037/a0017751
Sergent, C., Baillet, S., & Dehaene, S. (2005). Timing of the brain events underlying
access to consciousness during the attentional blink. Nature Neuroscience, 8(10),
1391–400. http://doi.org/10.1038/nn1549
Shapiro, K. L., Raymond, J. E., & Arnell, K. M. (1994). Attention to visual pattern
information produces the attentional blink in rapid serial visual presentation.
Journal of Experimental Psychology: Human Perception and Performance, 20(2),
357–371. http://doi.org/10.1037/0096-1523.20.2.357
Shapiro, K., Schmitz, F., Martens, S., Hommel, B., & Schnitzler, A. (2006). Resource
sharing in the attentional blink. NeuroReport, 17(2), 163–6.
http://doi.org/10.1097/01.wnr.0000195670.37892.1a
Shih, S. (2000). Recall of two visual targets embedded in RSVP streams of distractors
depends on their temporal and spatial relationship. Perception & Psychophysics,
62(7), 1348–1355. http://doi.org/10.3758/BF03212137
Smith, S. D., Most, S. B., Newsome, L. a, & Zald, D. H. (2006). An emotion-induced
attentional blink elicited by aversively conditioned stimuli. Emotion, 6(3), 523–7.
http://doi.org/10.1037/1528-3542.6.3.523 20 | Introduction and overview
Wang, L., Kennedy, B. L., & Most, S. B. (2012). When emotion blinds: A
spatiotemporal competition account of emotion-induced blindness. Frontiers in
Psychology, 3, 1–6. http://doi.org/10.3389/fpsyg.2012.00438
Williams, J. M. G., Mathews, A., & MacLeod, C. (1996). The emotional Stroop task
and psychopathology. Psychological Bulletin, 120(1), 3–24.
http://doi.org/10.1037/0033-2909.120.1.3
Wyble, B., & Swan, G. (2015). Mapping the spatiotemporal dynamics of interference
between two visual targets. Attention, Perception, & Psychophysics, 77(7), 2331-
2343. http://doi.org/10.3758/s13414-015-0938-x
Yiend, J. (2010). The effects of emotion on attention : A review of attentional
processing of emotional information. Cognition and Emotion, 24(1), 3–47.
http://doi.org/10.1080/02699930903205698
Introduction and overview | 21
Introduction and Overview Appendix: Distractor image
valence and arousal ratings
Most “distractor” images used throughout this thesis were used in previous emotion-induced blindness studies. Some additional images were supplemented in
Chapter 1, as outlined below. All distractor images were rated by independent participants, and participants rated the images on 9-point scales of valence (1=negative,
9=positive) and arousal (1=low arousal, 9=high arousal).
Images previously used in emotion-induced blindness tasks
Images that were used in previous EIB experiments were also previously rated by
12 individuals (6 female, 6 male; reported in Kennedy & Most, 2012). The set of images were chosen based on ratings of valence and arousal and were mostly gathered from the International Affective Picture System (IAPS; Lang, Bradley, & Cuthbert,
2008), with the rest supplemented by similar images taken from publicly available sources. This thesis included 56 negative (Chapters 2, 3, and 4), 56 neutral (Chapters 2 and 3), and 56 erotic (Chapter 4) images from that set of images. Here, I report the valence and arousal ratings only for images that were used in the current thesis.
Compared to neutral images, negative images were rated more negatively valenced (negative: M=1.737, SD=0.801, neutral: M=4.997, SD=0.289), t(11) =13.288, p<.001, dz=3.836, and more arousing (negative: M=6.039, SD=1.986; neutral:
M=3.182, SD=1.605), t(11) =4.791, p=.001, dz=1.384. Compared to neutral images, erotic images were rated as more positively valenced (M=6.058, SD=1.129), t(11)
=2.934, p=.014, dz=0.991, and more arousing (M=6.355, SD=0.889), t(11) =6.533, p<.001, dz=1.886. As expected, negative and erotic images differed on valence, t(11) 22 | Introduction and overview
=11.769, p<.001, dz=3.397, but were rated as similarly arousing, t(11) =0.487, p=.636, dz=0.150.
Additional negative distractor images in Chapter 1
Chapter 1 used a subset of negative images from the set of images described above, as well as supplementary images that were added to the negative distractor set.
The reason for deviating from the usual set was that we wanted a large set of images that were emotionally powerful, but that were not more emotionally intense than necessary. 24 images from the previous set (described above) were used. The additional images included 36 negative images (which were previously rated for another experiment in our lab), and another set of 120 negative images (rated for the purposes of the experiments in Chapter 1). We compared the ratings for these images to the original set of images.
The subset of 24 images from the original set of images were rated as negatively valenced (M=1.902, SD=0.809) and arousing (M=5.958, SD=1.866). Compared to the same set of neutral images described above (previously used in emotion-induced blindness tasks), the negative images comprising this subset were still rated more negatively valenced, t(11) =12.611, p<.001, dz=3.641, and more arousing t(11) =5.075, p<.001, dz=1.465.
The new set of 36 images was chosen from a much larger set of 525 images of varying arousal and valence signatures. 302 participants rated these 525 images on
Mechanical Turk, with each participant rating a different subset (of 42-49 images) of the total image set. Demographic information for participants was not collected. The 36 images were rated as negatively valenced (M=2.756, SD=1.513) and mildly arousing
(M=4.733, SD=2.230). Compared to the neutral images reported in the first set of Introduction and overview | 23
images (previously used in emotion-induced blindness tasks), these 36 images were rated to be more negative, t(46.388) =18.562, p<.001, d=2.057, and more arousing, t(310) =2.383, p=.018, d=0.798. Note that when comparing negative distractors from one set of images to neutral distractors from the original set of images (using a
Summary t-test), analyses were corrected when equal variances could not be assumed, as determined by the Hartley test for equal variance.
The final 120 images were rated by another group of 14 participants (6 female, 8 male). The images were also rated as negatively valenced (M=3.063, SD=0.609) and arousing (M=5.771, SD=0.1.228). Compared to the neutral distractors in the first set of images, they were also rated as more negative, t(19.162) =10.582, p<.001, d=4.057, and more arousing, t(24) =4.560, p<.001, d=1.812.
24 | Introduction and overview
Chapter 1:
Eye-tracking evidence of localized competition in emotion-induced blindness
Briana L. Kennedy, Daniel Pearson, David J. Sutton, Tom Beesley, and Steven B. Most
26 | Spatial specificity in EIB regardless of eye-gaze
Abstract
Emotional distractors can impair the ability to report targets in their temporal wake. This emotion-induced blindness (EIB) seems to be spatially localized, contributing to a theoretical account of EIB that suggests spatiotemporal competition between meaningful distractors and subsequent targets. However, the localized effect could instead be explained simply by where participants look during the task when the distractor appears. We employed gaze-contingent eye-tracking to position the distractor in the stream that participants were either fixating or not fixating. Target perception impairments were localized to the distractor location regardless of where participants fixated, consistent with a spatiotemporal competition account. In a second experiment, emotional distractors were rendered task-relevant by having participants identify them at the end of each trial. In this case target perception impairments occurred across both streams but were still strongest at the distractor location, suggesting that the roles of perceptual competition and memory encoding in EIB are additive.
Spatial specificity in EIB regardless of eye-gaze | 27
Emotional distractors can disrupt awareness of subsequent targets, an effect known as “emotion-induced blindness” (Arnell, Killman, & Fijavz, 2007; Most, Chun,
Widders, & Zald, 2005; Most & Wang, 2011; Wang, Kennedy, & Most, 2012). EIB is phenomenally similar to the attentional blink (AB; Chun & Potter, 1995; Raymond,
Shapiro, & Arnell, 1992), in which participants have difficulty reporting a second target within a rapid serial presentation (RSVP) when it appears shortly after a first target. In both EIB and AB, attention to a first stimulus impairs the ability to report a subsequent stimulus. However, while the AB has been proposed to reflect processing at a late, central stage (e.g., Chun & Potter, 1995; Di Lollo, Kawahara, Ghorashi, & Enns, 2005;
Shapiro, Raymond, & Arnell, 1994), EIB has been proposed to reflect relatively early spatiotemporal competition between a target and distractor, with emotional distractors dominating due to tendencies to prioritize emotional information (Wang et al., 2012).
The spatiotemporal competition account for EIB is grounded in findings that emotional distractors (relative to neutral distractors) primarily impair perception of temporally neighboring targets that appear in their same location, but not of targets that appear in a different location from the distractors (Most & Wang, 2011). This pattern is different from that observed in the AB, which typically involves impairment across multiple streams (Lunau & Olivers, 2010; Shih, 2000; but see Kristjansson & Nakayama, 2002).
The spatially specific impairment in EIB is observed when participants monitor two simultaneous rapid serial streams of images for a single target image. EIB occurs when the distractor and target appear in the same location as each other but not when they appear in different locations (Most & Wang, 2011). The spatiotemporal competition account can accommodate the spatially specific impairment in EIB.
However, a plausible alternative explanation is that such a pattern arises as a function of where participants look. In this scenario, participants may only be registering one 28 | Spatial specificity in EIB regardless of eye-gaze
stream of images at a time (and neglecting the other stream), which would result in a
Figure 5. Alternative account predictions. Predicted results according to an account that the spatially specific impairment in emotion-induced blindness results from participants fixating at only one stream at a time. See text for details.
pattern of performance strikingly similar to the results that were observed for EIB (see
Figure 5). This alternative explanation assumes that unless stimuli are fixated and attended, the images will not be processed. While this assumption goes against findings to show that emotional stimuli are processed even when not fixated or goal-relevant
(e.g., MacLeod, Mathews, & Tata, 1986), it is possible that the fast presentation of complex stimuli presented in two simultaneous streams may make the task demands too difficult to monitor both streams at the same time.
To illustrate the alternative account, consider the case in which participants fixate only one of two streams of images and the distractor appears in that stream (such that the distractor is “fixated”; see middle panel of Figure 5). In this case, when participants are fixating at the location of the distractor, targets appear in the same stream as the distractor and should elicit the typical pattern of emotion-induced blindness (like single stream versions of this effect), while targets that appear in the Spatial specificity in EIB regardless of eye-gaze | 29
opposite stream would likely be missed altogether and performance accuracy for reporting the target would be at chance (because participants are not fixating the stream where the target appears). Alternatively, consider the case in which participants are fixating the stream of images where the distractor does not appear (left panel in Figure
5). Targets that appear in the same stream as the distractor will likely be missed and performance would be at chance (because participants are not attending to that stream), but targets that appear in the opposite stream from the distractor will be well-reported, since participants were fixating to that stream and likely did not process the distractor in the other stream. Altogether, if performance was based on participants fixating to just one stream at a time, averaging across conditions where the distractor appeared in the stream participants were fixating and not fixating would yield results that only make it seem that impairment from emotional stimuli is limited to the “same-stream condition”.
Thus, the predictions in this account match the spatially specific pattern usually observed in two-stream versions of EIB. Lending credence to this alternative account, average baseline performance (when no distractor is presented) in correctly reporting the target rotation in the two-stream EIB task tends to be around 75% (e.g., Most &
Wang, 2011). As chance performance is 50% and perfect performance is 100%, baseline performance should average to around 75% if performance is a result of participants fixating only to one stream. As such, the current research cannot differentiate between the accounts of fixating to one stream at a time and the spatiotemporal competition account.
To tease apart these two potential accounts of the spatially specific impairment in EIB, we employed an eye-tracker and used gaze-contingent stimulus presentation to manipulate where distractors and targets appeared in relation to participants’ eye-gaze.
Gaze-contingent eye-tracking is a way to manipulate where stimuli appear depending on 30 | Spatial specificity in EIB regardless of eye-gaze
where a person is fixating. By placing the distractor in a specific position relative to a participant’s fixation location, we were able to control, at the very point at which the distractor appeared, whether it was placed in the fixated stream of images or the non- fixated stream of images. While the spatiotemporal competition account would predict
EIB both in the fixated and non-fixated stream conditions, the alternative fixate-to-one- stream account would instead predict EIB only in the fixated stream condition, not in the non-fixated stream condition (as illustrated in Figure 5).
Eye movements that occur after the appearance of the distractor are also likely to be informative for understanding the mechanisms underlying the spatially specific impairment in EIB. It is possible, for example, that attention may gravitate toward the emotional distractors after they occur in the trial, thus affecting the eye position when the target subsequently appeared. Previous studies have found gaze to move toward task-irrelevant but salient stimuli (Le Pelley, Pearson, Griffiths, & Beesley, 2015;
Theeuwes, Kramer, Hahn, Irwin, & Zelinsky, 1999). Emotional stimuli also seem to engage attention and are hard to disengage from (Fox, Russo, Bowles, & Dutton, 2001;
Mogg & Bradley, 1998; Vromen, Lipp, & Remington, 2014), and eye-tracking evidence suggests that individuals tend to fixate first and longer at emotional stimuli compared to neutral stimuli (Calvo & Lang, 2004; Nummenmaa, Hyönä, & Calvo, 2006). We therefore also collected eye-gaze throughout the entire trial, as a way to examine how participants shifted their eye-gaze following the presentation of the emotional distractors.
Experiment 1 tested the hypothesis that EIB would persist in trials when distractors and targets appeared in the same stream, regardless of where the participant was fixating - as predicted by the spatiotemporal competition account for EIB. This was in contrast to the prediction of the fixate-to-one-stream account, which would Spatial specificity in EIB regardless of eye-gaze | 31
predict EIB when the distractor and target appeared in the same stream only in the fixated distractor condition (and not in the non-fixated distractor condition). We further predicted that participants would shift their gaze toward negative distractors after they were presented.
Experiment 1
Method
Participants. Sixty-two participants were recruited from the community via the
University of New South Wales Paid Sona system (mean age=23.6 years, SD=5.8; 32 female, 30 male). Participants were compensated $15 for completing the experiment.
All participants gave informed consent and the experiment was approved by the
University of New South Wales Human Research Ethics Approval Panel.
Materials and Procedure. The experiment was conducted using a Tobii TX-300 eyetracker. The monitor was run with a refresh rate 60Hz, and the eyetracker had a
300Hz temporal and 0.15˚ spatial resolution. Stimuli were presented and responses made through the Psychophysics Toolbox for Matlab (Brainard, 1997; Pelli, 1997).
Head position was fixed via a chin rest ~60 cm away from the screen. Importantly, participants were allowed to move their eyes freely between the two streams of images throughout the experiment.
The experiment was composed of 20 blocks of 18 trials (360 trials in total). On every trial, participants saw a fixation point in the center of the screen for 500ms, then a blank screen for 200ms, and were then presented with two simultaneous, rapid streams of images (see Figure 6). Images were presented against a black background, and one image per stream was presented for one “frame”. There were 12 frames per trial, presented at a rate of 100 ms/frame. The two streams were vertically separated by 100 32 | Spatial specificity in EIB regardless of eye-gaze
pixels (each 50 pixels from the vertical center of the screen). Stimuli were images sized to 320 pixels wide and 240 pixels high.
Figure 6. Schematic of a partial trial in Experiment 1. Participants reported the direction of the one rotated picture that appeared in either of two simultaneously presented RSVP streams (presentation rate 100ms/frame). The distractor item (a colored negative image or colored scene image) appeared either one frame or two frames before the target, and either in the same stream or different stream from the target.
Every trial contained one target image, one distractor image, and 22 filler images to make up the remaining images in the two streams. 252 grayscale images of upright landscape and architectural scenes served as the filler images. The target image was always a colored landscape image that was rotated 90 degrees to the left or right. This target image came from a bank of eighty-four “target” images, and these images were rotated both clockwise and counterclockwise to appear in either direction.
An additional 160 images served as the “distractors”. Distractors were also colored images, but were not rotated. There were 80 negative, emotional distractors
(colored images depicting medical injuries, threatening animals, or grotesque scenes), and 80 “featural” distractors (colored images depicting upright landscape or Spatial specificity in EIB regardless of eye-gaze | 33
architectural scenes). Featural distractors were different images than those used as the filler images, but represented similar content and were also collected from publicly available sources. This was a deviation from previous emotion-induced blindness studies, which usually use “neutral” images – neutral images of people or animals - to compare performance with emotional distractors (e.g., Most & Wang, 2011). The reason for the change was to minimize the amount of potentially “meaningful” content displayed in the images. The neutral distractors do usually impair target performance in the direction of spatial localization (particularly at very early lags; Most & Wang, in revision), likely because they also represent “meaningful” content (people or animals) compared to the filler items. Featural distractors in this experiment did not represent meaningfully different content from the filler images, but did differ because of their color feature and thus were still distracting without representing any kind of meaningful content, because it shared this color feature with the target. This was important as a way to compare two physically salient stimuli types (colored negative and colored featural) to purely isolate the effects of distraction by the “meaningful” content in the negative distractors. Negative, emotional distractors were gathered from the
International Affective Picture System (IAPS; Lang, Bradley, & Cuthbert, 2008) and from publicly available sources.
There were an equal number of trials with negative and featural distractors.
There were also 40 additional trials with “no distractor” (2 per block) – another “filler” image was placed in the stream where a distractor would have been presented on these trials. Depending on the trial, the distractor appeared at serial position 3 through 7, and the target appeared either one position (lag-1) or two positions (lag-2) after the distractor. We expected performance to be impaired by negative distractors at both lag 34 | Spatial specificity in EIB regardless of eye-gaze
1 and lag 2 based on previous EIB studies (Kennedy & Most, 2015). Every distractor was presented once in the experiment at lag-1, and once at lag-2.
The placement of the distractor was manipulated in relation to the participant’s point of fixation. Depending on the trial, the distractor was presented either in the same stream that participants were fixating or in the opposite stream (non-fixated). The position of the target was manipulated orthogonally to this factor: on half of all trials the target would appear in the same stream as the distractor and on the remaining half it would appear in the opposite stream.
At the end of the trial, participants indicated the direction that the target image was rotated. Participants heard a bell through headphones if they answered correctly but heard nothing if they answered incorrectly.
Before starting the experiment, participants engaged in a 5-point calibration procedure on the eye-tracker. They then started the EIB task, first with 8 practice trials to get used to the task. Practice trials did not have any distractors, started at
200ms/frame, and progressed to the experiment speed of 100ms/frame. Practice trials were not included in the analyses.
Gaze-contingent analysis
Participants’ eye-gaze was tracked throughout the entire experiment. On every distractor-present trial, the distractor placement was determined by where the participant was fixating. (On baseline “no distractor” trials, the distractor was placed either in the fixated or non-fixated stream.) This was achieved by measuring eye position during the 200-ms immediately prior to the distractor onset. The algorithm then searched backwards through this period of eye-gaze data for a block of 50ms of “valid” eye-gaze (i.e., data without missing samples due to blinks). This 50ms could therefore Spatial specificity in EIB regardless of eye-gaze | 35
be attributed to an area of interest that was either one of the streams of images or the background. A participant was determined to be fixating at one stream of images if gaze was detected within 25 pixels of the image above or below the center of the screen.
In this case, the distractor was then presented in one of the two streams, as determined by the trial type (same or different stream).
If participants were fixating in an otherwise blank region of the screen (in the center of the screen, or to the left or right of the images) during the time when the contingency took place, the trial was discounted and the participant was considered to not have fixated at either stream. In the case that participants were not fixating at either stream, the distractor appeared randomly in one of the two streams. These trials were excluded from the analyses.
Results
Since only trials with valid gaze-contingency were included in the analyses, there was a degree of variability in the number of trials contributing to the analyses across participants. The median number of gaze-contingent trials across all conditions per participant was 255.5 trials (mean=229.3 trials; SD=75.2 trials), with a range from
62 to 346 trials out of the total 360 trials (first quartile=176 trials; third quartile=283.5 trials). All participants had data for each of the trial conditions. Data were collapsed across lags 1 and 2.1
1 The results were the same when we included lag into the analyses. A 2 (distractor fixation: fixated vs non-fixated) X 2 (distractor-target relationship: same stream vs different stream) X 2 (lag: 1 vs 2) X 2 (distractor type: negative vs featural) revealed significant effects of distractor fixation, F(1,61)=17.268, 2 2 p<.001, ηp =.221, distractor-target relationship, F(1,61)=21.870, p<.001, ηp =.264, and distractor type, 2 F(1,61)=48.330, p<.001, ηp =.442, but as predicted, no significant effect of lag, F(1,61)=0.006, p=.939, 2 ηp <.001. Like the analyses collapsed across lag, there was a significant interaction between distractor- 2 target relationship X distractor type, F(1,61)=66.231, p<.001, ηp =.521, and distractor fixation X stream, 2 F(1,61)=53.715, p<.001, ηp =.468, and no significant interaction between distractor fixation X distractor 2 type, F(1,61)=0.019, p=.892, ηp <.001, or distractor fixation X distractor-target relationship X distractor 2 type interaction, F(1,61)=1.411, p=.240, ηp =.023. 36 | Spatial specificity in EIB regardless of eye-gaze
Target Performance Accuracy
Figure 7. Experiment 1 target accuracy. In Experiment 1, impairment from emotional distractors was localized both when the distractor was fixated and non-fixated. EIB was observed when targets appeared in the same stream as distractors – regardless of whether the participants were fixating on the distractor stream or when fixating at the opposite stream. When the target and distractor appeared in different streams, no EIB was observed. Error bars represent standard error.
We used target accuracy (correctly reporting the direction of the rotated target)
as our primary measurement of interest (see Figure 7). A 2 (distractor fixation: fixated
vs. non-fixated) X 2 (distractor-target relationship: same stream vs. different stream) X
2 (distractor type: negative vs. featural) ANOVA collapsed across lag revealed a
2 significant effect of distractor fixation, F(1,61)=14.825, p<.001, ηp =.196, with
generally better target accuracy when the distractors were in the non-fixated stream.
That is, when distractors appeared in the stream that participants were not fixating,
participants were better able to report the target rotation. There was also a significant
2 main effect of distractor-target relationship, F(1,61)=21.341, p<.001, ηp =.259, with
worse overall accuracy when the target was positioned in the opposite stream to the
distractor compared to when the target was in the same stream as the distractor. We
will discuss this main effect more in the following paragraphs. The main effect of
Spatial specificity in EIB regardless of eye-gaze | 37
2 distractor type was also significant, F(1,61)=53.770, p<.001, ηp =.469, such that negative distractors elicited worse performance than featural distractors.
The interaction between distractor-target relationship and distractor type was
2 significant, F(1,61)=74.367, p<.001, ηp =.549, as predicted, with greater emotion- induced impairment when the distractor and target appeared in the same stream than when they appeared in opposite streams. There was also a significant interaction between distractor fixation and distractor-target relationship, F(1,61)=55.289, p<.001,
2 ηp =.475, with better performance when the target appeared in the stream that was fixated when the distractor appeared (also discussed below). There was no significant interaction between distractor fixation and distractor type, F<1, or between all three
2 factors, F(1,61)=1.233, p=.271, ηp =.020. The non-significant 3-way interaction suggests that the distractor-target relationship and distractor type interaction was similar both when the distractor was fixated and non-fixated, consistent with a spatiotemporal competition account for emotion-induced blindness.
EIB was localized to the coordinates of the distractor regardless of eye-gaze.
Both when the distractor was fixated and non-fixated, negative distractors impaired performance significantly more than featural distractors when the target appeared in the same stream (fixated: t(61) =7.622, p < .001, dz =0.947, non-fixated: t(61) =7.862, p < .001, dz =0.999). When the distractor and target appeared in different streams, there was no difference in the impairment from negative distractors compared to featural distractors in either the fixated, t(61) =1.296, p = .200, dz =0.165, or non-fixated, t(61) =0.044, p =
.965, dz =0.006, conditions.
An influence of distractor fixation was also revealed in several conditions of our experiment. Data from the “no distractor” conditions (trials in which no colored distractor was present) were not included in the ANOVA described above but did reveal 38 | Spatial specificity in EIB regardless of eye-gaze
two important insights. First, they demonstrated that performance was impaired in both negative and featural distractor conditions compared to the “no distractor” baseline performance – suggesting that both distractor types impaired target performance compared to when there was no distractor in the stream. They also suggested that baseline performance was affected by where participants were fixating, as performance in the two baseline conditions differed significantly, t(61) = 4.618, p < .001, dz = 0.586.
In “no distractor” trials, participants performed better when the target appeared in the stream participants were fixating soon before the target was presented (M=92.8%,
SD=10.5%), compared to when the target appeared in the opposite stream they were fixating (M=85.6%, SD=13.4%). Likewise, as mentioned above, there was a significant main effect of distractor fixation, a main effect of distractor-target relationship, and also a distractor fixation by distractor-target relationship interaction. Notably, when the target was in a different stream from the distractor, performance was worse when the distractor was fixated (negative: t(61) = 6.242, p < .001, dz =0.793, featural: t(61) =
5.575, p < .001, dz =0.708). This was not surprising, since when the distractor and target appeared in different streams, a non-fixated distractor indicates that participants were fixating at the stream the target would soon appear, while the opposite is true for a fixated distractor. As such, we did seem to find an effect of where participants were fixating on ability to report the target, however the spatially specific impairment (EIB) was not accounted for simply by where participants were fixating.
Eye-gaze Data Processing
For our secondary analysis, we examined if participants systematically shifted their eye-gaze after the distractor was presented. In this analysis, shift in eye-gaze was referenced to the placement of the distractor, such that gaze shifted either toward or away from the center of the stream in which the distractor appeared. As such, we could Spatial specificity in EIB regardless of eye-gaze | 39
determine where participants shifted their gaze respective of where the distractor appeared.
Similar to behavioral results, we only analyzed valid trials in which the gaze- contingency occurred. Eye-gaze shift was defined as the difference between the
Figure 8. Experiment 1 post-distractor eye-gaze results. For simplicity, data for the figure were collapsed across distractor-target relationship. Participants shifted their gaze more toward featural distractors than negative distractors. Error bars represent standard error. See text for details.
average gaze before the distractor presentation and the average eye-gaze after the distractor presentation. The average x- and y-coordinates of the participant’s gaze during two time periods were used to determine the average gaze. Eye-gaze before the distractor presentation was defined by the averaged gaze data from the time period of the gaze-contingency window (~0-200ms before the distractor was presented). Eye- gaze after the distractor presentation was defined by the averaged eye-gaze data from the time of the distractor presentation to the remainder of the trial afterward.2 Eye-gaze
2 We chose to examine the entire remaining trial in order to allow enough time for eye-movements to eventuate. Indeed, at the longest, the temporal distance between distractor presentation and the end of the trial was only 900ms, leaving little time for subsequent eye-movements to occur. Still, one might argue that this analysis leaves room for contamination of other eye-movements that are not a direct effect of the 40 | Spatial specificity in EIB regardless of eye-gaze
shift was then calculated by taking the difference in movement along the y-axis between the period before the distractor and the period after the distractor. This change reflected the average number of vertical pixels participants shifted their eye-gaze respective of the distractor location (since the two RSVP streams were separated vertically). We then averaged trials within each condition, to see if participants systematically shifted their eye-gaze in a certain direction when the distractor was presented.
The amount of average eye-gaze shift toward the distractor was used as the dependent variable in this analysis. Note that this shift in gaze is the mean of each participant’s mean shift in gaze. Thus, while the average movement may be small, this variable represents an average of trials ranging from when participants made a substantial change in gaze and trials when the participant did not move the eye-gaze much at all. A 2 (distractor fixation) X 2 (distractor-target relationship) X 2 (distractor type) ANOVA revealed a significant main effect of distractor type, F(1,61)=7.244,
2 p=.009, ηp =0.106. Participants were more likely to shift their eye-gaze toward the distractor if it was a featural distractor, compared to if it was a negative distractor (see
Figure 8). However, they did not always shift their eye-gaze toward the distractor.
Participants generally shifted their eye-gaze away from the distractors when they were non-fixated, but still shifted eye-gaze more toward the distractors when they were featural than when negative. Nevertheless, there was no significant main effect of
2 distractor fixation, F(1,61)=1.754, p=.190, ηp =0.028. There was also no significant
distractor alone. To address this possibility, we also analyzed the average eye-gaze change between the contingency window and 0-100ms after distractor presentation revealed no significant main effects or interactions between factors (via a 2 (distractor fixation) X 2 (distractor-target relationship) X 2 (distractor type) ANOVA, ps>.05). Nevertheless, even at this early window, the direction of eye-gaze change depended on distractor fixation, such that participants tended to look more toward fixated distractors, and further away from non-fixated distractors. Thus, even though the effect was not significant, the pattern of eye-gaze change mimicked the results at the later window, suggesting that eye- gaze changes may need time to eventuate.
Spatial specificity in EIB regardless of eye-gaze | 41
2 main effect of distractor-target relationship, F(1,61)=1.541, p=.219, ηp =0.025, nor any significant interaction between factors (ps>.05).
Discussion
Emotion-induced blindness (EIB) was observed when the distractor and target appeared in the same stream but not when they appeared in different streams, regardless of where participants were fixating when the distractor was presented. This is consistent with a spatiotemporal competition account for EIB, such that the spatially specific impairment in EIB cannot be attributed to where participants were fixating when the distractor appeared.
We did observe some effects of where participants were fixating. Participants were more likely to correctly report targets that appeared at the stream they were fixating than targets that appeared in the opposite stream they were fixating. However, the spatially specific impairment was observed above and beyond the influence of where participants fixated.
The shift in participants’ gaze after the distractor presentation differed based on the type of distractor. Participants shifted their gaze more toward featural distractors than negative distractors – which was different from our prediction that participants would shift their gaze more toward the negative distractors. In some ways, however, this finding is consistent with accounts that participants attempt to “suppress” the distracting negative distractors (Kennedy, Rawding, Most, & Hoffman, 2014). It may be that participants attempt to shift their gaze away in an effort to suppress the salient distractors from disrupting their main task. The shift of gaze, however, could also not explain the spatially localized impairment in target accuracy. If participants were shifting their gaze further away from negative distractors, its effect on subsequent 42 | Spatial specificity in EIB regardless of eye-gaze
targets would be the same no matter if targets appeared in the same or different streams
(since they looked away from negative distractors both when the distractor was fixated and non-fixated).
Experiment 2
While EIB demonstrates a spatially localized impairment, several attentional blink
(AB) studies demonstrate that performance impairment from a task-relevant target will spread across spatial locations (Lunau & Olivers, 2010; Shih, 2000). One quality that differentiates EIB and AB tasks is the task-relevance of the first attention-grabbing stimulus: in the AB participants have to identify the first target (T1), whereas in EIB distractors are best ignored. As such, the spatially specific impairment from emotional distractors may hinge on their task-irrelevance – perhaps if distractors are relevant to the task, they would also impair across space like AB patterns. Previous research demonstrates that task-relevant emotional distractors impair target detection more than task-irrelevant distractors in a traditional, one-stream, emotion-induced blindness task
(explored more in Chapter 3). This also suggests that task-relevant emotional stimuli may impact additional or different perceptual mechanisms relative to task-irrelevant ones.
According to one model of attentional dynamics within rapid serial visual presentations, perceptual failures can stem from several information processing bottlenecks. For example, stimuli that appear close in time and in the same location compete with each other in a mutually suppressive manner. In this case, stimuli with particular salience (such as emotional stimuli) can gain the competitive edge (Wyble &
Swan, 2015; see also Keysers & Perrett, 2002). This “competitive interference” yields spatially localized perceptual deficits. In contrast, when stimuli are selected for encoding into visual working memory, this process causes a suppression of attention Spatial specificity in EIB regardless of eye-gaze | 43
across the visual field (Wyble & Swan, 2015). It may be that relative absence of spatially localized effects in the AB stem from the fact that people are required to encode T1 into memory – something they are not required to do in EIB.
In Experiment 2, we included a recognition test for the distractors in order to render distractors relevant to the task and to encourage encoding of them into memory.
The aim of Experiment 2 was to determine if the spatially localized impairment from emotional distractors remains regardless of their task-relevance, or if task-relevant emotional distractors mimic the attentional blink phenomenon, instead making the pattern of impairment spread across space.
Method
Participants. Fifty-nine participants completed Experiment 2 and were recruited through the community via the University of New South Wales “Paid Sona” system
(mean age=25.4 years, SD=6.8; 35 female, 24 male). Participants were compensated
$15 for completing the study. Data from three participants (two male) were excluded from the analyses: two performed at or below chance, while the other fixated at one of the streams in only seven trials throughout the entire experiment, providing too little data for the contingency manipulation to incorporate into the analysis. All participants gave informed consent and the experiment was approved by the University of New
South Wales Human Research Ethics Approval Panel.
Materials and Procedure. Experiment 2 was designed the same as Experiment 1 with some exceptions. In Experiment 2, due to an oversight in experimental design, the “no distractor” condition always presented the target in the distractor fixated stream. This change was not important for the main analyses, but did differ from Experiment 1, such 44 | Spatial specificity in EIB regardless of eye-gaze
that the baseline performance was only measured when participants fixated at the stream the target would soon appear. This was only for the baseline condition.
In contrast to Experiment 1, participants completed a memory test for the colored distractors at the end of each block of 18 trials. Participants were told to remember the colored distractor in each trial. This change rendered the distractors task-relevant, as in the typical AB. There were 160 additional images (80 negative and 80 featural) in
Experiment 2 that served as foils in the memory tests. These images matched the negative distractors and the featural distractors in content type and emotional quality, but were never presented in the EIB trials. At the end of each block, participants saw a screen with 16 negative images arranged in a square shape made up of 8 images that had actually appeared in that block, and 8 foils that had not. Participants were given the instructions that “Eight of these pictures appeared in the most recent block. Please click on them.” When a participant chose an image, it was surrounded with a white border, and they could not choose it again. After they chose the 8 negative images that they believed had been presented, participants were shown a screen with 16 colored scene images, and were instructed to do the same (again, 8 images had actually appeared in that block as featural distractors, and 8 foils had not). The next block began after the memory test.
Results
The median number of gaze-contingent trials per participant was 266.5 trials
(mean=248.6 trials, SD=82.4 trials), with a range from 57 to 357 trials out of the total
360 trials (first quartile=187 trials; third quartile=314.5 trials). All participants had data Spatial specificity in EIB regardless of eye-gaze | 45
for each of the trial conditions. Like Experiment 1, data were collapsed across lags 1
and 2.3
Target Performance Accuracy
Like Experiment 1, the main variable of interest in Experiment 2 was accuracy
in reporting the target’s rotation (see Figure 9). A 2 (distractor fixation: fixated vs. non-
fixated) X 2 (distractor-target relationship: same stream vs. different stream) X 2
(distractor type: negative vs. featural) ANOVA collapsed across lag revealed a
Figure 9. Experiment 2 target accuracy. In Experiment 2, EIB was observed when targets appeared in the same stream as distractors – both when participants fixated at the distractor stream and when fixating at the opposite stream. When the target and distractor appeared in different streams, EIB was also observed. However, the emotion-induced impairment was greater when distractors and targets appeared in the same stream, compared to when they appeared in opposite streams. Error bars represent standard error.
2 significant effect of distractor fixation, F(1,55)=18.050, p<.001, ηp =.247, like
Experiment 1, with better target performance when the distractor was non-fixated than
when fixated. Also similar to Experiment 1, there was a main effect of distractor-target
2 relationship, F(1,55)=18.890, p<.001, ηp =.256, with worse performance when the
3 Consistent with Experiment 1, lag did not affect the main findings in Experiment 2. A 2 (distractor fixation: fixated vs non-fixated) X 2 (distractor-target relationship: same stream vs different stream) X 2 (lag: 1 vs 2) X 2 (distractor type: negative vs featural) revealed significant main effects of distractor 2 fixation, F(1,55)=17.895, p<.001, ηp =.245, distractor-target relationship, F(1,55)=21.038, p<.001, 2 2 ηp =.277, and distractor type, F(1,55)=82.467, p<.001, ηp =.600, but no significant effect of lag, 2 F(1,55)=0.232, p=.632, ηp =.004. 46 | Spatial specificity in EIB regardless of eye-gaze
distractor and target appeared in different streams. The main effect of distractor type
2 was also significant, F(1,55)=82.573, p<.001, ηp =.600, with worse performance after negative distractors compared to featural distractors.
As predicted, and consistent with Experiment 1, there was also a significant interaction between distractor-target relationship and distractor type, F(1,55)=15.229,
2 p<.001, ηp =.217 - with greater emotion-induced impairment when distractors and targets appeared in the same stream compared to when they appeared in different streams. The distractor fixation by distractor-target relationship interaction was also
2 significant, F(1,55)=43.577, p<.001, ηp =.442, similar to Experiment 1, to reflect an effect of distractor fixation on target accuracy – targets were better reported when they appeared in the stream participants were fixating when the distractor appeared. There was again no significant interaction between distractor fixation and distractor type,
2 2 F(1,55)=1.763, p=.190, ηp =.031, or all three factors, F(1,55)=0.703, p=.406, ηp =.013.
Like Experiment 1, the impairment from negative compared to featural distractors was greater when the distractor and target appeared in the same stream compared to different stream. As predicted, there were significant differences in the same stream conditions in both the distractor fixated, t(55) =8.176, p < .001, dz =1.092, and distractor non-fixated condition, t(55) =6.747, p < .001, dz =0.902. However, in contrast to Experiment 1, when the distractor and target were in different streams, the negative distractors also significantly impaired performance compared to the featural distractors in both the distractor fixated, t(55) =3.725, p < .001, dz =0.498, and distractor non-fixated conditions, t(55) =3.509, p=.001, dz =0.469. Thus, while the emotion- specific impairment was greater when the distractors appeared in the same stream as the distractor, the negative distractors also impaired more when they appeared in different streams from the distractor. Spatial specificity in EIB regardless of eye-gaze | 47
Performance in the baseline “no distractor” condition was generally higher than performance in the distractor-present trials (M=93.2, SD=7.8), confirming that both featural and negative distractors impaired performance (ps>.05).
Like Experiment 1, the main effect of distractor fixation, main effect of distractor-target relationship, and the distractor fixation by distractor-target relationship interaction suggests that performance benefited when participants were fixating at the location in which the target was soon presented. This was again particularly pronounced in conditions when the target was in a different stream from the distractor, with better performance when participants were fixating the stream where the target would soon appear, compared to when they fixated the opposite stream from where the target would soon appear (negative: t(55) =5.668, p < .001, dz =0.758; featural: t(55)
=6.385, p < .001, dz =0.853). Thus, performance was affected by where participants were fixating just before the target appeared, but like in Experiment 1, this did not seem to affect the emotional, spatially specific impairment.
48 | Spatial specificity in EIB regardless of eye-gaze
Memory Performance
We next examined the results from the memory tests for distractors, to see if the different trial types affected memory for the distractors in the streams (Figure 10).
Memory accuracy was calculated as the percentage of correct responses on the memory test (chance performance was 4/8=50%, perfect performance was 8/8=100%).
Participants did not remember distractors with great accuracy in either the negative
(M=58.6%, SD=6.7%) or featural (M=49.5%, SD=4.9%) condition. Nevertheless, there
Figure 10. Experiment 2 distractor memory performance. Participants remembered distractors better when they were negative compared to when they were featural distractors. They also remembered distractors better when the distractors were fixated, compared to non-fixated. Error bars represent standard error.
was a difference in memory performance based on trial condition. A 2 (distractor fixation) X 2 (distractor-target relationship) X 2 (distractor type) revealed significant
2 main effect of distractor fixation, F(1,55)=25.895, p<.001, ηp =0.320, with better memory for distractors that were fixated. There was also a significant main effect of
2 distractor type, F(1,55)=107.933, p<.001, ηp =0.662, with better memory for negative distractors. There was no significant main effect of distractor-target relationship, F<1.
There was also a significant distractor type by target fixation interaction, Spatial specificity in EIB regardless of eye-gaze | 49
2 F(1,55)=11.854, p=.001, ηp =0.177. Participants better remembered negative distractors when they were fixated compared to non-fixated, but did not better remember featural distractors when fixated compared to non-fixated. No other interaction between factors reached statistical significance (ps>.05). Together, these results demonstrate that negative distractors were remembered better than featural distractors, and negative distractors in the fixated stream were remembered better than those in the non-fixated stream.
Eye-gaze Data Processing As in Experiment 1, we examined if participants systematically shifted their eye- gaze position after the distractor was presented respective of where the distractor
Figure 11. Experiment 2 post-distractor eye-gaze results. For simplicity, data for the figure were collapsed across distractor-target relationship. Participants shifted their gaze more toward featural distractors than negative distractors. They also shifted their gaze more toward fixated distractors, and further away from non-fixated distractors. Error bars represent standard error. See text for details. 50 | Spatial specificity in EIB regardless of eye-gaze
appeared. Again, the amount of average eye-gaze shift toward the distractor over the remaining trial was used as the dependent variable in this analysis.4 Like in Experiment
1, a 2 (distractor fixation) X 2 (distractor-target relationship) X 2 (distractor type)
ANOVA revealed a significant main effect of distractor type, F(1,55)=10.243, p=.002,
2 ηp =0.157. Participants were more likely to shift their eye-gaze toward the distractor if it was a featural distractor, compared to if it was a negative distractor (see Figure 11).
In Experiment 2, there was also a significant main effect of distractor fixation,
2 F(1,55)=10.820, p=.002, ηp =0.164. Participants were more likely to shift their gaze toward the location of a fixated distractor, and away from the location of a non-fixated distractor. There was no significant main effect of distractor-target relationship,
2 F(1,55)=1.552, p=.218, ηp =0.027, nor any significant interaction between factors
(ps>.05).
Comparing target performance accuracy between experiments
In Experiment 2, participants demonstrated impaired performance after emotional distractors both when the distractor and target appeared in the same stream and when they appeared in different streams. Taken together with the results of
Experiment 1, this suggests the task-relevancy of distractors impacted performance to further spread the emotional impairment across space, in addition to the spatial specific impairment usually demonstrated in EIB.
To examine if the differences between the experiments were significant, we compared target performance accuracy in Experiment 1 and Experiment 2. A 2
4 As in Experiment 1, analysis of average eye-gaze change between the contingency window and 0-100ms after distractor presentation revealed a similar, though weaker trend. In Experiment 2, a 2 (distractor fixation) X 2 (distractor-target relationship) X 2 (distractor type) ANOVA revealed a significant main 2 effect of distractor fixation (F=11.751, p<.001, ηp =0.176), such that participants significantly moved their eyes toward fixated distractors, and away from non-fixated distractors. All other main effects and interactions were non-significant (ps>.05). We take this to reflect that the eye-gaze changes observed may need time to eventuate (particularly differences elicited by distractor type), but seem to develop soon after the distractor presentation.
Spatial specificity in EIB regardless of eye-gaze | 51
(distractor fixation) X 2 (distractor-target relationship) X 2 (distractor type) X 2
(experiment) ANOVA revealed no significant main effect of experiment,
2 F(1,116)=0.847, p=.359, ηp =0.007, but did reveal a significant distractor-target relationship X distractor type X experiment interaction, F(1,116)=5.831, p=.017,
2 ηp =0.048, suggesting that the spatially localized impairments following negative distractors differed across experiments. Subsequent ANOVAs revealed that the distractor type X experiment interaction was significant in the different stream
2 conditions, F(1,116)=7.542, p=.007, ηp =0.061, but not the same stream conditions,
F(1,116)<1.
The main difference between experiments was the task-relevancy of the distractors. We take these data to suggest that whether task-relevant or not, emotion- induced impairment was observed when distractors and targets appeared in the same stream – and to the same extent – suggesting a competition at early representational level. However, when the distractor was relevant to the task, emotion-induced impairment was also observed when the distractor appeared in the opposite stream to fixation, but was not observed when the distractor was task-irrelevant – suggesting that task-relevance may additively impose a competition between distractor and targets at later memory stages.
Discussion
In Experiment 2, participants searched for targets that appeared in either of the two streams, and a task-relevant distractor appeared either in the same or opposite stream shortly before it. Similar to Experiment 1, negative distractors elicited greater impairments when the distractor and target appeared in the same stream than when they appeared in different streams, regardless of where participants were fixating. However, likely due to the task-relevancy of distractors in Experiment 2, EIB was also observed 52 | Spatial specificity in EIB regardless of eye-gaze
when the distractor and target appeared in different streams. These results suggest that the impairment from task-relevant emotional distractors may represent an additive effect of spatially localized representational competition and task-relevance.
Interestingly, this additional impairment from task-relevancy was specific to emotional distractors and not featural distractors, consistent with the finding that emotional distractors were remembered better than featural ones. While a basic additive account of task-relevancy might predict an additional decrease in performance after both for the negative and featural distractors, the memory performance for featural distractors suggest that they may not have posed any competition in memory, as they were poorly remembered later in the experiment. Thus, it may be that only negative distractors were additionally distracting with task-relevance because of their memorability. The task-relevance also specifically impaired performance to a greater extent in the different stream, rather than same stream, conditions. This is consistent with a framework that distractors and targets compete for early representation when presented in the same space, but not when they appear in different locations. If that competition has not taken place, then intact target representations may be susceptible to later influences of competition: in this case, interference in a central, working memory process with the distractor. However, no further impairment will occur for targets that compete at an early level with the distractors, simply because there is no representation to compete in a later process of working memory consolidation. Note that this is consistent with results of Chapter 3, which suggest that targets presented at an early lag after an emotional distractor are not impacted by task relevance. Targets presented at later lags, however, when the competition is less strong, may be able to gain competitional strength when task-relevant. Spatial specificity in EIB regardless of eye-gaze | 53
Participants shifted their eye-gaze after the distractor presentation differently depending on the type of trial. Similar to Experiment 1, participants shifted their gaze more toward featural distractors than negative distractors. In Experiment 2, participants also shifted their gaze more toward distractors that they were fixating, and shifted their gaze further away from distractors they were not fixating. That is, participants shifted their gaze more toward the stream that they were fixating when the distractor appeared.
While highly speculative, it could be that participants were in a state of “temporary loss of control” after the presentation of a task-relevant distractor (e.g., Di Lollo et al.,
2005), and with the dual task of remembering the distractor and reporting the subsequent target, participants attempted to focus on information that was most available to them at the time. These eye-gaze movements still cannot explain the behavioral target accuracy results, but may be informative in the understanding of the later processing stages after emotional distraction.
General Discussion
Previous studies demonstrate that EIB is a spatially localized effect – emotional distractors impair performance for targets that appear in their same location, but do not impair performance for targets that appear in a different location (Most & Wang, 2011).
This pattern of data has been used to support a spatiotemporal competition account of
EIB, which suggests that emotional distractors and subsequent targets compete for representation (Wang et al., 2012). Until now, an alternative account for these findings has been that the spatially localized pattern is a result of participants only fixating to one stream at a time. In this study, we tested whether the spatially specific pattern observed in EIB persists regardless of fixation. A spatiotemporal competition account would predict EIB to occur when the distractor and target appear in the same location no 54 | Spatial specificity in EIB regardless of eye-gaze
matter where the participant is fixating, while the alternative account predicts EIB only when the distractor and target appear in the stream a participant is fixating.
We found that EIB was spatially localized independent of where participants were fixating, such that emotional distractors specifically impaired accuracy for targets that appeared in their same location regardless of whether they were fixating where the distractor appeared or the opposite location. These results rule out the alternative explanation that the pattern is an artifact of participants only fixating to one stream at a time, and support an account that EIB reflects a spatiotemporal competition between the distractor and target (Wang et al., 2012).
While we found greater EIB when the distractor and target appeared in the same location (compared to different locations) in both experiments, we further observed EIB when they appeared in different locations when participants were asked to encode the distractors into memory (Experiment 2). Studies of the phenomenally similar attentional blink studies typically find impairment to spread across spatial locations
(Lunau & Olivers, 2010; Shih, 2000; but see Kristjansson & Nakayama, 2002). By making distractors task-relevant in Experiment 2, they likely gained the properties of the first targets (T1s) in the AB, since participants were actively looking for them.
Notably, this was specific to negative distractors compared to featural distractors, despite both being task-relevant distractors. However, this likely remained specific to negative distractors because of their ability to be remembered as more distinct stimuli than landscapes.
Intriguingly, the contingent capture literature provides evidence of performance benefits when distractors are presented in the same stream as targets. In such experiments, distractors which share a defining property with the target (e.g., color) are less distracting when presented in the same stream as the target compared to when Spatial specificity in EIB regardless of eye-gaze | 55
presented in a peripheral stream (e.g., Moore & Weissman, 2011). According to these results, it may be surprising that such benefits are not observed in the present study for featural distractors, given that both distractors and targets shared the featural property of color in streams of black and white filler items. However, a difference between these tasks is the nature of distractors and targets. In the present study, distractors were often emotionally powerful and distracting because of their meaningful content. Featural distractors may also distract more because participants want to avoid anything other than the target, knowing it may be distracting because of possible emotionality.
Furthermore, the important, defining information about the target was its rotation, not its color feature. This would make the featural distractors depicting upright landscape images conflict with target information. With time, however, participants may learn to utilize featural distractors to cue the upcoming target. Indeed, in one study, featural distractors did improve target performance (Most & Wang, in revision), but the participants were recruited only after participation in another RSVP experiment.
Additional practice in the task may therefore lead to enhancement, rather than impairment, from these featural distractors.
Comparison between experiments enabled us to examine whether the degree of spatial localization differed significantly between experiments, but future studies should continue to examine the effect of task-relevancy on the patterns of emotion-induced blindness. Altogether, in Experiment 2, we found that emotional distractors seemed to act like the T1s in the attentional blink, while still impairing performance at the same location more. This suggests potentially additive impacts of spatiotemporal competition and suppression from memory encoding in EIB, which may indeed reflect two distinct mechanisms. 56 | Spatial specificity in EIB regardless of eye-gaze
In both experiments, after a distractor appeared, participants were more likely to shift their gaze toward featural distractors than they were to shift their gaze toward negative distractors. This may suggest that participants were attempting to suppress the negative distractors from interfering with their task. This would be consistent with previous event-related potential findings that emotional distractors seemed to elicit a PD component (related to suppressing salient distractors) in the emotion-induced blindness task (Kennedy et al., 2014). Furthermore, in Experiment 2, participants were more likely to shift their eye-gaze toward distractors that were fixated, and away from distractors that were not fixated. Phrased a different way, participants were more likely to shift their eyes toward the stream that they were fixating. These results were somewhat surprising but may suggest that participants attempted to focus on information that was most available after being distracted. The eye-gaze shifts cannot explain the behavioral results of spatially localized impairment, as participants reported targets differently when the distractor and target appeared in the same or opposite streams, despite shifting their gaze in a similar pattern. However, the shifts in gaze can inform the later processing stages in emotional distraction, such that – at least in an emotion-induced blindness task – participants seem to avoid allocating further attention to negative distractors.
As predicted, spatially localized EIB was observed in both fixated and non- fixated streams. However, we did find some effects that were determined by where participants were fixating. Overall, targets were better identified when they appeared in the stream participants were fixating, and distractors were better remembered when they were fixated. Eye-gaze is widely used as a marker of attention. It is a measure specifically of overt attention (as opposed to covert attention, see Posner, 1980) and while research suggests that eye-gaze is guided by covert attention, both are still Spatial specificity in EIB regardless of eye-gaze | 57
separate systems (Hoffman, 1998). It is therefore possible that covert attention could underlie the patterns of impairment beyond that which we were able to capture with an eye-tracker. Nevertheless, the spatial specific pattern, eye-gaze shifts after the distractor, and additive nature of task-relevance across both streams support the scenario that participants were engaged with both streams throughout the trials.
Together, the results of this study suggest that in EIB, emotional distractors specifically impair the detection of targets that appear in their same location regardless of where participants are fixating, which is consistent with a spatiotemporal competition account (Wang et al., 2012). Future research should continue to explore the dynamics of this competition to understand its underlying mechanisms, particularly related to the specificity of EIB in space.
58 | Spatial specificity in EIB regardless of eye-gaze
References
Arnell, K. M., Killman, K. V, & Fijavz, D. (2007). Blinded by emotion: target misses
follow attention capture by arousing distractors in RSVP. Emotion, 7(3), 465–477.
http://doi.org/10.1167/4.8.359
Brainard, D. H. (1997). The Psychophysics Toolbox. Spatial Vision, 10(4), 433–436.
http://doi.org/10.1163/156856897X00357
Calvo, M. G., & Lang, P. J. (2004). Gaze patterns when looking at emotional pictures:
Motivationally biased attention. Motivation and Emotion, 28(3), 221–243.
http://doi.org/10.1023/B:MOEM.0000040153.26156.ed
Chun, M. M., & Potter, M. C. (1995). A two-stage model for multiple target detection in
rapid serial visual presentation. Journal of Experimental Psychology: Human
Perception and Performance, 21(1), 109–27. http://doi.org/10.1037/0096-
1523.21.1.109
Di Lollo, V., Kawahara, J., Ghorashi, S. M. S., & Enns, J. T. (2005). The attentional
blink: resource depletion or temporary loss of control? Psychological Research,
69(3), 191–200. http://doi.org/10.1007/s00426-004-0173-x
Fox, E., Russo, R., Bowles, R., & Dutton, K. (2001). Do threatening stimuli draw or
hold visual attention in subclinical anxiety? Journal of Experimental Psychology:
General, 130(4), 681–700. http://doi.org/10.1037/0096-3445.130.4.681
Hoffman, J. E. (1998). Visual attention and eye movements. In H. Pashler (Ed.),
Attention (pp. 119–153). Hove, United Kingdom: Psychology Press. Spatial specificity in EIB regardless of eye-gaze | 59
Kennedy, B. L., & Most, S. B. (2015). The rapid perceptual impact of emotional
distractors. Plos One, 10(6), e0129320.
http://doi.org/10.1371/journal.pone.0129320
Kennedy, B. L., Rawding, J., Most, S. B., & Hoffman, J. E. (2014). Emotion-induced
blindness reflects competition at early and late processing stages: An ERP study.
Cognitive, Affective & Behavioral Neuroscience, 14(4), 1485–1498.
http://doi.org/10.3758/s13415-014-0303-x
Kristjánsson, A., & Nakayama, K. (2002). The attentional blink in space and time.
Vision Research, 42, 2039–2050. http://doi.org/10.1016/S0042-6989(02)00129-3
Le Pelley, M. E., Pearson, D., Griffiths, O., & Beesley, T. (2015). When goals conflict
with values: Counterproductive attentional and oculomotor capture by reward-
related stimuli. Journal of Experimental Psychology: General, 144(1), 158–171.
http://doi.org/10.1037/xge0000037
Lunau, R., & Olivers, C. N. L. (2010). The attentional blink and lag 1 sparing are
nonspatial. Attention Perception Psychophysics, 72(2), 317–325.
http://doi.org/10.3758/APP.72.2.317
MacLeod, C., Mathews, A., & Tata, P. (1986). Attentional bias in emotional disorders.
Journal of Abnormal Psychology, 95(1), 15–20. http://doi.org/10.1037/0021-
843X.95.1.15
Mogg, K., & Bradley, B. P. (1998). A cognitive-motivational analysis of anxiety.
Behavior Research and Therapy, 36, 809–848. http://doi.org/10.1016/S0005-
7967(98)00063-1 60 | Spatial specificity in EIB regardless of eye-gaze
Moore, K. S., & Weissman, D. H. (2011). Set-specific capture can be reduced by pre-
emptively occupying a limited-capacity focus of attention. Visual Cognition,
19(4), 417-444. http://doi.org/10.1080/13506285.2011.558862
Most, S. B., Chun, M. M., Widders, D. M., & Zald, D. H. (2005). Attentional
rubbernecking: cognitive control and personality in emotion-induced blindness.
Psychonomic Bulletin & Review, 12(4), 654–61.
http://doi.org/10.3758/BF03196754
Most, S. B., & Wang, L. (2011). Dissociating spatial attention and awareness in
emotion-induced blindness. Psychological Science, 22(3), 300–5.
http://doi.org/10.1177/0956797610397665
Most, S.B., & Wang, L. (in revision). The cost of seeing the meaning: Conceptual
capture by distractors triggers localized target suppression.
Nummenmaa, L., Hyönä, J., & Calvo, M. G. (2006). Eye movement assessment of
selective attentional capture by emotional pictures. Emotion, 6(2), 257–268.
http://doi.org/10.1037/1528-3542.6.2.257
Pelli, D. G. (1997). The VideoToolbox software for visual psychophysics: transforming
numbers into movies. Spatial Vision, 10(4), 437–442.
http://doi.org/10.1163/156856897X00366
Posner, M. I. (1980). Orienting of attention. Quarterly Journal of Experimental
Psychology, 32, 3–25. http://doi.org/10.1080/00335558008248231
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 Spatial specificity in EIB regardless of eye-gaze | 61
Psychology: Human Perception and Performance, 18(3), 849–60.
http://doi.org/10.1037/0096-1523.18.3.849
Shapiro, K. L., Raymond, J. E., & Arnell, K. M. (1994). Attention to visual pattern
information produces the attentional blink in rapid serial visual presentation.
Journal of Experimental Psychology: Human Perception and Performance, 20(2),
357–371. http://doi.org/10.1037/0096-1523.20.2.357
Shih, S. (2000). Recall of two visual targets embedded in RSVP streams of distractors
depends on their temporal and spatial relationship. Perception & Psychophysics,
62(7), 1348–1355. http://doi.org/10.3758/BF03212137
Theeuwes, J., Kramer, A. F., Hahn, S., Irwin, D. E., & Zelinsky, G. J. (1999). Influence
of Attentional Capture on Oculomotor Control. Journal of Experimental
Psychology: Human Perception and Performance, 25(6), 1595–1608.
Vromen, J. M. G., Lipp, O. V., & Remington, R. W. (2014). The spider does not always
win the fight for attention: Disengagement from threat is modulated by goal set.
Cognition and Emotion, 29(7), 1185–1196.
http://doi.org/10.1080/02699931.2014.969198
Wang, L., Kennedy, B. L., & Most, S. B. (2012). When emotion blinds: A
spatiotemporal competition account of emotion-induced blindness. Frontiers in
Psychology, 3, 1–6. http://doi.org/10.3389/fpsyg.2012.00438
Wyble, B., & Swan, G. (2015). Mapping the spatiotemporal dynamics of interference
between two visual targets. Attention, Perception, & Psychophysics, 77(7), 2331–
2343. http://doi.org/10.3758/s13414-015-0938-x
62 | Spatial specificity in EIB regardless of eye-gaze
Chapter 2:
Affective stimuli are prioritized in attention regardless of categorical distinctiveness: An emotion-induced blindness study
Briana L. Kennedy and Steven B. Most
64 | Categorical distinctiveness in emotion-induced blindness
Abstract
Affective stimuli are prioritized in attention, whether their affective value stems from emotional content or a history of reward. The uniqueness of such stimuli within their experimental contexts might imbue them with an enhanced categorical distinctiveness that accounts for their impact on attention. Indeed, in emotion-induced blindness, categorically distinctive neutral pictures disrupt target perception, albeit to a lesser degree than do emotional pictures. Here, we manipulated the categorical distinctiveness of distractors in an emotion-induced blindness task. Participants searched within RSVP streams for a target that followed an emotional or a neutral distractor picture. In a categorically homogenous condition, all non-distractor items were exemplars from a uniform category, thus enhancing the distractor’s categorical distinctiveness. In a categorically heterogeneous condition, each non-distractor item represented a distinct category. Neutral distractors disrupted target perception only in the homogenous condition, but emotional distractors did so regardless of their categorical distinctiveness.
Categorical distinctiveness in emotion-induced blindness | 65
Although we are often able to control the focus of our attention, some properties of the environment seem to be particularly adept at grabbing our attention without our volition (i.e., “attentional capture”). Within the traditional visual cognition literature, research has tended to focus on physical properties that capture attention, such as motion (e.g., Franconeri & Simons, 2003; Hillstrom & Yantis, 1994), salient colors
(e.g., Theeuwes, 1994), and sudden onsets (e.g., Jonides & Yantis, 1988; Yantis &
Jonides, 1984). Attentional capture has also been of interest within the clinical literature, where the focus has typically been on the attention-capturing power of emotional stimuli and how their impact is moderated by individual differences (e.g.,
Fox, 1993; MacLeod, Mathews, Tata, 1986; Mogg & Bradley, 1999). Increasingly, these two traditions of research have found common ground as investigators have shown that affectively tinged stimuli – whether their affective value stems from emotional content or reward (or punishment) history – capture attention in well- established visual cognitive paradigms (see Bar-Haim et al., 2007, and Dolan, 2002, for reviews).
The evidence for attentional capture by affectively rich stimuli – that is, stimuli with negative or positive emotional significance – takes different forms across different tasks. Visual search tasks (Öhman et al, 2001) and cueing studies (Fox et al., 2001;
Mogg & Bradley, 1999) typically use response time as the critical variable, as typically does the widely used “dot probe”, in which people respond faster to a target that appears in a location previously occupied by an emotional stimulus than by a neutral stimulus
(e.g., MacLeod et al., 1986; Mogg & Bradley, 1999). In contrast, accuracy is used as the measure of interest when participants search for targets in rapid serial visual presentations (e.g., Anderson & Phelps, 2001; Most, Chun, Widders, & Zald, 2005).
Across the different tasks, data consistently suggest that affectively powerful stimuli are 66 | Categorical distinctiveness in emotion-induced blindness
prioritized by attention faster and better than neutral stimuli, sometimes with beneficial- and sometimes with detrimental- consequences for performance on a primary task. As an example of a detrimental impact on task performance, when people search for a single target picture embedded within rapid serial visual presentations of items (e.g., presented at a rate of 100-ms/item), their accuracy is robustly impaired when the target follows a task-irrelevant emotionally evocative distractor picture – an effect that reflects attentional prioritization by the emotional distractors and which has been called emotion-induced blindness (Arnell, Killman, & Fijavz, 2007; Kennedy & Most, 2012;
Most et al., 2005; Most, Smith, Cooter, Levy, & Zald, 2007; Wang, Kennedy, & Most,
2012).
Like stimuli with intrinsically emotional content, stimuli that have been associated with reward and/or punishment also capture attention. Stimuli that participants train to associate with reward capture attention in subsequent visual search tasks (Anderson, Laurent & Yantis, 2011a, 2011b; Chelazzi, Perlato, Santandrea, &
Libera, 2013; Failing & Theeuwes, 2014) and appear to be perceptually enhanced
(Hickey, Chelazzi, & Theeuwes, 2010). Stimuli linked with punishment also grab attention (Schmidt, Belopolsky & Theeuwes, 2014; Smith, Most, Newsome, & Zald,
2006). For example, Smith and colleagues (2006) conditioned participants to associate otherwise neutral images with aversive bursts of white noise, finding that these stimuli subsequently induced emotion-induced blindness. It may be that reward-linked and punishment-linked stimuli engage similar mechanisms as stimuli with intrinsic emotional content (Murray, 2007; but see Sakaki & Mather, 2012).
A reasonable interpretation of such findings is that attention prioritizes stimuli directly on the basis of their affective value. Alternatively, however, such stimuli might capture attention because participants categorize them differently than other items in the Categorical distinctiveness in emotion-induced blindness | 67
visual display. For example, in some experiments, the attention grabbing emotional item belongs to a different object category from the other stimuli, such as a snake among mushrooms (e.g., Öhman, Flykt, & Esteves, 2001). This is largely the case in previous emotion-induced blindness studies, where the emotional distractor has often been a picture of people or animals among pictures of landscapes (e.g., Kennedy et al.,
2014; Most et al., 2005). Thus, it is important to understand whether such apparent
“emotion” effects are modulated by distractors’ categorical novelty. Even in cases where the perceptual properties of a distractor are controlled with more rigor – such as studies in which a color that has been linked with reward captures attention more than a color that has not (Anderson, Laurent, & Yantis, 2011a) – the possibility remains that the affective stimulus is categorized differently from other stimuli. Such stimuli may, for example, be deemed as belonging to a “high value” category and the other stimuli to a “low value” category, with the high value stimuli capturing attention only once their categorical distinctiveness within the display is registered. Evidence suggests that the behavioral impact and neural processing of emotional- and categorical- oddballs can be hard to disentangle (e.g., Strange & Dolan, 2001, 2007; Fichtenholtz et al., 2004), raising the question as to whether what appears to be direct effects of affective value on attentional prioritization may instead be indirect and driven by an intermediate categorization process. Consistent with this possibility, previous work has demonstrated that categorical “oddballs” can grab attention even when perceptual properties are largely controlled (Strange & Dolan, 2001, 2007; see also Barnard et al.,
2004; Goodhew, Kendall, Ferber, & Pratt, 2014; Stein, Zwickel, Kitzmantel, Ritter, &
Schneider, 2010).
Within emotion-induced blindness experiments, there is some evidence that categorical distinctiveness might contribute to the effect: in a number of instances, the 68 | Categorical distinctiveness in emotion-induced blindness
rapid streams are composed of landscape/architectural photos, and both the emotional and non-emotional distractors depict people or animals – making both categorically distinctive within the stream (e.g., Kennedy & Most, 2012; Most et al., 2005).
Although emotional distractors impair target perception more than non-emotional distractors do, the non-emotional distractors consistently impair performance relative to baseline. And although one possibility could be that this reflects attentional prioritization by the perceptual distinctiveness of the distractors, scrambled versions of the distractors – which contain the same color and luminance as negative distractors – barely affect performance at all, suggesting that categorical (as opposed to perceptual) distinctiveness does indeed play a role. Although the difference in the impact of emotional- vs. non-emotional distractors could in itself indicate attentional prioritization by affective stimuli over and beyond the contribution of categorical distinctiveness, an alternative interpretation is that the cues that indicate categorical distinctiveness are additive, with affective uniqueness being an additional contributor to the sum.
In the present study, we sought to understand the degree to which attentional prioritization by affective stimuli might be modulated by their categorical novelty within a task; we examined attentional prioritization by emotional and non-emotional stimuli when their relative affective values were constant but their uniqueness in terms of object category was enhanced or diminished. Specifically, the distractors appeared in streams composed of object exemplars from a homogenous category (enhancing distractors’ categorical distinctiveness) or from a heterogeneous collection of object categories (diminishing distractors’ categorical distinctiveness). Because evidence suggests that performance impairment by non-emotional distractors in an emotion- induced blindness task is driven by their categorical distinctiveness, we predicted that this manipulation should modulate the target processing impairments these non- Categorical distinctiveness in emotion-induced blindness | 69
emotional distractors elicit. Our critical question was what the outcome of this manipulation would be on the impact of emotional distractors. If affective stimuli grab attention due, either in whole or in part, to their categorical distinctiveness, then they should elicit less target processing impairment in the heterogeneous condition than in the homogeneous condition. If, however, their power to grab attention derives primarily from their affective nature, then this manipulation should have no observable impact.
Experiment 3
Method
Participants. 69 University of Delaware undergraduates (mean age 19.25 years; 35 female, 34 male) participated for course credit. 37 participants were part of the heterogeneous condition and 32 were part of the homogeneous condition. All participants provided informed consent and the experiment was approved by the
University of Delaware Human Subjects Review Board.
Materials and Procedure. The experiment was divided into 4 blocks, each containing
84 trials. Every trial consisted of a rapid serial visual presentation (RSVP) stream of 21 images, with each image appearing for 150-ms and being immediately replaced by the next. Images in the stream were colored, 320x240 pixel photographs, and every trial included one critical distractor, one target, and 19 filler images. The target and filler images were drawn from a set of 21 different categories of everyday objects, wherein each category contained 21 exemplars (e.g., teapots, shoes, keys, etc.).
Every stream had a single image that served as the critical distractor. The critical distractor in any given stream could be emotionally negative, neutral, or simply another filler image (as a “baseline” condition). 56 negative and 56 neutral pictures were selected as distractors and depicted people or animals. These images were mostly 70 | Categorical distinctiveness in emotion-induced blindness
gathered from the International Affective Picture System (IAPS; Lang, Bradley, &
Cuthbert, 2001) based on ratings of emotional valence and arousal, and were supplemented by images taken from publicly available sources. The negative and neutral distractors were rated on 9-point scales of valence and arousal by a separate group of participants and have been used in several previous emotion-induced blindness experiments (e.g., Kennedy et al., 2014). The negative emotional distractors, which included emotionally powerful scenes such as medical trauma, threatening animals, and violence, were rated as both more unpleasant and more arousing than the neutral distractors, which depicted people with neutral expressions and non-threatening animals
(ps < .001).
The category images included 21 different object categories that represented common, everyday objects. Each of the 21 object categories contained 21 photographic exemplars collected from publicly available sources, totaling 441 category images in all.
In every stream, a blue border (10 pixels wide) surrounded one of the category images, which served to identify it as the target. The emotionally negative, neutral, or baseline distractor was presented randomly at a serial position between 3 and 8, and it preceded the target picture by 3, 5, 7, or 9 images in the rapid serial stream.
Participants were randomly assigned to one of two conditions. In the heterogeneous condition, participants searched for the blue-bordered target in streams that contained exemplars from different categories of common objects (e.g., teapot, shoe, house, key, etc.). In this condition, the heterogeneity of items in each stream minimized the categorical “oddball” nature of the distractor, as each item represented a different category. Determination of which object served as the target was pseudorandom: the target object type was cycled so as not to repeat until all other object types had been targets. Order of the objects and exemplars of each target type were Categorical distinctiveness in emotion-induced blindness | 71
random. In baseline streams, participants saw exemplars of all 21 object types. In streams with negative or neutral distractors, the distractor replaced one of the objects in the stream.
In the homogeneous condition, streams were composed of exemplars all of the same type of common object (e.g., all teapots, all shoes, all houses, etc.). Order of the exemplars was random, and the assignment of the object stream type was pseudorandom: it cycled so as not to repeat until all other object types had been stream types. In baseline streams, subjects saw all 21 exemplars of the same object type. In streams with negative or neutral distractors, the distractor replaced a random exemplar in the stream.
At the end of each stream, all images from the stream except for the critical distractor appeared in an array on the computer screen (160x120 pixels per image) and participants were instructed to identify, via mouse click, the one target image that had been surrounded by the blue border. Note that this aspect of the task – the search for an otherwise non-distinctive blue bordered item and the forced-choice selection of it from an array – marked a departure from most previous emotion-induced blindness experiments, where participants reported the orientation of the one rotated target in each stream. However, because participants needed to identify an item based on its unique identity from other items in the stream (instead of via a binary orientation decision), the search array was optimal for this experiment. The positions of images in the array were random (see Figure 12). Correct responses were followed by a bell sound via headphones; no sound accompanied incorrect answers. The next trial began immediately after participants made their target selection. 72 | Categorical distinctiveness in emotion-induced blindness
Figure 12. Example of part of an experimental trial in Experiment 3. Items were presented at a rate of 150ms/item in a) heterogeneous or b) homogeneous streams. The target is defined by the surrounding blue border, and participants identified the target at the end of the stream with a mouse press on the array of images. Sometime before the target, an emotionally negative, neutral, or baseline (no) distractor appeared some before it. See text for details.
Throughout the experiment, stimuli appeared against a gray background on a 19- inch CRT monitor with a refresh rate of 100 Hz, via the Psychophysics Toolbox for
Matlab (Brainard, 1997; Pelli, 1997). Screen resolution was set to 800x600 pixels, making the 320x240 pixel photographs 5.4”x4” stimuli. Participants sat at a comfortable distance from the computer screen and head position was not fixed. Before starting the experiment, participants were shown examples of emotional and neutral images and engaged in a short 16-trial practice session, with RSVP rates starting at 300- ms/item and increasing to the experiment presentation rate of 150-ms/item. The Categorical distinctiveness in emotion-induced blindness | 73
practice session did not include negative or neutral distractors. Participants were debriefed at the end of the experiment.
Results
Percentage accuracy in reporting the target served as the primary measure of interest. An overall 3 (Distractor Type: negative, neutral, none) x 4 (Lag-3 vs. Lag-5 vs.
Lag-7 vs. Lag-9) x 2 (Categorical Context: Heterogeneous vs. Homogeneous) mixed
ANOVA revealed a significant main effect of Distractor Type, F(2, 134)=7.721,
2 2 p<.001, ηp =.103, Lag, F(3, 201)=56.577, p<.001, ηp =.458, and Categorical Context,
2 F(1, 67)=96.944, p<.001, ηp =.591, as well as a significant interaction between the
2 three, F(6, 402)=3.435, p=.003, ηp =.049 (see Figure 13). There was a significant
2 interaction between Distractor Type and Lag, F(6, 402)=6.614, p<.001, ηp =.090, but no significant interaction between Distractor Type and Categorical Context, F(2,
2 134)=1.304, p=.275, ηp =.019, or between Lag and Categorical Context, F(3,
2 201)=1.005, p=.392, ηp =.015. 74 | Categorical distinctiveness in emotion-induced blindness
Figure 13. Experiment 3 results. Percent accuracies in a) heterogeneous and b) homogeneous streams, following negative, neutral, or baseline distractors, across lags 3, 5, 7, and 9.
Participants in the Heterogeneous condition performed better overall on the task
(M=83.3%, SD=10.8%) than participants in the Homogeneous condition (M=48.6%,
SD=18.1%). This makes sense, as targets in the homogenous condition looked similar to non-targets in the same stream. In the heterogeneous condition, each stream item not only looked relatively distinctive, but participants could attach a categorically broad semantic label (e.g., “teapot”) to the target item in the heterogeneous condition; in contrast, in the homogeneous condition, semantic labels would be relatively useless for demarcating the target and participants would have had to maintain detailed visual memories of targets’ features.
Because the Lag X Categorical Context interaction was qualified by a further interaction with Distractor Type, we examined this two-way interaction for each
Distractor Type separately. In streams with Negative Distractors, there was no
2 interaction between Lag and Categorical Context, F(3, 201)=0.679, p=.544, ηp =.010.
Similarly, in streams with Baseline distractors, there was no significant interaction
2 between Lag and Category Type, F(3, 201)=0.468, p=.705, ηp =.007. However, there Categorical distinctiveness in emotion-induced blindness | 75
was a significant interaction between Lag and Category Type in the Neutral streams,
2 F(3, 201)=7.329, p<.001, ηp =.099.
This effect within the neutral condition was primarily driven by performance at
Lag-3 (the earliest lag tested), where performance following neutral distractors resembled performance following negative distractors in the homogeneous condition but resembled baseline in the heterogeneous condition. At Lag-3 in the Homogeneous condition, there was a significant difference in accuracy between the neutral and baseline streams (Neutral: M=38.6%, SD=19.2%; Baseline: M=48.1%, SD=18.4%; t(31)= 4.974, p<.001) but not between neutral and negative streams (Negative:
M=38.4%, SD=18.9%; t(31)=.087, p=.932). In contrast, in the Heterogeneous condition at this lag, there was a significant difference between neutral and negative streams (Negative: M=71.9%, SD=15.5%; Neutral: M=81.4%, SD=11.8%; t(36)=5.834, p<0.001), but not between the neutral and baseline streams (Baseline: M=81.0%,
SD=13.8%; t(36)=0.248, p=.806). To hone in on the role of categorical distinctiveness in streams with negative distractors at lag-3, we ran a 2 (negative vs. baseline) X 2
(homogeneous vs. heterogeneous) ANOVA, which revealed no significant interaction,
2 F(1, 67)=0.040, p=.843, ηp =.001, further suggesting that categorical distinctiveness does not modulate prioritization of negative stimuli. This contrasted with the pattern observed for neutral compared to baseline distractors at lag-3, in which the 2 (neutral vs. baseline) X 2 (homogeneous vs. heterogeneous) ANOVA revealed a significant
2 interaction, F(1, 67)=16.561, p<.001, ηp =.198.
Discussion
Across numerous tasks and studies, affectively laden stimuli appear to be prioritized in attention, whether their affective quality stems from their intrinsic emotional content or their history of association with rewarding or aversive outcomes 76 | Categorical distinctiveness in emotion-induced blindness
(e.g., Anderson & Phelps, 2001; Anderson et al., 2011a; Hickey et al., 2010; Öhman et al., 2001). Such findings have deeply influenced theories of emotional information processing, its role in clinically relevant individual differences (e.g., Bar-Haim et al.,
2007; Fox, 1993; MacLeod, 1991; MacLeod et al., 1986; Mogg & Bradley, 1999), and neurobiological accounts of communication between classically emotion- and attention- linked regions (e.g., Dolcos & McCarthy, 2006; Hickey et al., 2010; Vuilleumier,
2005). Yet, one possibility is that rather than affective value directly capturing attention, such stimuli are prioritized by attention because they are categorized differently than the non- affective stimuli in the display – with the affective stimuli capturing attention because of their categorical uniqueness (Ferrari et al., 2010;
Weinberg et al., 2012).
In the present emotion-induced blindness study, we manipulated the context within which emotional and non-emotional distractors appeared in order to increase or decrease the degree to which they represented a unique object category while holding their relative affective values constant. Specifically, emotional and neutral distractors appeared in rapid streams composed either of categorically homogeneous filler items
(thereby enhancing the categorical oddball nature of the distractors) or of categorically heterogeneous filler items (thereby diminishing their categorical oddball nature). As predicted, this manipulation modulated the target perception impairments elicited by non-emotional distractors: in the heterogeneous streams, performance following non- emotional distractors did not differ from when there was no distractor, but in homogeneous streams non-emotional distractors caused as much impairment as emotional ones did. Critical to the primary motivation of this experiment, the manipulation of categorical context did not modulate performance following emotional distractors relative to baseline. (A main effect of categorical context, where Categorical distinctiveness in emotion-induced blindness | 77
performance following all distractor types was worse in the homogeneous condition, likely reflected the greater difficulty of selecting a target based on specific features than by semantic category.)5
It is worth noting that it is unclear if emotional stimuli would have continued to impact subsequent stimuli if they were no longer the only affectively powerful stimulus in the stream. It may be that the affective stimuli in this study assumed categorical novelty not only on the basis of their object identity, but also on the basis of their affective significance (i.e., belonging to a “high value” category while all other items in the stream belonged to a “low value” category), and that this in itself represented an intermediate categorization stage that drove their ability to be prioritized by attention.
Future studies could possibly address the issue of affective novelty by manipulating the number of affective stimuli in a stream.
Intriguingly, the impacts of categorical- and affective- uniqueness were not additive: whereas increasing distractors’ object category uniqueness modulated attentional prioritization by non-emotional items, it did not similarly modulate prioritization of emotional items. Further, when the categorical uniqueness of non- emotional distractors was enhanced, they appeared to be prioritized in attention as much as the emotional distractors did. This raises the question as to whether the consequences of affective uniqueness are different from those of object category
5 We attempted to replicate these findings in a within-subjects design, with participants receiving two blocks each of heterogeneous and homogeneous conditions (interleaved to minimize differences in practice across conditions). However, this design did not elicit differences in performance between the homogeneous and heterogeneous conditions; instead, in both conditions, negative distractors impaired target accuracy more than neutral distractors, and neutral distractors impaired target accuracy more than baseline distractors. We then reran the between-subjects version described in the main text, which successfully replicated our initial results. It seems likely that carryover effects stemming from the within- subjects, interleaved manipulation of condition interfered with differences in strategy that participants would otherwise employ when faced with the homogeneous and heterogeneous streams. See Appendix for more details regarding the methods and results of this replication experiment (which also served to verify the effect across universities in the USA and Australia).
78 | Categorical distinctiveness in emotion-induced blindness
uniqueness (in the words of one of this manuscript’s reviewers, it may be that “a categorical oddball is a categorical oddball is a categorical oddball”). It is hard to draw firm conclusions about this from the current data, as the main effect of categorical context (i.e., better performance in the heterogeneous than homogenous condition) raises the spectre of ceiling or floor effects that might obscure additive effects of categorical and affective uniqueness. It may also be that the emotional intensity of the distractor images we used impaired performance to such an extent as to obscure a possible additional contribution of categorical uniqueness. That is, if affective uniqueness is one cue that contributes to a stimulus’s overall categorical distinctiveness, it may be that there is little else needed to set a stimulus apart when its emotional resonance is very high. Less emotionally powerful distractors might have induced a lesser effect, allowing observation of such additive effects. In this light, it is noteworthy that some previous research has found additive effects of emotional quality and categorical uniqueness (Weinberg et al., 2012).
A further limitation of the current work is that the earliest lag that we tested was
Lag-3, which involved a lapse of 450ms between the onsets of the distractor and target.
Thus, it is impossible to rule out a stronger role of categorical distinctiveness (or an additive effect of categorical- and affective- uniqueness) at earlier stages of attentional engagement. However, previous work suggests that emotional stimuli tend to impair performance more than neutral stimuli at Lag-1 (and at only 100ms between the distractor and target; Kennedy & Most, under review; Most & Jungé, 2008), which suggests that affective properties can be prioritized by attention over and beyond the role of object category at earlier stages of processing as well. It also bears mentioning that whereas emotion-induced blindness studies typically index perceptual processing
(e.g., Kennedy & Most, 2012), the present study employed a memory test instead. Categorical distinctiveness in emotion-induced blindness | 79
Although this raises questions as to whether the impairments in this study reflected the same mechanisms as in previous emotion-induced blindness studies, in both cases the impairments stem from attentional prioritization of affective stimuli.
Overall, these findings demonstrate that while categorically distinct items can be prioritized by attention, the impact of affective stimuli is robust regardless of whether the categorical distinctiveness of such stimuli is minimized or enhanced. Future work should continue to investigate how the affective qualities of a stimulus can grab attention in different contexts, if different sources of oddball status (e.g., object category vs. affective) work in parallel or are additive, and if all types of affective value (e.g., intrinsic emotional content vs. associations with reward or punishment) operate in the same way.
80 | Categorical distinctiveness in emotion-induced blindness
References
Arnell, K.M., Killman, K.V., & Fijavz, D. (2007). Blinded by emotion: Target misses
follow attention capture by arousing distractors in RSVP. Emotion, 7, 465-477.
Anderson, A.K., & Phelps, E.A. (2001). Lesions of the human amygdala impair
enhanced perception of emotionally salient events. Nature, 411, 305-309.
Anderson, B.A., Laurent, P.A., & Yantis, S. (2011a). Value-driven attentional capture.
PNAS, 108(25), 10367-10371.
Anderson, B.A., Laurent, P.A., & Yantis, S. (2011b). Learned value magnifies salience-
based attentional capture. PLoS One, 6(11), e27926.
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),
1-24.
Barnard, P.J., Scott, S., Taylor, J., May, J., & Knightley, W. (2004). Paying attention to
meaning. Psychological Science, 15(3), 179-186.
Brainard, D. H. (1997). The Psychophysics Toolbox, Spatial Vision, 10, 433-436.
Chelazzi, L., Perlato, A., Santandrea, E., & Della Libera, C. (2013). Rewards teach
visual selective attention. Vision Research, 85, 58-72.
Dolan, R.J. (2002). Emotion, cognition, and behavior. Science, 298, 1191-1194.
Dolcos, F., & McCarthy, G. (2006). Brain systems mediating cognitive interference by
emotional distraction. The Journal of Neuroscience, 26(7), 2072-2079.
Failing, M.F., & Theeuwes, J. (2014). Exogenous visual orienting by reward. Journal of
Vision, 14(5):6, 1-9. Categorical distinctiveness in emotion-induced blindness | 81
Fichtenholtz, H.M., Dean, H.L., Dillon, D.G., Yamasaki, H., McCarthy, G., & LaBar,
K.S. (2004). Emotion-attention network interactions during a visual oddball task.
Cognitive Brain Research, 20, 67-80.
Fox, E. (1993). Allocation of visual attention and anxiety. Cognition and Emotion, 7,
207-215.
Fox, E., Russo, R., Bowles, R., and Dutton, K. (2001). Do threatening stimuli draw or
hold visual attention in subclinical anxiety? Journal of Experimental
Psychology: General, 130, 681–700.
Ferrari, V., Bradley, M.M., Codispoti, M., & Lang, P.J. (2010). Detecting novelty and
significance. Journal of Cognitive Neuroscience, 22(2), 404-411.
Franconeri, S.L., & Simons, D.J. (2003). Moving and looming stimuli capture attention.
Perception and Psychophysics, 65(7), 999-1010.
Goodhew, S.C., Kendall, W., Ferber, S., & Pratt, J. (2014). Setting semantics:
Conceptual set can determine the physical properties that capture attention.
Attention, Perception & Psychophysics.
Hickey, C., Chelazzi, L., & Theeuwes, J. (2010). Reward changes salience in human
vision via the anterior cingulate. The Journal of Neuroscience, 30(33), 11096-
11103.
Hillstrom, A.P., & Yantis, S. (1994). Visual motion and attentional capture. Perception
& Psychophysics, 55(4), 399-411.
Jonides, J., & Yantis, S. (1988). Uniqueness of abrupt visual onset in capturing
attention. Perception and Psychophysics, 43(4), 346-354.
Kennedy, B.L., & Most, S.B. (2012). Perceptual, not memorial, disruption underlies
emotion-induced blindness. Emotion, 12(2), 199-202. 82 | Categorical distinctiveness in emotion-induced blindness
Kennedy, B.L., & Most, S.B. (under review). The rapid perceptual impact of emotional
distractors.
Lang, P.J., Bradley, M.M., & Cuthbert, B.N. (2001). International Affective Picture
System (IAPS): Instruction manual and affective ratings (Tech. Rep. No. A-5).
Gainesville: University of Florida, Center for Research in Psychophysiology.
MacLeod, C. (1991). Half a century of research on the Stroop Effect: An integrative
review. Psychological Bulletin, 109(2), 163-203.
MacLeod, C., Mathews, A., and Tata, P. (1986). Attentional bias in emotional
distractors. Journal of Abnormal Psychology, 95, 15–20.
Mogg, K., and Bradley, B. P. (1999). Orienting of attention to threatening facial
expressions presented under conditions of restricted awareness. Cognition and
Emotion, 13, 713–740.
Most, S.B., Chun, M.M., Widders, D.M., & Zald, D.H. (2005). Attentional
rubbernecking: Cognitive control and personality in emotion-induced blindness.
Psychonomic Bulletin & Review, 12, 654-661.
Most, S.B., & Jungé, J.A. (2008). Don’t look back: Retroactive, dynamic costs and
benefits of emotional capture. Visual Cognition, 16, 262-278.
Most, S.B., Smith, S.D., Cooter, A.B., Levy, B.N., & Zald, D.H. (2007). The naked
truth: Positive, arousing distractors impair rapid target perception. Cognition and
Emotion, 21, 961-981.
Murray, E.A. (2007). The amygdala, reward, and emotion. Trends in Cognitive
Sciences, 11(11), 489-497.
Öhman, A., Flykt, A., and Esteves, F. (2001). Emotion drives attention: detecting the
snake in the grass. Journal of Experimental Psychology: General, 130, 466–478. Categorical distinctiveness in emotion-induced blindness | 83
Pelli, D. G. (1997) The VideoToolbox software for visual psychophysics: Transforming
numbers into movies. Spatial Vision, 10, 437-442.
Sakaki, M., & Mather, M. (2012). How reward and emotional stimuli induce different
reactions across the menstrual cycle. Social and Personality Psychology
Compass, 6, 1-17.
Schmidt, L.J., Belopolsky, A.V., & Theeuwes, J. (2014). Attentional capture by signals
of threat. Cognition and Emotion.
Smith, S.D., Most, S.B., Newsome, L.A., & Zald, D.H. (2006). An emotion-induced
attentional blink elicited by aversively conditioned stimuli. Emotion, 6(3), 523-
527.
Stein, T., Zwickel, J., Kitzmantel, M., Ritter, J., & Schneider, W.X. (2010). Irrelevant
words trigger an attentional blink. Experimental Psychology, 57(4), 301-307.
Strange, B.A., & Dolan, R.J. (2001). Adaptive anterior hippocampus responses to
oddball stimuli. Hippocampus, 11, 690-698.
Strange, B.A., & Dolan, R.J. (2007). β-adrenergic modulation of oddball responses in
humans. Behavioral and Brain Functions, 3(29).
Theeuwes, J. (1994). Stimulus-driven capture and attentional set: Selective search for
color and visual abrupt onsets. Journal of Experimental Psychology: Human
Perception and Performance, 20(4), 799-806.
Vuilleumier, P. (2005). How brains beware: Neural mechanisms of emotional attention.
Trends in Cognitive Sciences, 9(12), 585-594.
Wang, L., Kennedy, B.L., & Most, S.B. (2012). When emotion blinds: A spatiotemporal
competition account of emotion-induced blindness. Frontiers in Psychology:
Special Topic on Emotion and Cognition, 3:438. 84 | Categorical distinctiveness in emotion-induced blindness
Weinberg, A., Hilgard, J., Bartholow, B.D., & Hajcak, G. (2012). Emotional targets:
Evaluative categorization as a function of context and content. International
Journal of Psychophysiology, 84, 149-154.
Yantis, S., & Jonides, J. (1984). Abrupt visual onsets and selective attention: Evidence
from visual search. Journal of Experimental Psychology: Human Perception and
Performance, 10(5), 601-621.
Categorical distinctiveness in emotion-induced blindness | 85
Chapter 2 Appendix: Methods and Results of Replication at the
University of New South Wales (Australia)
Method
Participants. 43 participants were recruited from the community using the University of New South Wales SONA system and were compensated $12 for their time (mean age
24.63 years; 30 female, 13 male). Twenty-one were assigned to the heterogeneous condition and 22 to the homogeneous condition. All participants gave informed consent and the experiment was approved by the University of New South Wales Human
Research Ethics Approval Panel.
Materials and Procedure. The materials and procedure were identical to those reported in the original experiment, with a few minor changes: Stimuli appeared against a black background on a 24-inch Benq LED monitor with a refresh rate of 120Hz.
Screen resolution was 1920x1080 pixels, with the 320x240 pixel photographs appearing as 3.46”x2.64” stimuli. The practice session was shortened to 8 trials, still starting at
300-ms/item and increasing to the experiment presentation rate of 150-ms/item.
Results 86 | Categorical distinctiveness in emotion-induced blindness
Figure 14. Results from Experiment 3 replication. Percent accuracies in a) heterogeneous and b) homogeneous streams, following negative, neutral, or baseline distractors, across lags 3, 5, 7, and 9.
The overall 3 (Distractor Type) x 4 (Lag) x 2 (Categorical Context) mixed
ANOVA revealed a significant main effect of Distractor Type, F(2, 82)=3.243, p=.044,
2 2 ηp =.073, Lag, F(3, 123)=25.382, p<.001, ηp =.382, and Categorical Context, F(1,
2 41)=26.162, p<.001, ηp =.390 (but no significant interaction between the three, F(6,
2 246)=1.083, p=.373, ηp =.026; see Figure 14). There was a significant interaction
2 between Distractor Type and Categorical Context, F(2, 82)=3.273, p=.043, ηp =.074, but no significant interaction between Distractor Type and Lag, F(6, 246)=1.001,
2 p=.425, ηp =.024, or between Lag and Categorical Context, F(3, 123)=0.293, p=.830,
2 ηp =.007.
As in the original experiment, participants in the Heterogeneous condition performed better overall on the task (M=75.9%, SD=22.8%) than participants in the
Homogeneous condition (M=44.1%, SD=20.2%). There were no interactions between
Lag and Categorical Context in either the Negative Distractor condition, F(3,
2 123)=0.835, p=.477, ηp =.020, the Baseline condition, F(3, 123)=0.614, p=.607, Categorical distinctiveness in emotion-induced blindness | 87
2 ηp =.015, or (in contrast to the original experiment) the Neutral Distractor condition,
2 F(3, 123)=1.013, p=.389, ηp =.024.
Critical to this replication, at Lag-3 in the Homogeneous condition, there was a significant difference in accuracy between the neutral and baseline streams (Neutral:
M=37.0%, SD=19.5%; Baseline: M=42.9%, SD=19.4%; t(21)= 2.824, p=.010) but not between neutral and negative streams (Negative: M=37.7%, SD=20.3%; t(21)=.252, p=.804). In contrast, in the Heterogeneous condition at this lag, there was a significant difference between neutral and negative streams (Negative: M=68.2%, SD=24.6%;
Neutral: M=73.3%, SD=23.2%; t(20)=2.140, p=.045), but not between the neutral and baseline streams (Baseline: M=70.6%, SD=21.1%; t(20)=1.563, p=.134). As in the original experiment, the 2 (negative vs. baseline) X 2 (homogeneous vs. heterogeneous)
ANOVA for lag-3 data revealed no significant interaction in the negative compared to
2 baseline trials, F(1, 41)=0.569, p=.455, ηp =.014. This contrasted with a significant 2
(neutral vs. baseline) X 2 (homogeneous vs. heterogeneous) interaction for lag-3 data in
2 the neutral relative to baseline trials, F(1, 41)=9.939, p=.003, ηp =.195.
88 | Categorical distinctiveness in emotion-induced blindness
Chapter 3:
Emotionally-infused stimuli disrupt visual awareness regardless of task goals
Briana L. Kennedy and Steven B. Most
90 | Emotion distracts regardless of task-relevance
Abstract
Emotional distractors can disrupt awareness for subsequent items, an effect known as “emotion-induced blindness” (EIB). In a typical EIB task, participants search for one target in a rapid stream of images and are explicitly instructed to ignore the attention-grabbing distractors. In the present experiment, we explored how task- relevancy of the emotional distractors modulates their impact in EIB. Participants viewed rapid streams of images, and two of those images were surrounded by borders.
All items in the stream were images of everyday objects, except the first bordered item
(T1), which could be an emotionally negative or neutral image. For half of the experiment, participants identified both bordered items (T1 and T2), but in the other half they identified only the second of the two bordered items (T2). Participants were generally better at identifying T2 when T1 was task-irrelevant compared to task- relevant. However, the emotional (relative to neutral) T1 images disrupted T2 identification regardless of its relevance, and the disruption lasted longer when the emotional T1 was task-relevant. Thus, emotional value appears to disrupt awareness for subsequent items even when task-irrelevant.
Emotion distracts regardless of task-relevance | 91
Stimuli can be prioritized by attention even if they are irrelevant to the current task (Folk, Leber, & Egeth, 2002; Folk, Remington, & Johnston, 1992; Most, Scholl,
Clifford, & Simons, 2005; Raymond, Shapiro, & Arnell, 1992; Yantis & Egeth, 1999).
For example, task-irrelevant emotional stimuli can impair performance on visual tasks
(Arnell, Killman, & Fijavz, 2007; Butler & Klein, 2009; Dolcos & McCarthy, 2006;
Mogg & Bradley, 1998; Most et al., 2005; Öhman, Flykt, & Esteves, 2001; Williams,
Mathews, & MacLeod, 1996). In an “emotion-induced blindness” task (EIB; Most et al., 2005), participants search for a single target in a rapid stream of images, and are explicitly instructed to ignore task-irrelevant, emotionally powerful distractors. Even though participants should ignore these distractors to correctly identify their target, the emotionally powerful distractors still impair target detection. Thus, EIB demonstrates how emotional stimuli can engage and hold attention even when they are task- irrelevant.
Notably, in our environment we often encounter emotionally powerful stimuli that are relevant to the task at hand, but which may still distract from other important information in our environment. Do these task-relevant emotional stimuli impair the awareness for subsequent items in the same way that task-irrelevant emotional stimuli do? That is, can the competition between target and distractor be modulated by whether people choose to regard the distractor as relevant or irrelevant? Based on previous studies in the literature, there are reasons to predict that emotional stimuli will be more distracting when they are task-relevant, and there are reasons to predict that they will be less distracting.
It is worth noting that here the term “relevancy” specifically refers to task- relevancy. Many theories suggest that emotional stimuli are powerfully distracting because they are biologically relevant (e.g., Bradley, 2009; Ferrari et al., 2013; Fox, 92 | Emotion distracts regardless of task-relevance
Griggs, & Mouchlianitis, 2007; Öhman et al., 2001; Sakaki, Niki, & Mather, 2012).
While the emotional images in our study can be considered relevant to the participant because of their biological implications, the current study specifically explored task- relevancy as determined by top-down goals.
Studies of the attentional blink (AB) seem to indirectly suggest that task- relevance and emotional salience contribute additively to the degree to which an item disrupts perception of other things. The AB is a phenomenon in which perception of a task-relevant target (T2) is impaired if it follows too soon after another task-relevant target (T1; Chun & Potter, 1995; Raymond et al., 1992). As such, it shares phenomenal qualities with EIB – in both tasks, a stimulus impairs the awareness of subsequently presented items (though their mechanistic qualities may be different; see Wang,
Kennedy, & Most, 2012). In cases where a standard 2-target version of the AB is adapted for use with emotional stimuli, an emotional T2 is better detected than a neutral
T2 (Anderson & Phelps, 2001; Anderson, 2005). More relevant to EIB studies, emotional T1 targets cause a greater AB than do neutral T1s (Ihssen & Keil, 2009;
Schwabe & Wolf, 2010; Schwabe et al., 2011). That is, whereas T1 in a standard AB task impairs perception of T2 by virtue of its task-relevance, it does so even more when
T1 is emotional, together suggesting that the impacts of task-relevance and emotional salience may be additive. Recent research also suggests that when participants have to identify the emotional qualities of a stimulus to complete their task, the emotional stimulus impairs more than when its emotionality is task-irrelevant (Dodd, Vogt,
Turkileri, & Notebaert, 2016; Stein, Zwickel, Ritter, Kitzmantel, & Schneider, 2009).
To our knowledge, only one study has compared target performance when emotional distractors were task-relevant to when task-irrelevant in an emotion-induced blindness task, and it did so indirectly (Mathewson, Arnell, & Mansfield, 2008). In this Emotion distracts regardless of task-relevance | 93
study, participants had to identify either one (T1-irrelevant) or two (T1-relevant) words that were designated as targets. T1 was usually an emotional word (represented in a different font when it was relevant), while T2 was always a word that described a color
(e.g., “blue”). When T1 was a taboo word, it elicited emotion-induced impairment in identifying T2 both when it was relevant and irrelevant. There was also a correlation which showed that words that were most distracting in the T1-relevant task were also most distracting in the T1-irrelevant task. However, T2 accuracy performance was not directly compared across task types, and the presentation of stimuli across task types was not the same (with differing font for the T1 word). Also, like other studies using words as stimuli, negative words did not impair performance, which is unlike many other emotion-induced blindness studies that use images. Thus, the type of stimuli used in the present experiment and the direct comparison across tasks further probe the question of task-relevancy in the context of other emotion-induced blindness work.
If task-relevance and emotionality are additive, it is unclear if the effect would be observed in amplitude of the impairment (with even worse percentage accuracy) or in its duration of impairment (with impairment across a longer amount of time).
Evidence suggests that it is particularly difficult to disengage attention from emotional stimuli even when they are task irrelevant (e.g., Fox, Russo, Bowles, & Dutton, 2001;
Georgiou et al., 2005), suggesting that task-relevance may increase the duration of the impairment. Recent literature further suggests that emotional distractors are more difficult to disengage from when they are task-relevant, compared to task-irrelevant
(Müller, Rothermund, & Wentura, 2015; Vromen, Lipp, & Remington, 2014). For example, Vromen and colleagues found that an emotional stimulus (a silhouette shape of a spider) that appeared at a primed location was harder to disengage from than neutral targets. However, this delayed disengagement was specifically found when the 94 | Emotion distracts regardless of task-relevance
emotional stimulus was a possible target item (and not when participants knew it could not be a target stimulus). This is consistent with suggestions that task-relevant emotional stimuli are more difficult to disengage from than task-irrelevant emotional stimuli. The results in Chapter 1 also allude to an additive effect of task-relevancy: when emotional distractors were made task-relevant, they impaired the detection of targets that appeared in different spatial locations (Experiment 2). This differed from typical findings with task-irrelevant emotional distractors, which tend to only impair stimuli that appear in their same location (Experiment 1; Most & Wang, 2011).
Event-related potential (ERP) evidence also suggests that the impacts of task- relevancy and emotionality are additive, and that task-relevancy increases the duration of impact. The late positive potential (LPP) is a component that is believed to be an index of sustained attention, and is present between 350-750ms after a stimulus that engages this attention (Schupp et al., 2000; Weinberg, Hilgard, Bartholow, & Hajcak,
2012). Both task-relevancy and emotionality can elicit the LPP (Ferrari, Codispoti,
Cardinale, & Bradley, 2008; Weinberg, Hilgard, Bartholow, & Hajcak, 2012; Wiens,
Sand, Norberg, & Andersson, 2011), and the impact of task-relevancy and emotionality appear to drive this component additively (Ferrari et al., 2013; Weinberg et al., 2012).
In particular, while both elicit the LPP component, relevancy of an item seems to modulate the LPP in the component’s later stages (Weinberg et al., 2012), which suggests that relevant stimuli engage more elaborative processing.
While such evidence gives us reason to predict a greater impairment
(particularly in duration) from emotional distractors when they are relevant, other evidence raises the possibility that the impairment will be lessened when the distractor is relevant. For example, in one study, an ERP waveform associated with distractor suppression correlated with the magnitude of EIB (Kennedy, Rawding, Most, & Emotion distracts regardless of task-relevance | 95
Hoffman, 2014). This waveform (likely the PD component; Kiss, Grubert, Petersen, &
Eimer, 2012; Sawaki, Geng, & Luck, 2012; Sawaki & Luck, 2010) has been linked to the suppression of salient distractors. Thus, emotion-induced blindness might dissipate when participants are instructed not to ignore the distractor.
By testing the impact of task-relevant distractors, this study also provided an opportunity to test memory for the emotional distractors typically used in the emotion- induced blindness task. Emotional stimuli tend to be remembered better than non- emotional stimuli (see Dolan, 2002; Kensinger, 2007; McGaugh, 2006; Wallace, 1965 for reviews). Moreover, emotional T1s are usually remembered better than neutral T1s in the attentional blink (Ihssen & Keil, 2009), though sometimes there is no difference in performance between them (Schwabe & Wolf, 2010; Stein et al., 2009). Mathewson and colleagues also found that the most distracting words in their attentional blink and emotion-induced blindness tasks were also the best remembered (Mathewson, et al.,
2008). Thus, previous research generally supports the prediction that emotional distractors would be well-remembered compared to neutral distractors.
The aim of the current study was to assess how task-relevancy might moderate the impact of emotional stimuli in the EIB task. We tested performance at both Lag-2
(when T1 and T2 appeared two serial positions apart) and Lag-4 (when T1 and T2 appeared four serial positions apart). Lag-2 is early in the window of EIB, whereas
Lag-4 represents a time later in the impairment window. Because Lag-4 occurs later in the window of EIB, it provides a valuable index of attentional disengagement. Given previous literature suggesting that task-relevant emotional stimuli are harder to disengage from (Müller et al., 2015; Vromen et al., 2014), we predicted that performance at Lag-4 would be more impaired in the relevant, compared to irrelevant, 96 | Emotion distracts regardless of task-relevance
condition. We also predicted that task-relevant emotional distractors would be better remembered than neutral distractors.
Experiment 4
Method
Participants. Sixty-five undergraduates from the University of New South Wales
(mean age=20.7 years; 37 female, 28 male) participated in exchange for course credit.
All participants gave informed consent and the experiment was approved by the
University of New South Wales Human Research Ethics Approval Panel. Data from one female participant were excluded from analyses due to performance accuracy at chance level.
Materials and Procedure. Stimuli were presented and responses gathered via the
Psychophysics Toolbox for Matlab (Brainard, 1997; Pelli, 1997). Participants sat at a comfortable distance from the computer screen and head position was not fixed.
Stimuli appeared against a black background on a 24-inch Benq LED monitor with a refresh rate of 120Hz. Screen resolution was 1920x1080 pixels. Emotion distracts regardless of task-relevance | 97
The experiment included four self-contained blocks and a total of 336 trials (84 trials per block). Each trial consisted of a rapid serial visual presentation (RSVP) of 18 images presented at a rate of 150 ms/item. This presentation rate is slower than in most
EIB experiments due to the nature of the task: picking items out of a line-up requires more elaborate encoding than simply reporting target rotation. The stream of images
Figure 15. Schematic of a partial trial sequence in Experiment 4. Participants viewed a rapid sequence of 18 items, and two of these items were surrounded by colored borders (T1 and T2). Depending on the trial, participants indicated both T1 and T2 (T1-relevant condition), or only T2 (T1-irrelevant condition). Participants made their response by clicking on the appropriate items in the test array at the end of the trial. See text for details.
was presented in the center of the screen against a black background. Each RSVP included two images (T1 and T2) surrounded by colored borders, which identified them as targets. A red border surrounded one of the images, while a green border surrounded the other. Both borders were 10 pixels wide. Order of the red and green borders was counterbalanced across participants, such that some participants had a green bordered
T1 and red bordered T2 throughout the experiment, while other participants had a red bordered T1 and green bordered T2. 98 | Emotion distracts regardless of task-relevance
Stimuli were colored, 320x240 pixel photographs. T1 could be emotionally negative, neutral, or a “baseline” image. All other images (including T2) were exemplars of one of 21 different, everyday objects. Each of the 21 object categories contained 21 photographic exemplars collected from publicly available sources, totalling 441 category object images in all. Each RSVP stream was made up of random exemplars from 18 different categories, such that there was no more than one exemplar of any category in each stream (Figure 15). This made it easier for participants to identify items based on their categorical membership, rather than specific perceptual qualities (see Kennedy & Most, 2015).
“Baseline” T1 images were randomly selected from the same set of object images and was always an exemplar from a different category than all other items in the stream. 52 graphic, aversive images (e.g., medical injuries, threatening animals, mutilated body parts, and violence) served as the negative T1 stimuli, while 52 neutral
T1 images (also depicting people or animals) were used as a comparison. Negative and neutral images were the same as those used in Experiment 3.
Depending on the trial, T1 appeared randomly at serial position 3, 4, 5, 6, 7, or
8. Based on the condition of the trial, T2 appeared as either the second (lag-2) or fourth
(lag-4) item after T1. At the end of the trial, all of the images that had appeared in the stream appeared in an array (160x120 pixels per image), in addition to three “T1 foils” that were of the same emotional category as the T1 from that trial (Figure 15). Thus, when T1 was negative, there would be four negative image options for T1 – the actual
T1 and three negative T1 foils. When T1 was a “baseline” image, it represented a new category compared to all other items in the stream (and thus all objects in the entire array were from a different object category). In the baseline condition, the four baseline image options were the actual T1 and three other items of different object categories not Emotion distracts regardless of task-relevance | 99
included in the trial. The four possible options for T1 were presented in a row at the top of the screen, and the 17 possible options for T2 were presented in a square shape on the lower portion of the screen. Participants used a mouse to click on the items that had been targets, and heard a ding noise if they made a correct target selection. The next trial began immediately after participants made their response.
Critically, we manipulated the task-relevance of the first bordered image (T1).
For half of the experiment, participants identified both of the bordered items (T1- relevant condition), and in the other half of the experiment they only identified T2 (T1- irrelevant condition). In both conditions, the response screen was the same, but participants could only choose a T1 option when it was relevant. The relevancy conditions were blocked and counterbalanced across participants, such that some participants completed the T1-relevant condition for the first two blocks and T1- irrelevant condition for the second two blocks, and vice versa for the remaining participants. Instructions were provided at the beginning of each block as to whether T1 was relevant or not, and the instructions on the test array screen also differed by condition. In the T1-relevant condition, if a participant was assigned a red T1 and green
T2, instructions on the test array screen read: “Click on both the red and green bordered items” and in the T1-irrelevant condition: “Click on only the green bordered item.” The same instructions were presented when participants were assigned a green T1 and red
T2, except with opposite color names listed.
Before beginning the experiment, participants were shown examples of negative and neutral images used in the experiment to ensure informed consent. They then engaged in a 6-trial practice session, with RSVP rates starting at 200-ms and slowly increasing to the experiment presentation rate of 150-ms. T1 was always relevant in the 100 | Emotion distracts regardless of task-relevance
practice session, and practice trials only included baseline trials. Participants were debriefed after the experiment.
Results
We used accuracy in identifying T1 and T2 as our primary measurement of interest. Chance performance in identifying T1 was 25% (1 out of 4), and chance performance in identifying T2 was 5.88% (1 out of 17).
T2 performance accuracy
We first explored performance accuracy in identifying T2 (see Figure 16). A 2
(Relevancy: T1-Relevant vs. T1-Irrelevant) X 2 (T1 Image Type: Negative vs. Neutral)
X 2 (Lag: 2 vs. 4) ANOVA revealed significant main effects of Relevancy, F(1,63) =
2 427.973, p < .001, ηp =0.872, with better accuracy identifying T2 when T1 was irrelevant than when relevant. There was also a significant effect of T1 Image Type,
2 F(1,63) = 88.741, p < .001, ηp =0.585, with better accuracy for T2 when T1 was neutral compared to negative. The main effect of Lag was also significant, F(1,63) = 429.123,
2 p < .001, ηp =0.872, with better performance at lag-4 than lag-2. There was also a
2 significant Relevancy X Lag interaction, F(1,63) = 143.096, p < .001, ηp =0.694, such that relevancy made a bigger impact at lag-4 than lag-2. The T1 Image Type X Lag
2 Interaction was also significant, F(1,63) = 18.223, p < .001, ηp =0.224, with a greater difference in performance between negative and neutral trials at lag-2 than at lag-4.
There was no Relevancy X T1 Image Type interaction, F(1,63) = 1.202, p = .277,
2 ηp =0.019. There was also a significant 3-way interaction between them, F(1,63) =
2 10.205, p = .002, ηp =0.139. Emotion distracts regardless of task-relevance | 101
Performance in identifying T2 in the T1-relevant, baseline condition was worse than when identifying T2 in the T1-relevant, negative and neutral conditions. This was likely because the T1 images in the baseline condition were object category images just like the filler items, as compared to a more salient, “oddball” person or animal T1 images in the negative and neutral conditions. As such, it was likely harder to identify the T1 images because they were more similar to the filler items in the stream, and thus made it more difficult to also identify T2. Because of the differences in difficulty between conditions, we did not directly compare the baseline condition with the negative and neutral conditions. Nevertheless, consistent with the time-course of the traditional AB, T2 accuracy was worse at lag-2 than at lag-4 after baseline T1s (lag-2:
Figure 16. Experiment 4 performance accuracy for T2. At lag-2, participants were impaired in identifying T2 after negative T1s (compared to neutral T1s) in both the T1- relevant and T1-irrelevant conditions. At lag-4, participants were impaired in the negative compared to neutral T1 conditions in the T1-relevant, but not T1-irrelevant condition. Error bars depict standard error. Note that the y-axis scale differs between the lag-2 and lag-4 graphs.
102 | Emotion distracts regardless of task-relevance
M=30.1%, SD=16.6%; lag-4 (M=64.0%, SD=20.1%), t(63) = 14.758, p<.001, dz=1.844. Baseline T2 performance in the T1-irrelevant condition (M=80.9%,
SD=14.4%) demonstrated that participants were better overall at identifying T2 when
T1 was irrelevant compared to when relevant. This demonstrates that purposeful attention to T1 impairs perception of T2.
Importantly, however, at lag-2 the negative, relative to neutral, T1 images disrupted T2 identification regardless of their relevance. At lag-2, identification of T2 after negative (relative to neutral) T1 images was impaired in both the T1-relevant condition (negative: M=24.9%, SD=17.8%; neutral: M=32.4%, SD=21.2%), t(63) =
4.876, p<.001, dz=0.609, and T1-irrelevant condition (negative: M=67.9%, SD=19.0%; neutral: M=78.5%, SD=14.8%), t(63) = 6.924, p<.001, dz=0.865. At lag-4, the negative T1s (relative to neutral T1s) impaired T2 perception in the T1-relevant condition (negative: M=63.1%, SD=21.7%; neutral: M=69.4%, SD=19.5%), t(63) =
3.534, p<.001, dz=0.442, but not in the T1-irrelevant condition (negative: M=84.0%,
SD=12.6%; neutral: M=83.6%, SD=14.5%), t(63) = 0.470, p=.640, dz=0.059.
Consistent with the 3-way interaction noted above, this suggests that negative T1s were prioritized by attention both when relevant and irrelevant, but that their impact on target perception persisted longer when they were relevant.
The results in the Lag-4 conditions suggest that task-relevant negative distractors impair performance at lag-4, but task-irrelevant negative distractors do not. One could argue that the results we observe demonstrate a ceiling effect at lag-4. However, this seems unlikely for two reasons. The standard deviations in both negative and neutral, irrelevant lag-4 conditions are comparable, which argues against a ceiling effect.
Moreover, in the context of previous literature (Ferrari et al., 2008; Müller et al., 2015; Emotion distracts regardless of task-relevance | 103
Vromen et al., 2014; Weinberg et al., 2012; Wiens et al., 2011), disengagement of attention is the mechanism that we were expected to be modulated by task-relevancy.
T2 performance accuracy given T1
Common practice in attentional blink studies is to analyze the accuracy in identifying T2 given that T1 was correctly reported (Chun & Potter, 1995). By considering T2 performance in this way, we could be more confident that participants were attending to T1 when relevant on any given trial. On average, participants correctly reported T1 correctly in most trials (M=141.2 out of the 168 trials when T1 was present, SD=16.6 trials). The results were virtually the same when accounting for
T1 accuracy (see Figure 17). A 2 (Relevancy) X 2 (T1 Image Type) X 2 (Lag) ANOVA
Figure 17. Performance accuracy for T2 given correct T1 identification. Results were similar to analyses that did not consider T1 accuracy. Like Figure 2, error bars depict standard error, and note that the y-axis scale differs between the lag-2 and lag-4 graphs. 104 | Emotion distracts regardless of task-relevance
2 revealed significant main effects of Relevancy, F(1,63) = 499.817, p < .001, ηp =0.888,
2 T1 Image Type, F(1,63) = 155.308, p < .001, ηp =0.711, and Lag, F(1,63) = 386.307, p
2 < .001, ηp =0.860, all in the same direction as previously noted. There was also a
2 significant Relevancy X Lag interaction, F(1,63) = 122.566, p < .001, ηp =0.660, and
2 T1 Image Type X Lag Interaction, F(1,63) = 15.225, p < .001, ηp =0.195. Unlike analyses that did not consider T1 accuracy, there was a Relevancy X T1 Image Type
2 interaction, F(1,63) = 7.716, p = .007, ηp =0.109, such that overall, negative distractors - more so than neutral distractors - impaired T2 accuracy more when relevant than irrelevant. The 3-way interaction between factors also remained significant, F(1,63) =
2 12.181, p = .001, ηp =0.162.
T1 performance accuracy
Intriguingly, negative distractors impaired target processing more than did neutral distractors despite not being remembered as well by participants. EIB emerged despite worse identification of negative than of neutral T1 images at both lag-2
(negative: M=87.0%, SD=11.3%; neutral: M=92.7%, SD=9.8%), t(63) = 5.801, p<.001, dz=0.725, and lag-4 (negative: M=87.9%, SD=11.0%; neutral: M=92.7%, SD=11.0%),
Figure 18. Performance accuracy for T1. Negative distractors were identified with less accuracy than neutral distractors. This is despite negative distractors having a greater impact on the subsequent T2 identification. Error bars depict standard error.
Emotion distracts regardless of task-relevance | 105
t(63) = 4.408, p<.001, dz=0.899 (see Figure 18). In the framework of competition between emotional items and subsequent targets, these data suggest that negative distractors might win the competition for attention without necessary consequences for memory encoding.
Discussion
The present study compared the impact of task-relevant and task-irrelevant emotionally negative stimuli in order to assess whether their power to be prioritized by attention was subject to people’s attempts to attend to or ignore them. We found that task-relevant and –irrelevant emotional stimuli impaired the detection of subsequent targets to a similar magnitude at lag-2, but that the impact of task-relevance was more apparent by lag-4. These results suggest that an emotionally negative stimulus that is task-relevant is harder to disengage from than a task-irrelevant emotionally negative stimulus, which is consistent with previous literature (e.g., Müller et al., 2015; Vromen et al., 2014).
Surprisingly, emotional T1s were recognized significantly worse than the neutral
T1s. This is inconsistent with previous findings that emotional items are remembered better than non-emotional stimuli (see Dolan, 2002; Kensinger, 2007; McGaugh, 2006;
Wallace, 1965 for reviews), as well as with attentional blink studies that have demonstrated either better memory for emotional T1s than neutral T1s (Ihssen & Keil,
2009, Mathewson et al., 2008) or no difference between them (Schwabe & Wolf, 2010;
Stein et al., 2009). One potentially important difference between these studies and the present experiment is that the stimuli in the earlier studies were words (Ihssen & Keil,
2009; Mathewson et al., 2008; Schwabe & Wolf, 2010) or faces (Stein et al., 2009), to which it may be easier to attach a sematic label. That is, it may be that the semantic content of words and faces are easier to grasp, whereas the stimuli used in the present 106 | Emotion distracts regardless of task-relevance
study relies more on a visual representation of stimuli. Alternatively, the task demands of the study, particularly with intermittent relevance of the first target in this experiment, may have rendered them less task-relevant than attentional blink studies. In a similar vein, recent research has demonstrated that task-relevant attributes of a task can be forgotten when probed, to suggest that some stimuli are attended to and used to compete a task, but may not be stored in memory (Chen & Wyble, 2015). One notable difference between the present study and that conducted by Chen and Wyble is that in our study, participants were required to remember the distractor, whereas in Chen and
Wyble’s experiment, participants did not have to necessarily remember the particular attribute to complete their task. However, because of the distracting nature of emotional distractors in this study, it may be that participants attempt to suppress them as a way to better identify their second target, similarly deprioritizing their content for later memory retrieval. In any case, the memory results here could provide clues to what stages in visual processing emotional distractors wield their impact in EIB. However, this result should be interpreted with caution, as the memorability of the images themselves, and the difference between the actual target item and other foil items were not experimentally balanced. Nevertheless, this finding should be explored in future studies, as it could be informative to understand the underlying mechanisms in emotion- induced blindness.
Notably, the present results appear to contradict previous studies in other ways as well. For example, some research suggests that emotionally powerful stimuli may only impair subsequent target accuracy when the emotional quality of a stimulus is task- relevant (Stein et al., 2009). Stein et al. found that emotional faces as T1 caused a greater blink for T2 when participants had to indicate that face’s emotional expression, but no greater impairment from emotion when the participant instead had to indicate the Emotion distracts regardless of task-relevance | 107
face’s gender. This research suggests that emotional faces only increased blink magnitude when the emotional aspects of the stimulus were relevant. EIB results generally stand in contrast to this finding, such that emotional stimuli impair perception even when they are completely task-irrelevant (Most et al., 2005). However, as Stein et al. note, some stimuli (faces) are likely less emotionally powerful than the stimuli we used in our study, leading to effects that are more or less pronounced depending on the saliency of the emotional stimuli. This may be an additional factor that differentiates the current study from that performed by Vromen and colleagues (Vromen et al., 2014). In that study, the researchers found little impairment from their emotional stimulus - a silhouette image depicting a spider - when it was task-irrelevant, but significant impairment from the same stimulus when it was task-relevant. We found instead that there was always impairment from emotionally negative stimuli, with greater impairment when items were task-relevant. Together, it seems that task-relevancy generally increases the impact of emotional distractors.
108 | Emotion distracts regardless of task-relevance
References
Anderson, A. K. (2005). Affective influences on the attentional dynamics supporting
awareness. Journal of Experimental Psychology: General, 134(2), 258–281.
http://doi.org/10.1037/0096-3445.134.2.258
Anderson, A. K., & Phelps, E. A. (2001). Lesions of the human amygdala impair
enhanced perception of emotionally salient events. Nature, 411(6835), 305–9.
http://doi.org/10.1038/35077083
Arnell, K. M., Killman, K. V, & Fijavz, D. (2007). Blinded by emotion: target misses
follow attention capture by arousing distractors in RSVP. Emotion, 7(3), 465–477.
http://doi.org/10.1167/4.8.359
Bradley, M. M. (2009). Natural selective attention: Orienting and emotion.
Psychophysiology, 46(1), 1–11. http://doi.org/10.1111/j.1469-
8986.2008.00702.x.Natural
Brainard, D. H. (1997). The Psychophysics Toolbox. Spatial Vision, 10(4), 433–436.
http://doi.org/10.1163/156856897X00357
Butler, B. C., & Klein, R. (2009). Inattentional blindness for ignored words: comparison
of explicit and implicit memory tasks. Consciousness and Cognition, 18(3), 811–
9. http://doi.org/10.1016/j.concog.2009.02.009
Chen, H., & Wyble, B. (2015). Amnesia for object attributes: Failure to report attended
information that had just reached conscious awareness. Psychological Science,
26(2), 203–210. http://doi.org/10.1177/0956797614560648
Chun, M. M., & Potter, M. C. (1995). A two-stage model for multiple target detection in
rapid serial visual presentation. Journal of Experimental Psychology: Human
Perception and Performance, 21(1), 109–27. http://doi.org/10.1037/0096-
1523.21.1.109 Emotion distracts regardless of task-relevance | 109
Dodd, H. F., Vogt, J., Turkileri, N., & Notebaert, L. (2016). Task relevance of
emotional information affects anxiety-linked attention bias in visual search.
Biological Psychology. http://doi.org/10.1016/j.biopsycho.2016.01.017
Dolan, R. J. (2002). Emotion, cognition, and behavior. Science, 298(5596), 1191–4.
http://doi.org/10.1126/science.1076358
Dolcos, F., & McCarthy, G. (2006). Brain systems mediating cognitive interference by
emotional distraction. The Journal of Neuroscience, 26(7), 2072–9.
http://doi.org/10.1523/JNEUROSCI.5042-05.2006
Ferrari, V., Bradley, M. M., Codispoti, M., & Lang, P. J. (2013). Detecting novelty and
significance. Journal of Cognitive Neuroscience, 22(2), 404–411.
http://doi.org/10.1162/jocn.2009.21244.Detecting
Ferrari, V., Codispoti, M., Cardinale, R., & Bradley, M. M. (2008). Directed and
motivated attention during processing of natural scenes. Journal of Cognitive
Neuroscience, 20(10), 1753–1761. http://doi.org/10.1162/jocn.2008.20121
Folk, C. L., Leber, A. B., & Egeth, H. E. (2002). Made you blink! Contingent
attentional capture produces a spatial blink. Perception & Psychophysics, 64(5),
741–753. http://doi.org/10.3758/BF03194741
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.
http://doi.org/10.1037/0096-1523.18.4.1030
Fox, E., Griggs, L., & Mouchlianitis, E. (2007). The detection of fear-relevant stimuli:
are guns noticed as quickly as snakes? Emotion, 7(4), 691–696.
http://doi.org/10.1037/1528-3542.7.4.691 110 | Emotion distracts regardless of task-relevance
Fox, E., Russo, R., Bowles, R., & Dutton, K. (2001). Do threatening stimuli draw or
hold visual attention in subclinical anxiety? Journal of Experimental Psychology:
General, 130(4), 681–700. http://doi.org/10.1037/0096-3445.130.4.681
Georgiou, G. A., Bleakley, C., Hayward, J., Russo, R., Dutton, K., Eltiti, S., & Fox, E.
(2005). Focusing on fear: Attentional disengagement from emotional faces. Visual
Cognition, 12, 1–13. http://doi.org/10.1080/13506280444000076.Focusing
Ihssen, N., & Keil, A. (2009). The costs and benefits of processing emotional stimuli
during rapid serial visual presentation. Cognition and Emotion, 23, 37–41.
http://doi.org/10.1080/02699930801987504
Kennedy, B. L., & Most, S. B. (2015). Affective stimuli capture attention regardless of
categorical distinctiveness: An emotion-induced blindness study. Visual
Cognition, 23(1-2), 105–117. http://doi.org/10.1080/13506285.2015.1024300
Kennedy, B. L., Rawding, J., Most, S. B., & Hoffman, J. E. (2014). Emotion-induced
blindness reflects competition at early and late processing stages: An ERP study.
Cognitive, Affective & Behavioral Neuroscience, 14(4), 1485-1498.
http://doi.org/10.3758/s13415-014-0303-x
Kensinger, E. A. (2007). Negative emotion enhances memory accuracy: Behavioral and
neuroimaging evidence. Current Directions in Psychological Science, 16(4), 213–
218. http://doi.org/10.1111/j.1467-8721.2007.00506.x
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. http://doi.org/10.1162/jocn_a_00127
Mathewson, K. J., Arnell, K. M., & Mansfield, C.A. (2008). Capturing and holding
attention: The impact of emotional words in rapid serial visual presentation.
Memory & Cognition, 36(1), 182-201. Emotion distracts regardless of task-relevance | 111
McGaugh, J. L. (2006). Make mild moments memorable: Add a little arousal. Trends in
Cognitive Sciences, 10(8), 345–7. http://doi.org/10.1016/j.tics.2006.06.001
Mogg, K., & Bradley, B. P. (1998). A cognitive-motivational analysis of anxiety.
Behavior Research and Therapy, 36, 809–848. http://doi.org/10.1016/S0005-
7967(98)00063-1
Most, S. B., Scholl, B. J., Clifford, E. R., & Simons, D. J. (2005). What you see is what
you set: sustained inattentional blindness and the capture of awareness.
Psychological Review, 112(1), 217–42. http://doi.org/10.1037/0033-
295X.112.1.217
Most, S. B., & Wang, L. (2011). Dissociating spatial attention and awareness in
emotion-induced blindness. Psychological Science, 22(3), 300–5.
http://doi.org/10.1177/0956797610397665
Müller, S., Rothermund, K., & Wentura, D. (2015). Relevance drives attention:
Attentional bias for gain- and loss-related stimuli is driven by delayed
disengagement. The Quarterly Journal of Experimental Psychology, 69(4), 752-
763. http://doi.org/10.1080/17470218.2015.1049624
Öhman, A., Flykt, A., & Esteves, F. (2001). Emotion drives attention: Detecting the
snake in the grass. Emotion, 130(3), 466–478. http://doi.org/10.1037/AXJ96-
3445.130.3.466
Pelli, D. G. (1997). The VideoToolbox software for visual psychophysics: transforming
numbers into movies. Spatial Vision, 10(4), 437–442.
http://doi.org/10.1163/156856897X00366
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 112 | Emotion distracts regardless of task-relevance
Psychology: Human Perception and Performance, 18(3), 849–60.
http://doi.org/10.1037/0096-1523.18.3.849
Sakaki, M., Niki, K., & Mather, M. (2012). Beyond arousal and valence: The
importance of the biological versus social relevance of emotional stimuli.
Cognitive, Affective & Behavioral Neuroscience, 12(1), 115–139.
http://doi.org/10.1016/j.biotechadv.2011.08.021.Secreted
Sawaki, R., Geng, J. J., & Luck, S. J. (2012). A Common Neural Mechanism for
Preventing and Terminating the Allocation of Attention. Journal of Neuroscience,
32(31), 10725–10736. http://doi.org/10.1523/JNEUROSCI.1864-12.2012
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.
http://doi.org/10.3758/APP
Schupp, H. T., Cuthbert, B. N., Bradley, M. M., Cacioppo, J. T., Ito, T., & Lang, P. J.
(2000). Affective picture processing: The late positive potential is modulated by
motivational relevance. Psychophysiology, 37(2), 257–261.
http://doi.org/10.1111/1469-8986.3720257
Schwabe, L., Merz, C. J., Walter, B., Vaitl, D., Wolf, O. T., & Stark, R. (2011).
Emotional modulation of the attentional blink: The neural structures involved in
capturing and holding attention. Neuropsychologia, 49(3), 416–425.
http://doi.org/10.1016/j.neuropsychologia.2010.12.037
Schwabe, L., & Wolf, O. T. (2010). Emotional modulation of the attentional blink: is
there an effect of stress? Emotion, 10(2), 283–288.
http://doi.org/10.1037/a0017751 Emotion distracts regardless of task-relevance | 113
Stein, T., Zwickel, J., Ritter, J., Kitzmantel, M., & Schneider, W. X. (2009). The effect
of fearful faces on the attentional blink is task dependent. Psychonomic Bulletin &
Review, 16(1), 104–109. http://doi.org/10.3758/PBR.16.1.104
Vromen, J. M. G., Lipp, O. V., & Remington, R. W. (2014). The spider does not always
win the fight for attention: Disengagement from threat is modulated by goal set.
Cognition and Emotion, 29(7), 1185-1196.
http://doi.org/10.1080/02699931.2014.969198
Wallace, W. P. (1965). Review of the historical, empirical, and theoretical status of the
von Restorff Phenomenon. Psychological Bulletin, 63(6), 410–424.
http://doi.org/10.1037/h0022001
Wang, L., Kennedy, B. L., & Most, S. B. (2012). When emotion blinds: A
spatiotemporal competition account of emotion-induced blindness. Frontiers in
Psychology, 3, 1–6. http://doi.org/10.3389/fpsyg.2012.00438
Weinberg, A., Hilgard, J., Bartholow, B. D., & Hajcak, G. (2012). Emotional targets:
Evaluative categorization as a function of context and content. International
Journal of Psychophysiology, 84(2), 149–154.
http://doi.org/10.1016/j.ijpsycho.2012.01.023
Wiens, S., Sand, A., Norberg, J., & Andersson, P. (2011). Emotional event-related
potentials are reduced if negative pictures presented at fixation are unattended.
Neuroscience Letters, 495(3), 178–182.
http://doi.org/10.1016/j.neulet.2011.03.042
Williams, J. M. G., Mathews, A., & MacLeod, C. (1996). The emotional Stroop task
and psychopathology. Psychological Bulletin, 120(1), 3–24.
http://doi.org/10.1037/0033-2909.120.1.3 114 | Emotion distracts regardless of task-relevance
Yantis, S., & Egeth, H. E. (1999). On the distinction between visual salience and
stimulus-driven attentional capture. Journal of Experimental Psychology: Human
Perception and Performance, 25(3), 661–676. http://doi.org/10.1037/0096-
1523.25.3.661
Chapter 4
Proactive deprioritization of emotional distractors enhances target perception
Briana L. Kennedy and Steven B. Most
116 | Proactive warning improves EIB
Abstract
Emotional distractors can impair our ability to report items that appear soon after them, an effect known as emotion-induced blindness (EIB). Can we deprioritize emotional distractors when forewarned that they will appear? To assess this possibility, we tested whether participants could overcome EIB when forewarned about the nature of an emotional distractor. On each trial, participants searched for one target (a rotated picture) presented within a rapid serial visual stream of upright images. An aversive, erotic, or neutral distractor could precede the target by either 200- or 400-msec. At the start of some trials, participants were informed which kind of distractor would appear on that trial, but in other trials they received no advance information. Results revealed that the provision of distractor information significantly improved target perception following both aversive and erotic distractors at the early SOA, and following aversive distractors at the later SOA. These results suggest that people can give targets a competitive edge in perceptual processing by proactively deprioritizing emotional distractors.
Proactive warning improves EIB | 117
In the competition for perceptual priority, emotional stimuli often dominate over non-emotional stimuli, even when they are task-irrelevant (Arnell, Killman, & Fijavz,
2007; Most, Chun, Widders, & Zald, 2005; Öhman, Flykt, & Esteves, 2001; Williams,
Mathews, & MacLeod, 1996). This is apparent in emotion-induced blindness (EIB), a phenomenon wherein emotional distractors spontaneously disrupt perception of a pre- defined, rapidly presented target that appears soon afterwards (Most et al., 2005). EIB has been suggested to stem from spatiotemporal competition between targets and emotional distractors (Most & Wang, 2011; Wang, Kennedy, & Most, 2012). Evidence suggests that it is possible to mitigate the degree to which emotional distractors dominate such perceptual competition by providing people with concrete details about features of the target – thereby enabling them to prepare a more vivid attentional template (Most et al., 2005; Most et al., 2006). But can people also buffer themselves against the attentional draw of emotional distractors when forewarned about the type of distractor that will appear?
Individuals can tune their attention to better overcome distraction. For example, people can tune attention to features of a target stimulus, and tend to be distracted by stimuli that have features that match target stimuli, but are not as distracted by stimuli that do not share target features (Folk & Remington, 1998; Folk, Leber, & Egeth, 2002;
Leber & Egeth, 2006). Attentional control refers to one’s ability to cope with conflict in order to guide attention toward goal-relevant information. People can apply attentional control both proactively and reactively (Braver, Gray, & Burgess, 2007; Braver, 2012).
Proactive control refers to the recruitment of cognitive resources in preparation for a cognitive challenge, whereas reactive control involves re-establishing control after it is disrupted. 118 | Proactive warning improves EIB
In studies of EIB, some participants can overcome EIB when they are given concrete information about the target they are looking for (Most, Chun, Johnson, &
Kiehl, 2006; Most et al., 2005). For example, when participants – at least those who scored low in a measure related to anxiety – were told that their target would be an image of a building, they performed better than when they were told that their target could be an image of either a building or a landscape (Most et al,. 2005). This evidence suggests that targets can gain a competitive edge when people are able to form a more detailed search template, enabling more effective preparatory visualization of it (Most et al., 2005). Bolstering this interpretation are findings that such instructions elicited greater activation in imagery-related neural regions, despite no difference in visual stimulation (Most et al., 2006). While these findings demonstrate that proactive attentional control can modulate the impact of emotional distraction when participants know what kind of target to expect, there are situations when people only know to brace themselves for a salient distraction. For example, when driving on an unfamiliar roadway, a warning sign indicating an accident ahead may help people brace themselves against the impulse to rubberneck without regard to whether they otherwise know what the road looks like beyond the bend. Thus, the question arises as to whether people can proactively prepare themselves to avoid emotional distraction even without knowing anything more about what their target will be.
While most emotion-induced blindness studies use negative emotional stimuli, positive emotional stimuli also induce deficits in being able to report subsequent targets
(Arnell et al., 2007; Most, Smith, Cooter, Levy, & Zald, 2007). Indeed, erotic stimuli have been found to elicit EIB similarly, if not more powerfully, compared to negative ones (Most et al., 2007). Moreover, Most and colleagues (2007) found that monetary incentive for good performance somewhat decreased EIB following negative images, Proactive warning improves EIB | 119
but not following erotic ones. This might suggest that participants can effectively recruit attentional control to overcome distraction by negative stimuli, but struggle to overcome distraction by erotic stimuli.
In one EIB study, participants knew the kind of distractor to expect. Presenting targets in rapid streams of words, Arnell and colleagues told participants the exact distractor words that they would see in subsequent trials (in blocks of all sexual words or all music related words), and compared their performance to a group of participants who did not know the words (Experiment 3, Arnell, Killman, & Fijavz, 2007). The task was for participants to identify the color word (e.g., “brown”) amongst other non-color words, and to ignore all other words, including the distractor before it (e.g., “orgasm”).
The researchers found that even when participants knew the emotional distractor words to expect, EIB was still observed, and to the same extent as those participants who did not know what to expect.
Thus, in the study by Arnell and colleagues (2007), prior knowledge about distractors did not help participants overcome EIB from positive stimuli. This might be indicative of the inability to use proactive attentional control to specifically overcome
EIB caused by positive emotional stimuli, or it may reflect the type of design that they used in their study. Notably, the stimuli used in that experiment (explicit words) are quite different than the types of very graphic images used in other EIB tasks, making it unclear if the effect may be attenuated with a stimulus that elicits stronger emotional responses. Indeed, in the study conducted by Arnell, Killman and Fijavz, (2007), they did not find strong EIB from negative, highly arousing words (in contrast to the negative, emotional stimuli that elicit an EIB using images; Most et al., 2005), and the stimuli in that study may not have been subject to attenuation by proactive control in the same way. Performance may have already been at ceiling levels, with little potential for 120 | Proactive warning improves EIB
proactive control to make a difference. Moreover, the design that they used was a blocked design, such that all trials with forewarning about the distractors appeared together and all trials without forewarning were tested in a separate block. Use of attentional control may differ in a more trial-by-trial basis of engaging proactive control.
The benefit of forewarning about upcoming distractors can also take time to develop. In a recent study using non-emotional distractors, Cunningham and Egeth found that by telling participants the kind of salient, singleton distractor to expect on a trial in a visual search task, participants were able to avoid attentional capture by that colored singleton in the service of improving their search (Cunningham & Egeth, 2016;
Cunningham & Egeth, 2015). That is, when participants knew there would be a certain colored distractor on the task, they were better able to overcome distraction caused by it than when they did not know the kind of distractor to expect. This help from top-down cues was learned throughout the experiment, such that the benefit from knowing the type of distractor gradually emerged with experience in the task.
There also seem to be individual differences in the ability to exercise proactive control. Event-related potential research demonstrates that individuals that are higher in negative affect produce a smaller pre-stimulus slow-wave before a task-relevant stimulus is presented (West, Choi, & Travers, 2010), suggesting that negative affect can disrupt processes used to prepare for an upcoming stimulus. Similarly, clinically depressed individuals have abnormal brain activity in the contingent negative variation
(an ERP component associated with anticipatory attention) when anticipating a stimulus compared to healthy controls (Vanderhasselt et al., 2014). Relatedly, individuals high in worry have difficulty in using attentional control more generally, evidenced by their poor performance in a flanker task when task-irrelevant emotional distractors were Proactive warning improves EIB | 121
presented compared to controls (Fox, Dutton, Yates, Georgiou, & Mouchlianitis, 2015).
Individual differences in the trait harm-avoidance also predicted the ability to utilize information about the target to aid their performance on an EIB task, as those high in harm-avoidance were unable to benefit from knowing more information about their target (Most et al., 2006). Furthermore, self-report measures of attentional control have been developed (Derryberry & Reed, 2002), with research suggesting that individuals who score high in attentional control are better able to overcome distraction from irrelevant emotional distractors (Peers & Lawrence, 2009).
The aim of the current study is to examine whether people can proactively inhibit attention to highly salient emotional distractors when they know what kind of distractor to expect. We tested this when the temporal distance between the distractor and target was 200ms (lag-2) and 400ms (lag-4). We predicted that participants would benefit from being forewarned when they knew that they would see a negative distractor, and likely see this benefit at both lag-2 and lag-4. We further predicted that the forewarning would benefit performance in trials with a positive distractor, but – based on previous findings (Most et al., 2007) – perhaps not to the same extent as negative distractors. Measures of depression, harm avoidance, and attentional control were also included and predicted to correlate with the ability to use proactive forewarnings to overcome emotional distraction.
Experiment 5
Method
Participants. 85 undergraduates from the University of New South Wales (mean age=19.4 years; 60 female, 25 male) participated in exchange for course credit. We employed a large sample in order to obtain enough power to observe possible individual 122 | Proactive warning improves EIB
differences in depression, harm avoidance, and attentional control. One female participant’s data were excluded because of low overall percentage accuracy (more than three standard deviations below the mean accuracy across the sample). All participants gave informed consent and the experiment was approved by the University of New
South Wales Human Research Ethics Approval Panel.
Materials and Procedure. Stimuli were presented and responses gathered via the
Psychophysics Toolbox for Matlab (Brainard, 1997; Pelli, 1997). Participants sat at a comfortable distance from the computer screen and head position was not fixed.
Stimuli appeared against a black background on a 24-inch Benq LED monitor with a refresh rate of 120Hz. Screen resolution was 1920x1080 pixels.
Stimuli were colored, 320x240 pixel photographs. Distractors were emotionally negative, emotionally positive (erotic), or “baseline” images. 56 negative and 56 erotic images were chosen based on ratings of valence and arousal and were mostly gathered from the International Affective Picture System (IAPS; Lang, Bradley, & Cuthbert,
2008), with the rest supplemented by similar images taken from publicly available sources. 252 filler images depicted architectural buildings and landscapes. Baseline distractors came from the same bank of landscape and architectural images as the filler images. Target images also depicted landscape and architectural buildings, but were rotated 90 degrees to the left or right (64 images rotated in both left and right directions). Images of each category were presented in random order, and all images in a category were shown once before the set was randomly ordered and presented again.
The experiment included four blocks of 96 trials (384 trials total). Blocks were made up of “mini-blocks” of forewarning conditions (“negative warning present”,
“erotic warning present”, and “warning absent”). This was to prevent participants from Proactive warning improves EIB | 123
having to alternate their attentional set too often throughout the experiment, but changed enough so that they could maintain their set for relatively short amounts of time.
Negative and erotic “warning present” mini-blocks were 8 trials long, while “warning absent” mini-blocks were 32 trials long. There were two of each type of mini-blocks per block. Negative “warning present” mini-blocks were composed entirely of negative trials, while erotic “warning present” mini-blocks were composed entirely of erotic trials. Half of the trials in the warning present mini-blocks were lag-2 and the other half lag-4, and their order was randomized. “Warning absent” mini-blocks included 8 negative, 8 erotic, and 16 baseline distractors, randomly ordered in the mini-block.
Again, half of trials for each image type in the warning absent mini-blocks were lag-2, while the other half were lag-4. This arrangement of trials and mini-blocks allowed for each block to be balanced for the number of trials in each condition.
At the start of every mini-block, participants would see a warning about the kind of distractors to expect. These would read “Ignore Gruesome,” “Ignore Erotic,” or
“Unknown” at the top of the screen in white, Arial font for each of the forewarning types. The warning first appeared for 300 milliseconds by itself, then shown with a fixation cross for an additional 100ms, and then continued to display throughout the trials for the entire mini-block. These warnings were always accurate, and their timing was intended to enable participants to establish proactive control before the trials in that mini-block began. 124 | Proactive warning improves EIB
Figure 19. Schematic of a partial trial sequence in Experiment 5. Warnings appeared in white text above the RSVP. See text for details.
Each trial consisted of a rapid serial visual presentation (RSVP) of 17 images presented at a rate of 100 ms/item. The stream of images was presented in the center of the screen (see Figure 19).
The distractor appeared randomly at serial position 3, 4, 5, 6, or 7. Participants indicated the direction they believed the target was rotated by using the respective arrow key on the keyboard, and heard a bell noise if they made a correct target selection. The next trial began one hundred milliseconds after participants made their response.
Three questionnaires were used to measure individual differences. The first measured depression (BDI-II – suicide ideation item removed; Beck, Steer, & Brown,
1996) in accordance with previous studies suggesting that individuals high in negative affect have difficulty using proactive attentional control. The TPQ-HA questionnaire was used to measure harm avoidance (Cloninger, Przybeck, & Svrakic, 1991), with the prediction that those scoring high would have difficulty using proactive control when exposed to something particularly harmful (the negative stimuli). The Attentional
Control Scale (Derryberry & Reed, 2002) was administered with the prediction that it would correlate with one’s ability to utilize proactive cues to aid performance. We examined potential correlations between each of these measures and the observed Proactive warning improves EIB | 125
differences in performance on “warning present” vs. “warning absent” trials (positive difference score = better performance with forewarning than without forewarning); these correlations were calculated separately for each of the lag x distractor type conditions.
We also included a questionnaire to assess how difficult the participants found the task. In the first four questions, participants were asked to rate on a seven point scale (from “not at all hard” to “very hard”) how hard it was to ignore the negative and erotic distractors when they could expect them and when they could not expect them.
They also reported how hard it was to find the target across the entire experiment. 64 participants completed this questionnaire, as we did not introduce it until data collection had already started.
Before beginning the experiment, participants were shown examples of negative and erotic images used in the experiment to ensure informed consent. They then completed the three individual differences questionnaires. After completing the questionnaires, participants began the EIB task, starting with an 8-trial practice session, which was always an “unknown” condition that included all distractor types. RSVP rates in the practice trial started at 200-ms and slowly increased to the experiment presentation rate of 100-ms. At the end of the EIB task, participants completed the questionnaire to rate how difficult they found the task.
Results 126 | Proactive warning improves EIB
Figure 20 shows the percentage accuracies for correctly responding to the target in each condition across all trials in Experiment 5. A 2 (Forewarning: Warning Present vs Warning Absent) X 2 (Lag: 2 vs 4) X 2 (Distractor Type: Negative vs Erotic)
ANOVA revealed a significant main effect of forewarning, F(1,83) = 25.090, p < .001,
2 ηp =0.232, with greater accuracy in warning present trials compared to warning absent
2 trials. There was a main effect of lag, F(1,83) = 262.601, p < .001, ηp =0.760, with worse performance at lag-2 than lag-4. And there was a main effect of distractor type,
Figure 20. Experiment 5 results. Emotion-induced blindness was observed in both negative and erotic conditions. Forewarning benefited target accuracy in the negative lag-2 and lag-4 conditions, and benefited target accuracy in the erotic lag-2 condition. Error bars depict standard error.
2 F(1,83) = 18.901, p < .001, ηp =0.185, with worse performance following erotic distractors than negative distractors. There was also a significant 3-way interaction
2 between them, F(1,83) = 6.996, p = .010, ηp =0.078, but no significant 2-way interactions between factors (ps>.05).
Performance at lag-2 was significantly impaired in both the negative and erotic conditions compared to baseline performance (M=88.5%, SD=7.3%). Forewarning about the distractors aided target identification at lag-2 in both negative (warning Proactive warning improves EIB | 127
present: M=76.7%, SD=12.0%; warning absent: M=74.2%, SD=11.2%), t(83) = 2.270, p=.026, dz=0.248 and erotic (warning present: M=74.2%, SD=12.8%; warning absent:
M=69.0%, SD=13.0%), t(83) =4.099, p < .001, dz=0.447, conditions. At lag-4, the forewarnings also significantly improved performance in the negative condition
(warning present: M=87.4%, SD=8.5%; warning absent: M=84.2%, SD=9.9%), t(83)
=3.522, p= .001, dz=0.384, but not in the erotic condition (warning present: M=83.9%,
SD=10.3%; warning absent: M=82.8%, SD=10.2%), t(83) =1.072, p=.287, dz=0.117.
The differences in performance between the warning present and warning absent conditions did not change across blocks. A 2 (Forewarning) X 2 (Lag) X 2 (Distractor
Type) X 4 (Block) revealed no significant interaction of block and forewarning,
2 F(1,249) = 0.075, p = .973, ηp =0.001.
There was also no systematic difference for early vs. late trial positions in a mini-block. We assessed this by comparing trials in the first half of a mini-block compared to the second half of a mini-block (Mini-Block Position). A 2 (Forewarning)
X 2 (Lag) X 2 (Distractor Type) X 2 (Mini-Block Position) revealed no interaction
2 between Mini-Block Position and Forewarning, F(1,83) = 0.251, p = .617, ηp =0.003.
We next assessed whether the difference in performance as a function of whether a warning was present or absent correlated with self-reported levels of depression, harm avoidance, or attentional control. Two participants did not complete every question in the BDI-II and thus were missing depression scores. Depression significantly correlated with warning-present minus warning-absent performance in the negative, lag-2 condition, r(82)=.263, p=.017. However, this correlation was opposite the direction that was predicted, such that individuals with higher levels of depression performed better with forewarning about the distractor. Depression and the difference in accuracy between warning-present and warning-absent did not significantly correlate 128 | Proactive warning improves EIB
in the negative lag-4 condition or either lags in the erotic condition. Harm avoidance significantly correlated with warning-present minus warning-absent performance in the erotic, lag-4 condition; participants with greater levels of harm avoidance were less likely to benefit from forewarnings - r(84)=-.227, p=.038, but this correlation was not present in any other condition. There were no reliable correlations between attentional control and the difference in performance in the warning present and warning absent conditions across all conditions (negative or erotic at either lag; ps>.05).
While participants’ average performance correlated strongly with their self- reported difficulty in identifying the direction of the rotated picture throughout the experiment, r(63)=-.412, p=.001, there were no significant correlations with the warning present minus warning absent performance and the difference between self-reported difficulty in overcoming warning present vs. warning absent trials with the different types of distractors (ps>.05).
Discussion
In Experiment 5, participants generally performed better when they knew what kind of distractor to expect, and this was observed in all conditions except for the erotic lag-4 condition. The lag-4 erotic condition may not have exhibited the additional benefit because performance was already at ceiling levels, such that the aid of proactive warning could not help overcome erotic distractors when performance had reached its maximal level after erotic distractors.
The benefit of forewarnings did not seem to build across the length of the experiment, nor during the time across mini-blocks, despite previous research to suggest that the benefit from proactive warning develops over time (Cunningham & Egeth,
2015). Individual differences in depression correlated with performance in the negative Proactive warning improves EIB | 129
lag-2 condition, such that individuals who reported higher levels of depression were best able to benefit from forewarnings about the distractors. It could be that in this design, participants with higher levels of depression could use the warnings as a way to better prepare for the distractors, whereas participants who scored low in depression were more likely to use attentional control more spontaneously. Greater levels of harm avoidance also correlated with worse benefit in performance for forewarnings, but this correlation was only present in the erotic lag-4 condition. No other individual differences, including attentional control, significantly correlated with performance.
Experiment 6
The results of Experiment 5 suggest that people are capable of bracing themselves against emotional distractors when they know the nature of the emotional distractor (e.g., negative or erotic). But what if they have less specific forewarning and are simply told to brace themselves for something that might be graphic without knowing the valence? In Experiment 6, we assessed whether the enhancement in performance from a warning about a distractor type would be observed if the instructions about the distractors were more general. We simply instructed participants that they would see “graphic” stimuli as their warning for both negative and erotic distractors. Encouraged by the correlation especially between depression and performance in the negative, lag-2 condition, we also collected individual difference data to see if the results would replicate in Experiment 6.
Method
Participants. 85 undergraduates from the University of New South Wales (mean age=20.4 years; 58 female, 27 male) participated in exchange for course credit. One female participant’s data were excluded because of low overall percentage accuracy 130 | Proactive warning improves EIB
(performing more than 3 standard deviations below the sample mean accuracy). All participants gave informed consent and the experiment was approved by the University of New South Wales Human Research Ethics Approval Panel.
Materials and Procedure. The materials and procedure used in Experiment 6 were the same as those used in Experiment 5 with the following exceptions. The types of mini-blocks were now simply “warning present” and “warning absent.” Like
Experiment 5, warning present blocks were still 8 trials long and warning absent blocks were 32 trials long. However, in Experiment 6, there were 4 warning present mini- blocks and 2 warning absent mini-blocks per block. In Experiment 6, each warning present mini-block was composed of both negative and erotic distractor types (four images of each type, with two of each kind of lag per image type). Trials with negative and erotic distractors were randomly ordered in the mini-block. The kind of mini-block could not repeat more than twice in a block before changing to the other type of mini- block. The warning for the warning present trials now read “Ignore Graphic,” while the warning absent warning remained as “Unknown.”
Results Proactive warning improves EIB | 131
Figure 21. Experiment 6 results. Emotion-induced blindness was observed in both negative and erotic conditions. Forewarning benefited target accuracy in the negative lag-2 condition, and target accuracy in the erotic lag-2 and lag-4 conditions. Error bars depict standard error.
A 2 (Forewarning: Warning Present vs Warning Absent) X 2 (Lag: 2 vs 4) X 2
(Distractor Type: Negative vs Erotic) ANOVA revealed a significant main effect of
2 forewarning, F(1,83) = 14.129, p < .001, ηp =0.145, with better accuracy in warning present than warning absent conditions. There was a main effect of lag, F(1,83) =
2 428.121, p < .001, ηp =0.838, with worse performance at lag-2 than lag-4. And there
2 was a main effect of distractor type, F(1,83) = 10.534, p = .002, ηp =0.113, with worse performance in erotic than negative conditions, but no significant 3-way interaction
2 between them, F(1,83) = 0.468, p = .496, ηp =0.006, nor any significant 2-way interactions between them, (ps>.05).
Baseline performance was similar to that in Experiment 5 (M=89.9%,
SD=5.8%), and both negative and erotic distractors impaired performance at lag-2.
Forewarnings about the distractors aided target identification at lag-2 in both negative
(warning present: M=75.7%, SD=11.4%; warning absent: M=73.3%, SD=11.8%, t(83)
=2.066, p=.042, dz=0.225) and erotic (warning present: M=72.7%, SD=12.8%; warning absent: M=70.0%, SD=11.9%, t(83) =2.454, p = .016, dz=0.268) conditions. At lag-4, 132 | Proactive warning improves EIB
the warnings did not improve performance in the negative condition (warning present:
M=87.9%, SD=8.2%; warning absent: M=87.6%, SD=8.7%, t(83) =0.392, p= .696, dz=0.043), but they did improve performance in the erotic condition (warning present:
M=86.9%, SD=8.2%; warning absent: M=85.1%, SD=8.7%, t(83) =2.135, p=.036, dz=0.233; see Figure 21).
In Experiment 6, there were also no systematic differences in performance in the warning present versus warning absent conditions across blocks, or across mini-blocks.
A 2 (Forewarning) X 2 (Lag) X 2 (Distractor Type) X 4 (Block) revealed no significant
2 interaction of block and forewarning, F(1,249) = 1.037, p = .377, ηp =0.012, and a 2
(Forewarning) X 2 (Lag) X 2 (Distractor Type) X 2 (Mini-Block Position) revealed no interaction between Mini-Block Position and Proactive Type, F(1,84) = 0.472, p = .494,
2 ηp =0.006.
Unlike Experiment 5, performance in warning present compared to warning absent trials did not significantly correlate with levels of depression, harm avoidance, or attentional control, (ps>.05). Like Experiment 5, participants were very good at assessing their own performance at finding the target throughout the entire experiment, r(84)=-.430, p<.001. However, the self-reported difficulty overcoming distractors in warning present versus warning absent trials did not correlate with the difference between warning present and warning absent performance after either distractor type
(ps>.05).
Discussion
The results from Experiment 6 replicated the general pattern of Experiment 5 at lag-2, suggesting that knowing that a graphic stimulus will appear helps people avoid being distracted by it. At lag-4, proactive warning improved performance after erotic Proactive warning improves EIB | 133
distractors, but not after negative distractors. While highly speculative, the term
“graphic” in Experiment 6 may have been more intuitively associated with erotic images than negative images to participants, contributing to this difference. It is worth noting that performance at lag-4 in both Experiment 5 and Experiment 6 was close to ceiling levels across conditions, likely making differences between the warning present and warning absent conditions difficult to observe at this later lag.
Experiment 7
Although Experiments 5 and 6 suggested that people are able to brace themselves when forewarned against distraction by emotional stimuli, one possibility is that participants may simply have been motivated to try harder on trials that they expected to be difficult. In other words, warnings that a trial would contain an emotional distractor may have doubled as forewarning that the trials would be harder.
To tease apart these two possibilities, we ran a third experiment to explore how participants might be able to use proactive effort without any knowledge about the distractor. In Experiment 7, we told participants that some trials may be more difficult than others, although in actuality this was not the case.
Method
Participants. 85 undergraduates from the University of New South Wales (mean age=20.2 years; 43 female, 42 male) participated in exchange for course credit. The data of two participants (one female and one male) were excluded from the analyses because of low overall accuracy (three standard deviations below average performance). All participants gave informed consent and the experiment was approved by the University of New South Wales Human Research Ethics Approval Panel. Participants were compensated with course credit for their participation in the experiment. 134 | Proactive warning improves EIB
Materials and Procedure. The materials and procedure used in Experiment 7 were the same as those used in Experiment 6 with the following exceptions. When the warning was present, the forewarning read “Difficult” and when the warning was absent, the forewarning read “Normal”. Warning-present mini-blocks were made up of 12 trials, while warning absent mini-blocks were made up of 24 trials. There were four warning- present mini-blocks per block, and two warning-absent mini-blocks per block. Both warning present and warning absent mini-blocks were made up of an equal number of negative, erotic, and baseline distractor trials (four of each type in warning present mini- blocks, and 8 of each type in warning absent mini-blocks). Note that unlike
Experiments 5 and 6, baseline trials were in both warning present and warning absent mini-blocks. This change was made to maintain the proportion of distractor types, and thus actual difficulty, to be the same in both warning present and warning absent trial types. Thus, there was no difference in actual difficulty between the “Difficult” and
“Normal” trials. The eight practice trials were all “Normal” trials.
Results
A 2 (Forewarning: Warning Present vs Warning Absent) X 2 (Lag: 2 vs 4) X 2
(Distractor Type: Negative vs Erotic) ANOVA revealed a significant main effect of
2 forewarning, F(1,82) = 11.398, p < .001, ηp = 0.122, with better performance in warning-present than warning-absent conditions. There was a main effect of lag,
2 F(1,82) = 296.814, p < .001, ηp = .784, with worse performance at lag-2 than lag-4.
2 There was also a main effect of distractor type, F(1,82) = 10.545, p = .002, ηp = 0.114, with worse performance following erotic distractors than negative distractors. There was also a significant Forewarning X Distractor Type interaction, F(1,82) = 9.330, p =
2 .003, ηp = 0.102, with a greater influence of forewarning in the erotic conditions Proactive warning improves EIB | 135
compared to the negative conditions. All other 2-way interactions, and the 3-way interaction between factors, were non-significant (ps>.05).
Figure 22. Experiment 7 results. When participants were told a trial may be more difficult, performance generally improved, though not to the extent of forewarnings in Experiments 5 and 6. Error bars depict standard error.
Unlike Experiments 5 and 6, baseline trials appeared in both warning present trials and warning absent trials (to maintain the same difficulty in both trial types).
Thus, baseline performance in Experiment 7 was measured in both warning present
(M=89.2%, SD=7.8%) and warning absent (M=88.8%, SD=7.9%) conditions, and there was no significant difference between them, t(82) =0.648, p=.519, dz=0.100. (For simplicity, Figure 22 displays the averaged baseline performance across the entire experiment.)
At lag-2, warnings that the trials were more difficult did not improve performance after negative distractors (warning present: M=76.3%, SD=13.7%; warning absent:
M=75.2%, SD=11.8%, t(82) =0.896, p=.373, dz=0.098), but did help performance after erotic distractors (warning present: M=73.3%, SD=12.4%; warning absent: M=70.3%,
SD=12.1%, t(82) =2.356, p = .021, dz=0.269). At lag-4, the difficulty warning
“marginally” improved performance in the negative condition (warning present: 136 | Proactive warning improves EIB
M=87.2%, SD=9.3%; warning absent: M=85.5%, SD=9.7%, t(82) =1.896, p= .062, dz=0.208), but significantly improved in the erotic condition (warning present:
M=87.0%, SD=8.9%; warning absent: M=85.1%, SD=9.0%, t(82) =2.017, p=.047, dz=0.221; see Figure 22). Like in Experiments 5 and 6, there was no difference in performance in the forewarning conditions across blocks, or across mini-blocks. A 2
(Forewarning) X 2 (Lag) X 2 (Distractor Type) X 4 (Block) revealed no significant
2 interaction of block and forewarning, F(1,246) = 0.253, p = .859, ηp =0.003, and a 2
(Forewarning) X 2 (Lag) X 2 (Distractor Type) X 2 (Mini-Block Position) revealed no interaction between Mini-Block Position and Forewarning, F(1,82) = 2.035, p = .157,
2 ηp =0.024.
Performance in warning present compared to warning absent trials did not correlate with levels of depression or attentional control (ps>.05). There was, however, a significant correlation between harm avoidance in the negative, lag-2 condition and the difference between warning present and warning absent performance, r(83)=-.281, p=.010. Higher harm avoidance correlated with less benefit from forewarning in the negative, lag-2 condition, but this correlation was not observed in any other condition.
Surprisingly, in Experiment 7, participants did not very accurately report their own performance on the task generally via the self-report measure, r(83)=-.140, p=.207.
There were also no significant correlations between self-reported difficulty and warning present versus warning absent performance for either distractor type (ps>.05).
Comparison across Experiments.
The results of Experiment 7 generally demonstrated an improvement in the
“difficult” compared to “normal” trials. However, we further explored if this enhancement in Experiment 7 was comparable to the enhancement from forewarnings Proactive warning improves EIB | 137
in the previous two experiments. A 3 (Experiment: 5 vs 6 vs 7) X 2 (Forewarning:
Warning Present vs Warning Absent) X 2 (Lag: 2 vs 4) X 2 (Distractor Type: Negative vs Erotic) ANOVA revealed a significant interaction between all factors, F(2,244) =
2 11.386, p < .001, ηp = 0.085, suggesting that the pattern across experiments may not be the same. The Forewarning X Experiment interaction was also significant, F(2,244) =
2 36.187, p < .001, ηp = 0.229, suggesting that the benefit from forewarnings was not the same across experiments.
To compare performance across experiments at lag-2, we ran a Forewarning X
Distractor Type X Experiment ANOVA and found that the Forewarning X Experiment
2 interaction was significant, F(2,244) = 49.105, p < .001, ηp = .287. A post hoc
Fischer’s LSD test revealed significant mean differences between Experiment 7 and
Experiments 5 (Mean Difference=4.506, SE=1.563, p=.004) and 6 (Mean
Difference=5.119, SE=1.563, p=.001). Thus, the effect of forewarning in Experiment 7 was significantly different from the effect of forewarning in both Experiment 5 and
Experiment 6. In both cases, thinking the trial was difficult (Experiment 7) was not as beneficial to lag-2 performance as knowing about the distractor type (Experiment 5 and
6). However, there was no significant mean difference between Experiments 5 and 6
(Mean Difference=0.614, SE=1.539, p=.690), suggesting that there was no difference in the effect of forewarning between Experiments 5 and 6.
The Forewarning X Experiment interaction was not significantly different across
2 the three experiments at lag-4, F(2,244) = 2.217, p = .111, ηp = 0.018, likely because of the generally strong performance at lag-4 across experiments and condition types.
Discussion 138 | Proactive warning improves EIB
In Experiment 7, we found that participants performed better when they believed that the trial type was more difficult. While utilizing an effort-based proactive control, having information about the trial being more difficult did improve performance in the
EIB task, but the enhancement was not as effective in overcoming EIB as in
Experiments 5 and 6 - particularly at lag-2.
There was a correlation between harm avoidance and the benefit from forewarnings in the lag-2, negative condition. However, this was only for the negative, lag-2 condition.
General Discussion
Across three experiments, our data suggest that forewarnings about emotional distractors is effective in attenuating EIB. Participants were better able to overcome distraction by emotional stimuli in an EIB task when they had knowledge about the kind of distractor to expect; this was both when the description of the distractor emotionality was specific (Experiment 5) and more general (Experiment 6). In Experiment 7, participants also performed better on trials that they were simply told were more difficult, but this effort-based proactive control was significantly less effective than when participants knew that distractors would be emotional. These results suggest that people can tune attention against emotional distractors.
Interestingly, our results differ from a past EIB experiment that demonstrated that knowing emotionally distracting words did not improve performance in correctly identifying target words after them (Arnell et al., 2007). While at first glance these results conflict, the stimuli used in the present study are likely more emotionally powerful than the emotional words used in the previous study. This seems probable, particularly in light of the lack of EIB after negative words (whereas negative emotional Proactive warning improves EIB | 139
images produced a large degree of EIB in our study). The blocked design used in that earlier study may also be a reason for different results when compared to the more frequent recruitment of proactive control in the design that we employed.
We also did not find that the benefit of proactive control developed dynamically over the course of the experiment, unlike other recent attentional studies (Cunningham
& Egeth, 2015; Cunningham & Egeth, 2016). Instead, in our experiment, foreknowledge of the distractor emotionality aided performance from the first block and continued throughout the experiment. Moreover, the distractor similarly impacted performance if it appeared in either the early or late positions of the “mini-blocks.” It may be that we did not observe an effect over time because the differences in stimuli or task demands dictate different mechanisms or timing through which proactive control exerts its influence.
The individual difference measures that we collected did not reliably predict the relative effectiveness of proactive control. This was somewhat surprising, given previous findings of an inverse relationship between negative affect and proactive control (Fox et al., 2015; Vanderhasselt et al., 2014). Our sample, however, was not a clinical sample, and negative affect in our sample may not have been severe enough to yield observable differences. It is also notable that the task-demands in the current task are different from those used in the past: the short exposures to the emotional stimuli may also explain why individual differences did not appear in this study. While previous EIB results do show that individual differences in negative affect can predict differences in the pattern (Kennedy & Most, 2015; Most et al., 2005), they generally did not seem to change the impact of proactive control in this study. The absence of a correlation between trait attentional control and proactive control also contrasts with findings from a previous study (Peers & Lawrence, 2009). Generally, harm avoidance 140 | Proactive warning improves EIB
also did not reliably predict performance, unlike previous studies that have demonstrated a relationship (Most et al., 2006).
Participants were very accurate on self-report measures to report their overall performance accuracy in the experiment. In both Experiments 5 and 6, participants predicted their performance with great accuracy, suggesting that participants were aware of when they were distracted (likely informed by the feedback about correct responses). Interestingly, this was not the case in Experiment 7, perhaps because participants were unsure what made some trials more “difficult” than others – and therefore were unsure about their overall performance. Nevertheless, their self-report measures did not correlate with the difference between warning present and warning absent trials, suggesting that participants did not realize the benefit of forewarning on their performance.
The results from these experiments carry important implications for curtailing emotional distraction in everyday life. Whether warned about approaching a distracting car accident while driving, or warned about a fire-alarms before they sound, having fair warning about the presence of emotional distraction seems to help maintain attention on an ongoing task.
Proactive warning improves EIB | 141
References
Arnell, K. M., Killman, K. V, & Fijavz, D. (2007). Blinded by emotion: target misses
follow attention capture by arousing distractors in RSVP. Emotion, 7(3), 465–477.
http://doi.org/10.1167/4.8.359
Beck, A., Steer, R., & Brown, G. (1996). Manual for the Beck Depression Inventory-II.
San Antonio, TX: Psychological Corporation.
Brainard, D. H. (1997). The Psychophysics Toolbox. Spatial Vision, 10(4), 433–436.
http://doi.org/10.1163/156856897X00357
Braver, T. S. (2012). The variable nature of cognitive control: A dual-mechanisms
framework. Trends in Cognitive Sciences, 16(2), 106–113.
http://doi.org/10.1016/j.tics.2011.12.010.The
Braver, T. S., Gray, J. R., & Burgess, G. C. (2007). Explaining the many varieties of
working memory variation: Dual mechanisms of cognitive control. In Variation in
working memory. (pp. 76–106). http://doi.org/10.3758/s13423-011-0165-y
Cloninger, C. R., Przybeck, T. R., & Svrakic, D. M. (1991). The Tridimensional
Personality Questionnaire: U.S. normative data. Psychological Reports, 69, 1047–
1057. http://doi.org/10.2466/pr0.1991.69.3.1047
Cunningham, C., & Egeth, H. E. (2015). Loosening the Snare: Top-down goals
overcome singleton driven attentional capture. Journal of Vision, 15(12), 320.
http://doi.org/10.1167/15.12.320
Cunningham, C. & Egeth, H. E. (2016). Taming the white bear: Initial costs and
eventual benefits of distractor inhibition. Psychological Science.
http://doi.org/10.1177/0956797615626564 142 | Proactive warning improves EIB
Derryberry, D., & Reed, M. A. (2002). Anxiety-related attentional biases and their
regulation by attentional control. Journal of Abnormal Psychology, 111(2), 225–
236. http://doi.org/10.1037/0021-843X.111.2.225
Folk, C. L., Leber, A. B., & Egeth, H. E. (2002). Made you blink! Contingent
attentional capture produces a spatial blink. Perception & Psychophysics, 64(5),
741–753. http://doi.org/10.3758/BF03194741
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), 847–58.
http://doi.org/10.1037/0096-1523.24.3.847
Fox, E., Dutton, K., Yates, A., Georgiou, G. A., & Mouchlianitis, E. (2015). Attentional
control and suppressing negative thought intrusions in pathological worry.
Clinical Psychological Science, 3(4), 593–606.
http://doi.org/10.1177/2167702615575878
Kennedy, B. L., & Most, S. B. (2015). The Rapid Perceptual Impact of Emotional
Distractors. Plos One, 10(6), e0129320.
http://doi.org/10.1371/journal.pone.0129320
Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (2008). International affective picture
system (IAPS): Affective ratings of pictures and instruction manual. Technical
Report A-8. Retrieved from ftp://dhcp-129-105-171-
164.psych.northwestern.edu/OpenShare/ESPN/IAPS 1-16/IAPS 1-
16/IAPSmanual.pdf
Leber, A. B., & Egeth, H. E. (2006). It’s under control: Top-down search strategies can
override attentional capture. Psychonomic Bulletin & Review, 13(1), 132–8.
http://doi.org/10.3758/BF03193824 Proactive warning improves EIB | 143
Most, S. B., Chun, M. M., Johnson, M. R., & Kiehl, K. a. (2006). Attentional
modulation of the amygdala varies with personality. NeuroImage, 31(2), 934–44.
http://doi.org/10.1016/j.neuroimage.2005.12.031
Most, S. B., Chun, M. M., Widders, D. M., & Zald, D. H. (2005). Attentional
rubbernecking: cognitive control and personality in emotion-induced blindness.
Psychonomic Bulletin & Review, 12(4), 654–61.
http://doi.org/10.3758/BF03196754
Most, S. B., Smith, S. D., Cooter, A. B., Levy, B. N., & Zald, D. H. (2007). The naked
truth: Positive, arousing distractors impair rapid target perception. Cognition &
Emotion, 21(5), 964–981. http://doi.org/10.1080/02699930600959340
Öhman, A., Flykt, A., & Esteves, F. (2001). Emotion drives attention : Detecting the
snake in the grass. Emotion, 130(3), 466–478. http://doi.org/10.1037/AXJ96-
3445.130.3.466
Peers, P. V, & Lawrence, A. D. (2009). Attentional control of emotional distraction in
rapid serial visual presentation. Emotion, 9(1), 140–145.
http://doi.org/10.1037/a0014507
Pelli, D. G. (1997). The VideoToolbox software for visual psychophysics: transforming
numbers into movies. Spatial Vision, 10(4), 437–442.
http://doi.org/10.1163/156856897X00366
Vanderhasselt, M.-A., De Raedt, R., De Paepe, A., Aarts, K., Otte, G., Van Dorpe, J., &
Pourtois, G. (2014). Abnormal proactive and reactive cognitive control during
conflict processing in major depression. Journal of Abnormal Psychology, 123(1),
68–80. http://doi.org/10.1037/a0035816 144 | Proactive warning improves EIB
West, R., Choi, P., & Travers, S. (2010). The influence of negative affect on the neural
correlates of cognitive control. International Journal of Psychophysiology, 76(2),
107–117. http://doi.org/10.1016/j.ijpsycho.2010.03.002
Williams, J. M. G., Mathews, A., & MacLeod, C. (1996). The emotional Stroop task
and psychopathology. Psychological Bulletin, 120(1), 3–24.
http://doi.org/10.1037/0033-2909.120.1.3
General discussion | 145
General Discussion
Summary
Emotional distractors impair the awareness of a task-relevant stimulus presented in close temporal proximity – a phenomenon known as emotion-induced blindness. The present series of experiments aimed to 1) verify a spatiotemporal competition account of emotion-induced blindness and 2) assess whether people are able to use strategies to bias the competition more in favor of subsequent targets.
Emotion-induced blindness is thought to occur because of a representational competition between the emotional distractor and task-relevant target (Wang, Kennedy,
& Most, 2012). Evidence for this account comes from its spatially localized nature, such that the emotional distractors primarily impair awareness of targets that appear in their same location, but not of targets that appear in a different location (Most & Wang,
2011; Wang et al., 2012). The experiments in Chapter 1 (Experiments 1 and 2) confirmed the spatially localized pattern in emotion-induced blindness and ruled out eye-gaze as a potential explanation for the pattern of impairment. Regardless of where participants were looking when the distractor appeared, they were impaired at reporting targets that appeared in the same space as emotional distractors compared to neutral distractors, but no such difference was observed when the distractor and target appeared in different locations.
When the emotional distractors were made task-relevant, emotional impairment was instead observed across space, though the impairment was still greater when the distractor and target appeared in the same location (Experiment 2). The impact of emotional distractors on target perception likely reflects converging simultaneous mechanisms. Emotional salience likely makes the distractors overwhelm at a level of 146 | General discussion
“competitive interference” (which seems to be spatially specific), whereas task- relevance additionally engages encoding mechanisms that suppress attention across the visual field. This evidence suggests that the emotion-specific impairment in emotion- induced blindness and task-relevancy may be additive.
To examine if the role of context can shift the weight of competition toward or away from emotional distractors, categorical context in EIB was explored in Experiment
3. Targets appeared in RSVP streams that were either categorically heterogeneous
(images depicting different semantic categories) or categorically homogeneous (images depicting the same semantic category). The nature of the streams emphasized the categorical distinctiveness of distractors – such that they were more distinct in homogeneous streams than in heterogeneous streams. While neutral distractors only impaired target detection when they were more categorically distinct, emotional distractors impaired target detection in both categorically homogeneous and heterogeneous contexts. The oddball nature of the distractor’s semantic category therefore cannot explain the impact of emotional distraction, though may explain the impact of neutral distractors used in the typical emotion-induced blindness task.
Experiment 4 explored whether the competition between the distractor and targets can be modulated by whether participants consider the distractor to be relevant or irrelevant to their current task. This was done by comparing performance in emotion- induced blindness when the distractor was to be reported (task-relevant) to when the distractor was not reported (task-irrelevant). When distractors were task-relevant, emotional distractors impaired target accuracy similarly to task-irrelevant distractors at lag-2, but impaired accuracy for a longer duration (through lag-4) than when they were task-irrelevant. This suggests that if people consider an emotional stimulus as relevant General discussion | 147
or irrelevant to their task, relevance does not modulate the attention grabbing power of emotional distractors, though it does modulate how long the distractors hold attention.
The competition was also modulated when participants were told the kind of emotional distractor to expect in a given trial, and participants were generally better able to overcome distractors and better report the subsequent target (Chapter 4). The benefit from being told the kind of distractor to expect was found both when participants were told about the specific to the type of emotional distractor to expect (i.e., “gruesome” or
“erotic”; Experiment 5) and when the description was more general (i.e., “graphic”;
Experiment 6). Participants also performed better in trials when they believed they were more difficult than other trials (Experiment 7), but not to the same degree as when they had foreknowledge about the emotional content of the distractor. Thus, the opportunity to prepare and brace oneself against an emotional distractor helped participants override their distracting power, above and beyond their extra effort put into those trials.
Altogether, the present study supported a spatiotemporal competition account in emotion-induced blindness, and demonstrated that the competition between emotional distractors and task-relevant targets can be modulated in an emotion-induced blindness paradigm. This is suggestive of the mechanisms involved in emotion-induced blindness, and also identifies ways that emotional distraction may be overcome.
Understanding emotion-induced blindness in the context of other attentional tasks
Emotion-induced blindness (EIB) is a phenomenon that sits at the intersection several fields, and is thus related to many different emotion and/or attentional tasks.
Here I compare results from these EIB studies to the two types of tasks which seem the 148 | General discussion
most comparable - the attentional blink and spatial emotion-attention tasks – before making the case for EIB as a paradigm that should be considered for more research endeavors.
EIB and the attentional blink. In the attentional blink, participants have difficulty reporting the second of two targets when they are presented close in time (e.g., Chun &
Potter, 1995). Traditional attentional blink theories suggest that it reflects central interference or a bottleneck in relatively late processing stages (Chun & Potter, 1995; Di
Lollo, Kawahara, Ghorashi, & Enns, 2005; Shapiro, Raymond, & Arnell, 1994), such as consolidation of targets into visual working memory (Chun & Potter, 1995), disruption to a filter that discriminates targets from non-targets (Di Lollo et al., 2005), or retrieval of targets from memory (Shapiro et al., 1994).
As mentioned previously, the spatially specific impairment in emotion-induced blindness suggests that the impairment occurs at a stage earlier than those traditionally posited for the attentional blink (Wang et al., 2012). This is because a central bottleneck would predict impairment throughout space – such that both targets are perceptually encoded, but are not consolidated into memory. Indeed, consistent with this differentiation, the attentional blink typically disrupts detection of the second target regardless of the spatial relationship with the first target (Lunau & Olivers, 2010; Shih,
2000; but see Kristjansson & Nakayama, 2002).
The results of the present study are consistent with the theory that emotion- induced blindness occurs because of a biased competition between the emotional distractor and subsequent items (Wang et al., 2012). Despite a plausible alternative account that spatially specific effect observed in EIB was a result of participants only General discussion | 149
attending to one stream at a time, eye-movements were not able to account for the spatial specific impairment (Chapter 1).
Interestingly, recent theories of the attentional blink deviate somewhat from more traditional theories. Rather than a central bottleneck, these theories suggest that the attentional blink results from a representational competition between the two targets
(Akyürek & Hommel, 2005; Akyürek et al., 2012; Hommel & Akyürek, 2005; Potter,
Staub, & Connor, 2002; Wyble, Bowman, & Nieuwenstein, 2009). In this framework, the two targets compete for representation when they are processed close together and in the same “episode” (Akyürek et al., 2012; Potter et al., 2002; Wyble et al., 2009).
Evidence for a competition account comes from findings regarding lag-1 sparing. Lag-
1 sparing is a common pattern observed in the attentional blink, in which accuracy in reporting the second target in the attentional blink is relatively unimpaired if it appears immediately after the first target (Chun & Potter, 1995; Dell’Acqua, Sessa, Jolicœur, &
Robitaille, 2006; Di Lollo et al., 2005; Olivers, van der Stigchel, & Hulleman, 2007).
Participants sometimes reverse the order in reporting the two targets (Akyürek &
Hommel, 2005), or integrate the two targets into a single representation (Akyürek &
Hommel, 2005; Akyürek et al., 2012; Hommel & Akyürek, 2005). This evidence suggests a representational competition between the two targets, since participants seem to be integrating the two items together into a single representation. Thus, these theories suggest that in the attentional blink, the target representations are grouped together when presented close enough in time, and that these representations compete for access.
While the spatially specific impairment in emotion-induced blindness suggests different underlying mechanisms than the attentional blink, the possibility of competition in both phenomena is an intriguing possibility. It could be that the 150 | General discussion
competition between distractor and target in emotion-induced blindness occurs at a different processing level than that for the attentional blink, such as a competition in perceptual processes rather than working memory processes. Alternatively, both could be competing for the same attentional resources, but at different rates or amounts.
Indeed, a recent model considers the attentional blink to involve a series of sources of interference (Wyble & Swan, 2015), so EIB may share some of these sources
(particularly related to early interference from competition) but not others.
Is emotion-induced blindness entirely separable from the attentional blink?
Probably not. They certainly share phenomenal qualities and even neural signatures
(Kennedy, Rawding, Most, & Hoffman, 2014). However, there are persistent and sizable patterns that distinguish them, which suggest that aspects of their mechanisms work differently. Future research should continue to explore how the mechanisms of these two phenomena compare, particularly in the framework of representational competition.
EIB and spatial emotion-attention tasks. Like emotion-induced blindness, spatial attention tasks have also been used to investigate mechanisms involved in emotional distraction. In the spatial attention literature, the processing step of attentional disengagement has been commonly implicated in emotional distraction (Fox, Russo,
Bowles, & Dutton, 2001; Notebaert, Crombez, Damme, De Houwer, & Theeuwes,
2011; Vromen, Lipp, & Remington, 2014). Indeed, some evidence suggests that emotional stimuli do not capture attention any more than other stimuli in the environment, but are instead more difficult to disengage attention from (Fox et al.,
2001). General discussion | 151
In the current study, disengagement was modulated based on the participants’ strategies. When participants regarded emotional distractors as relevant to a task, they seemed to have more difficulty disengaging from them (Experiment 4), whereas knowing the kind of distractor to expect generally made it easier (Experiment 3).
Participants were also more likely to continue looking at the location they were attending after emotional distractors were presented (Chapter 1), all suggesting emotional distractors impair the ability to disengage attention.
The findings in this thesis support the notion that emotional distractors are difficult to disengage from. However, attentional disengagement cannot fully explain the emotion-induced blindness phenomenon. This is evidenced not only by the spatially localized impairment, but also from previous findings that demonstrate impairment in target accuracy when targets appear just before an emotional distractor (at lag-minus-1;
Most & Jungé, 2008). A disengagement model would not predict these findings, but a competition model does. Whether presented before or after a distractor, targets seem to compete with the representations of the highly salient distractor. Across the context of different emotion-attention tasks, the distinction in contributing underlying mechanisms should be continually examined in order to gain complete understanding of how emotion and attention interact.
At first glance, the spatial localization of emotion-induced blindness runs somewhat counter to findings from the dot-probe task - a popular spatial attention task that examines emotional distraction. In emotion-induced blindness, emotional distractors impair the perception of items that appear in their same location, whereas in the dot-probe task, emotional distractors benefit perception of stimuli that appear in their same location (Fox et al., 2001; MacLeod et al., 1986; Mogg & Bradley, 1998). A key finding to distinguish these differences is that the spatial pattern of EIB changes 152 | General discussion
when no additional items appear in a stream after the target. In this case, the targets remain unmasked, as they typically do in dot probe experiments, and participants’ target accuracy benefits at the location of the emotional distractors (Most & Wang, 2011).
This may be because the absence of subsequent, masking stream items alleviates ambiguity regarding the temporal order of distractors and targets.
Emotion-induced blindness is likely representative of multiple levels of attentional processing. When left unmasked, target accuracy benefits from emotional distractors in the same space, suggesting that the benefits of spatial attention can be observed using an adapted version of the emotion-induced blindness task. However, in the typical design, the spatiotemporal competition of distractors likely renders any effect from spatial attention obsolete because the representation for the target is disrupted earlier than this stage. This is particularly important to consider when comparing results from different experimental paradigms studying emotion and attention.
The role of EIB in future research.
Recently, Firestone and Scholl (in press) published an impassioned paper refuting evidence for top-down influences on perception. In their article, they outlined a six- point checklist to avoid “pitfalls” present in current research on the topic, and suggested that all items on the checklist be addressed before considering a method to validly address top-down influences. Their skepticism considered, EIB seems to uphold this checklist and warrant consideration as such evidence.
The first “pitfall” is that many of the methods in question are not designed to be disconfirmatory. In this pitfall, studies are often unable to both observe an effect when it is present, and not observe an effect when the theory would predict it to not be General discussion | 153
present. For example, Firestone and Scholl cite the finding that participants rate the darkness of a room to be darker when they are thinking about immoral concepts
(Banerjee, Chatterjee, & Sinha, 2012). However, Firestone and Scholl found that in the same situation, participants also choose a darker swatch to represent the darkness of the room they are in, which should not be the case (as the same swatch in a dark room would also appear darker; Firestone & Scholl, 2014). This is a pitfall that is particularly applicable to studies of perceptual judgement. The format that one provides an answer should be representative of one’s perceptual experience. In the case of EIB, participants are not trying to match judgement with perception, as the dependent variable is usually accuracy in reporting the target rotation (or in some cases, identification of the target).
Thus, EIB is not readily susceptible to this pitfall, because it is not based on perceptual judgement, but rather the awareness of a target.
The second pitfall is when post-perceptual judgment is confused for perception.
This refers to situations when participants infer a property of a stimulus rather than report what they actually see. For example, participants may see a stimulus to be a certain size, but judge its size to be bigger or smaller. In EIB, participants are not making judgements about the emotional distractors or targets, they are simply reporting the direction they saw the target to be rotated. The spatially localized effect in EIB also suggests that it is involved in relatively early processing stages, such that distractors and targets seem to compete for representation (Wang et al., 2012). This evidence suggests that post-perceptual judgment is not driving the EIB effect.
The third pitfall is that task demands and response bias can make the participant respond in a way they believe they should. This is likely not the case in EIB, however, particularly demonstrated in the two-stream versions of EIB (Most & Wang, 2011;
Chapter 1). If participants were simply responding worse to targets after negative 154 | General discussion
distractors because they thought they should, the emotion-specific impairment should resonate across space, rather than be specific to the location of the distractor.
The fourth pitfall is that low-level differences in the experimental stimuli drive differences that are otherwise interpreted. For example, stimuli used to evoke an emotional response may do so because of low-level properties like color and luminance
(i.e., emotional images may not impair because they are emotional, but because they are more red, since they often depict blood). In EIB, performance in emotional trials is compared to performance after other distractor types. This effect still holds when comparing emotional distractors with scrambled versions of themselves, suggesting low-level properties do not drive the effect (Most, Chun, Widders, & Zald, 2005).
Moreover, EIB has been demonstrated with emotionally neutral stimuli that have been conditioned to be emotionally power (Smith, Most, Newsome, & Zald, 2006). As such, the result does not seem to be dependent on low-level properties of the stimuli.
The fifth pitfall is that peripheral attentional effects are confused as being reflective of perception. However, peripheral attention does not seem to drive EIB, evidenced in its spatial specific impairment. Regardless of where participants were looking when a distractor appeared, there was no difference in the spatially localized impairment from emotional distractors (Chapter 1). Thus, peripheral attentional effects do not seem to explain the patterns observed in EIB. Note that this is a subtle point, and
Firestone and Scholl remain ambiguous about what they mean by “peripheral”. EIB surely implicates attention, but attention operates throughout the process of building visual percepts. Evidence from the current studies suggest that EIB involves attention at core points in building a conscious representation, not at peripheral points that simply select what gets input into visual computations. General discussion | 155
The sixth pitfall is when a measure tests memory instead of perception, but where the investigators interpret the results as reflective of perception. EIB occurs even when participants have to report the target as soon as they see it, making memory an unlikely mechanism responsible for EIB (Kennedy & Most, 2012). Its spatially localized impairment also suggests that memory cannot explain the results (Most & Wang, 2011;
Chapter 1), as memory interference would predict impairment across space.
Taken together, EIB seems to occur at a perceptual level, and thus is may constitute at least one phenomenon that truly illustrates an impact of emotion on perception. Depending on how one splits hairs, definition of “perception” may be more strictly defined to exclude these results of EIB, but as it stands, EIB does not seem to fall within the six pitfalls outlined above.
The emotion-attention field currently seems largely focused on spatial-attention tasks like the dot-probe to study emotional distraction. The consequences of this are amplified in the context of Attentional Bias Modification (ABM) – a modification of the dot-probe – which is a widely employed tool to help people learn to overcome distraction by emotional stimuli (MacLeod & Mathews, 2012). In ABM, people are trained to look away from emotional stimuli by always responding to a target that appears in a location opposite to an emotional probe, and ABM is a technique being applied for clinical purposes. However, the mechanisms that ABM taps into may not be core to the drivers of clinical disorders. In a recent study, Onie and Most (in preparation) have demonstrated that both the dot probe and EIB predict negative affect
(an aggregate score combining self-reported anxiety, depression, and stress) but do not predict each other, indicating that their relationship to negative affect is via different routes. That study also found that EIB predicted persistent negative thought
(rumination and worry), whereas the dot-probe did not. Thus, it seems that not all 156 | General discussion
emotion-attention interactions are tapping into the same underlying mechanisms, which is important to consider both to understand the way they operate, and also when developing techniques to help people overcome emotional distraction.
In short, emotion-induced blindness appears to reflect a mechanism of attention- emotion interaction that is unique within the field, with implications for developing a richer understanding of individual differences in emotional information processing. The current EIB findings point the way towards several lines of future research, which I outline next.
Directions for future research
What’s “emotional” about emotion-induced blindness? Throughout this study, emotion-induced blindness was mostly induced using highly arousing, negative images
(and, in Chapter 4, highly arousing, erotic images, which are typically rated positively by both men and women; Bradley, Codispoti, Cuthbert, & Lang, 2001). This was usually compared with performance following emotionally “neutral” stimuli, which were also perceptual and/or categorical oddballs compared to the other items in the
RSVP streams. Based on dimensional accounts of emotion – where emotions can be understood along the orthogonal dimensions of arousal and valence – arousal is likely involved in producing the effect. This is demonstrated by the performance decrements after both high arousal negative and positive images (Most et al., 2007).
It is important to note that the literature contains challenges to dimensional approaches to emotion. There is an ongoing debate in the emotional literature about how to characterize emotions, existing either along dimensional space (Izard, 2007) or as discrete entities (Barrett, 1998). While arousal is a likely driver, the literature does give reason to consider different discrete emotions carefully. Studies demonstrate that General discussion | 157
different emotions can have distinct consequences on emotion-attention tasks - above and beyond the influence of arousal and valence dimensions. For example, disgusting compared to fearful images (Cisler, Olatunji, Lohr, & Williams, 2009; van Hooff,
Devue, Vieweg, & Theeuwes, 2013), and threatening compared to merely negative images (Kveraga et al., 2015), seem to produce different patterns in attentional tasks, particularly early in processing.
One emotion-induced blindness study has examined performance after fearful compared to disgusting images but found that both distractor types impaired performance similarly (Ciesielski, Armstrong, Zald, & Olatunji, 2010). Ciesielski and colleagues also reported the valence and arousal ratings provided by an independent group of participants and found that even though disgust images were rated more negatively valenced and more arousing than fearful images, they impaired performance similarly. However, this finding may not tell the whole story, as the subjective ratings
(especially from another group of participants) may not yield the most appropriate metric for comparing images of different discrete emotions in emotion-induced blindness. Moreover, the stimuli chosen would be particularly important, with dimensions of emotion that are very similar (indeed, from personal experience collecting stimuli, disgusting images are often also fearful). Viewing emotional images at a fast rate in a rapid sequence of images may also impact the emotionality of images.
To address these concerns, our lab has recently collected ratings for images presented at fast compared to slow rates, with the images representing a wide range of emotional dimensions. The analysis is ongoing and lies outside the scope of this dissertation, but will further examine differences in stimulus types and emotional categories to better understand their influence in rapidly presented tasks like emotion-induced blindness. 158 | General discussion
As it stands, emotionally arousing (whether positive or negative) stimuli do seem to elicit an emotion-induced blindness effect.
The stimuli used in emotion-induced blindness tasks typically depict emotional content themselves. Emotion-induced blindness has also been observed for neutral distractors that were conditioned to be aversive (Smith et al., 2006). The mechanisms involved, however, may differ when stimuli gain their “emotionality” based on extrinsic association, rather than their intrinsic properties. For example, in a study using neutral distractors previously paired with shock, emotion-induced blindness was observed, but did not produce a spatially localized effect (Le, Most, Kennedy, & White, in preparation). Thus, stimuli with learned associations with a negative outcomes do not appear to engage entirely overlapping mechanisms with intrinsically emotional stimuli, at least in a short experimental session.
Another open question in this literature is how stimuli that have been associated with reward – but which do not depict emotional content themselves - are regarded in an emotion-induced blindness task. Anderson and others have demonstrated that task- irrelevant stimuli can hold attention and distract from an ongoing task when they were previously linked to reward (Anderson, Laurent, & Yantis, 2011; Anderson, 2013,
2015; Le Pelley, Pearson, Griffiths, & Beesley, 2015; Müller, Rothermund, & Wentura,
2015). Similarly, in a dot-probe task, participants had difficulty disengaging from stimuli previously linked with reward or loss (Müller et al., 2015). These results suggest that rewarding or loss-affiliated stimuli may influence attention in ways similar to emotional stimuli.6
6 Note that this is a question is currently being explored in collaboration with the Mike LePelley lab at UNSW Australia. General discussion | 159
The paradigms in the reward literature also provide a way to discover how stimuli become meaningful to the point that they distract attention. Many emotional stimuli
(e.g., guns, knives) are emotionally evocative because of their learned meaning. How does this association happen? What makes them persist? And once learned, can we deprioritize them again?
Many questions remain, and further research should continue to explore the aspect of “emotion” that is driving emotion-induced blindness.
The impact of emotional “oddballs.” In Experiment 3, emotional distractors impaired performance regardless of their categorical distinctiveness from other items in the stream. That is, the context in which distractors appeared did not change the impact of emotional distractors, even though it did change the impact of neutral distractors. In some ways, this was surprising. Previous research has demonstrated that the behavioral and neural responses stemming from detection of categorical oddballs are similar to those in response to emotionally powerful stimuli. For example,β–adrenergic receptors are commonly thought to be involved in emotional processing (e.g.,
McGaugh, 2006), and research demonstrates that these β–adrenergic receptors are also involved in the processing of categorical oddballs (Strange & Dolan, 2001, 2007). This evidence suggests that at least some of the mechanisms involved in processing oddballs are similar to those observed in emotional processing.
The present thesis suggests that emotional distractors impair perception for nearby targets for reasons beyond their categorical difference. That said, in the design of typical emotion-induced blindness tasks, the emotional distractors are still “oddballs” compared to other stimuli in the stream, in the sense that they are the only emotional item in the stream. This aspect of the task was not directly tested and remains an 160 | General discussion
interesting question for future work. What should one predict, for example, if other filler items in the RSVP stream are also emotional, rendering the critical distractor a non-oddball: will this attenuate the impact of an emotional distractor? And will an emotional context – with emotional filler items dominating the stream - increase the impact of a non-emotional oddball distractor? While categorically oddball emotional distractors do not impair target detection any more than categorically similar emotional distractors, the emotional nature of distractors as an oddball property is a direction that future research should explore.
The memorability of emotional distractors in emotion-induced blindness.
Surprisingly, in Experiment 4, we found that when distractors were made relevant to the task, emotional distractors were remembered worse than non-emotional distractors.
Distractors were rendered task-relevant by asking participants to try to remember them, and it was surprising to find memory performance to be so poor. This finding runs contrary to a rich literature demonstrating that emotional stimuli tend to be remembered better than non-emotional stimuli (see Dolan, 2002; Kensinger, 2007; McGaugh, 2006;
Wallace, 1965 for reviews). It also runs contrary to previous attentional blink tasks that show emotional T1s to be remembered either better (Ihssen & Keil, 2009) or the same
(Schwabe & Wolf, 2010; Stein, Zwickel, Ritter, Kitzmantel, & Schneider, 2009) as neutral T1s.
The memory results here could provide exciting information about the stages in visual processing impacted by emotional distractors in EIB. It is also consistent with the idea that stimuli can be attended to but not encoded into memory. In a recent study, task-relevant attributes of stimuli, which were necessarily attended to and used to complete a task, were not remembered when tested (Chen & Wyble, 2015). This suggests that attributes of a stimulus can be explicitly attended and used to complete a General discussion | 161
task, but still not stored in memory. In our experiment, the task itself was to remember the stimulus, and thus this differs from the study by Chen and Wyble (in which remembering the particular attribute was not part of the task). However, it does provide converging evidence that attended stimuli are not necessarily encoded into memory.
Our results may also be influenced by specifics of our task design or the stimuli we selected in the experiment. Most attentional blink tasks that find strong memory for emotional T1s use stimuli that are words (Ihssen & Keil, 2009; Schwabe & Wolf, 2010) or faces (Stein et al., 2009), which are arguably less emotionally powerful and easier to attach a semantic label to than the graphic images used throughout this study (Note that along the same lines, the ability to attach a semantic label generally benefited performance in Chapter 2). Indeed, in an emotion-induced blindness task using words, target accuracy was predicted by the memorability of the distractors (Arnell, Killman, &
Fijavz, 2007). Additionally, the design of the study in Experiment 4, with the distractor task-relevant for only half of the experiment, may also, in a small way, deprioritize the importance to encode it into memory, but enough to remember it above chance. Any of these factors may be informative for the difference between remembering the emotional item with great accuracy, and having worse memory for them than neutral items.
It is worth noting that memory for distractors in Experiment 2 was better for negative distractors than “featural” distractors, but this comparison is substantially different than that of Experiment 4: featural distractors were landscape images, the same content as the filler items (making them arguably hard to differentiate), and they were not tested until the end of each block (instead of at the end of every trial).
In sum, the memory results for the first target in Experiment 4 were surprising and curious but should be taken with appropriate skepticism. The stimuli that we used were not normed on their memorability. More research should aim to investigate 162 | General discussion
memory for these items, particularly because of their strong competitive nature, and the possible distinction between their memorability and their prioritization by the visual system. Is it the case that the types of distractors we used in these experiments can impair the perception of nearby targets without being well-remembered? At this point, without better control of the stimulus properties and test format, this remains an empirical question.
Using attentional control to overcome distraction. The ability to use attentional control to overcome distraction has been observed in a variety of different attentional tasks (e.g., Cunningham & Egeth, 2015; Folk & Remington, 1998; Folk, Leber, &
Egeth, 2002; Leber & Egeth, 2006; Peers & Lawrence, 2009). Braver and colleagues postulate that attentional control works in a dual network, either proactively to prioritize task-goal information ahead of distraction, or reactively to recover from distraction that has already happened (Braver, Gray, & Burgess, 2007; Braver, 2012).
In Chapter 4, we demonstrated that proactive control can modulate the impact of emotional distractors. Participants were better able to identify their target when they knew what kind of distractor to expect. This was both when they knew the type of emotional stimulus to expect (i.e., gruesome or erotic) and when they were told only that the distractors would be “graphic” image. What’s more, the benefit from these warnings extended beyond participant’s effort on what they believed to be “harder” trials. Overall, we found that people are able to exercise proactive control of attention in order to deprioritize distractors and benefit perception of subsequent targets. Such findings complement previous emotion-induced blindness work that had demonstrated that participants could reduce EIB when they had concrete information about what their target would look like (Most, Chun, Johnson, & Kiehl, 2006; Most et al., 2005). General discussion | 163
The current findings stand in contrast to a study that found no modulation from warning about the emotionality of distractors (Arnell et al., 2007). The stimuli and design differences of the two studies may be the reason for differing results. Arnell and colleagues used words as their emotional stimuli, which are likely less emotionally powerful than the type of scene images that we used here. They also used long blocks in their design, rather than more frequent changes between trial types, which may have rendered attentional control mechanisms less effective or less likely to be used.
Our results provide a platform for future work related to the ways in which participants might proactively bias the competition against emotional distractors, and they can provide insight into the mechanisms by which advance knowledge helps people overcome emotional distraction. For example, would participants benefit in their baseline performance just by thinking an emotional image would occur, even if it never appears? That is, does the enhancement from proactive knowledge stem from biasing attentional competition away from the distractor, or from biasing toward the target?
Similarly, would being told a positive image would occur help overcome a negative one? In both scenarios the distractor is emotionally powerful, but is the attentional set to the distractor specific to the type of emotion? As it stands, being prepared for an emotional stimulus seems to benefit individuals in performing their task, and ties in with a literature that demonstrates the power of attentional control.
The influence of individual differences. Individual differences were generally beyond the scope of this thesis, but they are important to consider, particularly if an overarching goal is to help individuals with a maladaptive propensity to prioritize emotional information. We only used measures of individual differences in Experiment 5, as previous attentional control literature particularly predicts differences in the ability to use proactive control in individuals with different levels of depression (Fox, Dutton, 164 | General discussion
Yates, Georgiou, & Mouchlianitis, 2015; Vanderhasselt et al., 2014), harm-avoidance
(Most et al., 2006), and trait attentional control (Peers & Lawrence, 2009). While we did observe modest individual differences in depression in one experiment (Experiment
5), the current thesis does not provide stable, consistent findings regarding individual differences in emotion-induced blindness. This is despite previous research which has demonstrated individual differences to play a key role in emotion-attention interactions.
Anxiety is often found to modulate attention to negative stimuli (Bar-Haim,
Lamy, Pergamin, & Ijzendoorn, 2007; Ferneyhough, Kim, Phelps, & Carrasco, 2013;
Fox, Russo, & Georgiou, 2007). In a meta-analysis of studies examining attentional bias to threat, effects were particularly pronounced in individuals with anxiety, but not often observed in individuals with low anxiety (Bar-Haim et al., 2007, but see
Bockstaele et al., 2014). Similarly, depressed individuals have also shown different patterns in attention (Moran, Mehta, & Kring, 2012; Rokke, Arnell, Koch, & Andrews,
2002). For example, Rokke et al. found that the attentional blink was larger and longer for dysphoric participants (Rokke et al., 2002).
One possible reason why we did not observe individual differences when we tested for them may be that we tested sub-clinical samples. Additionally, the task demands of our study may not have been conducive to observing differences, showing stimuli for only a short amount of time. The very intense nature of the distractors used in this study may have also washed out any individual differences from proactive attentional control, whereas less intense emotional distractors may have allowed for individual differences to emerge. Nevertheless, previous emotion-induced blindness studies have demonstrated individual differences to play a role in EIB (Kennedy &
Most, 2015; Most et al., 2006), and future research should continue to explore General discussion | 165
individual differences in the framework of emotional distraction and bias toward prioritizing such distracting stimuli.
Understanding the neural underpinnings in emotion-induced blindness. As mentioned above, the spatially specific pattern (Chapter 1) suggests a relatively early disruption in emotion-induced blindness, possibly via perceptual competition. To identify the mechanisms underlying emotion-induced blindness, it needs to accommodate for “top-down” influences, evidenced in this thesis by the improved performance when participants were informed about the type of distractor on any given trial (Chapter 4).
In an event-related potentials (ERP) study, participants completed a single-stream, emotion-induced blindness task (Kennedy et al., 2014). Similar to the attentional blink, there was a trade-off in the amplitudes of N2 and P3b components to the distractor and target (Sergent, Baillet, & Dehaene, 2005; Shapiro, Schmitz, Martens, Hommel, &
Schnitzler, 2006). The N2 is associated with attentional selection (Sergent et al., 2005), while the P3b is a component attributed to consolidation of information into working memory (Vogel & Luck, 2002). In the EIB study, a posterior positivity - likely suggestive of attentional suppression (Kiss, Grubert, Petersen, & Eimer, 2012; Sawaki,
Geng, & Luck, 2012; Sawaki & Luck, 2010) - was also observed soon after emotionally powerful distractors. Together, these results suggest that competition between emotional distractors and target representations may span both early selection stages and later memory consolidation stages. EIB may additionally stem from attempts to suppress emotional distractors, with that suppression carrying over and hurting target perception. 166 | General discussion
If emotional distractors compete for representation with nearby targets, how is such competition instantiated at the neural level? Biased competition theories have traditionally focused on the competition of neural responses between attended visual stimuli (Desimone & Duncan, 1995). For example, neural responses are greater for items when presented alone than when presented with other stimuli in an array (Beck &
Kastner, 2005; Scalf, Torralbo, Tapia, & Beck, 2013), and this competition can be observed in visual areas, sometimes as early as V1, with the competition gradually increasing from V2-V4, and eventually to higher levels of cortex (Martinez-Trujillo &
Treue, 2004; Reynolds & Heeger, 2009; Treue & Martinez-Trujillo, 1999).
Some have attributed the results of RSVP studies to be a result of biased competition (Keysers & Perrett, 2002). Keysers and Perrett suggest that items in an
RSVP occur in the same space and so close in time that their neural responses overlap and create a competition. Applied to emotion-induced blindness, responses to the emotional distractor and subsequent target compete for representation because of their spatial and temporal overlap. Note that, because EIB is presented in an RSVP, all items in the stream likely compete with each other, and biased competition is likely involved in identifying the target without the distractors’ influence. Emotional distractors disrupt above and beyond this biased competition, but in what way? The emotional aspects likely serve as a different source of bias that counters the top-down biasing mechanism.
Thus, competition is integral in our understanding of emotion-induced blindness.
But what might foster this competition from emotional stimuli? The amygdala has long been implicated in emotional processing, and seems to be central in networks of cognitive processing (e.g., McGaugh, 2004; Pessoa, 2015). The role of the amygdala remains an open question, hypothesized to enhance selective processing to particularly General discussion | 167
salient stimuli, including but not exclusive to emotional and rewarding stimuli
(Adolphs, 2009; Pessoa, 2015).
Previous evidence suggests the amygdala plays a role in emotion-induced blindness (Most et al., 2006). In an fMRI study, Most and colleagues found that there was generally greater activity in the amygdalae and rostral anterior cingulate cortex
(ACC) in trials with emotional distractors compared to neutral distractors in an emotion-induced blindness task. The rostral ACC has been implicated in the resolution of interference from emotional stimuli in cognitive processing (Bush, Luu, & Posner,
2000). Intriguingly, the role of the amygdala and rostral ACC in EIB were particularly pronounced in individuals with high harm-avoidance (a trait that involves worrying, being fearful of uncertain situations, shyness, and fatigue; Cloninger, Przybeck, &
Svrakic, 1991), and depended on what they knew about the target (Most et al., 2006). If participants did not have a particular attentional set to identify their target, individuals high in harm-avoidance demonstrated high activity in the amygdalae. If instead they were able to use a specific attentional set, these participants high in harm-avoidance had significantly greater activity in the rostral ACC. Those with low harm-avoidance showed no significant activation differences in either the amygdalae or rostral ACC.
Taken together, emotion-induced blindness seems to involve the amygdala, particularly in participants that score high in harm-avoidance and are prone to the power of emotional distraction (Most et al., 2006). Top-down knowledge about the target can attenuate this activation and recruit the rostral ACC to modulate the emotional processing. This finding provides insight into the mechanisms underlying emotion- induced blindness, and how top-down knowledge might shift the competitional bias in favor of the target instead of the distractor. While this research suggests that the amygdala plays a role in emotion-induced blindness, future research is required to 168 | General discussion
specifically pinpoint the neural mechanisms involved in emotional distraction, particularly on a trial-by-trial basis.
A recent account, known as the Glutamate Amplifies Noradrenergic Effects
(GANE) model, accommodates for the priority of emotional stimuli while also accounting for arousal’s impact on other neutral stimuli in the environment, and it provides another potential neural basis for the results. This model suggests that arousal elicits the release of glutamate, which then modulates norepinephrine levels via the locus coeruleus-norepinephrine system (Mather, Clewett, Sakaki, & Harley, in press).
Together, these modulations enhance the gain for representations that are “high priority” at the expense of “low priority” representations. “Priority” can be determined by a variety of factors, such as top-down (e.g., task relevance) or bottom-up (e.g., perceptual saliency) factors, but ultimately implies information that is most important to the visual system at the time of arousal (Mather & Sutherland, 2011). In this framework, arousal enhances representations for whatever is high priority at that time, and impairs representation for what is low priority, and the amygdala is important in coordinating this modulation throughout broader areas of the brain involved in cognition (see also
Pessoa & Adolphs, 2010). While highly speculative, when applied to emotion-induced blindness, this could translate to the prioritizing of emotional distractors (as high priority stimuli because of their salience), at the expense of other items in the stream
(including the subsequent target). Future research should continue to link the behavioral findings of emotion-induced blindness and the neuroscience involved.
Conclusions
In emotion-induced blindness, emotional distractors impair the ability to report subsequent targets that appear in the same location. This thesis aimed to verify a spatiotemporal competition account of emotion-induced blindness and to determine if General discussion | 169
strategies can bias the competition between emotional distractors and subsequent targets. The results demonstrated that the spatiotemporal competition account in emotion-induced blindness is a viable explanation for emotion-induced blindness and that categorical distinctiveness of distractors in their context cannot explain the emotion-specific impairment. In doing so, the present results further suggest that emotion-induced blindness reflects a relatively novel form of emotion-driven attentional bias, which has to date been the focus of only scant research. The impact from emotional distractors was modulated based on whether participants considered them task-relevant, and when participants knew to expect an emotional distractor. Together, these results support a spatiotemporal competition account of emotion-induced blindness and suggest that context and strategies can shift the balance of such competition.
170 | General discussion
References
Adolphs, R. (2009). Fear, faces, and the human amygdala. Current Opinion in
Neurobiology, 18(2), 166–172. http://doi.org/10.1016/j.conb.2008.06.006.Fear
Akyürek, E., Eshuis, S. A. H., Nieuwenstein, M. R., Saija, J. D., Başkent, D., &
Hommel, B. (2012). Temporal target integration underlies performance at lag 1 in
the attentional blink. Journal of Experimental Psychology: Human Perception and
Performance, 38(6), 1448–1464. http://doi.org/10.1037/a0027610
Akyürek, E., & Hommel, B. (2005). Target integration and the attentional blink. Acta
Psychologica, 119, 305–314. http://doi.org/10.1016/j.actpsy.2005.02.006
Anderson, B. A. (2013). A value-driven mechanism of attentional selection stimulus
value. Journal of Vision, 13(3), 1–16. http://doi.org/10.1167/13.3.7.doi
Anderson, B. A. (2015). The attention habit: How reward learning shapes attentional
selection. Annnals of the New York Academy of Sciences, 1–16.
http://doi.org/10.1111/nyas.12957
Anderson, B. A., Laurent, P. A., & Yantis, S. (2011). Value-driven attentional capture.
PNAS, 108(25). http://doi.org/10.1073/pnas.1104047108
Arnell, K. M., Killman, K. V, & Fijavz, D. (2007). Blinded by emotion: target misses
follow attention capture by arousing distractors in RSVP. Emotion, 7(3), 465–477.
http://doi.org/10.1167/4.8.359
Banerjee, P., Chatterjee, P., & Sinha, J. (2012). Is it light or dark? Recalling moral
behavior changes perception of brightness. Psychological Science, 23(4), 407–
409. http://doi.org/10.1177/0956797611432497
Bar-Haim, Y., Lamy, D., Pergamin, L., & van Ijzendoorn, M. H. (2007). Threat-related
attentional bias in anxious and nonanxious individuals: A meta-analytic study.
Psychological Bulletin, 133(1), 1–24. http://doi.org/10.1037/0033-2909.133.1.1 General discussion | 171
Barrett, L. F. (1998). Discrete Emotions or Dimensions? The Role of Valence Focus
and Arousal Focus. Cognition and Emotion, 12(4), 579–599.
Beck, D. M., & Kastner, S. (2005). Stimulus context modulates competition in human
extrastriate cortex. Nature Neuroscience, 8(8), 1110–1116.
http://doi.org/10.1038/nn1501
Bockstaele, B. Van, Verschuere, B., Tibboel, H., De Houwer, J., Crombez, G., &
Koster, E. H. W. (2014). A review of current evidence for the causal impact of
attentional bias on fear and anxiety. Psychological Bulletin, 140(3), 682–721.
http://doi.org/10.1037/a0034834
Bradley, M. M., Codispoti, M., Cuthbert, B. N., & Lang, P. J. (2001). Emotion and
motivation I: Defensive and appetitive reactions in picture processing. Emotion,
1(3), 276–498. http://doi.org/10.1037/1528-3542.1.3.276
Braver, T. S. (2012). The variable nature of cognitive control: A dual-mechanisms
framework. Trends in Cognitive Sciences, 16(2), 106–113.
http://doi.org/10.1016/j.tics.2011.12.010.The
Braver, T. S., Gray, J. R., & Burgess, G. C. (2007). Explaining the many varieties of
working memory variation: Dual mechanisms of cognitive control. In Variation in
working memory. (pp. 76–106). http://doi.org/10.3758/s13423-011-0165-y
Bush, G., Luu, P., & Posner, M. I. (2000). Cognitive and emotional influences in
anterior cingulate cortex. Trends in Cognitive Sciences, 4(6), 215–222.
Chen, H., & Wyble, B. (2015). Amnesia for object attributes: Failure to report attended
information that had just reached conscious awareness. Psychological Science,
26(2), 203–210. http://doi.org/10.1177/0956797614560648
Chun, M. M., & Potter, M. C. (1995). A two-stage model for multiple target detection in
rapid serial visual presentation. Journal of Experimental Psychology: Human 172 | General discussion
Perception and Performance, 21(1), 109–27. http://doi.org/10.1037/0096-
1523.21.1.109
Ciesielski, B. G., Armstrong, T., Zald, D. H., & Olatunji, B. O. (2010). Emotion
modulation of visual attention: Categorical and temporal characteristics. Plos
One, 5(11). http://doi.org/10.1371/journal.pone.0013860
Cisler, J. M., Olatunji, B. O., Lohr, J. M., & Williams, N. L. (2009). Attentional bias
differences between fear and disgust: Implications for the role of disgust in
disgust-related anxiety disorders. Cognition & Emotion, 23(4).
http://doi.org/10.1080/02699930802051599.Attentional
Cloninger, C. R., Przybeck, T. R., & Svrakic, D. M. (1991). The Tridimensional
Personality Questionnaire: U.S. normative data. Psychological Reports, 69, 1047–
1057. http://doi.org/10.2466/pr0.1991.69.3.1047
Dell’Acqua, R., Sessa, P., Jolicœur, P., & Robitaille, N. (2006). Spatial attention freezes
during the attention blink. Psychophysiology, 43(4), 394–400.
http://doi.org/10.1111/j.1469-8986.2006.00411.x
Desimone, R., & Duncan, J. (1995). Neural mechanisms of selective visual attention.
Annual Review of Neuroscience, 18, 193–222.
http://doi.org/10.1146/annurev.ne.18.030195.001205
Di Lollo, V., Kawahara, J., Ghorashi, S. M. S., & Enns, J. T. (2005). The attentional
blink: resource depletion or temporary loss of control? Psychological Research,
69(3), 191–200. http://doi.org/10.1007/s00426-004-0173-x
Dolan, R. J. (2002). Emotion, cognition, and behavior. Science, 298(5596), 1191–4.
http://doi.org/10.1126/science.1076358 General discussion | 173
Firestone, C., & Scholl, B. J. (in press). Cognition does not affect perception:
Evaluating the evidence for “top-down” effects. Behavioral and Brain Sciences.
http://doi.org/10.1017/S0140525X15000965
Firestone, C., & Scholl, B. J. (2014). “Top-down” effects where none should be found:
The El Greco fallacy in perception research. Psychological Science, 25(1), 38–46.
http://doi.org/10.1177/0956797613485092
Ferneyhough, E., Kim, M. K., Phelps, E. A., & Carrasco, M. (2013). Anxiety modulates
the effects of emotion and attention on early vision. Cognition & Emotion, 27(1),
166–176. http://doi.org/10.1080/02699931.2012.689953.Anxiety
Folk, C. L., Leber, A. B., & Egeth, H. E. (2002). Made you blink! Contingent
attentional capture produces a spatial blink. Perception & Psychophysics, 64(5),
741–753. http://doi.org/10.3758/BF03194741
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), 847–58.
http://doi.org/10.1037/0096-1523.24.3.847
Fox, E., Dutton, K., Yates, A., Georgiou, G. A., & Mouchlianitis, E. (2015). Attentional
control and suppressing negative thought intrusions in pathological worry.
Clinical Psychological Science, 3(4), 593–606.
http://doi.org/10.1177/2167702615575878
Fox, E., Russo, R., Bowles, R., & Dutton, K. (2001). Do threatening stimuli draw or
hold visual attention in subclinical anxiety? Journal of Experimental Psychology:
General, 130(4), 681–700. http://doi.org/10.1037/0096-3445.130.4.681 174 | General discussion
Fox, E., Russo, R., & Georgiou, G. A. (2007). Anxiety modulates the degree of
attentive resources required to process emotional faces. Cognitive, Affective, &
Behavioral Neuroscience, 5(4), 396–404. http://doi.org/10.3758/CABN.5.4.396
Hommel, B., & Akyürek, E. (2005). Lag-1 sparing in the attentional blink: Benefits and
costs of integrating two events into a single episode. The Quarterly Journal of
Experimental Psychology: Section A, 58(8), 1415–1433.
http://doi.org/10.1080/02724980443000647
Ihssen, N., & Keil, A. (2009). The costs and benefits of processing emotional stimuli
during rapid serial visual presentation. Cognition and Emotion, 23, 37–41.
http://doi.org/10.1080/02699930801987504
Izard, C. E. (2007). Basic emotions, natural kinds, emotion schemas, and a new
paradigm. Psychological Science, 2(3), 260–280. http://doi.org/10.1111/j.1745-
6916.2007.00044.x
Kennedy, B. L., & Most, S. B. (2012). Perceptual, not memorial, disruption underlies
emotion-induced blindness. Emotion, 12(2), 199–202.
http://doi.org/10.1037/a0026380
Kennedy, B. L., & Most, S. B. (2015). The rapid perceptual impact of emotional
distractors. Plos One, 10(6), e0129320.
http://doi.org/10.1371/journal.pone.0129320
Kennedy, B. L., Rawding, J., Most, S. B., & Hoffman, J. E. (2014). Emotion-induced
blindness reflects competition at early and late processing stages: An ERP study.
Cognitive, Affective & Behavioral Neuroscience, 14(4), 1485–1498.
http://doi.org/10.3758/s13415-014-0303-x General discussion | 175
Kensinger, E. A. (2007). Negative emotion enhances memory accuracy: Behavioral and
neuroimaging evidence. Current Directions in Psychological Science, 16(4), 213–
218. http://doi.org/10.1111/j.1467-8721.2007.00506.x
Keysers, C., & Perrett, D. I. (2002). Visual masking and RSVP reveal neural
competition. Trends in Cognitive Sciences, 6(3), 120–125.
http://doi.org/10.1016/S1364-6613(00)01852-0
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. http://doi.org/10.1162/jocn_a_00127
Kristjánsson, A., & Nakayama, K. (2002). The attentional blink in space and time.
Vision Research, 42, 2039–2050. http://doi.org/10.1016/S0042-6989(02)00129-3
Kveraga, K., Boshyan, J., Adams, R. B., Mote, J., Betz, N., Ward, N., … Barrett, L. F.
(2015). If it bleeds, it leads: separating threat from mere negativity. Social
Cognitive and Affective Neuroscience. http://doi.org/10.1093/scan/nsu007
Le, L., Most, S. B., Kennedy, B. L., & White, E. (in preparation). Localized conceptual
interference is driven by intrinsic, not extrinsic, emotional value: An emotion-
induced blindness study.
Le Pelley, M. E., Pearson, D., Griffiths, O., & Beesley, T. (2015). When goals conflict
with values: Counterproductive attentional and oculomotor capture by reward-
related stimuli. Journal of Experimental Psychology: General, 144(1), 158–171.
http://doi.org/10.1037/xge0000037
Leber, A. B., & Egeth, H. E. (2006). It’s under control: Top-down search strategies can
override attentional capture. Psychonomic Bulletin & Review, 13(1), 132–8.
http://doi.org/10.3758/BF03193824 176 | General discussion
Lunau, R., & Olivers, C. N. L. (2010). The attentional blink and lag 1 sparing are
nonspatial. Attention Perception Psychophysics, 72(2), 317–325.
http://doi.org/10.3758/APP.72.2.317
MacLeod, C., & Mathews, A. (2012). Cognitive bias modification approaches to
anxiety. Annual Review of Clinical Psychology, 8, 189–217.
http://doi.org/10.1146/annurev-clinpsy-032511-143052
Martinez-Trujillo, J. C., & Treue, S. (2004). Feature-based attention increases the
selectivity of population responses in primate visual cortex. Current Biology, 14,
744–751. http://doi.org/10.1016/j
Mather, M., Clewett, D., Sakaki, M., & Harley, C. W. (in press). Norepinephrine ignites
local hot spots of neuronal excitation: How arousal amplifies selectivity in
perception and memory. Behavioral and Brain Sciences.
http://doi.org/10.1017/S0140525X15000667
Mather, M., & Sutherland, M. R. (2011). Arousal-Biased Competition in perception and
memory. Perspectives on Psychological Science, 6(2), 114–133.
http://doi.org/10.1177/1745691611400234
McGaugh, J. L. (2004). The amygdala modulates the consolidation of memories of
emotionally arousing experiences. Annual Review of Neuroscience, 27, 1–28.
http://doi.org/10.1146/annurev.neuro.27.070203.144157
McGaugh, J. L. (2006). Make mild moments memorable: Add a little arousal. Trends in
Cognitive Sciences, 10(8), 345–7. http://doi.org/10.1016/j.tics.2006.06.001
Moran, E. K., Mehta, N., & Kring, A. M. (2012). Emotional responding in depression:
Distinctions in the time course of emotion. Cognition and Emotion, 26(7), 1153-
1175. http://doi.org/10.1080/02699931.2011.638909 General discussion | 177
Most, S. B., Chun, M. M., Johnson, M. R., & Kiehl, K. A. (2006). Attentional
modulation of the amygdala varies with personality. NeuroImage, 31(2), 934–44.
http://doi.org/10.1016/j.neuroimage.2005.12.031
Most, S. B., Chun, M. M., Widders, D. M., & Zald, D. H. (2005). Attentional
rubbernecking: cognitive control and personality in emotion-induced blindness.
Psychonomic Bulletin & Review, 12(4), 654–661.
http://doi.org/10.3758/BF03196754
Most, S. B., & Jungé, J. A. (2008). Don’t look back: Retroactive, dynamic costs and
benefits of emotional capture. Visual Cognition, 16(2), 262–278.
http://doi.org/10.1080/13506280701490062
Most, S. B., & Wang, L. (2011). Dissociating spatial attention and awareness in
emotion-induced blindness. Psychological Science, 22(3), 300–305.
http://doi.org/10.1177/0956797610397665
Müller, S., Rothermund, K., & Wentura, D. (2015). Relevance drives attention:
Attentional bias for gain- and loss-related stimuli is driven by delayed
disengagement. The Quarterly Journal of Experimental Psychology, 69(4), 752–
763. http://doi.org/10.1080/17470218.2015.1049624
Notebaert, L., Crombez, G., Damme, S. Van, De Houwer, J., & Theeuwes, J. (2011).
Signals of threat do not capture, but prioritize, attention: A conditioning approach.
Emotion, 11(1), 81–89. http://doi.org/10.1037/a0021286
Olivers, C. N. L., van der Stigchel, S., & Hulleman, J. (2007). Spreading the sparing:
against a limited-capacity account of the attentional blink. Psychological
Research, 71(2), 126–39. http://doi.org/10.1007/s00426-005-0029-z
Onie, S. & Most, S. B. (in preparation). Distinct mechanisms of attention bias: Dot
probe and emotion-induced blindness uniquely predict negative affect. 178 | General discussion
Peers, P. V, & Lawrence, A. D. (2009). Attentional control of emotional distraction in
rapid serial visual presentation. Emotion, 9(1), 140–145.
http://doi.org/10.1037/a0014507
Pessoa, L. (2015). Précis on The Cognitive-Emotional Brain. Behavioral and Brain
Sciences, 38, 1–66. http://doi.org/10.1017/S0140525X14000120
Pessoa, L., & Adolphs, R. (2010). Emotion processing and the amygdala: from a “low
road” to “many roads” of evaluating biological significance. Nature Reviews
Neuroscience, 11(11), 773–783. http://doi.org/10.1038/nrn2920
Potter, M. C., Staub, A., & Connor, D. H. O. (2002). The time course of competition for
attention: Attention is initially labile. Journal of Experimental Psychology:
Human Perception and Performance, 28(5), 1149–1162.
http://doi.org/10.1037//0096-1523.28.5.1149
Reynolds, J. H., & Heeger, D. J. (2009). The normalization model of attention. Neuron,
61(2), 168–185. http://doi.org/10.1016/j.neuron.2009.01.002.The
Rokke, P. D., Arnell, K. M., Koch, M. D., & Andrews, J. T. (2002). Dual-task attention
deficits in dysphoric mood. Journal of Abnormal Psychology, 111(2), 370–379.
http://doi.org/10.1037//0021-843X.111.2.370
Sawaki, R., Geng, J. J., & Luck, S. J. (2012). A common neural mechanism for
preventing and terminating the allocation of attention. Journal of Neuroscience,
32(31), 10725–10736. http://doi.org/10.1523/JNEUROSCI.1864-12.2012
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.
http://doi.org/10.3758/APP General discussion | 179
Scalf, P. E., Torralbo, A., Tapia, E., & Beck, D. M. (2013). Competition explains
limited attention and perceptual resources: implications for perceptual load and
dilution theories. Frontiers in Psychology, 4, 1–9.
http://doi.org/10.3389/fpsyg.2013.00243
Schwabe, L., & Wolf, O. T. (2010). Emotional modulation of the attentional blink: is
there an effect of stress? Emotion, 10(2), 283–288.
http://doi.org/10.1037/a0017751
Sergent, C., Baillet, S., & Dehaene, S. (2005). Timing of the brain events underlying
access to consciousness during the attentional blink. Nature Neuroscience, 8(10),
1391–400. http://doi.org/10.1038/nn1549
Shapiro, K. L., Raymond, J. E., & Arnell, K. M. (1994). Attention to visual pattern
information produces the attentional blink in rapid serial visual presentation.
Journal of Experimental Psychology: Human Perception and Performance, 20(2),
357–371. http://doi.org/10.1037/0096-1523.20.2.357
Shapiro, K., Schmitz, F., Martens, S., Hommel, B., & Schnitzler, A. (2006). Resource
sharing in the attentional blink. Neuroreport, 17(2), 163–6.
http://doi.org/10.1097/01.wnr.0000195670.37892.1a
Shih, S. (2000). Recall of two visual targets embedded in RSVP streams of distractors
depends on their temporal and spatial relationship. Perception & Psychophysics,
62(7), 1348–1355. http://doi.org/10.3758/BF03212137
Smith, S. D., Most, S. B., Newsome, L. a, & Zald, D. H. (2006). An emotion-induced
attentional blink elicited by aversively conditioned stimuli. Emotion, 6(3), 523–7.
http://doi.org/10.1037/1528-3542.6.3.523 180 | General discussion
Stein, T., Zwickel, J., Ritter, J., Kitzmantel, M., & Schneider, W. X. (2009). The effect
of fearful faces on the attentional blink is task dependent. Psychonomic Bulletin &
Review, 16(1), 104–109. http://doi.org/10.3758/PBR.16.1.104
Strange, B. A., & Dolan, R. J. (2001). Adaptive anterior hippocampal responses to
oddball stimuli. Hippocampus, 11, 690–698. http://doi.org/10.1002/hipo.1084
Strange, B. A., & Dolan, R. J. (2007). β -adrenergic modulation of oddball responses in
humans. Behavioral and Brain Functions, 13; 3:29. http://doi.org/10.1186/1744-
9081-3-29
Treue, S., & Martinez-Trujillo, J. C. (1999). Feature-based attention influences motion
processing gain in macaque visual cortex. Nature, 399, 575–579.
http://doi.org/10.1038/21176
Van Hooff, J. C., Devue, C., Vieweg, P. E., & Theeuwes, J. (2013). Disgust- and not
fear-evoking images hold our attention. Acta Psychologica, 143, 1–6.
http://doi.org/10.1016/j.actpsy.2013.02.001
Vanderhasselt, M.-A., De Raedt, R., De Paepe, A., Aarts, K., Otte, G., Van Dorpe, J., &
Pourtois, G. (2014). Abnormal proactive and reactive cognitive control during
conflict processing in major depression. Journal of Abnormal Psychology, 123(1),
68–80. http://doi.org/10.1037/a0035816
Vogel, E. K., & Luck, S. J. (2002). Delayed working memory consolidation during the
attentional blink. Psychonomic Bulletin & Review, 9(4), 739–43. Retrieved from
http://www.ncbi.nlm.nih.gov/pubmed/12613677
Vromen, J. M. G., Lipp, O. V., & Remington, R. W. (2014). The spider does not always
win the fight for attention: Disengagement from threat is modulated by goal set.
Cognition and Emotion, 29(7), 1185–1196.
http://doi.org/10.1080/02699931.2014.969198 General discussion | 181
Wallace, W. P. (1965). Review of the historical, empirical, and theoretical status of the
von Restorff Phenomenon. Psychological Bulletin, 63(6), 410–424.
http://doi.org/10.1037/h0022001
Wang, L., Kennedy, B. L., & Most, S. B. (2012). When emotion blinds: A
spatiotemporal competition account of emotion-induced blindness. Frontiers in
Psychology, 3, 1–6. http://doi.org/10.3389/fpsyg.2012.00438
Wyble, B., Bowman, H., & Nieuwenstein, M. (2009). The attentional blink provides
episodic distinctiveness: Sparing at a cost. Journal of Experimental Psychology:
Human Perception and Performance, 35(3), 787–807.
http://doi.org/10.1037/a0013902
Wyble, B., & Swan, G. (2015). Mapping the spatiotemporal dynamics of interference
between two visual targets. Attention, Perception, & Psychophysics, 77(7), 2331–
2343. http://doi.org/10.3758/s13414-015-0938-x