ORIGINAL RESEARCH
published: 07 Decembe 2016
doi: 10.3389/ nhum.2016.00619
A ec i e P iming by Eye Gaze
S imuli: Beha io al and
Elec ophysiological E idence
Tingji Chen ,Mikko J. Pel ola ,Lo a J. Ran a and Ja i K. Hie anen *
Human In o ma ion P ocessing Labo a o y, School o Social Sciences and Humani ies/Psychology, Uni e si y o Tampe e,
Tampe e, Finland
Edi ed by:
A hu M. Jacobs,
F ee Uni e si y o Be lin, Ge many
Re iewed by:
La s Kuchinke,
In e na ional Psychoanaly ic
Uni e si y Be lin, Ge many
Benny Bjö n B iesemeis e ,
In e na ional Psychoanaly ic
Uni e si y Be lin, Ge many
*Co espondence:
Ja i K. Hie anen
[email p o ec ed]
Recei ed: 23 Sep embe 2016
Accep ed: 21 No embe 2016
Published: 07 Decembe 2016
Ci a ion:
Chen T, Pel ola MJ, Ran a LJ and
Hie anen JK (2016) A ec i e P iming
by Eye Gaze S imuli: Beha io al and
Elec ophysiological E idence.
F on . Hum. Neu osci. 10:619.
doi: 10.3389/ nhum.2016.00619
The p esen s udy employed he a ec i e p iming pa adigm and measu emen s o
e en - ela ed po en ials (ERPs) o in es iga e implici a ec i e eac ions elici ed by gaze
s imuli. Pa icipan s ca ego ized posi i e and nega i e wo ds p imed by di ec gaze,
a e ed gaze and closed eyes. The beha io al esponse ime (RT) esul s indica ed ha
di ec gaze implici ly elici ed mo e posi i e a ec i e eac ions han did closed eyes.
Analyses o he ERP esponses o he a ge wo ds e ealed a p iming e ec on he
N170 and an in e ac ion on la e posi i e po en ial (LPP) esponses, and cong uen ly
wi h he beha io al esul s, sugges ed ha , compa ed o closed eyes, di ec gaze was
a ec i ely mo e cong uen wi h posi i e wo ds and mo e incong uen wi h nega i e
wo ds. The p iming e ec on he N170 esponse indica ed ha gaze s imuli in luenced
he subsequen a ec i e wo d p ocessing a an ea ly s age o in o ma ion p ocessing.
In conclusion, he p esen beha io al and elec ophysiological e idence sugges s ha
di ec gaze au oma ically ac i a es mo e posi i e a ec i e eac ions han closed eyes.
Keywo ds: eye gaze, a ec i e p iming, e en - ela ed po en ial, sel - a ings, a ec i e e alua ion
INTRODUCTION
Gaze and eye con ac a e c ucial social signals, which p o ide in o ma ion abou indi iduals’
di ec ion o a en ion and in en ions. A gaze looking owa ds he pe cei e is likely o be
in e p e ed as signaling a endency o app oach and in en ion o communica e, whe eas eyes
looking away may signal a endency o a oid (A gyle and Cook, 1976). Eye gaze no only signals
he sende ’s app oach-a oidance endencies, bu i also ac i a es co esponding mo i a ional
endencies in he pe cei e s. Seeing ano he pe son’s di ec gaze, as compa ed o a e ed gaze
o closed eyes, elici s g ea e ela i e le -side on al Elec oencephalog aphic (EEG) ac i i y,
associa ed wi h app oach endency (Ha mon-Jones, 2003; Hie anen e al., 2008; Pönkänen e al.,
2011b; Kylliäinen e al., 2012). S udies in es iga ing he skin conduc ance esponse (SCR), a
measu e o physiological a ousal, ha e shown g ea e SCRs in esponse o di ec gaze as compa ed
o a e ed gaze o closed eyes (Nichols and Champness, 1971; Hie anen e al., 2008; Helminen e al.,
2011). These indings indica e ha eye gaze is a powe ul s imulus elici ing a ec i e-mo i a ional
esponses in he pe cei e .
In addi ion o elici ing physiological a ousal and mo i a ional esponses, eye gaze has
also been shown o elici a ec i e esponses in he pe cei e . Howe e , s udies measu ing
explici esponses, such as sel - a ed a ec i e alence, ha e indica ed somewha con adic ing
indings o bo h nega i e and posi i e a ec i e esponses o ano he indi idual’s di ec gaze.
F on ie s in Human Neu oscience | www. on ie sin.o g 1Decembe 2016 | Volume 10 | A icle 619
Chen e al. A ec i e P iming by Eye Gaze S imuli
Di ec gaze can signal dominance and agg ession, hus elici ing
a e si e esponses in bo h animals and humans (Eme y, 2000;
Skuse, 2003). Acco dingly, s udies in es iga ing sel - epo ed
a ec i e esponses o eye gaze ha e e ealed ha , o emo ionally
neu al aces, di ec gaze is e alua ed as less posi i e han a e ed
gaze and closed eyes (Hess e al., 2007; Hie anen e al., 2008;
Pönkänen e al., 2011a). On he o he hand, di ec gaze can
be pe cei ed as a signal o posi i e communica i e in en ion
(Kleinke, 1986). S udies in es iga ing sel - epo ed likabili y and
a ac i eness o aces ha e shown ha pa icipan s e alua e aces
wi h di ec gaze as mo e likable and a ac i e han aces wi h
a e ed gaze (Mason e al., 2005; Kuzmano ic e al., 2009; Ewing
e al., 2010). Some s udies using physiological measu emen s
o indi ec beha io al measu emen s, sugges he e alua ion o
di ec gaze in a mo e posi i e manne (Kampe e al., 2001;
Lawson, 2015). Fo example, using he Implici Associa ion Tes ,
Lawson (2015) showed a obus p e e ence o aces looking
owa ds he pe cei e , han looking away. Thus, i can be
specula ed ha he na u e o a ec i e esponses o gaze di ec ion
may depend on whe he explici o implici a ec i e eac ions
a e measu ed. As compa ed wi h implici esponses, explici
esponses a e mo e suscep ible o ask demands, mo i a ional
biases, and o he op-down in luences (Ho mann e al., 2005). By
con as , implici esponses, which do no ely on in ospec i e
expe ience, may be less suscep ible o mo i a ional in luence, and
may e lec pe cei e s’ ins inc ual esponse o a di ec gaze.
In suppo o his specula ion, we ecen ly showed ha
implici and explici measu es, indeed, esul ed in di e en ial
a ec i e e alua ions (Chen e al., 2016). We used he a ec i e
p iming pa adigm (Fazio e al., 1986) o in es iga e au oma ic
(implici ) a ec i e esponses elici ed by eye gaze. In a ypical
a ec i e p iming expe imen , pa icipan s a e equi ed o make
speeded e alua ions o a ec i e a ge s p eceded by b ie ly
p esen ed a ec i e p imes. Typically, he esponse la ency o
a ge s is sho e o he a ec i ely cong uen p ime- a ge
pai s compa ed o a ec i ely incong uen pai s (Fazio e al.,
1986). This inding has been in e p e ed o indica e ha he
p esen a ion o he p ime au oma ically elici s he associa ed
a ec i e e alua ion and acili a es he decoding o a ec i ely
cong uen a ge s (Fazio e al., 1986; Fazio, 2001). Since he
seminal wo k by Fazio e al. (1986), he a ec i e p iming
e ec has been epo ed in a la ge numbe o s udies and
demons a ed o be a eplicable and obus phenomenon (Fazio,
2001; Klaue and Musch, 2003). In ou p e ious a ec i e p iming
s udy, he gaze s imuli we e b ie ly p esen ed as p imes and
immedia ely ollowed by posi i e and nega i e wo ds as a ge s.
The pa icipan s e alua ed he alence o he wo ds as as as
possible. Responses o posi i e wo ds we e as e a e di ec
gaze han a e closed eyes, while esponses o nega i e wo ds
we e as e a e closed eyes han a e di ec gaze. These esul s
we e in e p e ed o indica e ha he pe cep ion o di ec gaze
au oma ically ac i a es mo e posi i e e alua ions han closed
eyes (Chen e al., 2016). Ins ead, he explici a ings showed
ha di ec gaze was e alua ed as less posi i e han closed
eyes.
Howe e , wi h he beha io al esponse ime (RT) da a alone,
i is impossible o know a which s age o p ocessing gaze
p imes s a o exe hei in luence on a ge p ocessing. The
high empo al esolu ion o e en - ela ed po en ials (ERPs),
which p o ide a con inuous ime-window on neu al p ocesses
du ing s imulus p esen a ion, may shed ligh on his ques ion.
Impo an ly, p e ious s udies ha e demons a ed ha some
ERP componen s a e modula ed by he a ec i e cong uence
be ween a ge s and p eceding con ex s. In es iga ion o hese
componen s may p o ide indica ions abou he a ec i e
(in)cong uence be ween gaze s imuli and a ec i e wo ds on he
neu al le el.
In he p esen s udy, we employed he a ec i e p iming
pa adigm (as in Chen e al., 2016) and eco ded ERPs in esponse
o he a ge wo ds p eceded by eye gaze. P e ious s udies
employing a simila me hodology ha e sugges ed ha P1 is he
ea lies componen o be modula ed in a ec i e p iming s udies,
wi h la ge P1 ampli udes in esponse o a ec i ely incong uen
s. cong uen a ge s (Hie anen and As ikainen, 2013; Sianipa
e al., 2015). The ollowing N170 componen , wi h ime window
a ound 150–200 ms a e a ge onse (ea lie o ace and la e
o wo d a ge s), has also been epo ed o be modula ed in
a ec i e p iming s udies wi h wo ds (Comesaña e al., 2013)
and aces (Hie anen and As ikainen, 2013; Hinojosa e al.,
2015) as a ec i e a ge s. Fo example, Hie anen and As ikainen
(2013) showed ha happy aces p eceded by pic u es o posi i e
emo ional scenes elici ed la ge N170 ampli udes, whe eas he
N170 ampli udes we e la ge o sad aces p eceded by nega i e
scenes. The N170 componen e lec s ea ly isual p ocessing o
aces, wo ds, and objec s (Rossion e al., 2000). In he con ex o
a ec i e p iming, modula ion o he N170 esponse by a ec i e
p imes may indica e inc eased ac i i y in isual p ocessing a eas
when he a ec i e con en s o he p ime and a ge ma ch. A
ew s udies ha e shown p iming e ec s on he ea ly pos e io
nega i i y (EPN), wi h enhanced EPN ampli udes (i.e., mo e
nega i e mean ac i i y a a ound 250–400 ms), in esponse o
incong uen s. cong uen a ge s (Hie anen and As ikainen,
2013; Rampone e al., 2014). The midline N400 componen
is sugges ed o e lec seman ic in eg a ion o wo ds wi h
he p eceding con ex (Ku as and Hillya d, 1980) and i has
been epo ed o be modula ed by he a ec i e cong uence
be ween wo ds and he p eceding con ex (Zhang e al., 2006,
2010; Ede e al., 2011). Typically, g ea e N400 ampli udes
a e obse ed o a ec i ely incong uen han o cong uen
pai s. La e posi i e po en ial (LPP) modula ion by a ec i e
cong uence has been epo ed in p iming s udies wi h wo ds
(Zhang e al., 2010), aces (We heid e al., 2005; Hie anen and
As ikainen, 2013), and scenes (He ing e al., 2011) as a ec i e
a ge s. The esul s ha e e ealed g ea e LPP ampli udes a
a ound 400–700 ms o a ec i ely incong uen han cong uen
a ge s. EPN and LPP a e bo h sugges ed o e lec selec i e
a en ion o emo ionally signi ican s imuli (Schupp e al.,
2004).
In he p esen s udy, we aimed o: (i) eplica e ou p e ious
esul s o implici and explici a ec i e esponses o eye
gaze; (ii) p o ide physiological e idence ela ed o a ec i e
cong uence o gaze-wo d pai s by in es iga ing hose ERP
componen s which ha e been demons a ed o be modula ed by
he a ec i e cong uence be ween a ge s and p eceding p imes;
F on ie s in Human Neu oscience | www. on ie sin.o g 2Decembe 2016 | Volume 10 | A icle 619
Chen e al. A ec i e P iming by Eye Gaze S imuli
and (iii) by measu ing ERP esponses, examine he ea lies
s age o p ocessing a which gaze s imuli s a o exe hei
in luence on he p ocessing o a ec i e wo ds. We expec ed
o eplica e he beha io al esul s o ou p e ious s udy (Chen
e al., 2016). Namely, in he a ec i e p iming ask, posi i e wo ds
would be esponded o as e a e di ec gaze han a e closed
eyes, while nega i e wo ds would be esponded o as e a e
closed eyes han a e di ec gaze (i.e., posi i e wo ds being
a ec i ely mo e cong uen wi h di ec gaze han closed eyes).
In he sel - a ing ask, on he o he hand, di ec gaze would be
e alua ed as less posi i e compa ed o closed eyes. Based on he
p e ious ERP s udies (Zhang e al., 2010; He ing e al., 2011;
Hie anen and As ikainen, 2013), we expec ed a ec i e p iming
e ec s on he P1, N170, EPN, N400 and LPP componen s. To
be speci ic, P1, EPN, N400 and LPP esponses o posi i e wo ds
we e expec ed o be la ge when p eceded by closed eyes s.
di ec gaze, whe eas hese esponses o nega i e wo ds would
be la ge when p eceded by di ec gaze s. closed eyes. The
N170 esponse o posi i e wo ds was expec ed o be g ea e when
p eceded by di ec gaze s. closed eyes, whe eas N170 o nega i e
wo ds would be g ea e when p eceded by closed eyes s. di ec
gaze.
MATERIALS AND METHODS
Pa icipan s
Thi y- wo na i e Finnish speake s (17 emales; 19–32 yea s,
mean = 25 yea s), wi h no mal o co ec ed- o-no mal ision,
pa icipa ed. Th ee pa icipan s we e excluded om he ERP da a
analyses due o echnical e o s. All pa icipan s we e in o med
abou he expe imen al p ocedu e, and a signed consen was
ob ained om each pa icipan . The pa icipan s we e gi en a
mo ie icke o hei pa icipa ion. The esea ch p o ocol was
app o ed by he E hics Commi ee o he Tampe e egion.
S imuli
Simila o he s udy by Chen e al. (2016), he p imes
we e g ayscale images showing a ec angula shaped a ea
o he eye egion o anima ed aces. Fou male and ou
emale aces wi h di ec gaze, a e ed gaze (20◦le / igh ),
and closed eyes (Figu e 1A) we e c ea ed by using a 3D
anima ion so wa e, Digi al A Zone [Daz] 3D S udio1. The
size o he p imes was 2.0◦×6.9◦ e ically and ho izon ally,
espec i ely. The a ge s we e 48 posi i e and 48 nega i e Finnish
wo ds selec ed om he s udy by Söde holm e al. (2013)
(Figu e 1B). Acco ding o he a ings made by 20–30-yea -
old pa icipan s (Söde holm e al., 2013), posi i e and nega i e
wo ds di e ed signi ican ly in alence a ings, (94)= 34.56,
p<0.001 (Mposi i e = 5.62; Mnega i e = 2.55), bu no in
a ousal (p= 0.579), wo d leng h (p= 0.691), o absolu e
o ela i e su ace equency (bo h ps = 0.787). The a ge s
we e displayed in Calib i on and he size o he wo ds
p esen ed on he sc een was 1.5◦ e ically and 3.8◦–6.9◦
ho izon ally.
1h p://www.daz3d.com/
Design and P ocedu e
The expe imen consis ed o a p iming ask and an explici
a ing ask. The p iming ask had a 3 (gaze: di ec , a e ed,
closed) ×2 (wo d alence: posi i e s. nega i e) design, including
six blocks wi h 96 ials in each. Each o he 96 a ge s
(48 posi i e and 48 nega i e wo ds) we e pai ed wice wi h
a di ec gaze, a e ed gaze and closed eyes p ime ac oss he
blocks. Thus, in o al, he expe imen con ained 96 ials
o each p ime × a ge condi ion. Each expe imen al ial
consis ed o he ollowing e en s in sequence (Figu e 1C): a
ixa ion c oss (500 ms), a gaze p ime (100 ms), a blank sc een
(200 ms), and a wo d a ge p esen ed un il he pa icipan ’s
esponse. S imulus-onse -asynch ony (SOA) was 300 ms. The
pa icipan s we e sea ed 70 cm away om he sc een and
equi ed o classi y he a ge wo ds as posi i e o nega i e,
as quickly and accu a ely as possible, by p essing he ‘‘+’’ o
‘‘−’’ key ( esponse key loca ion was coun e balanced ac oss
pa icipan s). A e he esponse, he e was a 2000-ms in e al
be o e he nex ial. Be ween he blocks, he e was a 1-min
b eak.
In he explici a ing ask, he gaze pic u es we e shown o
he pa icipan s one by one. The pa icipan s we e equi ed o
e alua e he a ec i e alence and a ousal o each gaze s imulus
on he 9-poin Sel -Assessmen Manikin (SAM; B adley and
Lang, 1994) scales (1 = unpleasan /calm, 9 = pleasan /a ousing).
Each gaze s imulus was p esen ed on he sc een, i s oge he
wi h he a ousal scale and hen wi h he alence scale. Each ace
emained on he sc een un il he pa icipan esponded on bo h
scales.
Analyses o he Beha io al Da a
Recogni ion accu acy and RTs we e calcula ed om he p iming
ask. T ials wi h inco ec esponses o he wo d alence (5.54%)
and ials wi h esponse la encies sho e han 2.5 s anda d
de ia ions (SDs) below o abo e each pa icipan ’s mean (2.88%)
we e excluded. Log10- ans o m was applied o co ec o
non-no mal dis ibu ion. A 3 (gaze: di ec , a e ed, closed) ×2
(wo d alence: posi i e s. nega i e) epea ed measu es analysis
o a iance (ANOVA) was pe o med on bo h ecogni ion
accu acy and RT da a. Fo he a ing ask, one-way ANOVAs
we e conduc ed o es di e ences in he alence and a ousal
a ings o di ec gaze, a e ed gaze and closed eyes. All s a is ical
analyses we e pe o med using he SPSS package. Huynh-
Feld co ec ion was applied when app op ia e. Bon e oni
co ec ion was pe o med o all mul iple compa isons. Fo
he sake o cla i y, unco ec ed deg ees o eedom a e
epo ed.
EEG Reco ding and Analyses
Elec oencephalog aphic (EEG) ac i i y was con inuously
eco ded om 64 elec odes moun ed in an elec ode cap
(ac iCAP), ampli ied wi h a QuickAmp ampli ie , and
moni o ed wi h B ain Vision Reco de (B ain P oduc s
GmbH, Munich, Ge many). The analog signal was ampli ied
26.55 imes and he sampling a e was se o 1000 Hz. Using
a common a e age e e ence, he EEG signals we e e e enced
F on ie s in Human Neu oscience | www. on ie sin.o g 3Decembe 2016 | Volume 10 | A icle 619
Chen e al. A ec i e P iming by Eye Gaze S imuli
FIGURE 1 | Examples o he p ime (A) and a ge (B) s imuli. The sequence o e en s on a single ial is illus a ed in (C).
online o an a e age o all scalp elec odes. Addi ionally,
e ical elec o-oculog am (VEOG) was eco ded wi h a
pai o elec odes placed abo e and below he le eye. The
FT9 elec ode was used o eco ding ho izon al elec o-
oculog am (HEOG). Elec ode impedances we e educed unde
30 k.
O line, he EEG signal was digi ally il e ed wi h a 0.5–30 Hz
band-pass il e (24 dB/oc slope) and ocula -co ec ed using
he G a on/Coles algo i hm (G a on e al., 1983). Then, he
co ec ed signal was segmen ed in o 800-ms long epochs s a ing
100 ms p io o a ge onse . The mean ol age du ing he 100-ms
p e- a ge pe iod was used o baseline co ec ion. Because we
we e mainly in e es ed in ERPs a e he a ge , using p e- a ge
baseline, ins ead o p e-p ime baseline, may be mo e app op ia e
as he p e- a ge ERPs may be shi ed di e en ly by di e en gaze
p imes (Hie anen and As ikainen, 2013). A i ac s we e de ec ed
based on he ollowing c i e ia: ol age s ep o e 50 µV/ms,
ampli ude exceeding ±100 µV, and ac i i y lowe han 0.5
µV. T ials con aining a i ac s (8.0%) we e ejec ed. A e age
wa e o ms in each expe imen al condi ion we e calcula ed o
each pa icipan .
Selec ion o he elec ode si es and ime windows was based
on isual inspec ion o he a e aged wa e o ms and he indings
om p e ious s udies (Rossion e al., 2000; Kissle e al., 2009;
Ede e al., 2011; Ci on, 2012; Hie anen and As ikainen, 2013).
Fo example, acco ding o he p e ious li e a u e, N170 and
EPN ha e an occipi o empo al scalp dis ibu ion (Rossion e al.,
2000; Ci on, 2012), N400 is acqui ed om midline elec odes
(Ede e al., 2011; Hie anen and As ikainen, 2013) and LPP
peaks o e he pa ie al lobe (Kissle e al., 2009). The e o e, he
peak ampli ude (an a e age o ampli udes wi hin ±5 ms o he
de ec ed peak) and la ency o P1, N170 and he mean ampli ude
o he EPN we e de e mined om occipi o empo al channels
P7 and P8, wi hin a ime window 80–160 ms, 160–260 ms and
300–400 ms, espec i ely. The mean ampli ude o N400 was
de e mined a midline elec odes Cz and CPz, wi hin he ime
in e al o 250–400 ms. Mean LPP ampli ude was analyzed om
channels CP1, CP2, P1, P2, CPz and Pz, wi hin a ime window o
400–700 ms. All ERP da a we e analyzed using epea ed measu es
ANOVA wi h gaze and wo d alence as wi hin-subjec ac o s.
RESULTS
Beha io al Da a
The ANOVA o he esponse accu acy da a showed a signi ican
main e ec o wo d alence, F(1,31)= 26.41, p<0.001, η2
p= 0.460.
Response accu acy o posi i e wo ds (98%) was signi ican ly
highe han ha o nega i e wo ds (91%). Fo he RTs, he
ANOVA showed a main e ec o wo d alence, F(1,31)= 46.84,
p<0.001, η2
p= 0.602. Pa icipan s esponded as e o posi i e
wo ds (M= 636 ms) han o nega i e wo ds (M= 695 ms).
Impo an ly, he analyses showed he expec ed in e ac ion
be ween eye gaze and wo d alence, F(2,62)= 5.36, p= 0.010,
η2
p= 0.148. Fo posi i e wo ds, a one-way ANOVA showed a
main e ec o gaze di ec ion, F(2,62)= 6.31, p= 0.003, η2
p= 0.169.
Pai wise compa isons showed ha esponses o posi i e wo ds
we e as e a e di ec gaze (M= 629 ms) han a e closed eyes
(M= 646 ms, p= 0.002, Cohen’s d= 0.69). The RT o posi i e
wo ds a e a e ed gaze (M= 634 ms) did no di e signi ican ly
om he RTs o posi i e wo ds p eceded by di ec gaze (p= 0.330,
d= 0.29) and closed eyes (p= 0.265, d= 0.33). No e ec o gaze
F on ie s in Human Neu oscience | www. on ie sin.o g 4Decembe 2016 | Volume 10 | A icle 619
Chen e al. A ec i e P iming by Eye Gaze S imuli
FIGURE 2 | Means and s anda d e o s o esponse imes (RTs) o
posi i e and nega i e a ge s p eceded by di ec gaze, a e ed gaze
and closed eyes p imes (∗∗ p<0.01).
di ec ion was ound on he RTs o nega i e wo ds (p= 0.401;
Figu e 2).
ERP Da a
Figu e 3 shows he g and a e age ERP opog aphies o he P1,
N170, EPN, N400 and LPP componen s wi hin he espec i e
ime windows.
P1
Fo he P1 ampli udes, he main e ec o gaze was signi ican ,
F(2,56)= 15.97, p<0.001, η2
p= 0.363. P1 ampli udes we e la ge
a e closed eyes (2.58 µV) han di ec gaze (1.77 µV, p<0.001,
d= 0.92) and a e ed gaze (1.97 µV, p<0.001, d= 0.90) p imes
(Figu e 4). The e was no signi ican main e ec o wo d alence
(p= 0.900) o an in e ac ion be ween gaze and wo d alence
(p= 0.454). Fo P1 la ency, he analyses showed signi ican main
e ec s o gaze, F(2,56)= 4.73, p= 0.013, η2
p= 0.144. Speci ically, he
la ency o P1 was sho e a e a e ed gaze (118 ms) han closed
eyes (122 ms, p = 0.017, d= 0.56) p imes. The P1 la ency a e
di ec gaze (120 ms) did no di e om ha a e a e ed gaze o
closed eyes (p = 0.197, d= 0.36; p = 0.854, d= 0.21, espec i ely).
The in e ac ion be ween gaze and wo d alence on P1 la ency was
no signi ican (p = 0.347).
N170
The analyses o he N170 ampli udes e ealed a signi ican
in e ac ion be ween gaze and wo d alence, F(2,56)= 3.89,
p= 0.026, η2
p= 0.122 (Figu e 4). Sepa a e one-way ANOVAs
we e conduc ed o in es iga e he e ec o gaze p imes on
posi i e and nega i e wo ds. Fo he posi i e wo ds, a one-way
ANOVA showed a main e ec o gaze, F(2,56)= 5.20, p= 0.008,
η2
p= 0.157. Pai wise compa isons showed signi ican ly la ge
N170 ampli udes in esponse o posi i e wo ds a e di ec gaze
(−8.68 µV) han a e closed eyes p imes (−8.01 µV, p= 0.005,
d= 0.66). The N170 ampli udes o posi i e wo ds p eceded by
a e ed gaze did no di e om hose p eceded by di ec gaze
FIGURE 3 | G and a e age e en - ela ed po en ial (ERP) opog aphies
o he P1, N170, ea ly pos e io nega i i y (EPN), N400 and la e posi i e
po en ial (LPP) componen s a he indica ed ime windows.
F on ie s in Human Neu oscience | www. on ie sin.o g 5Decembe 2016 | Volume 10 | A icle 619

Chen e al. A ec i e P iming by Eye Gaze S imuli
FIGURE 4 | G and a e age ERPs o he P1, N170, EPN and LPP componen s o posi i e and nega i e wo ds p imed by di ec gaze, a e ed gaze and
closed eyes a he de e mined elec ode si es (P7 and P8 o P1, N170 and EPN; CP1, CP2, P1, P2, CPz and Pz o LPP). The ol age maps show he ERP
di e ence (di ec gaze minus closed eyes) o posi i e (le ) and nega i e ( igh ) wo ds in he LPP ime window.
F on ie s in Human Neu oscience | www. on ie sin.o g 6Decembe 2016 | Volume 10 | A icle 619
Chen e al. A ec i e P iming by Eye Gaze S imuli
(p= 1.000, d= 0.06) o by closed eyes (p= 0.111, d= 0.43).
Fo nega i e wo ds, he e was no e ec o gaze (p= 0.689). No
signi ican e ec s we e ound in he analyses o N170 la ency.
EPN
The EPN analyses showed a main e ec o gaze p ime,
F(2,56)= 8.89, p<0.001, η2
p= 0.241 (Figu e 4). O e all, he
EPN ampli ude was la ge (i.e., mo e nega i e) a e a e ed gaze
(−3.52 µV) han di ec gaze (−3.18 µV, p= 0.008, d= 0.64) and
closed eyes (−2.98 µV, p<0.001, d= 0.73) p imes. No o he
signi ican e ec s we e ound.
N400
The analyses o he N400 mean ac i i y showed ha he main
e ec s o gaze and wo d alence we e signi ican . O e all, he
N400 ampli ude was smalle (i.e., mo e posi i e) a e a e ed
gaze (2.24 µV) han closed eyes (1.78 µV, p= 0.017, d= 0.60),
while he N400 ampli udes a e di ec gaze (2.13 µV) did no
di e om hose a e a e ed gaze and closed eyes (p= 0.940,
d= 0.18; p= 0.154, d= 0.39, espec i ely), F(2,56)= 5.31,
p= 0.008, η2
p= 0.159. Finally, he N400 was smalle o posi i e
(2.30 µV) han nega i e wo ds (1.80 µV), F(1,28)= 5.44, p= 0.027,
η2
p= 0.163.
LPP
The analyses showed a ma ginal main e ec o wo d alence,
F(1,28)= 4.01, p= 0.055, η2
p= 0.125. Posi i e wo ds (3.41 µV)
elici ed a la ge LPP han did nega i e wo ds (3.07 µV).
Impo an ly, he analyses showed a signi ican in e ac ion
be ween gaze and wo d alence, F(2,56)= 3.30, p= 0.044,
η2
p= 0.105 (see Figu e 4). Sepa a e one-way ANOVAs o
bo h posi i e and nega i e wo d condi ions we e no signi ican
(p= 0.296; p= 0.117, espec i ely). Howe e , as indica ed by
he ERP wa e o ms in Figu e 4, he LPP ampli udes o nega i e
wo ds p imed by di e en eye gaze condi ions seemed o di e .
Thus, we pe o med pai ed - es s and he esul s sugges ed
la ge LPP esponses o nega i e wo ds p eceded by di ec gaze
(3.23 µV) s. closed eyes (2.83 µV, p= 0.033, d= 0.42).
Explici Ra ings
Fo he alence a ings, he analyses showed a ma ginal main
e ec o gaze, F(2,62)= 3.09, p= 0.053, η2
p= 0.091. Closed eyes
(M= 5.85) and a e ed gaze (M= 5.72) we e a ed sligh ly
mo e posi i e han di ec gaze (M= 5.45), bu hese di e ences
we e no signi ican (p= 0.086, d= 0.42; p= 0.204, d= 0.34,
espec i ely). The e was a signi ican di e ence be ween gaze
s imuli in a ousal a ings, F(2,62)= 7.22, p= 0.002, η2
p= 0.189.
Di ec gaze (M= 3.94) was a ed mo e a ousing han a e ed
gaze (M= 3.32, p= 0.002, d= 0.70) and closed eyes (M= 3.27,
p= 0.021, d= 0.52; Figu e 5).
DISCUSSION
The p esen s udy employed a ec i e p iming and sel - a ing
asks o in es iga e implici and explici a ec i e e alua ion o
eye gaze. ERPs in esponse o he a ge wo ds we e eco ded
FIGURE 5 | Means and s anda d e o s o he alence and a ousal
a ings o he gaze s imuli (∗p<0.05; ∗∗ p<0.01).
du ing he a ec i e p iming ask in o de o examine he ime
cou se o he p iming e ec s. The esul s o he a ec i e p iming
ask showed an in e ac ion be ween eye gaze and wo d alence in
he beha io al RT da a, and e ealed a p iming e ec o posi i e
wo ds, i.e., posi i e wo ds we e e alua ed as being posi i e as e
a e di ec gaze han a e closed eyes. No p iming e ec was
ound o he nega i e wo ds. The sel - a ing esul s showed ha
di ec gaze was e alua ed as less posi i e han closed eyes. The
ERP analyses showed a clea p iming e ec on he N170 esponse
and an in e ac ion be ween eye gaze and wo d alence on he LPP
esponse: a g ea e N170 ampli ude o posi i e wo ds a e di ec
s. closed eyes, and a sligh ly g ea e LPP o nega i e wo ds a e
di ec s. closed eyes.
The beha io al esul s success ully eplica e hose o ou
p e ious s udy (Chen e al., 2016), and indica e ha di ec gaze
implici ly elici s mo e posi i e a ec i e eac ions han closed
eyes, whe eas di ec gaze is explici ly e alua ed as less posi i e
han closed eyes. Chen e al. (2016) sugges ed ha because
humans ha e a undamen al need o belong, and a e mo i a ed
o o m and main ain in e pe sonal ela ionships (Baumeis e
and Lea y, 1995), di ec gaze, as a signal o indi iduals’ app oach
endency and in en ion o social communica ion, ul ils hese
needs and hus is au oma ically e alua ed as a posi i e signal.
Howe e , when i comes o explici e alua ions, when indi iduals
become awa e o o he people’s gaze and s a o e alua e hei
co esponding eelings, hei esponses may become con olled
and he ini ial and au oma ic esponses a e a enua ed and
supp essed. They may espond o he gaze s imuli in a e lec i e
and analy ic way, and may s a o analyze he eason o he
eye con ac . Consequen ly, a di ec gaze may e oke a eeling o
unce ain y and, he e o e, diminish posi i e eelings (Chen e al.,
2016).
Impo an ly, he ERP esul s showed an in e ac ion be ween
eye gaze and wo d alence on he N170 and LPP esponses.
Speci ically, a p iming e ec was ound o posi i e wo ds on he
N170: esponses o posi i e wo ds we e g ea e when p eceded
by di ec gaze as compa ed o closed eyes. P e ious esea ch has
F on ie s in Human Neu oscience | www. on ie sin.o g 7Decembe 2016 | Volume 10 | A icle 619
Chen e al. A ec i e P iming by Eye Gaze S imuli
shown g ea e N170 esponses o a ge s p eceded by a ec i ely
cong uen compa ed o incong uen p imes (Hie anen and
As ikainen, 2013; Hinojosa e al., 2015). A plausible explana ion
is ha he enhanced N170 in esponse o a ec i ely cong uen
s. incong uen pai s may e lec he in eg a i e ac i a ion o
a ec i e alence by he con ex s (p imes) and he a ge s imuli,
and he e ec o his ac i a ion on isual p ocessing (Hie anen
and As ikainen, 2013; Diéguez-Risco e al., 2015). Thus, he
N170 modula ion indica es ha di ec gaze is a ec i ely mo e
cong uen wi h posi i e wo ds han closed eyes.
An in e ac ion be ween eye gaze and wo d alence was also
ound on he LPP wi h sligh ly g ea e LPP o nega i e wo ds
p eceded by di ec gaze as compa ed o closed eyes. P io
s udies ha e epea edly epo ed ha he LPP is modula ed
by a ec i e cong uence be ween he a ge s and p eceding
con ex s, wi h a ec i ely incong uen a ge s elici ing la ge
LPP ampli udes compa ed o a ec i ely cong uen a ge s
(We heid e al., 2005; Zhang e al., 2010; He ing e al., 2011;
Hie anen and As ikainen, 2013). An enhanced LPP in esponse
o a ec i ely incong uen a ge s is in e p e ed as e lec ing
inc eased a en ional esou ce alloca ion when he a ge s a e
unexpec ed o do no a ec i ely ma ch he p eceding p imes
(We heid e al., 2005; Zhang e al., 2010; Diéguez-Risco e al.,
2015). Thus, he p esen LPP esul indica es ha di ec gaze
is a ec i ely mo e incong uen wi h nega i e wo ds compa ed
wi h closed eyes. I should be no ed, howe e , ha he one-way
ANOVAs o bo h posi i e and nega i e wo d condi ions we e
no signi ican . This implies ha he e ec s on he LPP we e no
obus .
Acco ding o he abo e in e p e a ion o social needs being
ul illed by di ec gaze, one may aise a ques ion o why he
e ec o a e ed gaze did no signi ican ly di e om di ec
gaze in bo h implici a ec i e e alua ion and ERP measu es,
as a e ed gaze signals a endency o a oidance. A possible
explana ion o his esul migh be he use o s a ic gaze s imuli
as p imes, a he han dynamic gaze shi s o gaze o eal pe sons.
As compa ed wi h s a ic gaze, dynamic and eal gaze is mo e
ealis ic and ecologically alid, and may lead o ela i ely s onge
a ec i e esponses. S udies ha e shown ha dynamic s imuli,
ela i e o s a ic s imuli, enhance neu al esponses and in ensi y
judgmen s (Sa o e al., 2004; Weye s e al., 2006). A se ies o
s udies by Hie anen e al. (2008) has demons a ed ha gaze
s imuli p esen ed as pic u es o by a eal li e pe son elici
di e en ial beha io al and physiological esponses (Pönkänen
e al., 2011a,b). Fo example, Pönkänen e al. (2011a) epo ed
g ea e N170 and EPN ampli udes o di ec s. a e ed gaze
and closed eyes o he li e gaze s imuli, bu no o he gaze
pic u es. Simila ly, a g ea e N170 o di ec s. a e ed gaze was
epo ed in a s udy using dynamic gaze s imuli (Con y e al.,
2007). Thus, we specula e ha he employmen o mo e ealis ic
gaze s imuli may yield a clea e dis inc ion be ween di ec and
a e ed gaze.
In he p esen s udy, he p iming e ec on N170 ampli udes
was obse ed only wi h posi i e a ge wo ds and he e ec
on LPPs only wi h nega i e a ge wo ds. I is no ewo hy ha
he p iming e ec s shown on b ain po en ials a e no always
epo ed o bo h posi i e and nega i e a ge s. Fo example,
Zhang e al. (2010) epo ed p iming e ec s on LPP only o
posi i e wo ds; speci ically, LPP o posi i e wo ds was la ge
p eceded by nega i e s. posi i e pic u es. They specula ed ha
his was due o nega i e pic u es gene ally elici ing s onge
emo ional eac ions han posi i e pic u es. This explana ion
can also be applied o he p esen esul s. As compa ed wi h
closed eyes, he p esen as well as p e ious s udies show ha
di ec gaze is usually e alua ed as emo ionally mo e a ousing
and ac i a ing s onge au onomic esponses (Helminen e al.,
2011; Pönkänen e al., 2011a; Chen e al., 2016). As a esul ,
di ec gaze p imes may lead o ela i ely s onge p iming e ec s
han closed eyes p imes. Thus, o he N170 esponse, posi i e
wo ds p esen ed in he con ex o di ec gaze may elici a
s onge addi i e e ec o in eg a i e ac i a ion as compa ed
wi h nega i e wo ds p esen ed in he con ex o closed eyes.
Wi h espec o he LPP, nega i e wo ds p esen ed in he
con ex o di ec gaze may elici a s onge incong uency e ec
as compa ed wi h posi i e wo ds p esen ed in he con ex o
closed eyes. I should be no ed, howe e , ha he obse ed
in e ac ions be ween he gaze s imuli and he posi i e and
nega i e wo ds canno be explained by he di e en ial a ousal
o di ec gaze and closed eyes alone. We me ely sugges ha
he au oma ic a ec i e e alua ions and he ollowing a ec i e
p iming e ec s may also depend on he a ec i e a ousal elici ed
by he p imes.
Fo he beha io al esul s, we obse ed he p iming e ec
wi h posi i e wo ds bu no wi h nega i e wo ds. We ha e no
ob ious explana ion o his inding. I should be no ed ha , in
ou p e ious s udy (Chen e al., 2016), he beha io al p iming
e ec s we e obse ed o bo h posi i e and nega i e wo ds. The
p esen s udy had a simila expe imen al design as he p e ious
s udy, wi h he excep ions ha he p esen s udy included mo e
a ge wo ds and ials, only one p ime— a ge SOA (ins ead
o wo), and he p imes we e sup aliminally p esen ed (wi hou
an addi ional subliminal condi ion). Ne e heless, we belie e
ha his sligh di e ence in he esul s shall no e u e ou
conclusion.
The ERP analyses showed ha he p iming e ec s began
o be e lec ed on b ain ac i a ion as ea ly as a he le el
o he N170 componen . P io s udies in es iga ing language
comp ehension ha e indica ed ha he in luence o con ex ual
in o ma ion on lexical p ocessing occu s app oxima ely wi hin
he ini ial 200 ms ollowing s imulus onse (Se eno e al.,
2003). The p esen esea ch acco ds wi h and ex ends he ea lie
indings by demons a ing ha a gaze s imulus se ing as an
a ec i e con ex also exe s an in luence on a ec i e wo d
p ocessing a an ea ly s age o neu al p ocessing.
The p esen s udy showed ha he P1, EPN and
N400 componen s we e no modula ed by he a ec i e
cong uence be ween p imes and a ge s. A ew p io a ec i e
p iming s udies ha e epo ed p iming e ec s on he P1 and
EPN (Hie anen and As ikainen, 2013; Rampone e al., 2014;
Sianipa e al., 2015). The easons o he disc epan indings
may be ela ed o di e ences in he me hods be ween he
s udies. Fo example, he a ge s imuli used by Hie anen and
As ikainen (2013) we e acial exp ession pic u es, whe eas
isually p esen ed wo ds we e used in he p esen s udy. As
F on ie s in Human Neu oscience | www. on ie sin.o g 8Decembe 2016 | Volume 10 | A icle 619
Chen e al. A ec i e P iming by Eye Gaze S imuli
compa ed wi h a ec i e aces, he a ec i e meaning o wo ds
is on ogene ically lea ned and symbolic and, he e o e, he
p ocessing o a ec i e in o ma ion om wo ds and aces may
be di e en (Schach and Somme , 2009). Se e al a ec i e
p iming s udies wi h wo ds as a ge s ha e epo ed a la ge
N400 in esponse o a ec i ely incong uen a ge s (Zhang e al.,
2006, 2010; Ede e al., 2011). Howe e , some s udies ha e no
obse ed an N400 modula ion (He ing e al., 2011; Kissle
and Koessle , 2011), o ha e e en epo ed e e sed a ec i e
p iming e ec s (Aguado e al., 2013; Hie anen and As ikainen,
2013). He ing e al. (2011) conduc ed h ee a ec i e p iming
expe imen s wi h ei he a ec i e pic u e o wo d pai s and
epo ed a p iming e ec on he LPP, bu no on he N400.
They a gued ha hese esul s may sugges a dissocia ion in
he p ocesses o seman ic and a ec i e p iming: he LPP is
modula ed by a ec i e misma ch, whe eas he N400 may la gely
depend on he seman ic connec edness be ween he p imes and
a ge s.
A limi a ion ha should be add essed in u u e esea ch is
ela ed o isual low-le el di e ences be ween di e en gaze
s imuli. Fo example, di ec gaze has a highe scle a-i is con as
han closed eyes, and is geome ically mo e symme ic han
a e ed gaze. One may specula e ha he e ec s epo ed
in ou s udies we e d i en by low-le el ea u es o he eyes
when p esen ed b ie ly, a he han by social e alua ion as
we sugges ed. To examine his, u u e s udies could include
ace s imuli wi h on al and o a ed head o ien a ions, and
app op ia e con ol s imuli, such as sc ambled aces wi h di ec
gaze, a e ed gaze and closed eyes.
CONCLUSION
Co esponding o ou hypo heses, he p esen s udy ound ha
he beha io al da a success ully eplica ed p e ious indings
(Chen e al., 2016), indica ing ha di ec gaze is implici ly
e alua ed as mo e posi i e han closed eyes, whe eas i is
e alua ed as less posi i e in explici e alua ions. The ERP analyses
showed a p iming e ec on he N170 and an in e ac ion on
LPP esponses: a la ge N170 o posi i e wo ds a e di ec
s. closed eyes and a sligh ly la ge LPP o nega i e wo ds
a e di ec s. closed eyes. These esul s indica e ha di ec
gaze, as compa ed wi h closed eyes, is pe cei ed as mo e
a ec i ely cong uen wi h posi i e wo ds, and as mo e a ec i ely
incong uen wi h nega i e wo ds. The obse ed p iming e ec on
he N170 also indica es ha gaze s imuli in luence he subsequen
a ec i e wo d p ocessing a an ea ly s age o in o ma ion
p ocessing.
AUTHOR CONTRIBUTIONS
All he ou au ho s we e in ol ed in he expe imen al design,
da a collec ion and analysis. TC, MJP and JKH we e in ol ed in
manusc ip w i ing and da a in e p e a ion wo k.
FUNDING
This wo k was suppo ed by he GATE E asmus Mundus P ojec
( o TC) and he Academy o Finland (g an #275519 o MJP;
MIND p og am g an #266187 o JKH).
REFERENCES
Aguado, L., Dieguez-Risco, T., Méndez-Bé olo, C., Pozo, M. A., and
Hinojosa, J. A. (2013). P iming e ec s on he N400 in he a ec i e p iming
pa adigm wi h acial exp essions o emo ion. Cogn. A ec . Beha . Neu osci. 13,
284–296. doi: 10.3758/s13415-012-0137-3
A gyle, M., and Cook, M. (1976). Gaze and Mu ual Gaze. New Yo k, NY:
Camb idge Uni e si y P ess.
Baumeis e , R. F., and Lea y, M. R. (1995). The need o belong: desi e o
in e pe sonal a achmen s as a undamen al human-mo i a ion. Psychol. Bull.
117, 497–529. doi: 10.1037/0033-2909.117.3.497
B adley, M. M., and Lang, P. J. (1994). Measu ing emo ion: he sel -assessmen
Manikin and he seman ic di e en ial. J. Beha . The . Exp. Psychia y 25, 49–59.
doi: 10.1016/0005-7916(94)90063-9
Chen, T., Helminen, T. M., and Hie anen, J. K. (2016). A ec in he eyes: explici
and implici e alua ions. Cogn. Emo . doi: 10.1080/02699931.2016.1188059
[Epub ahead o p in ].
Ci on, F. M. M. (2012). Neu al co ela es o w i en emo ion wo d p ocessing: a
e iew o ecen elec ophysiological and hemodynamic neu oimaging s udies.
B ain Lang. 122, 211–226. doi: 10.1016/j.bandl.2011.12.007
Comesaña, M., Soa es, A. P., Pe ea, M., Piñei o, A. P., F aga, I., and Pinhei o, A.
(2013). ERP co ela es o masked a ec i e p iming wi h emo icons. Compu .
Hum. Beha . 29, 588–595. doi: 10.1016/j.chb.2012.10.020
Con y, L., N’Diaye, K., Tijus, C., and Geo ge, N. (2007). When eye c ea es
he con ac ! ERP e idence o ea ly dissocia ion be ween di ec and a e ed
gaze mo ion p ocessing. Neu opsychologia 45, 3024–3037. doi: 10.1016/j.
neu opsychologia.2007.05.017
Diéguez-Risco, T., Aguado, L., Albe , J., and Hinojosa, J. A. (2015). Judging
emo ional cong uency: explici a en ion o si ua ional con ex modula es
p ocessing o acial exp essions o emo ion. Biol. Psychol. 112, 27–38. doi: 10.
1016/j.biopsycho.2015.09.012
Ede , A. B., Leu hold, H., Ro he mund, K., and Schweinbe ge , S. R. (2011).
Au oma ic esponse ac i a ion in sequen ial a ec i e p iming: an ERP s udy.
Soc. Cogn. A ec . Neu osci. 7, 436–445. doi: 10.1093/scan/ns 033
Eme y, N. J. (2000). The eyes ha e i : he neu oe hology, unc ion and e olu ion
o social gaze. Neu osci. Biobeha . Re . 24, 581–604. doi: 10.1016/s0149-
7634(00)00025-7
Ewing, L., Rhodes, G., and Pellicano, E. (2010). Ha e you go he look? Gaze
di ec ion a ec s judgemen s o acial a ac i eness. Vis. Cogn. 18, 321–330.
doi: 10.1080/13506280902965599
Fazio, R. H. (2001). On he au oma ic ac i a ion o associa ed e alua ions: an
o e iew. Cogn. Emo . 15, 115–141. doi: 10.1080/02699930125908
Fazio, R. H., Sanbonma su, D. M., Powell, M. C., and Ka des, F. R. (1986). On
he au oma ic ac i a ion o a i udes. J. Pe s. Soc. Psychol. 50, 229–238. doi: 10.
1037/0022-3514.50.2.229
G a on, G., Coles, M. G. H., and Donchin, E. (1983). A new me hod o o -line
emo al o ocula a i ac . Elec oencephalog . Clin. Neu ophysiol. 55, 468–484.
doi: 10.1016/0013-4694(83)90135-9
Ha mon-Jones, E. (2003). Cla i ying he emo i e unc ions o asymme ical on al
co ical ac i i y. Psychophysiology 40, 838–848. doi: 10.1111/1469-8986.00121
Helminen, T. M., Kaasinen, S. M., and Hie anen, J. K. (2011). Eye con ac and
a ousal: he e ec s o s imulus du a ion. Biol. Psychol. 88, 124–130. doi: 10.
1016/j.biopsycho.2011.07.002
He ing, D. R., Taylo , J. H., Whi e, K. R., and C i es, S. L., J . (2011).
Elec ophysiological esponses o e alua i e p iming: he LPP is sensi i e o
incong ui y. Emo ion 11, 794–806. doi: 10.1037/a0022804
Hess, U., Adams, R. B., J ., and Kleck, R. E. (2007). Looking a you o looking
elsewhe e: he in luence o head o ien a ion on he signal alue o emo ional
acial exp essions. Mo i . Emo . 31, 137–144. doi: 10.1007/s11031-007-
9057-x
Hie anen, J. K., and As ikainen, P. (2013). N170 esponse o acial exp essions is
modula ed by he a ec i e cong uency be ween he emo ional exp ession and
F on ie s in Human Neu oscience | www. on ie sin.o g 9Decembe 2016 | Volume 10 | A icle 619