Error-related negativity predicts academic performance
JACOB B. HIRSH and MICHAEL INZLICHT
Department of Psychology, University of Toronto, Toronto, Ontario, Canada
Activity in the anterior cingulate cortex (ACC) has been linked to the processes of error detection and conflict
monitoring, along with the subsequent engagement of cognitive-control mechanisms. The error-related negativity
(ERN)isanelectrophysiologicalsignalassociatedwiththisACC monitoringprocess, occurringapproximately100 ms
afteranerrorismade. Thecurrentstudy examinedthepossibility thatindividualdifferencesinERN magnitudewould
predictperformanceoutcomesrelatedto cognitive control. Undergraduatestudents completedacolor-naming Stroop
task while their neural activity was recorded via electroencephalogram. Results indicated that a larger ERN following
errors was significantly correlated with better academic performance as measured by official student transcripts. A
with improved real-world performance.
Descriptors: Anterior cingulate cortex, Error-related negativity, Electrophysiology, Cognitive control, Academic
The anterior cingulate cortex (ACC) has been implicated in per-
formance monitoringandconflictdetectioninhumans (Carter et
al., 1998; MacDonald, Cohen, Stenger, & Carter, 2000; Rid-
derinkhof, Ullsperger, Crone, & Nieuwenhuis, 2004). During
situations of high response conflict, activity in the ACC engages
subsequentlyimproving behavioral regulation
Braver, Barch, Carter, & Cohen, 2001; Gehring & Knight,
2000; Kerns et al., 2004). An electrophysiological signal associ-
ated with this process is the error-related negativity (ERN), a
negative polarity wave peaking approximately 100 ms after
making an error (Gehring & Willoughby, 2002; Holroyd &
Coles, 2002; Yeung, Botvinick, & Cohen, 2004).
Recent analyses using twin data indicate that approximately
47% of the variance in ERN magnitude is due to shared genetic
factors (Anokhin, Golosheykin, & Heath, 2008), suggesting that
the ERN has trait-like properties. From an individual differences
perspective, a larger ERN should reflect improved action
monitoring and more effective engagement of top-down cogni-
tive-control mechanisms in the PFC (Hester, Fassbender, &
Garavan, 2004; Pailing, Segalowitz, Dywan, & Davies, 2002).
Improved PFC function, in turn, relates to a greater capacity for
control (Kane & Engle, 2002; Miller, 2000), allowing for the top-
down regulation of ongoing behavioral processes. Differences in
anisms when needed, as reflected by the ERN, should theoretically
relate to important self-regulatory differences. Indeed, larger
ERNs correlate with reduced impulsivity (Potts, George, Martin,
& Barratt, 2006) and improved emotional regulation (Compton et
al., 2008). Over time and across a large number of situations,
potentially manifest themselves in a variety of different life out-
comes (Barkley, 2001; Carver & Scheier, 1998; Robinson, 2007).
In the current study, we examined the real-world conse-
quences of effective performance monitoring by correlating the
magnitude of the ERN with undergraduate academic perfor-
mance. It was hypothesized that a greater ability to recruit cog-
nitive-control processes, as reflected in a larger ERN, would be
associated with better academic performance.
Participants were 31 undergraduate students at the University of
Toronto Scarborough (14 female). Right-handed participants
were selected to avoid physiological differences due to brain la-
Participants completed an informed consent form, along with a
demographics questionnaire detailing their age, gender, and the
number of years they had been speaking English. Electrophys-
iological responses were then measured via electroencephalo-
gram (EEG) as participants completed a standard color-naming
Stroop task. Color words were presented in colors that either
matched or conflicted with the semantic meaning of the words.
Grants from the Social Science and Humanities Research Council
and from the Canada Foundation for Innovation supported this re-
Address reprint requests to: Jacob B. Hirsh, Department of Psychol-
ogy, Sidney Smith Hall, 100 St. George St., Toronto, ON M5S 3G3,
Canada. E-mail: firstname.lastname@example.org
Psychophysiology, 47 (2010), 192–196. Wiley Periodicals, Inc. Printed in the USA.
Copyright r 2009 Society for Psychophysiological Research
Participants were instructed to press one of four colored buttons
on a response box that corresponded to the font color of the
stimulus word (red, green, blue, or yellow). Each word appeared
for 200 ms, with a maximum response window of 800 ms. An
inter-trial interval of 1000 ms was used. A practice session pre-
ceded 5 blocks of 48 trials each (32 congruent, 16 incongruent).
Students granted us permission to access their academic tran-
scripts during the consent process. Official transcripts were ob-
tained from the Office of the Registrar, and all identifying
completed at least one year of studies, with a mean of 11 courses
completed (SD56.47). Academic performance was quantified
as the average numeric course grade earned across all courses, as
indicated by official transcripts. This measure of performance
was used in all subsequent analyses. To ensure that this variable
was not confounded by differences in course difficulty, we ex-
amined the relationship between the average difficulty of a stu-
dent’s classes (as reflected in the mean of his or her course
averages) and the academic performance variable described
above. No relationship was found between these variables, in-
dicatingthat differences incourse selection were not significantly
influencing the obtained performance measure.
Electrophysiological processing. EEG was recorded from 32
Ag/AgCl sintered electrodes in a stretch-lycra cap. Vertical eye
movements (VEOG) were monitored via a supra- to sub-orbital
B.V., Enschede, Netherlands) with average-ear references and a
forehead ground. Electrode impedances were kept below 5 kO for
all recordings. EEG was corrected for VEOG artifacts using the
SOBI procedure (Tang, Liu, & Sutherland, 2005). Frequencies be-
shift). The signal was baseline corrected by subtracting the average
voltage occurring 400 to 200 ms pre-response. Movement artifacts
were detected with a ?75 mVand 175 mVthreshold. Correct and
incorrect trials were averaged separately with an epoch from 200
at the frontal midline electrode (Fz).
As predicted by models of self-regulation and cognitive control,
academic performance was correlated with ERN magnitude,
ERN responses, r5 ?.40, po.05 (see Figure 1 for scatter plot).
This correlation was significant despite statistically controlling
for the effects of participants’ gender, age, and experience with
the English language. A bootstrapped correlation analysis using
10,000 samples computed a 95% confidence interval ranging
from ?0.13 to ?0.60 (SE50.12), indicating thatthe effectwas
(r5 ?.37) or using the mean rather than minimum voltage be-
different methods of calculating the ERN.
Post-error slowing was used as a behavioral indicator of cog-
nitive control, calculated by subtracting each participant’s aver-
age reaction time on post-correct trials (M5539 ms, SD556
ms) from the average reaction time on post-error trials (M5556
ms, SD568 ms) – the average reaction time across all trials was
540 ms (SD556 ms). As expected, reaction times following er-
rors and correct trials were significantly different from each
other, t(30)52.46, po.05. In keeping with cognitive models of
error slowing, r5 ?.47, po.05, theoretically reflecting the en-
gagement of cognitive-control processes in the PFC following
errors (Gehring & Knight, 2000). This post-error slowing was in
turn associated with higher grades, r5.42, po.05. A mediation
analysis using the product of coefficients method recommended
in MacKinnon, Lockwood, Hoffman, West, and Sheets (2002)
confirmed that post-error slowing mediated the relationship be-
tween ERN size and academic performance (z051.68, po.01).
Overall accuracy rates for the Stroop task ranged from 81% to
98% (M591.7%, SD54.3%), but these rates did not signifi-
cantly differ between post-error (M592.4%, SD58.2%) and
post-correct (M591.6%, SD54.5%) trials. No relationship
was found between Stroop error rates and the ERN or GPA.
Table 1 displays a summary of the obtained correlations.
Headmaps of the ERN revealed the expected frontocentral
spatial distribution (see Figure 2). A single equivalent current
dipole model of the post-error ERPs identified a dorsal ACC
source (PAN coordinates [mm]: x512.7, y5 ?3.3, z537.2;
dipole strength554.74 nAm), accounting for 98.4% of the sig-
nalvariance. Although EEG lacksspatialprecision, theobtained
function (as opposed to the more rostral, emotional aspects;
Bush, Luu, & Posner, 2000).
Academic performance is a gateway to many important life out-
comes, influencing the career options that are available to a stu-
dent. At the broader societal level, achievement in academic
domains plays a vital role in sustaining cultural and scientific
innovation. The current study suggests that individuals who are
better able to monitor their performance and engage cognitive-
control mechanisms when needed enjoy greater success in
undergraduate programs. It further suggests that the ERN can
potentially be employed as a neural marker of this ability.
ERN & academic performance 193
Figure 1. Scatter plotofstudentgrades and ERN magnitude. Bothscales
reflect z-scores of the variables.
The current results are in keeping with previous research link-
ing the ERN to self-regulatory processes. For instance, a larger
ERN has been associated with better stress regulation (Compton
et al., 2008), reduced impulsivity (Potts et al., 2006; Ruchsow,
Spitzer, Gro ¨ n, Grothe, & Kiefer, 2005), and a lower incidence of
externalizing disorders (Hall, Bernat, & Patrick, 2007). Given
that the ERN appears related to the engagement of self-regula-
tory control systems, its relationship with academic performance
may come as no surprise. Indeed, self-regulatory processes are
goals (Covington, 2000; Pintrich & De Groot, 1990).
substantially heritable, approximately half of the variance is ac-
counted for by environmental factors (Anokhin et al., 2008).
Thus, while ERN sizeisrelatedtoacademicperformance, itdoes
not necessarily reflect an immutable cognitive ability. It is cer-
tainly possible that the engagement of cognitive-control mech-
anisms associated with the ERN can be improved through
training, something that should be explored in future research.
On a related note, it is not yet clear how the ERN is related to
general mental ability, which is one of the most effective predictors
of academic performance (Higgins, Peterson, Lee, & Pihl, 2007;
Neisser et al., 1996). One possibility is that some of the variance
captured by the ERN overlaps with standard tests of intelligence,
potentially explaining the observed association with academic per-
formance. Contrary to this explanation, however, is the fact that
performance on intelligence tests is related more to lateral PFC
activity (Duncan et al., 2000; Gray, Chabris, & Braver, 2003),
whereasthe ERN isassociatedwith ACC activity (Botvinicketal.,
2001; Kerns et al., 2004). Although these brain regions interact,
they appear to support two distinct aspects of cognitive control,
with the ACC serving an evaluative function which can then signal
the need for strategic, executive processes in the lateral PFC.
executive components of cognitive control is that these systems
might interact to predict performance outcomes. It is possible,
for instance, that intelligence would moderate the relationship
between the ERN and academic performance, such that engag-
ing the cognitive-control systems of the lateral PFC would only
lead to performance improvements when enough cognitive re-
sources were available for deployment. An error-monitoring
system would be unlikely to predict performance outcomes if
there were insufficient cognitive resources for making behavioral
corrections after error detection.
An alternative interpretation of the current results is that the
ERN may be reflecting increased motivation to perform well on
194 J. B. Hirsh & M. Inzlicht
Figure 2. Therelationbetweenacademicsuccessandthe ERN.(A) Event-relatedpotentialsatFzonerrortrialsforindividualswithhighandlowGrade
Point Averages, as derived from a tertiary split of the sample. (B) Spatial distribution of the ERN, quantified as the peak minimum voltage deflection
occurring between 50 and 150 ms after an error. (C) Headmap of correlations between GPA and ERN magnitude. (D) Source localization indicates an
anterior cingulate generator for the ERN.
Table 1. Correlations Between ERN, Grades, Post-Error Slowing,
and Stroop Errors
npo.05, two-tailed. PES5Post-Error Slowing; Errors5Number of
errors on Stroop task.
the task, rather than the capacity to engage cognitive-control
the motivational salience of a task leads to larger ERN magni-
tudes (Hajcak, Moser,Yeung,& Simons, 2005). Individualswho
displayalarger ERN onthe Strooptaskmayattainbettergrades
more likely to engage cognitive-control mechanisms and adjust
their behavioral responses following errors. This view is sup-
more likely to demonstrate a consistently large ERN, whereas
less conscientious respondents only displayed a large ERN when
additional monetary incentives were provided to motivate accu-
rate performance (Pailing & Segalowitz, 2004). Importantly,
conscientiousness is the best personality predictor of academic
achievement, independent of cognitive ability (Higgins et al.,
2007; Poropat, 2009), and is also related to higher overall levels
of performance motivation (Judge & Ilies, 2002). It is thus pos-
sible that the size of the ERN is reflecting some of the perfor-
mance motivation associated with conscientiousness, which
would explain the observed relationship with academic out-
comes. Future research could expand on these possibilities by
separately examining the contributions of cognitive, motiva-
tional, and personality factors.
It should also be noted that while larger ERN responses were
associated with improved performance outcomes in the popula-
tion examined in the current study, large ERNs are sometimes
associated with dysfunctional behavior. In particular, large
ERNs have been previously associated with anxiety, especially
among clinical populations (Olvet & Hajcak, 2008). It is thus
possible that larger ERNs may not always reflect improved self-
regulation and performance, particularly at the high ends of the
distribution. There may instead be an inverse U-shaped rela-
tionship with performance, much like the classic theory of op-
timum arousal (Yerkes & Dodson, 1908).
this topic, further research is needed to more clearly specify the
relationship between ERN magnitude, behavioral indicators of
cognitive control, and academic performance. In particular,
future studies would benefit from using larger sample sizes. The
sample size used in the current study is comparable to those used
in previous ERN studies, but larger samples would allow
for more detailed statistical analyses, including a closer look at
potential moderating variables that had somewhat restricted
range in this sample (e.g., age and gender). Similarly, while the
obtained confidence intervals for the correlation between the
ERN and academic performance did not approach zero, a larger
sample would help to provide a narrower estimate of the true
effect size, as the obtained intervals span a moderately large
Overall, the current study provides further evidence for the
real-world importance of effective performance monitoring. The
pursuit of academic goals requires the continual self-regulation
of learning, motivation, and cognitive effort. Individual differ-
ences in the extent to which self-regulatory resources can be mo-
bilized in response to errors appear to be an important predictor
of success in this domain.
brain function related to action monitoring. Psychophysiology, 45,
Barkley, R. (2001). The executive functions and self-regulation: An evo-
lutionary neuropsychological perspective. Neuropsychology Review,
Botvinick,M.M.,Braver,T. S.,Barch,D. M.,Carter,C.S.,&Cohen, J.
D. (2001). Conflict monitoring and cognitive control. Psychological
Review, 108, 624–652.
Bush, G., Luu, P., & Posner, M. I. (2000). Cognitive and emotional
influencesinanteriorcingulatecortex. Trendsin Cognitive Sciences, 4,
Carter, C., Braver, T., Barch, D., Botvinick, M., Noll, D., & Cohen, J.
(1998). Anterior cingulate cortex, error detection, and the online
monitoring of performance. Science, 280, 747.
Carver, C., & Scheier, M. (1998). On the Self-Regulation of Behavior.
Cambridge: Cambridge University Press.
Compton, R. J., Robinson, M. D., Ode, S., Quandt, L. C., Fineman, S.
L., & Carp, J. (2008). Error-monitoring ability predicts daily stress
regulation. Psychological Science, 19, 702–708.
Covington, M. (2000). Goal theory, motivation, and school achieve-
ment: An integrative review. Annual Reviews in Psychology, 51, 171–
Duncan, J., Seitz, R., Kolodny, J., Bor, D., Herzog, H., Ahmed, A.,
Gehring, W. J., & Knight,R. T. (2000). Prefrontal-cingulateinteractions
in action monitoring. Nature Neuroscience, 3, 516–520.
Gehring, W. J., & Willoughby, A. R. (2002). The medial frontal cortex
and the rapid processing of monetary gains and losses. Science, 295,
fluid intelligence. Nature Neuroscience, 6, 316–322.
Hajcak,G., Moser,J.,Yeung,N., &Simons,R.(2005).OntheERNand
the significance of errors. Psychophysiology, 42, 151–160.
Hall, J., Bernat, E., & Patrick, C. (2007). Externalizing psychopatho-
logy and the error-related negativity. Psychological Science, 18, 326–
Hester,R., Fassbender, C., & Garavan, H. (2004). Individualdifferences
in error processing: A review and reanalysis of three event-related
fMRI studies using the GO/NOGO task. Cerebral Cortex, 14, 986–
Higgins, D. M., Peterson, J. B., Lee, A. G. M., & Pihl, R. O. (2007).
Prefrontalcognitive ability, intelligence, Big Five personality, andthe
prediction of advanced academic and workplace performance. Jour-
nal of Personality and Social Psychology, 93, 298–319.
Holroyd, C. B., & Coles, M. G. H. (2002). The neural basis of human
error processing: Reinforcement learning, dopamine, and the error-
related negativity. Psychological Review, 109, 679–709.
Judge, T., & Ilies, R. (2002). Relationship of personality to performance
motivation: A meta-analytic review. Journal of Applied Psychology,
Kane, M., & Engle, R. (2002). The role of prefrontal cortex in working-
memory capacity, executive attention, and general fluid intelligence:
An individual-differences perspective. Psychonomic Bulletin & Re-
view, 9, 637–671.
Kerns, J., Cohen, J., MacDonald, A., Cho, R., Stenger, V., & Carter, C.
(2004). Anterior cingulate conflict monitoring and adjustments in
control. Science, 303, 1023–1026.
the role of the dorsolateral prefrontal and anterior cingulate cortex in
cognitive control. Science, 288, 1835.
MacKinnon, D., Lockwood, C., Hoffman, J., West, S., & Sheets, V.
(2002). A comparison of methods to test mediation and other inter-
vening variable effects. Psychological Methods, 7, 83–104.
Miller, E. K. (2000). The prefrontal cortex and cognitive control. Nature
Reviews Neuroscience, 1, 59–65.
Neisser, U., Boodoo, G., Bouchard, T., Boykin, A., Brody, N., Ceci, S.,
et al. (1996). Intelligence: Knowns and unknowns. American Psy-
chologist, 51, 77–101.
ERN & academic performance 195
Olvet, D., & Hajcak, G. (2008). The error-related negativity (ERN) and Download full-text
psychopathology: Toward an endophenotype. Clinical Psychology
Review, 28, 1343–1354.
and trait measure: Motivation, personality, and ERPs in response to
errors. Psychophysiology, 41, 84–95.
Pailing, P., Segalowitz, S., Dywan, J., & Davies, P. (2002). Error neg-
ativity and response control. Psychophysiology, 39, 198–206.
Pintrich, P., & De Groot, E. (1990). Motivational and self-regulated
learning components of classroom academic performance. Journal of
Educational Psychology, 82, 33–40.
Poropat, A. E. (2009). A meta-analysis of the five-factor model of per-
sonality and academic performance. Psychological Bulletin, 135,
Potts, G., George, M., Martin, L., & Barratt, E. (2006). Reduced pun-
ishment sensitivity in neural systems of behavior monitoring in im-
pulsive individuals. Neuroscience Letters, 397, 130–134.
Ridderinkhof, K., Ullsperger, M., Crone, E., & Nieuwenhuis, S. (2004).
Robinson, M. (2007). Gassing, braking, and self-regulating: Error self-
regulation, well-being, and goal-related processes. Journal of Exper-
imental Social Psychology, 43, 1–16.
Ruchsow, M., Spitzer, M., Gro ¨ n, G., Grothe, J., & Kiefer, M. (2005).
Error processing and impulsiveness in normals: Evidence from event-
related potentials. Cognitive Brain Research, 24, 317–325.
Tang, A., Liu, J., & Sutherland, M. (2005). Recovery of correlated neu-
ronal sources from EEG: The good and bad ways of using SOBI.
Neuroimage, 28, 507–519.
Yerkes, R., & Dodson, J. (1908). The relation of strength of stimulus to
rapidity of habit formation. Journal of Comparative Neurology, 18,
Yeung, N., Botvinick, M. M., & Cohen, J. D. (2004). The neuralbasis of
error detection: Conflict monitoring and the error-related negativity.
Psychological Review, 111, 931–959.
(Received February 5, 2009; Accepted March 5, 2009)
196J. B. Hirsh & M. Inzlicht