doi:10.

1093/scan/nsq027

SCAN (2011) 6, 277^285

Asymmetric frontal cortical activity and negative

affective responses to ostracism
Carly K. Peterson, Laura C. Gravens, and Eddie Harmon-Jones
Department of Psychology, Texas A&M University, 4235 TAMU, College Station, TX 77843, USA
Ostracism arouses negative affect. However, little is known about variables that influence the intensity of these negative affective responses. Two studies fill this void by incorporating work on approach- and withdrawal-related emotional states and their
associated cortical activations. Study 1 found that following ostracism anger related directly to relative left frontal cortical
activation. Study 2 used unilateral hand contractions to manipulate frontal cortical activity prior to an ostracizing event.
Right-hand contractions, compared to left-hand contractions, caused greater relative left frontal cortical activation during the
hand contractions as well as ostracism. Also, right-hand contractions caused more self-reported anger in response to being
ostracized. Within-condition correlations revealed patterns of associations between ostracism-induced frontal asymmetry and
emotive responses to ostracism consistent with Study 1. Taken together, these results suggest that asymmetrical frontal cortical
activity is related to angry responses to ostracism, with greater relative left frontal cortical activity being associated with
increased anger.
Keywords: ostracism; anger; sadness; asymmetrical frontal cortical activity; approach motivation

INTRODUCTION
We often learn much about the power of psychological
variables to influence important outcomes when we strip
the variables down to minimal manipulations (Prentice
and Miller, 1992). Attesting to the pervasiveness of our
liking for our own groups, the minimal group paradigm
revealed that simply assigning individuals to groups on
the bias of random criteria causes individuals to evaluate their own group more positively (Tajfel et al., 1971).
Attesting to the ease with which we can learn to
like things, the mere exposure paradigm revealed that
simply repeatedly presenting individuals with unreinforced
stimuli causes them to like those stimuli more (Zajonc,
1968). More recently, the Cyberball paradigm has revealed
that, when ostracized from a group, individuals will feel
negative emotions (Williams, 2007a), even when they
know the group is fictitious (Zadro et al., 2004). These
results suggest that individuals easily bond with others and
feel powerful negativeeven painfulfeelings when left out
(Eisenberger and Lieberman, 2004; MacDonald and Leary,
2005).
We might imagine, based on our own experiences,
that there are reliable individual differences in how bad
individuals feel when ostracized. However, we would
be wrong: research has been unable to find reliable

Received 15 December 2009; Accepted 22 February 2010

Advance Access publication 1 April 2010
The research described within this article was partially supported by grants from the National Science
Foundation to E.H.-J. (BCS 0350435; BCS 0643348).
Correspondence should be addressed to Carly Peterson, Department of Psychology, Texas A&M University,
4235 TAMU, College Station, TX 77843, USA. E-mail: [email protected].

predictors of the degree of negative affect individuals

feel when ostracized1 (Williams, 2007a). In fact, Williams
(2007b) has even said, we have measured these individual
differences [social anxiety, introversion/extraversion, secure
attachment, self-esteem, loneliness, individualism and agreeableness] and so far found no moderation on the reflexive
self-reports of distress, negative affect or need threat (p. 239;
bracketed information added from other parts of the article).
The present studies sought to address this issue by taking a
social neuroscience approach.
In ostracism research, negative affect usually describes
both anger and sadness, and these distinct emotions are typically combined and referred to as distress (Williams,
2007a, b). When these distinct emotions have been examined
separately following ostracism, anger is more affected than
sadness by the ostracism manipulation (Chow et al., 2008).
This ostracism-induced negative affect link has been shown
to be unaffected by a number of situational and individual
difference variables (see Williams, 2007a). Understanding
variables that influence the intensity of negative affective
responses to ostracism is important for predicting emotional
consequences of ostracism.
Subjective emotional responses associated with approach
vs withdrawal motivation are influenced by asymmetric
frontal cortical activity (Harmon-Jones, 2003; Schutter
et al., 2008; Carver and Harmon-Jones, 2009). Relative left
frontal cortical activity relates to approach-oriented
1
One individual difference that may relate to affective responses to ostracism is rejection sensitivity (Downey
and Feldman, 1996). However, most research on rejection sensitivity and affective responses is correlational.
Buckley et al. (2004) conducted an experiment that presented acceptance vs rejection feedback but found that
individuals high in rejection sensitivity reported more negative affect regardless of the acceptance-rejection
manipulation.

The Author (2010). Published by Oxford University Press. For Permissions, please email: [email protected]

278

SCAN (2011)

emotions such as anger (e.g. Harmon-Jones and Sigelman,

2001; van Honk and Schutter, 2006; Verona et al., 2009) and
desire (Harmon-Jones and Gable, 2009; Harmon-Jones et al.,
2008), whereas relative right frontal cortical activity relates to
withdrawal-oriented emotions such as sadness (Jones and
Fox, 1992; Jacobs and Snyder, 1996; Schmidt and Trainor,
2001), fear and disgust (Coan et al., 2001).
Study 1 was designed to test whether the pattern of
asymmetric frontal cortical activity to ostracism would
assist in predicting the degree and type of negative emotional
experience ostracism evokes. Based on past research, we
would predict that anger should be associated with
greater relative left frontal activation, whereas sadness
should be associated with greater relative right frontal
activation. This prediction would lead us to expect ostracism to not evoke greater relative left or greater relative
right frontal cortical activity (i.e. no main effect of ostracism on left vs right frontal activity), given that ostracism can evoke varying degrees of these negative affects
in individuals (e.g. Williams, 2007a). That is, while one individual may respond to ostracism with anger and thus
greater relative left frontal cortical activity, another may
respond to the same event with sadness and display the
opposite pattern of frontal asymmetry. Instead, our prediction would suggest that the pattern of asymmetric frontal activity would predict the subjective emotional
response. Consequently, this would benefit understanding
of reflexive emotional responses to ostracism by providing
a pattern of neural activation that predicts the extent
to which an individual feels distinct negative emotions
to being ostracized.
In Study 2, frontal asymmetry was manipulated via unilateral hand contractions, to assess whether asymmetric
frontal cortical activations were causally involved in the
emotional experiences associated with ostracism. This manipulation was predicated on past research that suggested
that unilateral hand contractions activate the contralateral
motor cortex and these motor cortex activations spread to
the dorsolateral prefrontal cortex and prime approach or
withdrawal motivational processes (Schiff and Lamon,
1994; Schiff et al., 1998; Harmon-Jones, 2006; Peterson,
et al., 2008). Based on this past research, it was predicted
that in Study 2 unilateral hand contractions would show the
same pattern of contralateral activation in the central, frontalcentral and frontal regions of the brain. Furthermore,
right-hand contractions should be associated with an increase in self-reported anger to ostracism, whereas left-hand
contraction should be associated with an increase in
self-reported sadness to ostracism. Finally, we expect to
find the same relationships as Study 1 between frontal cortical activity during ostracism and self-reported emotions.
Together, both studies aim to shed light on the emotional
consequences of ostracism, specifically in terms of the effect
of asymmetrical frontal cortical activity on the intensity and
type of responses.

C. K. Peterson et al.
STUDY 1
Method
Participants and procedure
Forty (20 male) right-handed introductory psychology students participated in exchange for partial course credit.
Instructions for Cyberball were presented on the computer
monitor. All participants were instructed to practice their
mental visualization skills during the game, and to pretend
as if they were playing the game in real life. All participants
were aware that the other players did not actually exist, as in
Zadro et al. (2004). The game was programmed in one of
two ways, and condition assignment was determined randomly. In the ostracism condition, participants were
included the first part of the game (approximately eight
throws) and then ostracized during the second half (approximately 16 throws). In the inclusion condition, participants
were included during the entire game. This was done because
research has shown that this type of inclusion does not affect
mood and thus is an appropriate control condition (Gerber
and Wheeler, 2009). EEG was recorded during the task.
When the game was over ( 4 min later), participants
completed a questionnaire assessing their perceived level of
anger, enjoyment and the four fundamental needs (belonging, control, meaningful existence and self-esteem; Zadro
et al., 2004). Responses were made on a 9-point scale
(1 not at all, 9 very much so). A manipulation check
was included to assess participants perceived level of inclusion during the game (i.e. what percent of the throws were
thrown to you? Circle your best guess with possible answers
ranging from 0% to 100% in 10% intervals). In this study,
participants also reported their sadness (sad, gloomy, down,
discouraged; Cronbachs  0.76), distress (distress, disgust,
afraid, nervous; Cronbachs  0.56) and positive affect
(happy, good mood, satisfied, glad, content, eager, excited,
interested; Cronbachs  0.87). All responses were made on
9-point scales (1 not at all, 9 very much so).
Data collection and reduction
EEG, rereferenced globally to the whole head from the left
ear, was recorded from 59 tin electrodes mounted in a
stretch-lycra electrode cap (Electro-Cap International,
Eaton, OH). Impedances were under 5000 V; homologous
sites were within 1000 V of each other. Signals were amplified (60-Hz notch filter), bandpass filtered (0.05100 Hz)
and digitized at 500 Hz. Signals were manually scored for
artefacts. Then, a regression-based eye movement correction
was applied (Semlitsch et al., 1986). All 1.024-s epochs were
extracted through a Hamming window. A fast Fourier transform extracted power within the lower alpha (810.25 Hz)
and upper alpha (10.2512.50 Hz) bands. Preliminary analyses revealed that effects were found in the upper alpha
band, and thus statistical analyses focus there. Power was
averaged across epochs during the two parts of the game.
Asymmetry indices were created for homologous sites (natural log right minus natural log left). Because alpha power is

Ostracism, asymmetric frontal cortical activity, and negative affect

SCAN (2011)

279

Table 1 Between-condition means (s.d.) of cyberball variables in Studies 1 and 2

Study 1

Percent included
Included
Belonging
Control
Self-esteem
Meaningful existence
Enjoyment
Anger
Sadness
Distress
Positive mood
Mid-frontal asymm
Lat-frontal asymm
Frontal-central asymm

Study 2

Inclusion

Ostracism

Left hand

Right hand

44.50
6.26
5.93
5.58
1.42
6.78
4.60
1.75
1.56
1.53
3.82
0.03
0.04
0.01

23.00
2.80
3.02
2.80
1.15
3.85
2.10
3.30
2.05
1.71
2.85
0.00
0.04
0.01

6.88***
9.22***
7.38***
6.31***
0.76
5.20***
3.78***
2.88**
1.65*
0.80
2.64***
0.39
0.04
0.26

26.43
3.29
3.24
2.86
1.50
4.29
3.36
2.43
2.63
1.73
2.66
0.20
0.29
0.07

20.00
2.50
3.06
2.92
0.97
3.58
3.25
4.17
2.65
1.81
2.62
0.05
0.04
0.05

1.66
1.63
0.39
0.15
1.51
1.07
0.14
2.04**
0.04
0.24
0.11
0.97
1.95**
0.13

(10.50)
(1.15)
(0.98)
(1.58)
(1.17)
(1.87)
(2.68)
(1.16)
(0.76)
(0.71)
(1.14)
(0.33)
(0.34)
(0.35)

(9.23)
(1.20)
(1.47)
(1.18)
(1.06)
(1.69)
(1.25)
(2.11)
(0.98)
(0.65)
(0.94)
(0.17)
(0.26)
(0.27)

(8.42)
(1.44)
(1.07)
(0.79)
(0.89)
(1.96)
(1.86)
(1.70)
(1.37)
(0.76)
(0.90)
(0.42)
(0.52)
(0.58)

(11.28)
(0.90)
(1.33)
(1.26)
(0.88)
(1.25)
(2.01)
(2.62)
(1.60)
(0.93)
(1.08)
(0.30)
(0.31)
(0.39)

Note: ***P < 0.01; **P < 0.05; *P < 0.06.

inversely related to cortical activity, higher scores indicate

greater left than right activity (Davidson et al., 2000). The
primary variable of interest was created by subtracting the
asymmetry index from the first half of the game (an inclusion period of all participants) from the asymmetry index
from the second half of the game (an ostracism period for
half of the participants and an inclusion period for the other
half of the participants). Higher scores indicated greater relative left frontal activity during the second half of the game.
Due to our interest in asymmetrical frontal cortical activity,
a mid-frontal (F1/2, F3/4), lateral frontal (F5/6, F7/8) and
frontalcentral (FC1/2, FC3/4 and FC5/6) asymmetry index
was created by combining the two mid-frontal electrodes,
the two lateral frontal electrodes and the three frontal central
electrodes. Degrees of freedom differ for some analyses because some participants did not complete all self-report
measures. Because predictions for all studies were directional, derived from theory and specified in advance, they were
evaluated using a one-tailed criterion of significance
(Rosenthal et al., 2000).

Results and discussion

Regression analyses in which condition (included vs ostracized, effect coded), frontal asymmetry and their interaction
predicted subjective anger response were used to test predictions. Mid-frontal asymmetry (F (1,36) 10.43, P < 0.01),
condition (F (1,36) 11.36, P < 0.01) and their interaction
(F (1,36) 13.39, P < 0.001) predicted anger. To follow up
the interaction, within-condition correlations between anger
and mid-frontal asymmetry were examined. As predicted,
within the ostracism condition, anger was significantly
related to relatively greater left mid-frontal activation from
the period of inclusion to the period of ostracism (r 0.62,
P < 0.01). Within the inclusion condition, anger did not

relate to asymmetrical mid-frontal activation (r < 0.14,

P > 0.56).
Similar effects emerged for frontalcentral asymmetry.
That is, frontalcentral asymmetry (F (1,36) 8.13,
P < 0.01), condition (F (1,36) 9.09, P < 0.01) and their
interaction marginally predicted anger (F (1,36) 3.40,
P 0.07). Within the ostracism condition, anger was significantly related to greater left frontalcentral activation
(r 0.50, P < 0.05). Within the inclusion condition, anger
did not relate to left frontalcentral activation (r < 0.26,
P > 0.27). No other cortical asymmetries interacted with condition to predict anger.2 See Figure 1 for a topographic map
displaying correlations between frontal asymmetry and anger
within the ostracism condition.
The same type of interaction regression analyses were performed for self-reported sadness, distress, meaningful existence, control, enjoyment, positive mood, self-esteem and
belonging. Only control produced a significant interaction
with mid-frontal (F (1, 36) 7.57, P < 0.01) and frontalcentral (F (1, 36) 8.08, P < 0.01) asymmetries. Withincondition correlations revealed that during ostracism,
control correlated inversely with mid-frontal (r 0.39,
P 0.09) and frontalcentral (r 0.53, P < 0.05) asymmetries, but during inclusion, control correlated directly with
mid-frontal asymmetry (r 0.53, P < 0.05).
Replicating past research, as compared to participants
who were included, participants who were ostracized reported being included in a smaller percentage of throws,
and they reported greater levels of anger and lower levels
of positive mood, enjoyment, belonging, control and
2
Two self-report anger measures were included. The first asked participants to rate their agreement with the
statement: I felt angry during the Cyberball game. The second was presented after this as part of a myriad
possible emotions (anger words were anger, irritated, and mad; Cronbachs  0.83). The two anger
measures were only marginally correlated and the second anger measure was not related to frontal
asymmetry. It is possible that the second anger measure was less sensitive because of its placement
behind the first anger measure.

280

SCAN (2011)

meaningful existence. Participants who were ostracized also

reported marginally greater levels of sadness compared to
participants who were included. There were no significant
condition differences in self-esteem, distress or frontal cortical asymmetries (Table 1).
The results of Study 1 suggest that reflexive emotional
responses to ostracism involve anger and the degree of
anger depends critically on the degree of relative left cortical
activation aroused by ostracism. Also, results suggest that
both condition and relative left frontal cortical activation
play a role in perceived control, so that during ostracism
increased left frontal activation was associated with
decreased control, whereas during inclusion the opposite
effect was found. Because this effect was not predicted, we
will see if it replicates in Study 2.
Study 2 extends the correlational results of Study 1 by
examining whether emotional consequences of ostracism,
specifically anger and sadness, are affected by manipulating
frontal asymmetry with unilateral hand contractions.
Furthermore, it is possible that the predicted interaction
involving sadness was not found in Study 1 due to
the small effect of ostracism on sadness and to restricted
range of frontal asymmetry scores. Perhaps the direct manipulation of frontal asymmetry will create greater variance
in the asymmetric frontal activity and allow the predicted effect to emerge. Finally, we altered the manipulation
of ostracism to include ostensibly real participants, to assess whether the effects generalize to another ostracism
paradigm.
STUDY 2
Method
Participants and procedure
Twenty-six3 right-handed introductory psychology students
(17 women, 9 men) participated in exchange for partial
course credit. Participants were brought to the lab under
the guise that they were participating in an experiment
examining how personality, brain activity and the muscular
system interact to affect cognitive performance during a task
with other ostensible participants in the laboratory.
Participants were told that the other individuals were
asked to wait in another location in order to avoid coming
into contact with him or her before the experiment began.
After obtaining consent, the experimenter asked the participant to step out into the hall so a photograph could be taken
3

Following earlier suggestions (e.g. Basso et al., 1994; Peterson et al., 2008; Stemmler, 2003), we excluded
participants who failed to show an asymmetric effect of the unilateral contraction manipulation on contralateral motor strip; data from 22 participants who failed to show greater relative left (right) activation in
contralateral central electrodes (average of C3/4 and C5/6) during unilateral right (left) contractions were
discarded. Two other participants were removed from analyses due to suspicion. Because of this loss of
participants and because the procedure of Study 2 involved deception, we conducted another experiment
(without EEG) using the procedure of Study 1 to assess whether the hand contraction manipulation affected
self-reported anger. It did. Self-reported anger was greater for individuals who made right-hand contractions
(M 5.41, s.d. 2.32) compared to individuals who made left-hand contractions (M 3.63, s.d. 1.95),
t(34) 2.5, P < 0.01.

C. K. Peterson et al.
of the participant. They explained that it would be used later
in the experiment; no other instructions were given. Then,
EEG and EMG were attached.
After EEG and EMG attachments, face-to-face contact
with the participant ceased and all instructions were given
via envelope, computer or intercom. Next, participants were
instructed to hold a toy ball in their right or left hand with
the palm facing up. They were asked to squeeze the ball as
hard as they could while the opposite hand remained flat
with the palm facing down. Hand contraction assignment
was determined randomly and experimenters were blind to
condition. Four 45-s contraction trials occurred with a 15-s
relaxation period between each trial. The same procedure
was used in Harmon-Jones (2006), Peterson et al. (2008)
and Schiff et al. (1998). EEG and forearm EMG were recorded during contractions.
In the current study, the participant was told that they
were playing Cyberball against two other participants. To
bolster the story that the other players were real, photographs of real individuals (gender matched) were shown
next to each cartoon player. In this version, all participants
were in the ostracism condition. All other instructions and
procedures for the game and post-game questionnaires were
identical to Study 1.

Data collection and processing

Forearm EMG was recorded by placing tin electrodes
(Electro-Cap International, Eaton, OH) on each forearm
flexor. One electrode was placed one-third of the distance
from the medial epicondyle of humerus to the styloid process of radius. The other electrode was placed 5 cm from the
first electrode along the same line. Impedance levels were
10 000 V or below.
EMG signals were amplified online (an analogue 60-Hz
notch filter was enabled) with Neuroscan Synamps (El Paso,
TX), bandpass filtered (0.05500 Hz) and digitized at
2500 Hz. Offline, the signals were visually scored and portions of the data that contained artifacts were removed prior
to being bandpass filtered (30500 Hz) and rectified. All
epochs 1.024 s in duration were extracted through a 20%
tapered (10% at each end) cosine window. A fast Fourier
transform was used to calculate the power spectra, which
was averaged across each period of the hand contractions.
Total power within 30500 Hz was obtained. Values were
then log transformed to normalize across participants.
The methods for processing EMG signals are similar to
those described by Fridlund and Cacioppo (1986). All
procedures for acquiring and processing EEG data were
identical to those used in Study 1. Also, as in Study 1, frontal
asymmetry was examined using asymmetry indexes where
the natural log of the left electrode was subtracted from
the natural log of the right electrode. As such, higher
scores indicate greater relative left than right activity
(Davidson et al., 2000).

Ostracism, asymmetric frontal cortical activity, and negative affect

SCAN (2011)

281

Fig. 1 Topographic map displaying correlations between relative left hemispheric activation during ostracism and anger in Study 1. The display is a left lateral head view.
Areas in red and orange correlated directly with anger.

Results and discussion

Between-group differences
EMG activity. EMG activity in the right and left forearms
was measured during hand contractions to ensure that participants were indeed squeezing the ball as instructed. Data
from two participants, both in the left-hand condition, were
lost due to equipment failure. Of the remaining 24 participants, all showed the predicted EMG asymmetry. A 2 (hand
contraction: right, left)  2 (forearm: right, left) ANOVA
revealed that the hand contractions differentially affected
EMG activity in the forearms, F(1,22) 424.58, P < 0.001.
Greater EMG activity occurred in the right forearm during
right-hand contractions (M 3.20, s.d. 1.18) compared
to left-hand contractions (M 2.42, s.d. 0.96),
t(22) 12.81, P < 0.001, whereas greater EMG activity

occurred in the left forearm during left-hand contractions

(M 2.68, s.d. 0.84) compared to right-hand contractions
(M 2.88, s.d. 0.99), t(22) 14.84, P < 0.001.
As a manipulation check, we examined correlations between left frontal cortical activation and EMG activity
during the contractions. As expected, greater relative
EMG activity in the right forearm compared to left forearm during contractions related to greater relative left
lateral frontal (r 0.40, P < 0.05) and frontalcentral (r 0.35, P < 0.05) activation during the hand
contractions.
Frontal asymmetry during hand contractions. As predicted, the hand contraction manipulation affected lateral
frontal (t(24) 2.39, P < 0.05) and frontalcentral
(t(24) 1.92, P < 0.05) activation. Participants who made

282

SCAN (2011)

C. K. Peterson et al.

Fig. 2 Topographic map displaying correlations between relative left hemispheric activation during ostracism and anger within the right-hand contraction condition in Study 2.
The display is a left lateral head view. Areas in red and orange correlated directly with anger.

right-hand contractions evidenced greater relative left (or

reduced relative right) lateral frontal activation (M 0.15,
s.d. 0.19) and frontalcentral activation (M 0.12,
s.d. 0.25) than those who made left-hand contractions
(M 00.07, s.d. 0.26 and M 0.10, s.d. 0.32, respectively). Hand contraction condition did not affect
mid-frontal asymmetry, t < 0.93.
Frontal asymmetry in response to ostracism. Change in
frontal EEG asymmetry from the period of inclusion to the
period of ostracism was next examined. Consistent with predictions, relative left lateral frontal activation was greatest
(and thus relative right lateral frontal activation was lowest)
for right-hand contractions compared to left-hand contractions. Relative left mid-frontal and frontalcentral activation
did not differ between right- and left-hand contractions.
Emotions and basic needs in response to
ostracism. Consistent with our primary prediction, selfreported anger in response to ostracism was significantly
greater after right-hand contractions compared to left-hand

contractions. See Table 1 for all other between-condition

effects involving emotions and basic needs, none of which
was significant (t < 1.66).
Relationships between frontal asymmetry and
emotive responses
Within the right-hand condition, ostracism-induced anger
was associated with greater relative left lateral frontal
(r 0.50, P < 0.05) and frontalcentral (r 0.50, P < 0.05)
activation (Figure 2). Within the left-hand contraction condition, ostracism-induced sadness was associated with
decreased relative left frontal central activation (r 0.52,
P < 0.05) (Figure 3). Also within the left-hand contraction
condition, ostracism-induced distress related to decreased
relative left mid-frontal (r 0.49, P < 0.05) and frontalcentral (r 0.60, P < 0.05) activation.
Study 2 extended Study 1 by showing that greater relative
left frontal cortical activity is causally involved in the anger
response to ostracism. Right-hand contractions, relative to
left-hand contractions, caused increased left frontal cortical
activity (during the contractions and during ostracism) and

Ostracism, asymmetric frontal cortical activity, and negative affect

SCAN (2011)

283

Fig. 3 Topographic map displaying correlations between relative left hemispheric activation during ostracism and sadness within the left-hand contraction condition in Study 2.
The display is a left lateral head view. Areas in blue correlated inversely with sadness.

increased self-reported anger. Furthermore, withincondition correlations replicated Study 1 in that left frontal
activation related directly to anger within the right-hand
condition. Adding to Study 1, Study 2 also showed that,
within the left-hand condition, relative right frontal activation related to sadness and distress.
We also examined whether frontal asymmetry mediated
the effect of hand contraction condition on anger using steps
stipulated by Baron and Kenny (1986) (all tests are one
tailed). Results revealed that (1) hand contractions affected
the anger response ( 0.38, P 0.03), (2) hand contractions affected lateral frontal asymmetry during ostracism
( 0.38, P 0.03) and (3) when both hand contractions
and lateral frontal asymmetry were entered as predictors,
hand contractions did not affect the anger response
( 0.21, P 0.28) whereas lateral frontal asymmetry did
( 0.42, P 0.02). These results suggest partial mediation
occurred. Full mediation was marginally supported (Sobels
test z 1.45, P 0.08). Thus, relative left lateral frontal activation during ostracism is partially responsible for the effect
of right-hand contractions on ostracism-induced anger.

GENERAL DISCUSSION
Study 1 found that ostracism-induced anger is directly correlated with increased relative left frontal activity.
Supporting these findings, Study 2 demonstrated that
right-hand, as compared to left-hand, contractions caused
greater relative left frontal cortical activation during the
hand contractions as well as during ostracism, and caused
greater self-reported anger in response to ostracism. Withincondition correlations revealed patterns of associations between frontal activation and angry responses to ostracism
consistent with Study 1. These results suggest that greater
relative left frontal activity is associated with increased anger
to ostracism. These results are consistent with the motivational direction model of asymmetric frontal cortical activity
(Harmon-Jones, 2004)
Another possible explanation for the ostracism-induced
anger is that the right-hand contractions are similar to the
muscle contractions one might use to prime an angry response and this prime (for right-handed individuals) caused
the anger response following ostracism. One way to test this

284

SCAN (2011)

would be to see if left-handed individuals show the same

increase in anger when contracting their left hands.
However, given that some left-handed individuals show
functional lateralizations similar to right-handed individuals,
whereas other left-handed individuals show opposite patterns of lateralizations from right-handed individuals, this
explanation would be difficult to test. On the other hand,
we took precautions in the methods of the current experiment to make the manipulation as dissimilar as possible
from an angry pose. The contractions used in the current
experiment involved squeezing a toy ball with the palm
facing upwards, an action quite dissimilar from the action
made in displaying a fist in anger. Perhaps most importantly,
this alternative explanation would be unable to explain the
correlation observed between relative left frontal activation
and anger as well as the partial mediation evidence.
Only subjective anger was affected by the ostracism manipulation of Study 1 and the hand contraction manipulation of Study 2. Sadness and distress were not significantly
affected by these manipulations. However, in Study 2, relative right frontal activation related directly with sadness and
distress, only within the left-hand contraction condition.
Study 1 did not reveal similar correlations when hemispheric
dominance was not manipulated. Perhaps the manipulated
increase in relative right frontal activation was necessary to
cause the association of relative right frontal activation and
sadness/distress to ostracism. It may also be possible that
these relationships may not have been evident in Study 1
due to restricted variance. That is, the left-hand contraction
condition of Study 2 demonstrated larger standard deviations in frontal asymmetry than the right-hand condition
of Study 2 and both conditions of Study 1. The left-hand
contractions apparently freed up additional relative
right-frontal variance, making that variable more able to
correlate with sadness and distress in Study 2. In addition,
it is possible that the distribution of males vs females in each
study may have contributed to these differences. That is, an
equal number of each sex participated in Study 1, whereas
twice as many females as males participated in Study 2. If
such an explanation is true, we should find the predicted
effects for females in Study 1; re-examination of the data
found that such was not the case (r < 0.46, P > 0.21).
Correlations between frontal asymmetry and ostracisminduced sadness and distress did not differ significantly for
males vs females (P > 0.15).
In the present studies, asymmetric frontal cortical activity
was assessed by combining individual channels into three
main indexes covering a large anterior area of the brain.
Consistent with recent fMRI research, this suggests that the
dorsolateral prefrontal cortexs involvement in motivational
direction spans a large area (Berkman and Lieberman, 2010).
Although frontal asymmetry effects were found at each
index, they were not consistent across studies. This variation
in effect location is not uncommon, as Allen et al. (2004)
reviewed evidence from 70 studies that showed that

C. K. Peterson et al.
relationships with frontal EEG asymmetry appear. . .at different regions at different times (p. 26).
This research illustrates the importance of examining distinct negative affects rather than clustering all negative affects into one index, which has been encouraged by factor
analytic studies that suggest that all negative affects load on
one factor. Recent research has revealed that one of the most
often used measures of negative affect splits into distinct
factors of anger vs fear/distress when individuals are actually
experiencing strong bouts of affect caused by distinct emotion manipulations (Harmon-Jones et al., 2009) as opposed
to how they felt over long periods of time (Watson, 2000).
The current research assists in understanding reflexive
emotional responses to ostracism noted by Williams
(2007a) and suggest some interesting avenues for further
research on moderators of such responses to ostracism.
Past research has suggested that ostracism-induced negative
affect emerges regardless of several moderators (Williams,
2007a). Perhaps neural measures, like EEG, provide more
direct assessments of approach and withdrawal motivation
that may relate better to reflexive emotion responses such as
anger than other measures used in past research.
In the end, this work illustrates two benefits of a social
neuroscience approach (Ochsner and Lieberman, 2001;
Adolphs, 2003; Harmon-Jones and Devine, 2003;
Lieberman, 2010): it generated a novel hypothesis derived
from a neuroscience approach that shed light on a problem
in social psychological research on ostracism, and it used
social psychological methods to further our understanding
of the role of a pattern of neural activation in psychological
processes.
REFERENCES
Adolphs, R. (2003). Investigating the cognitive neuroscience of social behavior. Neuropsychologia, 41, 11926.
Allen, J.J.B., Coan, J.A., Nazarian, M. (2004). Issues and assumptions on the
road from raw signals to metrics of frontal EEG asymmetry in emotion.
Biological Psychology, 67, 183218.
Baron, R.M., Kenny, D.A. (1986). The moderator-mediator variable
distinction in social psychological research: Conceptual, strategic, and
statistical considerations. Journal of Personality and Social Psychology,
51, 117382.
Basso, M.R., Schefft, B.K., Hoffmann, R.G. (1994). Mood-moderating
effects of affect intensity on cognition: Sometimes euphoria is not beneficial and dysphoria is not detrimental. Journal of Personality and Social
Psychology, 66, 36368.
Berkman, E.T., Lieberman, M.D. (2010). Approaching the bad and avoiding
the good: Lateral prefrontal cortical asymmetry distinguishes between
action and valence. Journal of Cognitive Neuroscience, 22, 19709.
Buckley, K., Winkel, R., Leary, M. (2004). Reactions to acceptance and
rejection: Effects of level and sequence of relational evaluation. Journal
of Experimental Social Psychology, 40, 1428.
Carver, C.S., Harmon-Jones, E. (2009). Anger is an approach-related affect:
Evidence and implications. Psychological Bulletin, 135, 183204.
Chow, R.M., Tiedens, L.Z., Govan, C.L. (2008). Excluded emotions: The
role of anger in antisocial responses to ostracism. Journal of Experimental
Social Psychology, 44, 896903.
Coan, J.A., Allen, J.J.B., Harmon-Jones, E. (2001). Voluntary facial
expression and hemispheric asymmetry over the frontal cortex.
Psychophysiology, 38, 91225.

Ostracism, asymmetric frontal cortical activity, and negative affect

Davidson, R.J., Jackson, D.C., Larson, C.L. (2000). Human electroencephalography. In: Cacioppo, J.T., Tassinary, L.G., Berntson, G.G., editors.
Handbook of psychophysiology,
2nd edn. New York: Cambridge
University Press, pp. 2752.
Downey, G., Feldman, S.I. (1996). Implications of rejection sensitivity for
intimate relationships. Journal of Personality and Social Psychology, 70,
132743.
Eisenberger, N.I., Lieberman, M.D. (2004). Why rejection hurts: A common
neural alarm system for physical and social pain. Trends in Cognitive
Science, 8, 294300.
Fridlund, A.J., Cacioppo, J.T. (1986). Guidelines for human electromyographic research. Psychophysiology, 23, 56789.
Gerber, J., Wheeler, L. (2009). On being rejected: A meta-analysis of experimental research on rejection. Perspectives on Psychological Science, 4,
46888.
Harmon-Jones, E. (2003). Clarifying the emotive functions of asymmetrical
frontal cortical activity. Psychophysiology, 40, 83848.
Harmon-Jones, E. (2004). Insights on asymmetrical frontal brain activity
gleaned from research on anger and cognitive dissonance. Biological
Psychology, 67, 5176.
Harmon-Jones, E. (2006). Unilateral right-hand contractions cause contralateral alpha power suppression and approach motivational affective
experience. Psychophysiology, 43, 598603.
Harmon-Jones, E., Sigelman, J. (2001). State anger and prefrontal brain
activity: Evidence that insult-related relative left-prefrontal activation is
associated with experienced anger and aggression. Journal of Personality
and Social Psychology, 80, 797803.
Harmon-Jones, E., Devine, P.G. (2003). Introduction to the special section
on social neuroscience: Promise and caveats. Journal of Personality and
Social Psychology, 85, 58993.
Harmon-Jones, E., Gable, P. (2009). Neural activity underlying the effect of
approach-motivated positive affect on narrowed attention. Psychological
Science, 20, 40609.
Harmon-Jones, E., Harmon-Jones, C., Fearn, M., Sigelman, J.D.,
Johnson, P. (2008). Action orientation, relative left frontal cortical activation, and spreading of alternatives: A test of the action-based model of
dissonance. Journal of Personality and Social Psychology, 94, 115.
Harmon-Jones, E., Harmon-Jones, C., Abramson, L.Y., Peterson, C.K.
(2009). PANAS positive activation is associated with anger. Emotion, 9,
18396.
Jacobs, G.D., Snyder, D. (1996). Frontal brain asymmetry predicts affective
style in men. Behavioral Neuroscience, 110, 36.
Jones, N.A., Fox, N.A. (1992). Electroencephalogram asymmetry during
emotionally evocative films and its relation to positive and negative affectivity. Brain and Cognition, 20, 28099.
Lieberman, M.D. (2010). Social cognitive neuroscience. In: Fiske, S.T.,
Gilbert, D.T., Lindzey, G., editors. Handbook of Social Psychology,
5th edn. New York, NY: McGraw-Hill, pp. 14393.

SCAN (2011)

285

MacDonald, G., Leary, M.R. (2005). Why does social exclusion hurt? The
relationship between social and physical pain. Psychological Bulletin, 131,
202223.
Ochsner, K.N., Lieberman, M.D. (2001). The emergence of social cognitive
neuroscience. American Psychologist, 56, 71734.
Peterson, C.K., Shackman, A.J., Harmon-Jones, E. (2008). The role of asymmetrical frontal cortical activity in aggression. Psychophysiology, 8692.
Prentice, D.A., Miller, D.T. (1992). When small effects are impressive.
Psychological Bulletin, 112, 16064.
Rosenthal, R., Rosnow, R.L, Rubin, D.B. (2000). Contrasts and effect sizes
in behavioral research: A correlational approach. New York: Cambridge
University Press.
Schiff, B.B., Lamon, M. (1994). Inducing emotion by unilateral contraction
of hand muscles. Cortex, 30, 24754.
Schiff, B.B., Guirguis, M., Kenwood, C., Herman, C.P. (1998). Asymmetrical
hemispheric activation and behavioral persistence: effects of unilateral
muscle contractions. Neuropsychology, 12, 52632.
Schutter, D.J.L.G., de Weijer, A.D., Meuwese, J.D.I., Morgan, B., van
Honk, J. (2008). Interrelations between motivational stance, cortical
excitability, and the frontal electroencephalogram asymmetry of emotion:
A transcranial magnetic stimulation study. Human Brain Mapping, 29,
57480.
Semlitsch, H.V., Anderer, P., Schuster, P., Presslich, O. (1986). A solution
for reliable and valid reduction of ocular artifacts, applied to the P300
ERP. Psychophysiology, 23, 695703.
Stemmler, G. (2003). Methodological considerations in the psychophysiological study of emotion. In: Davidson, R.J., Goldsmith, H.H.,
Scherer, K.R., editors. Handbook of affective sciences. New York: Oxford
University Press, pp. 22555.
Tajfel, H., Billig, M.G., Bundy, R.P., Flament, C. (1971). Social categorization and intergroup behavior. European Journal of Social Psychology, 1,
14978.
van Honk, J., Schutter, D.J.L.G. (2006). From affective valence to motivational direction: The frontal asymmetry of emotion revised. Psychological
Science, 17, 96365.
Verona, E., Sadeh, N., Curtin, J. (2009). Stress-induced asymmetric frontal
brain activity and aggression risk. Journal of Abnormal Psychology, 118,
13145.
Watson, D. (2000). Mood and temperament. New York: Guilford Press.
Williams, K.D. (2007a). Ostracism. Annual Review of Psychology, 58, 42552.
Williams, K.D. (2007b). Ostracism: The kiss of social death. Social and
Personality Psychology Compass, 1, 23647.
Zadro, L., Williams, K.D., Richardson, R. (2004). How low can you go?
Ostracism by a computer is sufficient to lower self-reported levels of
belonging, control, self-esteem, and meaningful existence. Journal of
Experimental Social Psychology, 40, 56067.
Zajonc, R.B. (1968). Attitudinal effects to mere exposure. Journal of
Personality and Social Psychology, 9, 127.