Immediate effects of exposure to positive and negative emotional stimuli on visual search characteristics in patients with unilateral neglect

Neuropsychologia 51 (2013) 2729–2739

Contents lists available at ScienceDirect

Neuropsychologia journal homepage: www.elsevier.com/locate/neuropsychologia

Immediate effects of exposure to positive and negative emotional stimuli on visual search characteristics in patients with unilateral neglect N. Oren a, N. Soroker b, L.Y. Deouell a,c,n a

Department of Psychology, The Hebrew University of Jerusalem, Jerusalem 91905, Israel Loewenstein Rehabilitation Hospital, Raanana, and Sackler Faculty of Medicine, Tel-Aviv University, Tel Aviv, Israel c Edmond and Lily Safra Center for Brain Sciences, The Hebrew University of Jerusalem, Jerusalem 91905, Israel b

art ic l e i nf o

a b s t r a c t

Article history: Received 27 February 2013 Received in revised form 19 August 2013 Accepted 23 September 2013 Available online 27 September 2013

The performance of patients with unilateral neglect (UN) in tasks demanding visual attention is characterized by contralesional disadvantage which is markedly unstable in magnitude. Such instability of the attentional system is seen very clearly in clinical practice and thus far has no satisfying explanation. Here we examined the immediate effect of exposure to non-lateralized emotional stimuli on UN patients' attentional bias and performance variability. We tested eight right-hemisphere damaged stroke patients with left-sided neglect and eight age-matched healthy subjects in a visual conjunctionsearch task, each trial performed immediately after viewing a centrally-presented picture, which was emotionally negative, positive or neutral. Both performance bias and variability in performing the search task was analyzed as a function of the valence of the picture, and a method for analyzing reaction time (RT) variance in a small sample is introduced. Overall, UN subjects, but not controls, were slower and more variable in their RT for left- compared to right-sided targets. In the UN group, detecting left-sided targets was significantly slower in trials that followed presentation of negative pictures as compared to positive pictures, regardless of the fact that both picture types were judged as equally arousing by the patients. Moreover, UN patients exhibited larger performance variance on the left then on the right, and negative emotional stimuli were associated with larger variance asymmetry than positive emotional stimuli. Examining the coefficient of variation pointed to a possible dissociation between the effects of emotional stimuli on the lateralized RT mean (reflecting attentional bias) and on the lateralized RT variance (reflecting system instability). We conclude that emotional stimuli affect the spatial imbalance of both performance speed and stability in UN patients. & 2013 Elsevier Ltd. All rights reserved.

Keywords: Unilateral spatial neglect Emotion Affect Attention Variance Stroke Performance instability

1. Introduction Unilateral neglect (UN) is characterized by failure of salient contralesional stimuli to activate an orienting response, attract attention and generate conscious awareness, a failure that cannot be fully accounted for by sensory or motor loss (Mesulam, 2002). UN is a strong predictor for unfavorable prognosis following right hemisphere stroke (Katz, Hartman-Maeir, Ring, & Soroker, 1999). Therefore its understanding has both theoretical and clinical-therapeutic implications. Although the hallmark of the syndrome – inattention and neglect – is strongly lateralized, there is a growing body of research examining the contribution of non-spatially lateralized deficits (Robertson, 1999; Robertson, Mattingley, Rorden, & Driver, 1998; Van

n Corresponding author at: Department of Psychology, The Hebrew University of Jerusalem, Jerusalem 91905, Israel. Tel.: þ 972 2 5881739. E-mail address: [email protected] (L.Y. Deouell).

0028-3932/$ - see front matter & 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.neuropsychologia.2013.09.033

Vleet & Robertson, 2006) and mechanisms (He et al., 2007; Husain & Rorden, 2003; Robertson, 2001). The spatial and the non-spatial components interact with each other and create the complex clinical picture of UN (Corbetta & Shulman, 2011; Husain & Rorden, 2003). Typically, UN patients miss or are slower to find targets on the left of a search array, relative to their detection accuracy and reaction time (RT) on the right side (Mesulam, 2002). Patients’ performance can be modified by various external manipulations, such as lateralized cues (Posner, Walker, Friedrich, & Rafal, 1984; Van Vleet & Robertson, 2006), non-lateralized cues affecting vigilance, arousal and alertness (Robertson et al., 1998; Thimm, Fink, Küst, Karbe, & Sturm, 2006; Van Vleet & Robertson, 2006) and by internal self-generated intentions (Robertson, 2001). The performance is also characterized by unexplained variability across and within patients (Mesulam, 2002): performance during a task is marked by large variance and inconsistency (Anderson, Mennemeier, & Chatterjee, 2000; Bartolomeo, Siéroff, Chokron, & Decaix, 2001) and performance on multiple administrations of the

2730

N. Oren et al. / Neuropsychologia 51 (2013) 2729–2739

same test may change during the course of the day (Small & Ellis, 1994), although not in all cases (Levy, Blizzard, Halligan, & Stone, 1995). Additionally, the patients’ neglect characteristics change over time (Hamilton, Coslett, Buxbaum, Whyte, & Ferraro, 2008), not necessarily in a linear fashion, with occasional relapses during the recovery phase (Jehkonen, Laihosalo, Koivisto, Dastidar, & Ahonen, 2007). Some of the instability in the functioning of the attentional systems, observed in UN patients, may reflect fluctuations in the patients’ affective state or processing of emotional stimuli, as emotions are a powerful motivational force (Damasio, 1999; Panksepp, 1998, 2007) that affect overt behavior, cognitive processing (Dolan, 2002; Rosler et al., 2005; Scherer, 2005) and consciousness (Damasio, 1999). The relationship between emotion and attention has been extensively studied in recent years, revealing a complex interaction (for reviews see Pourtois, Schettino, & Vuilleumier, 2013; Raymond, 2009; Vuilleumier & Driver, 2007; Yiend, 2010). In healthy subjects, many studies find enhanced processing or preferable response to emotionally non-neutral stimuli (e.g., Hartikainen, Ogawa, & Knight, 2000; Pereira et al., 2006; Rowe, Hirsh, & Anderson, 2007; Simon-Thomas & Knight, 2005; SimonThomas, Role, & Knight, 2005), with some exceptions (e.g., Fox, Russo, Bowles, & Dutton, 2001; Kitayama, 1991; Lipp, Derakshan, Waters, & Logies, 2004). Most studies indicate that negative stimuli are detected faster and more efficiently than neutral stimuli (Eastwood, Smilek, & Merikle, 2001; Fox et al., 2000; Hansen & Hansen, 1988; but see Lipp et al., 2004). Depending on the task, aversive or threatening stimuli may affect performance by several alternative mechanisms including withdrawing attention away from threatening stimuli (Bradley et al., 1997; Mather & Carstensen, 2003; Yiend, 2010), attraction of attention (Bar-Haim, Lamy, Pergamin, Bakermans-Kranenburg, & van IJzendoorn, 2007; Mogg et al., 2000; Wilson & MacLeod, 2003), difficulty to disengage (Koster, Crombez, Verschuere, & De Houwer, 2004) or a variation of the phenomena known as inhibition of return (IOR, Lupianez, Klein, & Bartolomeo, 2006; Posner et al., 1984), whereby attentional facilitation is followed by inhibition. Finally, it is also not settled yet whether emotional stimuli need attention resources in order to be processed (Okon-Singer, Tzelgov, & Henik, 2007; Pessoa, 2005, 2009; Pessoa & Adolphs, 2010; Pessoa, Kastner, & Ungerleider, 2002; Wiens, Sand, Norberg, & Andersson, 2011) or whether the emotional content may be perceived pre-attentively, perhaps by sub-cortical circuits (Vuilleumier & Driver, 2007). Some of the interactions between emotional processing and attention may be related to the fact that both functions seem to be asymmetrically distributed over the cerebral hemisphere. The ventral attention system, which is frequently affected in UN, is more prominent in the right hemisphere than on the left, and the right and left hemispheres compete for directing attention to the contralateral side via the dorsal attention system (Corbetta, Patel, & Shulman, 2008; Corbetta & Shulman, 2002, 2011). Emotional processing is also lateralized (Borod, 1992; Borod, Bloom, Brickman, Nakhutina, & Curko, 2002; Sherratt, 2007; Tsuchiya & Adolphs, 2007; Wager, Phan, Liberzon, & Taylor, 2003) although the exact nature of this lateralization is less clear. The “right hemisphere hypothesis” claims that the right hemisphere dominates processing and expression of emotions of all valences (Adolphs, Damasio, Tranel, & Damasio, 1996; Levine & Levy, 1986). In contrast, according to the “valence hypothesis” the right hemisphere supports processing of negative emotions, while the left hemisphere supports processing of positive emotions (Davidson, 1984, 1995; Davidson & Irwin, 1999). Under both accounts, it seems reasonable to assume that engaging in emotional processing may affect attention by altering the inter-hemispheric balance. The interplay between attention and emotions is especially relevant in UN (see review by Dominguez-Borras, Saj, Armony, &

Vuilleumier, 2012), which is considered an attentional deficit (Mesulam, 2002). Indeed, emotional left-side stimuli are extinguished less in simultaneous bilateral presentation (Fox, 2002; Grandjean, Sander, Lucas, Scherer, & Vuilleumier, 2008; Tamietto, Geminiani, Genero, & de Gelder, 2007; Vuilleumier et al., 2002; Vuilleumier & Schwartz, 2001a, 2001b), detected more in a unilateral presentation (Grabowska et al., 2011) and reduce the rightwards bias in a line-bisection task (Tamietto et al., 2005). It was suggested that these effects of emotional stimuli are due to attention mechanisms that are partly independent from other circuits controlling spatial and object-based attention mechanisms (Dominguez-Borras et al., 2012; Lucas & Vuilleumier, 2008; Vuilleumier, 2005). In the above studies the emotional stimuli were presented laterally, and the facilitation seen with emotional stimuli could be explained by postulating that emotional stimuli, more than neutral stimuli, attract spatial attention to their position in space, in a bottom-up manner. Alternatively, however, the emotional content might affect processing regardless of the spatial position of stimulus, for example due to different engagement of the two hemispheres as noted above. To examine the effect of emotional content regardless of spatial lateralization, stimuli need to be presented without spatial bias. Soto et al. (2009) examined the accuracy level in visuospatial tasks in three UN patients who listened to their preferred music, contrasted with non-preferred or silence. Listening to the preferred music ameliorated neglect. However, it should be noted that the two emotional music conditions were differentiated not just by valence but also by familiarity. While the preferred music was selected by each patient, based on personal preference, the nonpreferred music was selected by the experimenters. Therefore, the preferred music was not only enjoyable and pleasant but also familiar and predictable, as opposed to the non-preferred music which might have drawn more attention due to its novelty. Stimulus novelty influences perception (Schomaker & Meeter, 2012) the level of interest (Silvia, 2005), physiological response (Bradley, Lang, & Cuthbert, 1993) and processing style (Forster, Liberman, & Shapira, 2009) and may have affected the patients’ performance beyond its emotional effect. In a single patient, the lingering effect of music following its termination was tested using a positive, a negative and another positive block. Each block started with induction of mood: music-video of the patient's preferred artist in the positive blocks and a conversation on a disturbing subject in the negative block. In order to sustain the induced mood, a positive emotional picture was presented before each trial in the positive blocks, and similarly, negative pictures were presented in the negative block. This manipulation yielded lower accuracy levels in the search task in the negative block relative to the positive blocks. This block design leaves open the question of the effect of transient emotional stimuli. There is evidence that emotional stimuli of different valence can induce distinct affects on a trial by trial basis even when presented within a mixed block (e.g., Smith, Low, Bradley, & Lang, 2006). In the present study we sought to examine how non-lateralized visual emotional stimuli transiently influence UN patients’ performance in a subsequent visual search task, while focusing on the valence of the stimuli. We used pictures from the International Affective Picture System (IAPS: Lang, Bradley, & Cuthbert, 2005), a standardized pool of color pictures of various contents with norms for valence and arousal. We selected non-arousing pictures of three different valence levels: negative (low score), neutral (medium score) and positive (high score) and presented them in a random order. The pictures were presented centrally in order to prevent spatial attentional bias. All the stimuli were novel to the subjects and were presented once. A single trial of a conjunction search task was performed following each picture presentation.

N. Oren et al. / Neuropsychologia 51 (2013) 2729–2739

2731

Table 1 Demographic and lesion data of the UN patients. Patient

Sex

Age (years)

Education (years)

Lesion type and location

TAO (days)

AA DTa LB MD MN RG SM YBD Mean

M M F M F F M M –

66 74 69 66 57 44 60 60 62 (9.1)

12 7 15 14 10 17 19 12 13.2 (3.8)

H—CP, IHWM, F, T I—T–P junction area, Ins, CP, IHWM I—F, CP, IHWM I (H)—CP, IHWM I—T, CP, IHWM I (H)—F, T, P, Ins, CP, IHWM I—F, T, P, Ins, CP, IHWM I—T(infero-medial)-O region, Thalamus –

38 57 56 80 46 41 38 40 49.5 (14.5)

Abbreviations: M—male, F—female; I—ischemic stroke, H—hemorrhagic stroke, I (H)—ischemic infarction with hemorrhagic transformation; Regions: CP—capsularputaminal, IHWM—intra-hemispheric white matter, F—frontal, T—temporal, P—parietal, O—occipital; TAO—time of testing (days) after stroke onset. Standard deviation (SD) values appear in brackets. a Patient DT was left-handed, all others—right handed (see Section S2 in the Supplementary materials). In all the patients except YBD the damage was within the territory of the right middle-cerebral artery. YBD was affected in right posterior cerebral artery territory including the thalamus.

2. Methods

Table 2 Performance of the UN patients in tests.

2.1. Subjects Patient Eight right-hemisphere-damaged stroke patients with left-sided neglect, hospitalized at the Loewenstein Hospital (Raanana, Israel) for rehabilitation, were recruited for the study. There were three women and five men, at an age range of 44–74 (mean (SD): 62.0 (9.1) years), and an educational level of 13.2 (3.8) years of formal schooling (see Table 1 for demographic and lesion details). Inclusion criteria were: (1) first event of ischemic or hemorrhagic stroke affecting the right cerebral hemisphere, (2) negative neurologic or psychiatric past history, (3) no use of sedative or anti-depressant medications, (3) stable clinical and metabolic state, (4) left-sided UN revealed in at least one of the following formal tests — the behavioral inattention test (BIT: Wilson, Cockburn, & Halligan, 1987a, 1987b), line bisection, Mesulam–Weintraub cancellation test (Weintraub & Mesulam, 1985), or the computerized starry- night test (SNT: Deouell, Sacher, & Soroker, 2005). Patients’ performance in these tests is described in Table 2. Exclusion criteria were: (1) hemior quadrantanopia on confrontation test, (2) difficulty comprehending the instructions or completing the task. Three UN patients admitted to the ward during the patient recruitment period, who answered the inclusion criteria, had to be excluded for these reasons. All patients were right handed, except DT, who was left-handed, yet portrayed a typical left-side UN profile. Eight healthy age-matched control subjects (five women, three men) were recruited from among family members of the patients and acquaintances of the experimenters. The age range in the control group was 54–74 years (mean (SD): 64 (5.9) years). Educational level was 12.5 (4.4) years. The patient and control groups did not differ in age (t(14) ¼  0.52, p ¼ 0.611) or education (t(14) ¼0.382, p ¼0.708). All the control subjects were right handed, with a normal or corrected to normal vision and with no prior neurological or psychiatric problems. The study was approved by the Ethics (Helsinki) committee at the Loewenstein Rehabilitation Center, Raanana, Israel and all subjects, patients and controls, gave their informed consent to participate. Control subjects were compensated for their time with a small fee. 2.2. Materials The International Affective Picture system (IAPS; Lang et al., 2005) is a database of color pictures, each with a standardized score for valence (from 1, negative, to 9, positive, with 5 indicating neutral), and arousal level (from 1, calm, to 9, arousing, with 5 again indicating neutral). Numerous studies have used this database to examine affective influences in different populations (e.g., Backs, da Silva, & Han, 2005; Gavazzeni, Wiens, & Fischer, 2008; Gruhn & Scheibe, 2008; Ribeiro, Pompeia, & Bueno, 2005; Verschuere, Crombez, & Koster, 2001). We chose 96 non-arousing1 (i.e., arousal level between 4 and 6) color pictures form the IAPS, with three valence levels: 32 negative pictures (valence—mean (SD): 2.46 (0.25), range: 1.91–2.92; arousal: 5.29 (0.52), 4.06–5.99); 32 neutral pictures (valence: 4.82 (0.67), 4.01– 5.97; arousal: 5.03 (0.45), 4.25–5.7); and 32 positive pictures (valence: 7.61 (0.37), 7.01–8.34; arousal: 5.26 (0.45), 4.43–5.87) (see left side of Fig. 2: IAPS norms). The three categories were significantly different in valence, as indicated by a one-way analysis of variance (ANOVA), with valence as a dependent variable and affective category (negative, neutral, positive) as a between-item variable (F(2,93) ¼987.28, p o 0.001). Independent-sample t-tests with Bonferroni correction confirmed that

1 Pictures which we a priori judged as potentially offensive such as mutilation or nudity were not included.

AA DT LB MD MN RG SM YBD

BIT 97 125 71 127 140 129 142 130

LB 18.4 26.5 2.1 8.2 15 4.8 3.5  8.8

n

(9.33) (7.3)n (3.24) (11.61) (4.71)n (5.39)n (4.7)n (13.62)

MWCT (L-R)

SNT-C

SNT-F

2–22 12–19 0–6 14–20 13–23 29–30 23–30 26–30

þ þ

þ þ þ þ þ þ þ þ

þ þ þ

Abbreviations: BIT—Behavioral inattention test (Wilson et al., 1987a, 1987b). Scores below 130 are abnormal; LB—line bisection, mean signed displacement in mm (SD) of the subjective midpoint from the objective midpoint in lines of 180 mm. Positive numbers indicate rightward displacements. MWCT—Mesulam–Weintraub cancellation test (Weintraub & Mesulam, 1985), L/R—number of target stimuli (out of 30 on each side) detected on the left/right side of the page; SNT—computerized Starry night test (Deouell et al., 2005), F—feature search mode, C—conjunction search mode, þ —significantly longer reaction time for target stimuli in the three contralesional columns compared to the three ipsilesional columns (po0.05). n

Indicates significant difference from zero (t-test, po 0.05).

negative pictures had lower valence level than neutral (t(62) ¼  18.82, p o 0.001) or positive pictures (t(62) ¼  64.73, po 0.001), and that neutral pictures had lower valence level than positive pictures (t(62) ¼  20.6, po 0.001). Arousal level was similar in the three categories, as suggested by a one-way ANOVA, with arousal as a dependent variable and affective category (negative, neutral, positive) as a between-item variable (F(2,93) ¼2.81, p¼ 0.065). Analysis of content, luminance and spatial frequency of the three categories are presented in the Supplementary materials (Section S1). The pictures captured 10.61 of visual space with the observer seated 1 m from the screen. In the original database, some of the pictures had a black frame around the main figure. In order to maintain a constant size of all the pictures against the black background on which they were presented, the black frames were painted in gray in our study. Thirty two gray rectangles, the same size as the pictures, were also intermixed with the pictures. This category was meant to serve as a “low level” baseline, but was eventually discarded from analysis as there were considerable differences in chromaticity and luminance between it and the affective pictures. Additionally, there were inconsistent responses, in terms of RT, arousal and valence ratings, within and across subjects to this category, suggesting that each subject perceived this category differently (i.e., some as a neutral stimuli, others as negative), and thus that this was not an emotion-free baseline. Search task: The visual conjunction search task was a variation of the task used by Eglin, Robertson & Knight (1989). There were 20 stimuli in each array: 1 target and 19 distracters. The array was presented on a white screen, virtually divided into a 5  4 grid, which was not visible. A single stimulus appeared in a random position in each one of the virtual cells (Fig. 1). Thus the stimuli were equally spread across the array, in a non-organized fashion. There were two kinds of distractors. The first was a blue circle of 1.151 diameter, divided by a white line, 0.11 wide, with an arrow at its end (Fig. 1). The direction of each distractor’s arrow – pointing either to the upper left or the upper right – was randomly assigned. The second kind of the distractors consisted of red circles of

2732

N. Oren et al. / Neuropsychologia 51 (2013) 2729–2739

Fig. 1. A trial began with a fixation cross presented for 1.5 s, followed by a single centrally located emotional picture presented for 1 s. The picture was non-arousing and of negative, neutral or positive valence (example of each category is presented). A conjunction search array appeared immediately after. The subject had to detect the target (a red circle with an arrow at its end—encircled by a broken line in the figure for highlighting purpose only) and press a button as quickly as possible. When the array disappeared the subject had to state aloud the direction of the target’s arrow. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) 1.151 diameter, with no bisecting line. The target was a conjunction of both distractors—it was a red circle with a bisecting arrow. It could appear on any virtual row on the left or the right columns, but never in the middle column. For every subject, half of the pictures in each category were coupled with a right-sided target stimulus and half with a left-sided target stimulus. Additionally, half of the pictures were coupled with a target stimulus positioned in the upper half of the array, and half with a target stimulus in the lower half of the array, but the target’s vertical position was not a factor in the analysis. Moreover, since each picture was presented only once for every subject, and in order to avoid effects of any specific stimuli, the coupling between the picture preceding the array, the target column locations, and the side the target’s arrow was pointing at was counter balanced across the 8 subjects of each group (i.e., UN patients and control). 2.3. Procedure Subjects were tested individually, seated 1 m in front of a CRT computer screen (ViewSonic, USA). They were free to move their eyes and head, and all used their right (contralesional) hand to respond (see Section S2 in the Supplementary materials). There were 32 trials of each affect, as well as 32 trials with gray rectangles instead of a picture (see above). The trials were pseudo-randomly intermixed, with the constraint that no more than two consecutive trials were from the same affective category. A trial began with a white fixation cross presented at the center of a black screen for 1.5 s, followed by a centrally located emotional picture presented for 1 s (Fig. 1). The search array appeared immediately after, and remained on the screen until the subject responded or 20 s elapsed. Subjects had to press a button on a response box as fast as they could when they detected the

target. If the subject managed to detect the target and press the response button within the time limit, the array disappeared and a question was presented (in Hebrew): “Where did the target’s arrow point, right or left?” on which the subject replied orally. The verbal response was recorded by the experimenter who was blind to the correct response. RT was measured from the onset of the search array to the button press indicating target detection, while accuracy relates to the left/right verbal identification. An inter-trial black screen was presented before the beginning of the next trial, which was initiated once the subject indicated he or she was ready. If no response was made within the 20-s time frame, the array changed to the inter-trial black screen. To ensure the subject was looking at the pictures, the trials were presented in mini-blocks of four trials. Following each mini-block, the subject was asked a simple yes/no question regarding the last four pictures, on which the subject responded orally. An example for such a question is: “was there a human figure in the last four pictures?” In order to confirm the subjective experience of valence and arousal, following the experiment the subjects were asked to rate each picture’s valence and arousal level. We used the same procedure employed by Lang et al. (2005), except that the rating scales were vertical and not horizontal, in order to prevent neglect of the left side of the scale by the UN patients (see Section S3 and Fig. S1 in Supplementary materials). 2.4. Data analysis Analysis of subjects’ rating of the affective pictures was done using the same analysis that was used for the original norms (see above), for each group separately.

N. Oren et al. / Neuropsychologia 51 (2013) 2729–2739 A one-way ANOVA with the original affective category (negative, neutral, positive) as a between-item variable and either valence or arousal as a dependent variable was performed. Post-hoc paired comparisons with Bonferroni correction for multiple comparisons were carried out. Since the groups’ valence ratings maintained the original division into three affective categories, as in the original norms (see Section 3), further analyses were conducted using these three categories. Subjects’ performance in the visual search task was analyzed using three separate measures: accuracy level, RT mean and RT variance. All analyses were performed after discarding trials in which the subject could not locate the target in the 20 s time frame (UN patients only), or obviously not concentrated in the task (e.g., talked during the trial, fell asleep), or when the RT was less than 100 ms indicating a premature key press. Incorrect trials, in which the subject misidentified the target’s arrow direction, were excluded from RT mean and variance analyses. Accuracy level was calculated in each condition (three affective conditions (negative, neutral, positive) by two sides) as the number of the correct answers divided by the total number of trials in that condition. For the RT mean analysis we used a laterality index (LI) which was computed for every affective category in the following manner: LI¼ (RT left  RT right)/(RT leftþ RT right). A positive LI thus indicates longer RTs for targets on the left then on the right, while negative LIs indicate the opposite. To establish the presence of a significant performance lateralization, we first tested the LI, collapsed across categories, against zero, in each group. Next, to examine the effect of picture valence, the LI served as the dependent variable in a two-way mixed ANOVA, with group (UN patients, control) as a between subject variable, and affective condition (negative, neutral, positive) as a within subject variable. Results were further analyzed using post-hoc contrasts: negative vs. neutral; positive vs. neutral; and negative vs. positive. The non-orthogonal contrasts were a direct derivative of the hypotheses that all three affective conditions are different from each other, but with no conjectures as to the direction of those differences. The contrasts were 2-way mixed ANOVAs with factors affect and group, with the affect factor consisting of a duo of affects (e.g., negative vs. positive) for each ANOVA. Bonferroni correction was used to correct for the three comparisons. Statistically significant results were further analyzed separately within each group. Since there were differences between the patients, in terms of lesion location and time after onset of stroke, we also examined each UN patients’ raw RT separately. 2.4.1. Variance ratios (F-values) Anderson et al. (2000) were the first to point out that UN patients’ RT variance on the left side of space was larger than on the right side. Bartolomeo et al. (2001) examined how different manipulations affect the RT variance of 6 UN patients, using it as the dependent variable in an ANOVA analysis. This poses a statistical difficulty, since ANOVA requires the dependent variable to have a normal or a close to normal distribution, whereas variance in such a small sample has a chi-square distribution (Hays & Winkler, 1971). One way to address this issue is to perform log transformations in order to normalize a skewed chi-square distribution. Here we present an alternative method for assessing differences in variance in small samples that do not require the use of log transformation. We examined the differences between left and right target stimuli within each affective condition using the F statistic, i.e., the ratio of left over right variance estimates (F-value). F is a random variable created by the division of two independent chi-square variables, each divided by its degree of freedom (Dixon & Massey, 1969; Hays & Winkler, 1971). Put differently, division of the estimation of variance of two independent populations, each with a normal distribution, results in an F-value.2 In the present case, the samples are taken from the populations of reactions of an individual subject for the left- and right-sided targets, which are considered to be close to normally distributed. The responses are independent of each other since there was no pairing between the two sets of responses. F tables could thus be used to test the hypothesis that the variance was larger on the left than on the right (F-value4 1, one-tailed test in the UN group), at the single subject’s level. Since we did not have a prior assumption regarding the control group, we tested their results using a two-sided comparison. Moving beyond the single subject, the same left–right variance ratios where the dependent variable in a two-way mixed ANOVA, with group (UN patients, control) as a between subject variable, and affective condition (negative, neutral, positive) as a within subject variable. However, since variance rations (F-values) do not distribute normally, the statistical significance of the results was established in a non-parametric way. For the two main effects a permutation procedure (Manly, 1997) was used. Here, the distribution of the ANOVA statistics under the null hypothesis was constructed, by repeatedly creating shuffled datasets and subjecting them to the same mixed ANOVA as in the original procedure. Specifically, in each iteration the valence labels (negative, positive, neutral) were shuffled randomly within each subject, followed by shuffling of group labels between subjects. The shuffled data was then subjected to the same ANOVA and the resulting ANOVA statistics were logged. This was repeated 50,000 times to create the data driven null distribution of the ANOVA’s statistics, to which the statistics obtained from the original ANOVA were compared. If the actual statistic was larger

2 Note that this is the same F statistic commonly used in ANOVA, where the F-value is the ratio between the between- and within-cell variances.

2733

than 95% of the statistics obtained from the shuffled data, the result was considered significant. To establish the null distribution for the interaction effect, we followed the procedure recommended by Anderson and Ter Braak (2003). First, we calculated the differences between the true values and the values predicted by the additive general linear model (GLM) with only main effects (i.e., without interaction effect). The residuals represent the interaction effect plus noise. Second, since under the null hypothesis of no interaction the residuals are exchangeable in all directions, we shuffled the residuals without restriction, and inserted them back to the additive model. Thus, only the interaction effect was shuffled and the shuffled data set was tested in a mixed ANOVA. Repeating the shuffling and ANOVA procedure 50,000 time we thus created the null distribution for the interaction effect. Significant results were further analyzed using the same logic of analyses employed in the LI analysis. That is, each pair of affective conditions (i.e., negative vs. neutral; positive vs. neutral; and negative vs. positive) was examined with a separate 2-way mixed ANOVA. Here again, significance was tested using the residual procedure, except that Bonferroni correction was used to correct for the three comparisons. Statistically significant results were further analyzed separately within each group using the non-parametric Wilcoxon signed ranks test (Siegel & Castellan, 1988). This non-parametric test examines the relative magnitude and direction of differences between pairs of observations. 2.4.2. Affective coefficient of variation lateralization (ACVL) Since there is a linear relationship between RT’s mean and variance (Wagenmakers & Brown, 2007), the two measures may not be independent. Thus, we also examined the lateralization of coefficient of variation (CV), in order to confirm that the results observed in the variance examination were not merely a byproduct of differences in the mean. CV is a dimensionless number calculated by dividing the standard deviation (SD) with the mean and is considered a reliable measure for the sample variability (Reed, Lynn, & Meade, 2002). In the present study we calculated the CV in every condition (three affect conditions by two sides), and subtracted the right from the left CV in each affective condition, thus creating an affective coefficient of variation lateralization (ACVL) index. In ACVL, a positive number indicates larger CV on the left, and a negative indicates a larger CV on the right. To examine whether the affective conditions affected this measure consistently, we again used the Wilcoxon test (Siegel & Castellan, 1988).

3. Results 3.1. Rating of affective pictures Subjects rated each picture valence and arousal at the end of the experiment. Fig. 2 presents the ratings of patients and controls against the IAPS norms (Lang et al., 2005). UN patients’ valence rating was low for negative pictures (mean (SD): 3.07 (1.04)), medium for neutral pictures (4.7 (1.59)) and high for positive pictures (7.23 (1.05)). The three ratings differed significantly from one another (F(2,93) ¼ 82.93, p o0.001). Post-hoc paired comparisons with Bonferroni correction confirmed that negative pictures were given significantly lower valence ratings than neutral (t(62) ¼  4.19, p o0.001) and positive pictures (t(62) ¼  15.43, po 0.001), and that neutral pictures were given significantly lower valence ratings than positive pictures (t(62) ¼  7.5, p o0.001). The control group’s valence ratings were also in accordance with the original affective category (negative 2.3 (1.13); neutral 4.74 (1.42); positive 7.33 (0.96)). Again, the three ratings were significantly different from one another (F(2,93) ¼143.27, p o0.001). Post-hoc paired comparisons with Bonferroni correction confirmed that negative pictures were given significantly lower valence ratings than neutral pictures (t(62) ¼  7.56, p o0.001) and positive pictures (t(62) ¼  19.21, p o0.001), and that neutral pictures were given significantly lower valence ratings than positive pictures (t(62) ¼  8.52, po 0.001). UN patients’ arousal rating was around 5 (i.e., ‘neutral’ in the original IAPS classification) for all the three categories (negative 5.71 (0.55); neutral 5.42 (0.41); positive 5.26 (0.68)) with no significant difference between the ratings (F(2,93) ¼2.62, p¼ 0.087). The control subjects also rated the arousing level of the neutral and the positive pictures around 5 (neutral 5.5 (0.83); positive 5.15 (0.87)), but the negative pictures were rated as somewhat more arousing (6.28 (0.75)). These differences were significant (F(2,93) ¼14.95, p o0.001). Post-hoc independent

2734

N. Oren et al. / Neuropsychologia 51 (2013) 2729–2739

Fig. 2. Mean valence and arousal level of the pictures in the three affective categories, as rated by the UN patients and control groups, compared to the original IAPS norms (Lang et al., 2005). Error bars represent standard deviations.

sample t-test with Bonferroni correction in the control group showed that negative pictures were given higher arousal scores than neutral pictures (t(62) ¼ 3.78, po 0.001) and positive pictures (t(62) ¼5.41, po 0.001), but that the positive and neutral categories were rated similarly (t(62) ¼1.65, p ¼0.1). The results of the valence and arousal rating are presented in Fig. 2. 3.2. Accuracy level Subjects' task in the experiment was to detect the target in the search array within a 20 s time frame. There were only five incidents in which a left-sided target was not detected in that time frame, and only in the UN group (AA in two trials and LB, MN and YBD in one trial each). Following target detection, the subject had to report the direction of the target’s arrow. Four patients (MD, RG, SM, YBD) completed the task with no error and three with few mistakes (AA made one mistake on the right side; DT made four mistakes on the left side; LB made four mistakes on the right and two on the left side). Patient MN made three to eight mistakes in each condition (three affect conditions by two sides), and overall was correct in 57 of 95 trials. Although worse than the other subjects, the odds for achieving these results by chance are very low (0.032) using the sign test (Siegel & Castellan, 1988), based on a binominal distribution. We thus conclude that the patient understood the task and was performing it, although inaccurately. Six of the eight control subjects completed the task with no error, one subject had one error and another subject had two. Due to the near ceiling effect in both groups, we did not perform any additional analyses using the accuracy measure. 3.3. Reaction time—Mean 3.3.1. Group level As expected UN patients showed a significant rightward bias in RT, indicated by a positive LI which was significantly different from zero (mean across the three affective categories (SD): 0.32 (0.15); t(7) ¼6.05, p ¼0.001). The mean LI of the control group (0.03 (0.05)) was not significantly different from zero (t(7) ¼1.41, p¼ 0.19). To test whether the valence of the stimuli affected the spatial bias, a mixed ANOVA, with group (UN patients, control) as a between-subject variable and affect (negative, neutral, positive) as a within-subject variable, was conducted. Violation of sphericity was examined by Mauchlhy’s test and not confirmed (p ¼0.3). There was a significant main effect for group (F(1,14) ¼27.249, p o0.001), confirming that the UN patients’ LI was higher than that of the controls. There was no main effect for affective category (F(2,28) ¼1.997, p ¼ 0.155), but there was a significant interaction between group and affect (F(2,28) ¼3.786, p ¼0.035). This interaction was further analyzed by three contrasts that examined the differences between each pair of affective categories (negative vs.

Fig. 3. Average laterality index (LI) for the UN patients and control group. Error bars represent standard error (SE) within each affective condition. LI ¼(RT left  RT right)/(RT left þ RT right). Abbreviations: *p o 0.05.

positive; negative vs. neutral; positive vs. neutral) in the two groups, using 2-way mixed ANOVAs, with the within-subject error term of the original ANOVA, as the sphericity assumption was not violated. The threshold for significance was Bonferroni corrected for three comparisons. There was a significant interaction between affect (negative vs. positive) and group (UN vs. control) (F(1,28) ¼ 7.38, p ¼ 0.011). Additional analyses revealed that the LI in the negative condition was higher than the LI in the positive condition for the UN patients (F(1,28) ¼ 11.096, p ¼0.0024), but not for the control subjects (F(1,28) ¼0.25, p ¼0.61). None of the other contrasts was significant. The RT-based LI in each group is presented in Fig. 3. 3.3.2. Individual level We further examined the RT pattern of each UN patient (Fig. 4). While all patients presented prolongation of RT for left-sided target stimuli relative to right-sided targets, five patients (AA, DT, LB, MD and MN) also manifested a non-spatially lateralized component known to occur in the UN syndrome (Husain & Rorden, 2003; Robertson, 2001; Robertson et al., 1998)—they were also slow in detection of right-sided targets, relative to the other three patients and to the control subjects (in Fig. 4, compare their results to the black line, which represents the average RT of the control group for left and right target stimuli). The same five patients also missed items on the right side of the MWCT (Weintraub & Mesulam, 1985) (Table 2). These patients also portrayed a pattern of longer RT following negative compared to positive pictures for left-sided targets, and an opposite pattern on the right, suggesting that negative stimuli were associated with accentuated bias towards the right. The three patients in the second sub-group (RG, SM and YBD) were not different from the control subjects in RTs for right sided stimuli, and they rarely missed right sided items on the MWCT. The effect of the affective categories of the pictures was less consistent in these subjects on either side of space (see Fig. 4). 3.4. Reaction time—Variance 3.4.1. Variance ratios (F-values) To examine the system stability in detection of left- compared to right-sided target, we calculated the ratio (F-value) between left-side and right-side variance, in each affective condition. All but two UN patients, in one affective condition each (SM in the negative condition and MN in the positive condition), had F-values larger than 1, indicating larger variance in the speed of detecting left- than rightside targets (Table 3). The significance of these results was determined for each subject and condition using F tables. All patients but one (SM)

N. Oren et al. / Neuropsychologia 51 (2013) 2729–2739

Fig. 4. Average RT (in milliseconds) of UN patients (represented by initials), for left and right targets, in each affective condition. The UN patients group is divided tentatively into two sub-groups on the basis of demonstration (AA, DT, LB, MD, MN) or not (RG, SM, YBD) of a general slowness of RT, affecting also the right-sided stimuli (see text for explanation). The black line represents the average RT of the control group for right and left targets combined, serving as a reference for the assessment of UN patients’ performance. Abbreviations: Neg-Negative, Neu-Neutral, Pos-Positive.

Table 3 Variance ratios (F-values) for left over right target RTs, of UN patients and control subjects. Group

Subject

Affective condition Negative

UN patients

AA DT LB MD MN RG SM YBD Total number of significant results Control subjects

DIT EA HG JL RW VK YA YN Total number of significant results

Neutral

Positive

n

31.59 5.5n 73.98n 6.7n 32.72n 7.11n 0.45 62.62n 7

n

3.27 1.96 5.52n 1.94 1.08 6.14n 1.16 83.49n 4

1.53 1.12 1.29 1.04 0.3 4.78n 2.95n 9.8n 3

0.49 1.06 1.55 1.16 0.6 0.36 0.31n 0.7 1

0.38 0.32n 0.36 0.9 2.14 3.3n 0.67 0.7 2

1.27 4.16n 1.44 5.33n 11.18n 1.12 0.62 0.22n 4

Variance ratios (F-value ¼left variance/right variance) in each affective category, for UN patients and control subjects (represented by initials). n

Indicates significant F-values (p o 0.05).

had statistically significant F-values in the negative condition, whereas only three patients had a significant F-values in the positive condition (in the other five patients the F-values were in fact close to or smaller than 1). To test whether the valence of the stimuli affected variance ratios across the groups, a mixed ANOVA, with group (UN patients, control) as a between-subject variable and affect (negative, neutral, positive) as a within-subject variable, was conducted, and the significance of the results was tested using a permutation and a residual procedures (see Section 2). There was a significant main effect for group (F(1,14) ¼4.404, p¼0.004), confirming that the UN patients’ left–right variance ratios were higher than that of the controls. There was a marginal effect for affective category (F(2,28) ¼ 2.809, p¼0.064). These main effects were qualified by a significant interaction between group and affect (F(2,28) ¼4.05, p¼ 0.017). Examination of this interaction testing pairs of valence conditions revealed a significant

2735

Fig. 5. Affective coefficient of variation lateralization (ACVL) of UN patients (represented by initials), calculated by subtracting the right coefficient of variation (CV) form the left CV in every affective condition. Positive numbers indicate larger CV on the left than on the right.

interaction between negative and positive stimuli and group (F(1,28) ¼7.78, p¼0.007). A non-parametric Wilcoxon signed ranks test (Siegel & Castellan, 1988) indicated that in UN patient group the left–right variance ratios of the negative condition were significantly larger than the F-values of the positive condition (p¼ 0.012, twotailed), but not so in the control (p¼0.1). None of the other comparisons was significant. 3.4.2. Affective coefficient-of-variation lateralization (ACVL) Since a change in variance may be a trivial result of a change in mean RT (Wagenmakers & Brown, 2007), we computed the affective coefficient-of-variation lateralization index (ACVL) which normalizes the variance by the mean (see Section 2). The ACVL was larger following negative pictures than following positive pictures in all but one patient: SM (Fig. 5). At the group level, a trend for significance of this latter difference was found using the Wilcoxon signed ranks test (p ¼0.078). While this result does not meet the threshold for significance, we note the extremely large outlier result of patient SM, who had a very large negative ACVL in the negative affect condition, a direction opposite to all other patients. Excluding this patient, there is a significant effect across the group, with ACVL in the negative affect condition being larger then ACVL in the positive affect condition (Wilcoxon signed ranks test, p ¼0.016). It is interesting to note that not only was the ACVL larger in the negative than positive affective conditions, but in fact the ACVL indicated larger CV on the left then on the right when the affect was negative in all but one patient, while it indicated larger CV on the right then on the left when the affect was positive in all but one patient (note however that these are not the same seven patients in both cases). Taken together, these results corroborate the conclusion that the affective conditions influence performance variance beyond their effects on the mean RT (Fig. 5).

4. Discussion The hallmark of UN is a lateral bias of attention manifested by failure of salient stimuli on the side of space contralateral to the lesion side to induce an orienting response and gain access to conscious awareness (Mesulam, 2002). When such contra-lesional stimuli are detected, their processing shows marked disadvantage in terms of accuracy and speed, compared to the processing of ipsi-lesional stimuli. Current neglect theories recognize the importance of non spatially lateralized factors that interact with the lateral bias (Husain & Rorden, 2003; Robertson, 2001; Robertson et al., 1998). We considered the processing of centrally presented emotional stimuli as a distinct non-spatial (non-lateralized) factor,

2736

N. Oren et al. / Neuropsychologia 51 (2013) 2729–2739

and examined its interaction with the lateral bias. To induce processing of emotionally-laden stimuli without directly attracting attention to the left or right, the emotional stimuli in our study were presented centrally before every trial of a visual task demanding conjunction search. Overall, the current study presents a complex interaction between processing of affective stimuli and spatial attention. Centrally presented emotional pictures influenced the rightwards performance bias of UN patients. Exposure to pictures which the patients rated as inducing a negative emotion was characterized by strong rightwards attentional bias in an immediately following search task. In contrast, exposure to positive pictures caused “rebalancing” of the spatial attention asymmetry, as evidenced in both RT means and variance measures. It is unlikely that these effects are related to differences in arousal caused by the pictures. The pictures were chosen to have similar arousing values, and the UN patients own ratings of arousal did not significantly differ between the affective categories. For the control group, the negative pictures indeed yielded significantly higher arousal ratings; however, control subjects did not show any affective category effect to go along with the difference in arousal. Presentation of negative valence pictures to UN patients increased their RT laterality index relative to positive pictures, with the neutral images yielding an intermediate result. Moreover, in a subgroup of five patients, RT on the left side was longer in the negative compared to the positive condition, whereas the reverse was observed on the right side. These patients were specifically characterized by a prominent “non-spatial” component UN (Husain & Rorden, 2003; Robertson, 2001; Robertson et al., 1998), demonstrating generalized attenuation of processing speed and inaccuracy on both sides of space. Thus, it seems that relative to positive emotional stimuli, negative emotional stimuli exacerbate the lateralized bias in UN patients, especially in patients with a marked nonspatial attentional deficit component. Previous studies showed that UN patients’ RT to contraslesional stimuli is not only slower, but also more variable than on the ipsilesional side (Anderson et al., 2000; Bartolomeo et al., 2001). Elevation of RT variability was also reported in other clinical populations, such as attention-deficit/hyperactivity disorder (ADHD) (de Zeeuw et al., 2008; Epstein et al., 2011; Klein, Wendling, Huettner, Ruder, & Peper, 2006), autism (Geurts et al., 2008) and Alzheimer’s disease (Duchek et al., 2009). It seems that difficulty maintaining performance stability is a sensitive marker of impaired cerebral functioning, perhaps even more than performance level itself. Unilateral neglect is different from other populations of brain damaged patients in that the increase in variability is lateralized, consistent with the spatially-specific deficit. Previous works with UN patients (Bartolomeo et al., 2001) and other populations (Klein et al., 2006; Segalowitz, Poulsen, & Segalowitz, 1999) used ANOVA to assess the variance or CV. In small samples, such as in the present study, this could be an inadequate approach as variance has a chi-square distribution rather than a normal or close to normal distribution required for ANOVA (Dixon & Massey, 1969; Hays & Winkler, 1971). Hence, alternative statistical methods, such as the one employed in the present study or in previous work (Anderson et al., 2000) should be used. In the current study we reaffirmed the existence of left vs. right variance differences in UN patients using individual level and nonparametric group statistics. Our findings show that the ratios of left over right variance are larger in UN patients than controls. Additionally, in UN, but not in controls, negative pictures were associated with higher left–right imbalance in performance variance, compared to positive pictures. Importantly, this effect of emotional pictures, and presumably of emotional processing or emergent transient emotional states, on performance variability,

was present even when the effect on RT mean was taken into account by measuring the coefficient of variation. Although the ACVL results were statistically marginal due to an outlier subject in a small cohort, we note that these effects were found in seven out of eight patients studied. Therefore, like the mean RT lateralityindex measures, the variance measures show a distinct effect of negative and positive affective stimuli on UN patients’ lateral bias. The effect of transient emotional stimuli could be explained by induced changes in the level of attention to the search array. Patients might have withdrawn attention from the screen in response to negative stimuli and were therefore late to shift their attention back to the task displayed immediately after. Though some earlier findings indicate the existence of a bias away from threatening stimuli (e.g., Bradley et al., 1997; Mather & Carstensen, 2003; Yiend, 2010), this effect depends on stimulus threat level and task requirements (Cooper & Langton, 2006; Holmes, Green, & Vuilleumier, 2005). Alternatively, negative stimuli may have attracted attention, as was shown both in healthy subjects (e.g., Fox et al., 2001; Rosler et al., 2005) and in UN patients (e.g., Fox, Russo, & Dutton, 2002; Tamietto et al., 2005, 2007; Vuilleumier et al., 2002; Vuilleumier & Schwartz, 2001a, 2001b). Therefore, it might have been difficult for the patients to disengage from the processing of such stimuli, again delaying attention to the task display following the picture. It could also be suggested that the negative stimuli first attracted attention, but later an inhibition of return (IOR, Lupianez et al., 2006; Posner et al., 1984) hindered the response to the targets. However, under these purely attentional explanations, RT should have been prolonged on both sides of space, even if more so on the left. Instead, we observe shorter RT for rightsided targets following exposure to negative relative to positive pictures in most patients (Fig. 4). Thus, it seems that in the case of emotional processing induced by centrally-presented visual stimuli, as in the current study, or non-visual stimuli like music (Soto et al., 2009), a different mechanism may be employed. We hypothesize that there is a trade-off between two processes presumably operating in the right hemisphere: processing of negative stimuli (Borod, 1992; Borod et al., 2002; Davidson, 1984, 1995; Davidson & Irwin, 1999) and directing attention to the contralateral side of space (Corbetta et al., 2008; Corbetta & Shulman, 2002). According to the “valence hypothesis”, negative emotions are preferentially processed by the right hemisphere and positive emotions by the left (Borod, 1992; Borod et al., 2002; Davidson, 1984, 1995; Davidson & Irwin, 1999). Since emotional processing continues over time (Davis et al., 1995; Huang & Luo, 2006; Larson, Ruffalo, Nietert, & Davidson, 2005; Smith et al., 2006), it may consume attentional resources that are necessary for fast and consistent subsequent responses to the contralateral hemispace. Accordingly, when negative stimuli precede a search array, the damaged right hemisphere allocates its limited resources to their processing, and the ability to orient to the left side of the search array is further hampered. Positive stimuli, on the other hand, tax the left hemisphere (Borod, 1992; Borod et al., 2002; Davidson, 1984, 1995; Davidson & Irwin, 1999), and might allow some rebalancing of the left and right systems (Corbetta et al., 2008; Corbetta & Shulman, 2002). This so called “competitive” account is consistent with Pessoa & Adolphs’ (2010) view that processing of emotional stimuli does not rely purely on a fast sub-cortical route, but rather on cortical neuronal networks. Using affective visual stimuli and RT measures, we expand previous results obtained in the auditory modality using accuracy measure (Soto et al., 2009). Soto et al. examined the performance of three chronic UN patients in various visual tasks, either when listening to preferred individually-selected music or to nonpreferred music. In one subject the effect of music was maintained even when the music was played before rather than concurrently with the task, and affective visual images were presented before each trial, in a block design (as described in details in Section 1).

N. Oren et al. / Neuropsychologia 51 (2013) 2729–2739

Accuracy levels were higher in the positive condition than in the negative condition. There are some fundamental differences between the research reported by Soto et al. (2009) and our study, beyond the sensory modality used for stimulation of the emotional system. We tested a larger group of patients, thus being able to examine inter-subject variability in the response to emotional stimuli. Unlike Soto et al. (2009) who separated the presentation of negative and positive affective stimuli into different blocks, we intermixed stimuli with negative, neutral and positive valence in the same block, thus focusing on the immediate effects of emotional stimuli (Hamilton et al., 2008; Jehkonen et al., 2007; Small & Ellis, 1994). Importantly, in Soto et al. (2009), the preferred and non-preferred music differed not only in their emotional level, but also in familiarity (as explained in Section 1), and it is hard to know whether the valence of the music, or its level of novelty was responsible for the behavioral effects. In the current work all the emotional pictures were novel to the subjects and the UN subjects rated them as having neutral arousal levels. Finally, there is an ongoing debate whether music evokes the same emotions as other kind of emotional stimulants or not (see for example review by Juslin & Vastfjall, 2008). The results of the current study, and those obtained by Soto et al. (2009), indicate that exposure to stimuli inducing a negative emotion or requiring processing of negative emotional content worsens UN manifestations, relative to exposure to positive stimuli. These results seem to contradict earlier findings obtained with lateralized presentation of emotional stimuli to UN patients, where contralesional negative stimuli were shown to ameliorate performance in visual (Fox, 2002; Tamietto et al., 2005, 2007; Vuilleumier et al., 2002; Vuilleumier & Schwartz, 2001a, 2001b) and auditory extinction tasks (Grandjean et al., 2008), detection task (Grabowska et al., 2011), visual search (Lucas & Vuilleumier, 2008), line bisection (Tamietto et al., 2005) and identification tasks (Williams & Mattingley, 2004). Taking the effects of lateralized and non-lateralized stimulation together, the data suggest that in UN, exposure to negative emotional stimuli is likely to initiate two distinct operations: (a) consumption and further depletion of already limited cognitive resources needed for appropriate mobilization of attention in space; and (b) attraction of attention due to increased valence of emotionally laden stimuli. When stimuli presented in the contralesional hemispace carry a negative emotional meaning, they are more likely to attract attention, escape neglect and gain access to conscious awareness. With such lateralized presentation, the increment in stimulus salience overcomes the effect of resource depletion, hence reducing the rightward bias of attention (e.g., Vuilleumier & Schwartz, 2001a, 2001b). If, on the other hand, the emotional stimuli are displayed at a central location (i.e., do not attract attention to any one side) the cost of increased resource consumption is not counter balanced by attraction of attention to the contralesional space. The effect of emotional stimuli on the performance of UN patients is in accord with previous studies of healthy subjects indicating deteriorated performance following exposure to negative stimuli (e.g., Hartikainen et al., 2000; Kitayama, 1991; McKenna, 1986; McKenna & Sharma, 1995; Pereira et al., 2006; Van Strien & Valstar, 2004), yet discordant with other studies that showed improved performance following such stimulation (e.g., Hansen & Hansen, 1988; Simon-Thomas & Knight, 2005; SimonThomas et al., 2005). Raymond (2009) and Yiend (2010) suggested a multitude of factors that could determine the influence of emotion on performance. While settling these discrepancies is beyond the scope of this paper, we note that our findings were obtained in a specific setting, and further research is needed in order to fully understand how emotional stimuli influence spatial

2737

attention in other experimental conditions, both in UN patients and in healthy subjects. Emotions may change across several time scales, from milliseconds in fast presentations (Calvo & Aavero, 2009; Codispoti, Mazzetti, & Bradley, 2009; Kubota et al., 2000) to days and life periods in mood and mood-disorders (Benton & DeMaso, 2010). There are indications that different mechanisms underlie sustained and phasic emotional states (Damasio, 1999). For example, viewing emotional pictures had independent transient and sustained effects on visuomotor performance of healthy subjects (Pereira et al., 2006). There are evidences that performance of UN patients in tests used for diagnosis of the syndrome and assessment of its severity is characterized by marked variation seen on different time scales—from trial to trial during the testing session (Anderson et al., 2000; Bartolomeo et al., 2001); to changes in performance on repeated tests given during the day (Small & Ellis, 1994) or across weeks or months (Hamilton et al., 2008; Jehkonen et al., 2007). Here, we sought to find whether emotional processing and behavioral fluctuations are related in the short time scale. Specifically, we aimed to elucidate whether shortterm emotional processing affect the performance instability revealed from trial to trial within a testing session. Indeed, we found that exposure to centrally-presented stimuli associated with negative affect causes a transient increase in the rightward attentional bias immediately afterwards, with elongation of reaction time to left-sided stimuli, coupled with a significant increase in performance variability (system instability) in that side.

Acknowledgments The authors wish to thank both Prof. Ruma Falk and Assaf Breska from the Hebrew University of Jerusalem for their help in establishing the statistical methods of analyzing the RT variance, and Dr. Ricardo Tarrasch from Tel-Aviv University for useful advices. None of the authors have any conflict of interest.

Appendix A. Supplementary materials Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.neuropsychologia. 2013.09.033. References Adolphs, R., Damasio, H., Tranel, D., & Damasio, A. R. (1996). Cortical systems for the recognition of emotion in facial expressions. Journal of Neuroscience, 16(23), 7678–7687. Anderson, B., Mennemeier, M., & Chatterjee, A. (2000). Variability not ability: Another basis for performance decrements in neglect. Neuropsychologia, 38, 785–796. Anderson, M. J., & Ter Braak, C. J. F. (2003). Permutation tests for multi-factorial analysis of variance. Journal of Statistical Computation and Simulation, 73(2), 85–113. Backs, R. W., da Silva, S. P., & Han, K. (2005). A comparison of younger and older adults’ Self-Assessment Manikin ratings of affective pictures. Experimental Aging Research, 31(4), 421–440. 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. Bartolomeo, P., Siéroff, E., Chokron, S., & Decaix, C. (2001). Variability of response times as a marker of diverted attention. Neuropsychologia, 39, 358–363. Benton, T. D., & DeMaso, D. R. (2010). Mood disorders. In: R. J. Shaw, & D. R. DeMaso (Eds.), Textbook of pediatric psychosomatic medicine (pp. 77–100). Arlington, VA, US: American Psychiatric Publishing, Inc. Borod, J. C. (1992). Interhemispheric and intrahemispheric control of emotion: A focus on unilateral brain damage.pdf. Journal of Consulting and Clinical Psychology, 60, 339–348. Borod, J. C., Bloom, R. L., Brickman, A. M., Nakhutina, L., & Curko, E. a. (2002). Emotional processing deficits in individuals with unilateral brain damage. Applied Neuropsychology, 9, 23–36.

2738

N. Oren et al. / Neuropsychologia 51 (2013) 2729–2739

Bradley, B. P., Mogg, K., Millar, N., BonhamCarter, C., Fergusson, E., Jenkins, J., et al. (1997). Attentional biases for emotional faces. Cognition & Emotion, 11(1), 25–42. Bradley, M. M., Lang, P. J., & Cuthbert, B. N. (1993). Emotion, novelty, and the startle reflex—Habituation in humans. Behavioral Neuroscience, 107(6), 970–980. Calvo, M. G., & Aavero, P. (2009). Reaction time normative data for the IAPS as a function of display time, gender, and picture content. Behavior Research Methods, 41(1), 184–191. Codispoti, M., Mazzetti, M., & Bradley, M. M. (2009). Unmasking emotion: Exposure duration and emotional engagement. Psychophysiology, 46(4), 731–738. Cooper, R. M., & Langton, S. R. H. (2006). Attentional bias to angry faces using the dot-probe task? It depends when you look for it. Behaviour Research and Therapy, 44(9), 1321–1329. Corbetta, M., Patel, G., & Shulman, G. L. (2008). The reorienting system of the human brain: From environment to theory of mind. Neuron, 58(3), 306–324. Corbetta, M., & Shulman, G. L. (2002). Control of goal-directed and stimulus-driven attention in the brain. Nature Reviews Neuroscience, 3(3), 201–215. Corbetta, M., & Shulman, G. L. (2011). Spatial neglect and attention networks. In: S. E. Hyman, T. M. Jessell, C. J. Shatz, C. F. Stevens, & H. Y. Zoghbi (Eds.), Annual Review of Neuroscience, 34. Damasio, A. R. (1999). The feeling of what happens: Body and emotion in the making of consciousness. New York: Harcourt Brace & Company. Davidson, R. J. (1984). Affect, cognition and hemispheric specialization. In: C. E. Izard, J. Kagan, & R. Zajonc (Eds.), Emotions, cognition and behavior (pp. 320– 365). Cambridge: Cambridge Press. Davidson, R. J. (1995). Cerebral asymmetry, emotion and affective style. In: R. J. Davidson, & K. Hugdahl (Eds.), Brain asymmetry (pp. 361–387). Cambridge: MIT Press. Davidson, R. J., & Irwin, W. (1999). The functional neuroanatomy of emotion and affective style. Trends in Cognitive Sciences, 3, 11–20. Davis, J. W., Rahman, M. A., Smith, L. J., Burns, A., Senecal, L., McArthur, D., et al. (1995). Properties of human affect induced by static color slides (IAPS): Dimensional, categorical and electromyographic analysis. Biological Psychology, 41(3), 229. de Zeeuw, P., Aarnoudse-Moens, C., Bijlhout, J., Konig, C., Uiterweer, A. P., Papanikolau, A., et al. (2008). Inhibitory performance, response speed, intraindividual variability, and response accuracy in ADHD. [Article]. Journal of the American Academy of Child and Adolescent Psychiatry, 47(7), 808–816. Deouell, L. Y., Sacher, Y., & Soroker, N. (2005). Assessment of spatial attention after brain damage with a dynamic reaction time test. Journal of the International Neuropsychological Society: JINS, 11, 697–707. Dixon, W. J., & Massey, F. J. (1969). Introduction to statistical analysis (3rd ed.). Tokyo: McGraw-Hill. Dolan, R. J. (2002). Emotion, cognition, and behavior. Science (New York, N.Y.), 298, 1191–1194. Dominguez-Borras, J., Saj, A., Armony, J. L., & Vuilleumier, P. (2012). Emotional processing and its impact on unilateral neglect and extinction. Neuropsychologia, 50(6), 1054–1071. Duchek, J. M., Balota, D. A., Tse, C.-S., Holtzman, D. M., Fagan, A. M., & Goate, A. M. (2009). The utility of intraindividual variability in selective attention tasks as an early marker for Alzheimer’s disease. Neuropsychology, 23(6), 746–758. Eastwood, J. D., Smilek, D., & Merikle, P. M. (2001). Differential attentional guidance by unattended faces expressing positive and negative emotion. Perception & Psychophysics, 63(6), 1004–1013. Eglin, M., Robertson, L. C., & Knight, R. T. (1989). Visual Search Performance in the Neglect. Journal of Cognitive Neuroscience, 1, 372–385. Epstein, J. N., Langberg, J. M., Rosen, P. J., Graham, A., Narad, M. E., Antonini, T. N., et al. (2011). Evidence for higher reaction time variability for children with ADHD on a range of cognitive tasks including reward and event rate manipulations. Neuropsychology, 25(4), 427–441. Forster, J., Liberman, N., & Shapira, O. (2009). Preparing for novel versus familiar events: Shifts in global and local processing. Journal of Experimental Psychology —General, 138(3), 383–399. Fox, E. (2002). Processing emotional facial expressions: The role of anxiety and awareness. Cognitive, Affective & Behavioral Neuroscience, 2, 52–63. Fox, E., Lester, V., Russo, R., Bowles, R. J., Pichler, A., & Dutton, K. (2000). Facial expressions of emotion: Are angry faces detected more efficiently? Cognition and Emotion, 14(1), 61–92. 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. Fox, E., Russo, R., & Dutton, K. (2002). Attentional bias for threat: Evidence for delayed disengagement from emotional faces. Cognition & Emotion, 16(3), 355–379. Gavazzeni, J., Wiens, S., & Fischer, H. (2008). Age effects to negative arousal differ for self-report and electrodermal activity. Psychophysiology, 45(1), 148–151. Geurts, H. M., Grasman, R. P. P. P., Verté, S., Oosterlaan, J., Roeyers, H., van Kammen, S. M., et al. (2008). Intra-individual variability in ADHD, autism spectrum disorders and Tourette’s syndrome. Neuropsychologia, 46(13), 3030–3041. Grabowska, A., Marchewka, A., Seniów, J., Polanowska, K., Jednoróg, K., Królicki, L., et al. (2011). Emotionally negative stimuli can overcome attentional deficits in patients with visuo-spatial hemineglect. Neuropsychologia, 49(12), 3327–3337. Grandjean, D., Sander, D., Lucas, N., Scherer, K. R., & Vuilleumier, P. (2008). Effects of emotional prosody on auditory extinction for voices in patients with spatial neglect. Neuropsychologia, 46, 487–496. Gruhn, D., & Scheibe, S. (2008). Age-related differences in valence and arousal ratings of pictures from the International Affective Picture System (IAPS): Do ratings become more extreme with age? Behavior Research Methods, 40(2), 512–521.

Hamilton, R. H., Coslett, H. B., Buxbaum, L. J., Whyte, J., & Ferraro, M. K. (2008). Inconsistency of performance on neglect subtype tests following acute right hemisphere stroke. Journal of the International Neuropsychological Society: JINS, 14, 23–32. Hansen, C. H., & Hansen, R. D. (1988). Finding the face in the crowd: An anger superiority effect. Journal of Personality and Social Psychology, 54(6), 917–924. Hartikainen, K. M., Ogawa, K. H., & Knight, R. T. (2000). Transient interference of right hemispheric function due to automatic emotional processing. Neuropsychologia, 38(12), 1576–1580. Hays, W. L., & Winkler, R. L. (1971). Statistics: Probability, inference and decision (2nd ed.). New York: Holt, Rinehart & Winston. He, B. J., Snyder, A. Z., Vincent, J. L., Epstein, A., Shulman, G. L., & Corbetta, M. (2007). Breakdown of functional connectivity in frontoparietal networks underlies behavioral deficits in spatial neglect. Neuron, 53, 905–918. Holmes, A., Green, S., & Vuilleumier, P. (2005). The involvement of distinct visual channels in rapid attention towards fearful facial expressions. Cognition and Emotion, 19(6), 899–922. Huang, Y. X., & Luo, Y. J. (2006). Temporal course of emotional negativity bias: An ERP study. Neuroscience Letters, 398(1–2), 91–96. Husain, M., & Rorden, C. (2003). Non-spatially lateralized mechanisms in hemispatial neglect. Nature Reviews Neuroscience, 4, 26–36. Jehkonen, M., Laihosalo, M., Koivisto, A. M., Dastidar, P., & Ahonen, J. P. (2007). Fluctuation in spontaneous recovery of left visual neglect: A 1-year follow-up. European Neurology, 58(4), 210–214. Juslin, P. N., & Vastfjall, D. (2008). Emotional responses to music: The need to consider underlying mechanisms. [Review]. Behavioral and Brain Sciences, 31(5). Katz, N., Hartman-Maeir, A., Ring, H., & Soroker, N. (1999). Functional disability and rehabilitation outcome in right hemisphere damaged patients with and without unilateral spatial neglect. Archives of Physical Medicine and Rehabilitation, 80, 379–384. Kitayama, S. (1991). Impairment of perception by positive and negative affect. Cognition & Emotion, 5(4), 255–274. Klein, C., Wendling, K., Huettner, P., Ruder, H., & Peper, M. (2006). Intra-subject variability in attention-deficit hyperactivity disorder. Biological Psychiatry, 60 (10), 1088–1097. Koster, E. H. W., Crombez, G., Verschuere, B., & De Houwer, J. (2004). Selective attention to threat in the dot probe paradigm: Differentiating vigilance and difficulty to disengage. Behaviour Research and Therapy, 42(10), 1183–1192. Kubota, Y., Sato, W., Murai, T., Toichi, M., Ikeda, A., & Sengoku, A. (2000). Emotional cognition without awareness after unilateral temporal lobectomy in humans. The Journal of Neuroscience, 20(19), RC97. Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (2005). International Affective Picture System (IAPS): Affective ratings of pictures and instruction manual (pp. Technical Report A-6). Larson, C. L., Ruffalo, D., Nietert, J. Y., & Davidson, R. J. (2005). Stability of emotionmodulated startle during short and long picture presentation. Psychophysiology, 42, 604–610. Levine, S. C., & Levy, J. (1986). Perceptual asymmetry for chimeric faces across the life-span. Brain and Cognition, 5(3), 291–306. Levy, D., Blizzard, R. A., Halligan, P. W., & Stone, S. P. (1995). Fluctuations in visual neglect after stroke. European Neurology, 35(6), 341–343. Lipp, O. V., Derakshan, N., Waters, A. M., & Logies, S. (2004). Snakes and cats in the flower bed: Fast detection is not specific to pictures of fear-relevant animals. Emotion, 4(3), 233–250. Lucas, N., & Vuilleumier, P. (2008). Effects of emotional and non-emotional cues on visual search in neglect patients: Evidence for distinct sources of attentional guidance. Neuropsychologia, 46, 1401–1414. Lupianez, J., Klein, R. M., & Bartolomeo, P. (2006). Inhibition of return: Twenty years after. Cognitive Neuropsychology, 23(7), 1003–1014. Manly, B. F. J. (1997). Randomization, bootstrap and Monte Carlo methods in biology. Chapman & Hall. Mather, M., & Carstensen, L. L. (2003). Aging and attentional biases for emotional faces. Psychological Science, 14(5), 409–415. McKenna, F. (1986). Effects of unattended emotional stimuli on color-naming performance. Current Psychology, 5(1), 3–9. McKenna, F., & Sharma, D. (1995). Intrusive cognitions: An investigation of the emotional Stroop task. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21(6), 1595–1607. Mesulam, M. M. (2002). Functional anatomy of attention and neglect: From neurons to networks. The cognitive and neural bases of spatial neglect (pp. 33–46)New York: Oxford University Press33–46. Mogg, K., McNamara, J., Powys, M., Rawlinson, H., Seiffer, A., & Bradley, B. P. (2000). Selective attention to threat: A test of two cognitive models of anxiety. Cognition & Emotion, 14(3), 375–399. Okon-Singer, H., Tzelgov, J., & Henik, A. (2007). Distinguishing between automaticity and attention in the processing of emotionally significant stimuli. Emotion, 7(1), 147–157. Panksepp, J. (1998). Affective neuroscience: The foundations of human and animal emotions. New York: Oxford University Press. Panksepp, J. (2007). Affective neuroscience and the ancestral sources of human feelings consciousness and cognition (pp. 173–188)London: Academic Press Inc.173–188 Pereira, M. G., Volchan, E., de Souza, G. G. L., Oliveira, L., Campagnoli, R. R., Pinheiro, W. M., et al. (2006). Sustained and transient modulation of performance induced by emotional picture viewing. Emotion, 6(4), 622–634. Pessoa, L. (2005). To what extent are emotional visual stimuli processed without attention and awareness? Current Opinion in Neurobiology, 15(2), 188–196.

N. Oren et al. / Neuropsychologia 51 (2013) 2729–2739

Pessoa, L. (2009). How do emotion and motivation direct executive control? Trends in Cognitive Sciences, 13(4), 160–166. 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), 773–782. Pessoa, L., Kastner, S., & Ungerleider, L. G. (2002). Attentional control of the processing of neutral and emotional stimuli. Cognitive Brain Research, 15(1), 31–45. Posner, M., Walker, J., Friedrich, F., & Rafal, R. (1984). Effects of parietal injury on covert orienting of attention. Journal of Neuroscience, 4, 1863. Pourtois, G., Schettino, A., & Vuilleumier, P. (2013). Brain mechanisms for emotional influences on perception and attention: What is magic and what is not. Biological Psychology, 92(3), 492–512. Raymond, J. (2009). Interactions of attention, emotion and motivation. Attention, 176, 293–308. Reed, G. F., Lynn, F., & Meade, B. D. (2002). Use of coefficient of variation in assessing variability of quantitative assays. Clinical and Diagnostic Laboratory Immunology, 9(6), 1235–1239. Ribeiro, R. L., Pompeia, S., & Bueno, O. F. A. (2005). Comparison of Brazilian and American norms for the International Affective Picture System (IAPS). Revista Brasileira De Psiquiatria, 27(3), 208–215. Robertson, I. H. (1999). Cognitive rehabilitation: Attention and neglect. Trends in Cognitive Sciences, 3, 385–393. Robertson, I. H. (2001). Do we need the “lateral” in unilateral neglect? Spatially nonselective attention deficits in unilateral neglect and their implications for rehabilitation. NeuroImage, 14, S85–S90. Robertson, I. H., Mattingley, J. B., Rorden, C., & Driver, J. (1998). Phasic alerting of neglect patients overcomes their spatial deficit in visual awareness. Nature, 395, 169–172. Rosler, A., Ulrich, C., Billino, J., Sterzer, P., Weidauer, S., Bernhardt, T., et al. (2005). Effects of arousing emotional scenes on the distribution of visuospatial attention: Changes with aging and early subcortical vascular dementia. Journal of the Neurological Sciences, 230, 109–116. Rowe, G., Hirsh, J. B., & Anderson, A. K. (2007). Positive affect increases the breadth of attentional selection. [Article]. Proceedings of the National Academy of Sciences of the United States of America, 104(1), 383–388. Scherer, K. R. (2005). What are emotions? And how can they be measured? Social Science Information Sur Les Sciences Sociales, 44(4), 695–729. Schomaker, J., & Meeter, M. (2012). Novelty enhances visual perception. PloS One, 7, 12. Segalowitz, N., Poulsen, C., & Segalowitz, S. (1999). RT coefficient of variation is differentially sensitive to executive control involvement in an attention switching task. Brain and cognition, 40(1), 255–258. Sherratt, S. (2007). Right brain damage and the verbal expression of emotion: A preliminary investigation. Aphasiology, 21, 320–339. Siegel, S., & Castellan, J. N. (1988). Nonparametric statistics for the behavioral sciences. Singapore: McGraw-Hill. Silvia, P. J. (2005). What is interesting?—Exploring the appraisal structure of interest. Emotion, 5(1), 89–102. Simon-Thomas, E. R., & Knight, R. T. (2005). Affective and cognitive modulation of performance monitoring: Behavioral and ERP evidence. Behavioral Neuroscience, 5, 362–372. Simon-Thomas, E. R., Role, K. O., & Knight, R. T. (2005). Behavioral and electrophysiological evidence of a right hemisphere bias for the influence of negative emotion on higher cognition. Journal of Cognitive Neuroscience, 17, 518–529. Small, M., & Ellis, S. (1994). Brief remission periods in visuospatial neglect— Evidence from long-term follow-up. European Neurology, 34(3), 147–154. Smith, J. C., Low, A., Bradley, M. M., & Lang, P. J. (2006). Rapid picture presentation and affective engagement. Emotion, 6(2), 208–214. Soto, D., Funes, M. J., Guzmán-García, A., Warbrick, T., Rotshtein, P., & Humphreys, G. W. (2009). Pleasant music overcomes the loss of awareness in patients with visual neglect. Proceedings of the National Academy of Sciences of the United States of America, 106, 6011–6016.

2739

Tamietto, M., Geminiani, G., Genero, R., & de Gelder, B. (2007). Seeing fearful body language overcomes attentional deficits in patients with neglect. Journal of Cognitive Neuroscience, 19, 445–454. Tamietto, M., Latini Corazzini, L., Pia, L., Zettin, M., Gionco, M., & Geminiani, G. (2005). Effects of emotional face cueing on line bisection in neglect: A single case study. Neurocase: Case Studies in Neuropsychology, Neuropsychiatry, and Behavioural Neurology, 11, 399–404. Thimm, M., Fink, G. R., Küst, J., Karbe, H., & Sturm, W. (2006). Impact of alertness training on spatial neglect: A behavioural and fMRI study. Neuropsychologia, 44, 1230–1246. Tsuchiya, N., & Adolphs, R. (2007). Emotion and consciousness. Trends in Cognitive Sciences, 11, 158–167. Van Strien, J. W., & Valstar, L. H. (2004). The lateralized emotional Stroop task: Left visual field interference in women. Emotion, 4(4), 403–409. Van Vleet, T. M., & Robertson, L. C. (2006). Cross-modal interactions in time and space: Auditory influence on visual attention in hemispatial neglect. Journal of Cognitive Neuroscience, 18, 1368–1379. Verschuere, B., Crombez, G., & Koster, E. (2001). The international affective picture system: A Flemish validation study. Psychologica Belgica, 41(4), 205–217. Vuilleumier, P. (2005). How brains beware: Neural mechanisms of emotional attention. Trends in Cognitive Sciences, 9, 585–594. Vuilleumier, P., Armony, J. L., Clarke, K., Husain, M., Driver, J., & Dolan, R. J. (2002). Neural response to emotional faces with and without awareness: Event-related fMRI in a parietal patient with visual extinction and spatial neglect. Neuropsychologia, 40, 2156–2166. Vuilleumier, P., & Driver, J. (2007). Modulation of visual processing by attention and emotion: Windows on causal interactions between human brain regions. Philosophical Transactions of the Royal Society of London Series B, Biological Sciences, 362, 837–855. Vuilleumier, P., & Schwartz, S. (2001a). Beware and be aware: Capture of spatial attention by fear-related stimuli in neglect. Neuroreport, 12, 1119. Vuilleumier, P., & Schwartz, S. (2001b). Emotional facial expressions capture attention. Neurology, 56, 153. Wagenmakers, E.-J., & Brown, S. (2007). On the linear relation between the mean and the standard deviation of a response time distribution. Psychological Review, 114(3), 830–841. Wager, T. D., Phan, K. L., Liberzon, I., & Taylor, S. F. (2003). Valence, gender, and lateralization of functional brain anatomy in emotion: A meta-analysis of findings from neuroimaging. NeuroImage, 19, 513–531. Weintraub, S., & Mesulam, M. M. (1985). Mental state assessment of young and elderly adults in behavioral neurology. In: M. M. Mesulam (Ed.), Principles of behavioral neurology (1st ed.). Philadelphia: FA Davis. 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. Williams, M. a., & Mattingley, J. B. (2004). Unconscious perception of nonthreatening facial emotion in parietal extinction. Experimental Brain Research, 154, 403–406. Wilson, B., Cockburn, J., & Halligan, P. (1987a). Behavioral inattention test. Fareham, Hants, United Kingdom: Thames Valley Test Company. Wilson, B., Cockburn, J., & Halligan, P. (1987b). Development of a behavioral test of visuospatial neglect. Archives of Physical Medicine and Rehabilitation, 86(4), 98–102. Wilson, E., & MacLeod, C. (2003). Contrasting two accounts of anxiety-linked attentional bias: Selective attention to varying levels of stimulus threat intensity. Journal of Abnormal Psychology, 112(2), 212–218. Yiend, J. (2010). The effects of emotion on attention: A review of attentional processing of emotional information. In: J. De Houwer, & D. Hermans (Eds.), Cognition and emotion: Reviews of current research and theories (pp. 211–275). New York, NY, US: Psychology Press.