Neuroscience Letters 584 (2015) 219–223

Contents lists available at ScienceDirect

Neuroscience Letters
journal homepage: www.elsevier.com/locate/neulet

Electrophysiological correlates related to the conflict adaptation effect

in an emotional conflict task
Song Xue a,b,c,1 , Guofang Ren f,1 , Xia Kong b,c , Jia Liu d,e,∗∗ , Jiang Qiu b,c,∗
a
State Key Laboratory of Cognitive Neuroscience and Learning and IDG/McGovern Institute for Brain Research, Beijing Normal University, Beijing, China
b
Key Laboratory of cognition and personality (SWU), Ministry of Education, Chongqing, China
c
School of Psychology, Southwest University (SWU), Chongqing 400715, China
d
School of Psychology, Beijing Normal University, Beijing, China
e
Center for Collaboration and Innovation in Brain and Learning Sciences, Beijing Normal University, Beijing, China
f
Education Department of Anyang Normal University, Xian’ge Road, Anyang City, Henan 455002, China

h i g h l i g h t s

• ERPs related to conflict and conflict adaptation were investigated in an emotional conflict task.
• Behavioral data show an interference and conflict adaptation effect in reaction times.
• The frontal N450 is influenced by both interference and conflict adaptation.
• The parietal conflict SP is influenced by interference, only.

a r t i c l e i n f o a b s t r a c t

Article history: Previous studies have provided some evidence of the neural basis of the emotional conflict adaptation
Received 8 June 2014 effect. However, the neural time-course is largely unknown. Therefore, a face-word Stroop task was used
Received in revised form in the present study to explore the neural dynamics of the emotional conflict control effect, using event-
23 September 2014
related potentials (ERPs). The behavioral data showed a robust emotional conflict adaptation effect, and
Accepted 9 October 2014
there was an interaction between previous trials and current trials for RT. There were two ERP components
Available online 18 October 2014
(N450 and conflict SP) that might be related to trial congruency. The N450 results showed both a main
effect of current trial congruency and an interaction between previous trials and current trials, which
Keywords:
Conflict adaptation effect
might be related to successful conflict adaptation. The SP results only showed the main effect of current
Emotional conflict trial congruency, which might be associated with post-response monitoring.
Stroop task © 2014 Elsevier Ireland Ltd. All rights reserved.
Event-related potentials
ERPs

1. Introduction conflict-induced facilitation of response time (RT) on incongruent

trials that are preceded by incongruent trials during a conflict task
Conflict adaptation is the human ability to adjust executive [5,10]. This phenomenon has been found, not only in several cog-
control flexibly to resolve conflicts that arise in the cognitive con- nitive control paradigms, such as the Stroop task [20], the flanker
trol system [2]. In many conflict control tasks, such as the Stroop task [7], and the Simon task [19], but also as a cognitive index in
and Flanker tasks, participants have longer reaction times and clinical studies [9,13].
lower accuracy on the incongruent trials than on the congruent Conflict monitoring theory proposes that the anterior cingulate
trials, which is called the “interference effect.” On the other hand, cortex (ACC) is responsible for detecting a conflict then transmits
the conflict adaptation effect (CAE), or Gratton effect, refers to a information to the prefrontal cortex (PFC), which resolves the con-
flict [2]. It explains why behavioral adjustment on incongruent
trials preceded by incongruent trials (iI) is faster than adjustment
on incongruent trials preceded by congruent trials (cI) and why
∗ Corresponding author at: School of Psychology, Southwest University, Beibei,
adjustment on incongruent trials preceded by congruent trials (cI)
Chongqing 400715, China. Tel.: +86 23 6836 7942/23 6836 79442.
∗∗ Corresponding author at: School of Psychology, Beijing Normal University, China. is slower than adjustment on congruent trials preceded by con-
E-mail addresses: [email protected] (J. Liu), [email protected] (J. Qiu). gruent trials (cC). Kerns et al. [12] provided fMRI evidence using
1
These authors contributed equally to this work. the color-word Stroop task. Etkin et al. [8], who investigated the

http://dx.doi.org/10.1016/j.neulet.2014.10.019
0304-3940/© 2014 Elsevier Ireland Ltd. All rights reserved.
220 S. Xue et al. / Neuroscience Letters 584 (2015) 219–223

pattern of neural activity during an emotional conflict task, found fearful superimposed on a fearful face picture) or incongruent (e.g.,
the dorsolateral PFC and amygdala were involved in emotional the character for fearful superimposed on a happy face picture).
conflict detection and the rostral ACC was involved in conflict The stimuli were presented on a Dell 19-in monitor that was pro-
resolution. Egner et al. [6] directly compared emotional con- grammed by E-Prime 2.0.
flict adaptation with cognitive conflict adaptation using the same Participants performed a modified face-word Stroop task in a
paradigm, and observed that the right dorsolateral PFC was only quiet room with dim lighting. Participants were instructed to iden-
activated during the cognitive conflict task. In addition, Etkin et al. tify the emotions expressed by the target faces while ignoring the
[9] found that patients with generalized anxiety disorder had meaning of the words. They were also instructed to respond as
decreased cognitive control resulting from a problem with the ACC- quickly and accurately as possible by pressing a button correspond-
amygdala connection. ing to “fearful” faces (right index finger) or “happy” faces (right
Although these studies provided brain imaging evidence related middle finger). The stimuli were presented in pseudorandom order,
to emotional conflict adaptation, the electrophysiological corre- in four conditions: congruent trials followed by congruent trials
lates remain largely unknown. To our knowledge, Zhu et al. [24] and (cC), congruent trials followed by incongruent trials (cI), incongru-
Shen et al. [18] are the only studies that have used event-related ent trials followed by congruent trials (iC), and incongruent trials
potentials (ERPs) to investigate emotional conflict control. Unfortu- followed by incongruent trials (iI). There were a total of 240 tri-
nately, they did not investigate the conflict adaptation effect during als, which were divided into four blocks of 60 trials. Each block
emotional control tasks. Therefore, the aim of the present study consisted of an equal amount of cC, cI, iC, Ii trials. To avoid “repeti-
was to investigate the neural time-course of the conflict adaptation tion priming” [16] and “partial repetition” effects [11], there was no
effect during an emotional conflict task. immediate repetition of the same face with different words [6,8].
Several previous ERP studies have examined the neural basis Participants performed a practice block of 24 trials prior to the main
of the conflict adaptation effect on cognitive tasks, including the experiment. The order and timing of each trial was as follows: a fixa-
Stroop and Flanker tasks. For instance, Larson et al. [14] found that tion dot was presented for a duration of 500 ms, followed by a blank
a parietal conflict slow potential (conflict SP) was sensitive to the screen, which was presented for 300–500 ms. Then, the picture of
context of the previous trial, while the fronto-medial N450 only was the target face appeared for 1000 ms in the center of the screen. Par-
sensitive to current trial congruency in the Stroop task. Tang et al. ticipants had to respond within 1500 ms. Three inter-trial intervals
[21] suggested that the right-frontal N450 reflected conflict detec- (ITI) were used (800 ms, 1000 ms, and 1200 ms), which were varied
tion and that both the frontal and parietal conflict SPs reflected randomly; the mean ITI was 1000 ms (see Fig. 1).
adaptation to conflict resolution. In addition, Clayson and Larson
[4] found that N2 amplitudes showed a significant conflict adapta-
tion effect on the Flanker task, and that N2 conflict adaptation was 2.3. ERP recording and analysis
associated with executive function.
Accordingly, coupled with the fMRI findings on emotional con- Brain electrical activity was recorded from 64 scalp sites using
flict adaptation, we predicted that participants would show a tin electrodes mounted in an elastic cap (Brain Products GmbH,
significant behavioral conflict adaptation effect. At the ERP level, Gilching, Germany), with the reference electrodes placed on the
differences between two classical components (N450, conflict SP) left and right mastoids. The vertical electrooculogram (EOG) was
should be observed on incongruent trials and congruent trials. recorded with electrodes placed above and below the left eye.
In view of previous studies that revealed (a) different patterns All inter-electrode impedances were maintained below 5 k.
between emotional conflict control and cognitive conflict control, The electroencephalogram (EEG) and EOG signals were amplified
and (b) that emotion conflict detection and resolution occurred ear- using a 0.05–80 Hz bandpass filter and continuously sampled at
lier than cognitive conflict detection and resolution [18,24], we also 500 Hz/channel for offline analysis. Trials with eye movement arti-
predicted that the N450 would show the emotional conflict adap- facts (eye blinks and eye movements) were rejected offline. Trials
tation effect (i.e., that previous-trial congruency would influence were excluded from averaging that: had EOG artifacts (mean EOG
the current-trial). voltage exceeding ± 80 ␮V); were contaminated with artifacts due
to amplifier clipping, bursts of electromyographic activity; or had
peak-to-peak deflections exceeding ±80 ␮V.
2. Materials and methods
The focus of our analysis was the ERPs elicited by the correct
responses of emotional judgments (the ERPs were time-locked
2.1. Participants
with the onset of the target face picture). The averaged epoch
for ERPs was 900 ms, including 800 ms post-stimulus and 100 ms
Twenty healthy students recruited from Southwest University
pre-stimulus. ERPs were averaged for trials associated with cor-
in China participated in the study. Participants included ten male
rect responses; at least 45 trials were available in the congruent
and ten female students with a mean age of 21.7 years (range
and incongruent conditions for each subject. According to previous
18–25 years). All participants were right-handed, had normal or
studies, the scalp distribution of Stroop-related ERP components
corrected-to-normal vision, and had no history of neurological or
(N450 and SP) include the frontal-medial and posterior parietal
psychiatric illness. All the participants gave written informed con-
sites [14]. Therefore, the following sites were chosen for quan-
sent and were paid at the end of the experiment. The study was
titative analysis: the frontal-medial negative component in the
approved by the Southeast University ethics committee.
350–50 ms time-window was recorded at sites Fz, F3, F4, FCz, FC3,
FC4, Cz, C3 and C4; and the posterior parietal positive component
2.2. Stimuli and procedure in the 700–00 ms time-window was recorded at sites Pz, P1, P2,
POz, PO3 and PO4. The behavioral data were analyzed by two-way
The stimuli consisted of 10 pictures of fearful faces (5 male, repeated measures analysis of variance (ANOVA), with previous-
5 female) and 10 pictures of happy faces (5 male, 5 female) trial congruency (previous congruent vs. previous incongruent) and
selected from the Chinese affective picture system [17]. Two Chi- current trial congruency (current congruent vs. current incongru-
nese characters “ ” (which means “fear”) or “ ” (which means ent) as within factors. The ERP data were analyzed by three-way
“happy”) were superimposed on the faces in red. The words and repeated measures ANOVA, with previous-trial congruency, cur-
facial expressions were either congruent (e.g., the character for rent trial congruency, and electrode sites, as within-group factors.
 S. Xue et al. / Neuroscience Letters 584 (2015) 219–223 221

Fig. 1. Example of the conflict paradigm. Flow-chart showing the time order of trials and an example of trial combinations for conflict adaptation effects.

The p-values of all the analyses were adjusted for deviations using 4. Discussion
to Greenhouse Geisser correction.
The present study explored the patterns of spatiotemporal cor-
tical activation of the conflict adaptation effect during a modified
3. Results face-word Stroop task. The current results revealed a significant
interference effect on accuracy and both interference and conflict
3.1. Behavioral performance adaptation effects on RT. Moreover, the ERP results showed two
components: an N450 over the fronto-central scalp, and a con-
Two-way repeated measures ANOVAs (Previous Trial × Current flict SP over the posterior parietal scalp. The implications of these
Trial) were conducted on the RT and accuracy data. Response accu- findings will be discussed.
racy was similar and high across the four conditions: cC (97%), cI There was a significant interaction, at the behavioral level,
(95%), iC (98%), iI (96%). There was no interaction between previ- between previous trials and current trials, which showed the con-
ous and current trials [F(1,19) = .458, p > .5], but the main effect of flict adaptation effect. Behavior and ERP studies usually take the
current trials was statistically significance [F(1,19) = 6.057, p < .05]. interaction between previous trials and current trials as an index
The interaction between previous and current trials was sta- of the conflict adaptation effect, although fMRI studies often use
tistically significant for RT [F(1,19) = 7.741, p < .05]: mean RT was the difference between iI and cI trials (iI minus cI) to measure
586 ± 46 ms for cC trials; 623 ± 54 ms for cI trials; 597 ± 48 ms for the conflict adaptation effect in imaging data [5]. Therefore, in the
iC trials; and 620 ± 51 ms for iI trials. Post hoc tests revealed that the present study we selected this interaction to measure the conflict
difference between cC and cI trials [F(1,19) = 78.654, p < .001] and adaptation effect.
the difference between iC and iI trials [F(1,19) = 40.025, p < .001] The findings of the present study revealed a negative deflection
were significant. in the 350–550 ms time-window (N450) over the fronto-central
scalp. According to many previous studies, the N450 is an index
of the interference effect on the color-word Stroop task. It is prob-
3.2. Electrophysiological scalp data ably related to conflict monitoring [15,23] and conflict resolution
[17]. For instance, Liotti et al. [15] suggested that the N450 is asso-
There were two ERP components: one in the 350–550 ms ciated with conflict monitoring, whereas Qiu et al. [17] proposed
time-window and one in the 700–800 ms time-window. In the that it is associated with both conflict monitoring and conflict
350–550 ms time-window, the amplitudes for the incongru- resolution. In addition, Shen et al. [18] reported that the N450
ent conditions were larger than for the congruent conditions reflected conflict resolution and response selection in the face-
[F(1,19) = 12.504, p = .002], which yielded a negative ERP deflec- word task. In the present study, the N450 had a classical ERP
tion (N450). Moreover, there was an significant interaction component that was associated with the emotional conflict evoked
effect between previous and current trials on N450 amplitude by the incongruence between facial expression and word mean-
[F(1,19) = 5.197, p < .05]. Post hoc tests revealed that the differ- ing.
ence between the cC and cI trials was significant [F(1,19) = 18.300, Consistent with predictions, N450 amplitude showed an inter-
p < .001 vs. F(1,19) = 2.006, p > .05]. In the 700–800 ms time- action between previous trials and current trials, revealing that
window, the main effects of previous [F(1,19) = 7.823, p < .05] and the N450 was modulated by previous-trial context, and the N450
current trials were significant [F(1,19) = 4.512, p < .05], yet the exhibited a significant conflict adaptation effect. The present find-
interaction between previous and current trials did not reach signif- ings show that the amplitude of the N450 is more attenuated in the
icance [F(1,19) = .932, p > .3]. The main effect of electrode sites was iI trials than in the cI trials, which is in line with the conflict adap-
significant, but there was no other significant main or interaction tation effect. The adaptation leads to faster RTs and less control is
effect (Figs. 2 and 3). needed to resolve the conflict in the iI condition. We think the N450
222 S. Xue et al. / Neuroscience Letters 584 (2015) 219–223

Fig. 2. Grand-average ERP for congruent–congruent, congruent–incongruent, incongruent–congruent, and incongruent–incongruent trials at sites Fz, FCz and Cz. A topo-
graphical map of the differences in the incongruent–congruent waves at 350–550 ms.

is related to successful conflict adaptation in the present task. To a investigate whether the N450 showed the conflict adaptation effect
certain extent, previous studies have failed to resolve the role of the on emotional conflict control. We found that the N450(cI–cC) was
N450 in the conflict adaptation effect. Larson et al. [14] thought that larger than the N450(iI–iC) which was associated with RT(cI–cC) being
the N450 was not differentially affected by previous-trial context, longer than RT(iI–iC) . Previous research found that emotional con-
whereas Tang et al. [21] suggested that the N450 revealed the con- flict detection and conflict resolution occurred at earlier stages
flict adaptation effect in the look-do Stroop task. It is probable that [18,24]. As a consequence, conflict resolution is easier and less
this difference is the result of different experimental paradigms. cognitive control is needed in the iI than in the cI condition, and
In the present study, the face-word Stroop task was used to the neuronal correlate of less cognitive control is observable as an

Fig. 3. Grand-average ERP for congruent–congruent, congruent–incongruent, incongruent–congruent, and incongruent–incongruent trials at sites Pz and POz. A topographical
map of the differences in the incongruent–congruent waves at 700–800 ms.
 S. Xue et al. / Neuroscience Letters 584 (2015) 219–223 223

attenuated N450. Hence, we think the N450 in the present study Laboratory of Cognitive Neuroscience and Learning (CNLYB1310).
may be associated with successful emotional conflict adaptation. The authors declare no competing interests.
Additionally, the SP amplitude in the 700–800 ms time-window
that was evoked by the incongruent condition was greater than that References
evoked by the congruent condition over the posterior parietal scalp.
In line with previous studies, the SP was a late and conflict-sensitive [1] K. Bailey, R. West, C.A. Anderson, A negative association between video game
experience and proactive cognitive control, Psychophysiology 47 (2010) 34–42.
parietal positivity, which has been associated with resolving the [2] M.M. Botvinick, T.S. Braver, D.M. Barch, C.S. Carter, J.D. Cohen, Conflict moni-
conflict and post-response monitoring [1,3,22]. It probably reflects toring and cognitive control, Psychol. Rev. 108 (2001) 624.
a process in which participants in the present study identified the [3] A. Chen, K. Bailey, B.N. Tiernan, R. West, Neural correlates of stimulus and
response interference in a 2–1 mapping stroop task, Int. J. Psychophysiol. 80
facial expression and inhibited the interfering word meaning. Yet, (2011) 129–138.
compared with previous ERP studies [14,21] that investigated the [4] P.E. Clayson, M.J. Larson, Cognitive performance and electrophysiological
conflict adaptation effect, the present study did not find an inter- indices of cognitive control: a validation study of conflict adaptation, Psy-
chophysiology 49 (2012) 627–637.
action between previous trials and current trials. We think this can
[5] T. Egner, Congruency sequence effects and cognitive control, Cogn. Affect.
be attributed to the difference between emotional conflict and cog- Behav. Neurosci. 7 (2007) 380–390.
nitive conflict control. Previous research suggests that emotional [6] T. Egner, A. Etkin, S. Gale, J. Hirsch, Dissociable neural systems resolve con-
conflict control has a different neural activation pattern than cog- flict from emotional versus nonemotional distracters, Cereb. Cortex 18 (2008)
1475–1484.
nitive conflict control. For instance, Zhu et al. [24] found emotional [7] B.A. Eriksen, C.W. Eriksen, Effects of noise letters upon the identification of a
conflict occurred at an earlier stage, whereas Shen et al. [18] found target letter in a nonsearch task, Percept. Psychophys. 16 (1974) 143–149.
that selection response in the emotional conflict also occurred [8] A. Etkin, T. Egner, D.M. Peraza, E.R. Kandel, J. Hirsch, Resolving emotional con-
flict: a role for the rostral anterior cingulate cortex in modulating activity in
earlier than in cognitive conflict. In the present study, the N450 the amygdala, Neuron 51 (2006) 871–882.
indicated that successful conflict adaptation occurred at an ear- [9] A. Etkin, K.E. Prater, F. Hoeft, V. Menon, A.F. Schatzberg, Failure of anterior cin-
lier time. Therefore, we speculate that SP did not show the conflict gulate activation and connectivity with the amygdala during implicit regulation
of emotional processing in generalized anxiety disorder, Am. J. Psychiatry 167
adaptation effect in the present results because of the different cog- (2010) 545–554.
nitive processes involved in emotional conflict tasks and cognitive [10] G. Gratton, M.G. Coles, E. Donchin, Optimizing the use of information: strategic
conflict tasks. control of activation of responses, J. Exp. Psychol. Gen. 121 (1992) 480.
[11] B. Hommel, R.W. Proctor, K.-P.L. Vu, A feature-integration account of sequential
Therefore, the difference neural time-course between emotional effects in the Simon task, Psychol. Res. 68 (2004) 1–17.
conflict adaptation and cognitive conflict adaptation remains an [12] J.G. Kerns, J.D. Cohen, A.W. MacDonald, R.Y. Cho, V.A. Stenger, C.S. Carter, Ante-
open question. We intend to compare them with each other further rior cingulate conflict monitoring and adjustments in control, Science 303
(2004) 1023–1026.
using advanced methods and different experimental designs.
[13] M.K. Krug, C.S. Carter, Adding fear to conflict: a general purpose cognitive con-
trol network is modulated by trait anxiety, Cogn. Affect. Behav. Neurosci. 10
5. Conclusion (2010) 357–371.
[14] M.J. Larson, D.A. Kaufman, W.M. Perlstein, Neural time course of conflict adap-
tation effects on the Stroop task, Neuropsychologia 47 (2009) 663–670.
The incongruent condition in a face-word Stroop task elicited [15] M. Liotti, M.G. Woldorff, R. Perez III, H.S. Mayberg, An ERP study of the tempo-
more negative deflection than the congruent condition in the ral course of the Stroop color-word interference effect, Neuropsychologia 38
350–550 ms time-window (N450) and a more positive deflection in (2000) 701–711.
[16] U. Mayr, E. Awh, P. Laurey, Conflict adaptation effects in the absence of execu-
the 700–800 ms time-window (conflict SP). The N450 also showed tive control, Nat. Neurosci. 6 (2003) 450–452.
an interaction between previous trials and current trials, which [17] J. Qiu, Y. Luo, Q. Wang, F. Zhang, Q. Zhang, Brain mechanism of Stroop interfer-
might be associated with successful conflict adaptation, and the ence effect in Chinese characters, Brain Res. 1072 (2006) 186–193.
[18] Y. Shen, S. Xue, K. Wang, J. Qiu, Neural time course of emotional conflict control:
conflict SP might be related to post-response monitoring. These an ERP study, Neurosci. Lett. 541 (2013) 34–38.
results extend our understanding of the neural basis of emotional [19] J.R. Simon, K. Berbaum, Effect of conflicting cues on information processing: the
conflict adaptation. ‘Stroop effect’ vs. the ‘Simon effect’, Acta Psychol. (Amst.) 73 (1990) 159–170.
[20] J.R. Stroop, Studies of interference in serial verbal reactions, J. Exp. Psychol. 18
(1935) 643.
Acknowledgements [21] D. Tang, L. Hu, H. Li, Q. Zhang, A. Chen, The neural dynamics of conflict adapta-
tion within a look-to-do transition, PLoS ONE 8 (2013) e57912.
[22] R. West, Neural correlates of cognitive control and conflict detection in the
This research was supported by the National Natural Science Stroop and digit-location tasks, Neuropsychologia 41 (2003) 1122–1135.
Foundation of China (31070900; 31271087), the Program for New [23] R. West, R. Bowry, C. McConville, Sensitivity of medial frontal cortex to response
Century Excellent Talents in University (2011) by the Ministry of and nonresponse conflict, Psychophysiology 41 (2004) 739–748.
[24] X.-r. Zhu, H.-j. Zhang, T.-t. Wu, W.-b. Luo, Y.-j. Luo, Emotional conflict occurs at
Education and the Fundamental Research Funds for the Central Uni- an early stage: evidence from the emotional face-word Stroop task, Neurosci.
versities (SWU1209101), the Open Research Fund of the State Key Lett. 478 (2010) 1–4.