Thatcherization impacts the processing of own-race faces ...myweb.facstaff.wwu.edu/~jantzek/PDF/Hahn, Jantzen, Symons...To cite this article: Amanda C. Hahn, Kelly J. Jantzen & Lawrence

This article was downloaded by: [University of St Andrews], [Amanda Hahn]On: 20 July 2011, At: 08:12Publisher: Psychology PressInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Social NeurosciencePublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/psns20

Thatcherization impacts the processing of own-racefaces more so than other-race faces: An ERP studyAmanda C. Hahn a b , Kelly J. Jantzen a & Lawrence A. Symons aa Department of Psychology, Western Washington University, Bellingham, WA, USAb Department of Psychology, University of St Andrews, St Andrews, UK

Available online: 20 Jul 2011

To cite this article: Amanda C. Hahn, Kelly J. Jantzen & Lawrence A. Symons (2011): Thatcherization impacts the processingof own-race faces more so than other-race faces: An ERP study, Social Neuroscience, DOI:10.1080/17470919.2011.583080

To link to this article: http://dx.doi.org/10.1080/17470919.2011.583080

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching and private study purposes. Any substantial or systematicreproduction, re-distribution, re-selling, loan, sub-licensing, systematic supply or distribution in any form toanyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses shouldbe independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly inconnection with or arising out of the use of this material.

SOCIAL NEUROSCIENCE, 2011, iFirst, 1–13

Thatcherization impacts the processing of own-racefaces more so than other-race faces: An ERP study

Amanda C. Hahn1,2, Kelly J. Jantzen1, and Lawrence A. Symons1

1Department of Psychology, Western Washington University, Bellingham, WA, USA2Department of Psychology, University of St Andrews, St Andrews, UK

It has been suggested that differential use of configural processing strategies may underlie racially basedrecognition deficits (known as the “other-race effect”). By employing a well-known configural manipulation(Thatcherization, i.e., rotating the eyes and mouth by 180◦), we aimed to demonstrate, electrophysiologically,that configural processing is used to a greater extent when viewing same-race faces than when viewing other-racefaces. Face-related event-related potential (ERP) responses were measured for participants viewing normal andThatcherized faces of their own race (Caucasian) and of another race (African-American). The P1 and N170 com-ponents were modulated to a greater extent by Thatcherization for same-race faces, suggesting that the processingof these faces is, in fact, more reliant on configural information than other-race faces. Thatcherization also affectedthe P250 component more so for same-race faces independently of orientation. The race-dependent effects ofThatcherization as early as P1 suggest that configural encoding may be occurring much earlier than the well-citedN170.

Keywords: N170; P100 (P1); Thatcher; Race; Holistic processing.

1Also referred to as the “own-race bias”, “own-race effect,” or“own-race advantage.”

2The expertise theory is also referred to as the “contact theory”in some literature.

For many, the faces of one’s own race are easierto recognize than faces of another race (Chiroro &Valentine, 1995; Malpass & Kravitz, 1969; Meissner &Brigham, 2001; Walker & Hewstone, 2006). This phe-nomenon, known as the “other-race effect,”1 has beendemonstrated in a number of studies using behavioralparadigms (e.g., Bothwell, Brigham, & Malpass, 1989;Brigham & Malpass, 1985; Michel, Rossion, Han,Chung, & Caldara, 2006b; Tanaka, Kiefer, & Bukach,2004; Walker & Tanaka, 2003). More recent workutilizing eye-tracking (e.g., Blais, Jack, Scheepers,Fiset, & Caldara, 2008; Levin, 2000), functional imag-ing (Cunningham et al., 2003; Cunningham et al.,2004; Golby, Gabrieli, Chiao, & Eberhardt, 2001; Hartet al., 2000), and electrophysiological measures (Ito &Urland, 2003, 2005; James, Johnstone, & Hayward,2001; Walker, Silvert, Hewstone, & Nobre, 2008) hasindicated that own-race bias in recognition rates may

Correspondence should be addressed to: Amanda Hahn, Perception Laboratory, School of Psychology, St Mary’s Quad, South Street, StAndrews, Fife KY16 9JP, UK. E-mail: [email protected]

be the result of differential face processing basedon race.

The expertise theory2 posits that this differen-tial processing is the result of increased exposureto faces of one’s own race; specifically, it hasbeen suggested that when viewing own-race faces,observers are better able to extract configural infor-mation about the spatial relations between individualfacial features (Diamond & Carey, 1986; Lindsay,Jack, & Christian, 1991; Rhodes, Brake, Taylor, &Tan, 1989). Conversely, a lack of expertise withother-race faces may result in reduced ability toencode this configural information and a greaterneed to rely on featurally based processing strate-gies, thus affecting processing and recognition abil-ities (Michel, Caldara, & Rossion, 2006a; Michel,Rossion, Han, Chung, & Caldara, 2006b; Tanaka et al.,2004).

© 2011 Psychology Press, an imprint of the Taylor & Francis Group, an Informa businesswww.psypress.com/socialneuroscience DOI: 10.1080/17470919.2011.583080

Dow

nloa

ded

by [

Uni

vers

ity o

f St

And

rew

s], [

Am

anda

Hah

n] a

t 08:

12 2

0 Ju

ly 2

011

2 HAHN, JANTZEN, SYMONS

Configural manipulations, such as the well-established face inversion effect, have been invaluablefor investigating the mechanisms underlying face pro-cessing. In contrast to non-face objects, faces aresignificantly more difficult to process and recognizewhen inverted (Freire, Lee, & Symons, 2000; Rossionet al., 2000; Yin, 1969). Processing deficits follow-ing face inversion are thought to be due to config-ural disruption—that is, disruption of the ability toprocess features configurally or holistically (Farah,Tanaka, & Drain, 1995; Young, Hellawell, & Hay,1987)—because the configuration of individual fea-tures is processed in an orientation-specific manner(Oram & Perrett, 1992; Perrett, Heitanen, Oram, &Benson, 1992). For example, when a face is inverted,the horizontal and vertical distances between the eyesand nose are no longer apparent, because the eyesare no longer above the nose and mouth (Goffaux &Rossion, 2007). Recognition rates for faces are dispro-portionately impaired, because we rely on configuralprocessing to a greater extent when viewing facesthan when viewing non-face objects. Another com-monly used manipulation of configural informationis Thatcherization (Thompson, 1980), in which theeye and mouth regions of a face are rotated by 180◦,resulting in a grotesque appearance. This grotesque-ness is reduced by inversion, implying that role ofconfigural information in the phenomenon (Bartlett &Searcy, 1993; Boutsen & Humphreys, 2003; Boutsen,Humphreys, Praamstra, & Wartrick, 2006).

These techniques have also shed light on the other-race effect. Previous behavioral work utilizing theinversion effect (Rhodes, Brake, Taylor, & Tan, 1989;Vizioli, Foreman, Rousselet, & Caldara, 2010) andthe Thatcher effect (Murray, Rhodes, & Schuchinsky,2003) has demonstrated that configural disruptions arestronger for same-race faces than other-race faces,suggesting that configural processing strategies are uti-lized to a greater extent for faces of one’s own race.Moreover, the relative contribution of configural andfeatural processing changes as a function of experienceduring development (Carey & Diamond, 1977), andexpertise in identifying specific objects (Gauthier &Tarr, 1997), indicating that experience plays an impor-tant role. The other-race effect has been shown todevelop as a function of age, with adults demonstrat-ing a clear bias for recognizing own-race faces thatis absent in children (Chance, Turner, & Goldstein,1982; Goodman et al., 2007). These findings suggestthat experience with own-race faces is responsible forthe development of differential processing strategies.Most notably, other-race studies using part (featural)versus whole (configural) recognition paradigms havealso shown that individuals living in an other-race

environment can develop the expertise necessary touse configural processing strategies similar to whenviewing faces of their own race (Tanaka et al., 2004).

Electroencephalography (EEG) is a valuable toolfor investigating the neural basis of face processingand expertise. A growing number of studies haverevealed several “face-related,”3 event-related poten-tial (ERP) components typically centered over occip-itotemporal scalp locations that include the P100 (P1),N170, P250 (P2), and late positive component (LPC)(Bentin, Allison, Puce, Perez, & McCarthy, 1996). Theearly components (i.e., P1 and N170) reflect the struc-tural encoding stage of face processing, while latercomponents (i.e., P250 and LPC) reflect stimulus cat-egorization and/or attention to motivationally relevantinformation (e.g., race, gender, identity). These com-ponents can be affected by facial manipulations suchas inversion, contrast reversal, and other configuralalterations (Bentin et al., 1996; Eimer, 1998, 2000a,2000b; Itier, Herdman, George, Cheyne, & Taylor,2006; Itier & Taylor, 2002, 2004; Linkenkaer-Hansonet al., 1998). The P1, P2, and LPC components are notconsidered face-specific because their peak amplitudeand latencies typically do not differ between faces andother objects (cf. Itier & Taylor, 2004). However, theyare still of interest in understanding the neural basis offace processing.

The early P1 component (100–150 ms) is sensi-tive to the physical characteristics of stimuli, partic-ularly low-level visual properties such as luminanceand contrast (Halgren, Raij, Marinkovic, Jousmaki,& Hari, 2000; Rebai, Poiroux, Bernard, & Lalonde,2001; Rossion, Joyce, Cottrell, & Tarr, 2003), andmay be modulated by affective responses to stim-uli (Halit, de Haan, & Johnson, 2000; Pizzagalli,Regard, & Lehmann, 1999; Pizzagalli, Lehmann,Koenig, Regard, & Pascual-Marqui, 2000; Pizzagalliet al., 2002). Although not considered face-specific,the P1 has been shown to be sensitive to facialmanipulations such as inversion and contrast rever-sal (Jacques & Rossion, 2007; Itier & Taylor, 2002,2004; Linkenkaer-Hansen et al., 1998). Itier and Taylor(2004) argue that the P1 likely reflects the holistic pro-cessing of a face as a face while the “face-specific”N170 is involved in processing specific face config-uration (i.e., the spatial relations between individualfeatures).

Of particular interest is the “face-specific” N170.This negative deflection reaches peak amplitude

3Note that these components exist for non-face stimuli as well.As such, they are not considered to be face-specific, but ratherindices of visual processing.

Dow

nloa

ded

by [

Uni

vers

ity o

f St

And

rew

s], [

Am

anda

Hah

n] a

t 08:

12 2

0 Ju

ly 2

011

THATCHERIZATION IMPACTS THE PROCESSING 3

approximately 150–200 ms post-stimulus, and isconsistently larger in amplitude to faces than toother objects (Bentin et al., 1996; Bötzel, Schulze, &Stodieck, 1995; McCarthy, Puce, Belger, & Allison,1999). On the basis of its sensitivity to inversion, ithas been suggested that the N170 reflects a structuralencoding stage of face processing (Eimer, 2000a).

The N170 component seems to be particularly sen-sitive to configural disruptions, suggesting that it mayact as an index of configural processing. Inversion hasbeen demonstrated to increase the amplitude and delaythe latency of the N170 (Bentin et al., 1996; Rebai etal., 2001). It is possible that increased N170 amplitudemay represent an increase in the difficulty of extract-ing the configural information from a face. Similarly,delays to the N170 latency may reflect the increasedtime necessary to successfully process the altered stim-ulus. It is also possible that these N170 responses arenot face-specific, but are due to increased experiencewith a class of visual objects (i.e., faces). For example,in trained greeble4 experts, viewing greebles resultsin a larger N170 response and can interfere with theN170 to faces when faces and greebles are presentedtogether (Rossion, Kung, & Tarr, 2004), suggestingthat the N170 may be expertise-specific rather thanface-specific.

The late occurring ERP components reflect a moremeaning-based processing of visual stimuli. The P250component (230–300 ms) is related to semantic pro-cessing of face (and other) stimuli and likely reflectshigher-level processing such as familiarity (Marzi &Viggiano, 2006). The LPC (300–600 ms) is alsoinvolved in higher-order processing and may reflectattentional (Ashley, Vuilleumier, & Swick, 2004),emotional (Eimer & Holmes, 2002; Schutter, deHaan, & van Honk, 2004; Werheid, Alpay, Jentzch,& Sommer, 2005), or aesthetic (Höfel & Jacobsen,2007; Johnston & Oliver-Rodriguez, 1997; Oliver-Rodriguez, Guan, & Johnston, 1999) responses tostimuli.

Evidence from several ERP studies suggests thatThatcherization impacts the face-related ERP compo-nents, and that the N170 component, in particular, issusceptible to this face-distortion technique (Boutsenet al., 2006; Carbon, Schweinberger, Kaufmann, &Leder, 2005; Milivojevic, Clapp, Johnson, & Corballis,2003). However, this type of distortion may havebroader effects than those well documented in theN170 literature. Milivojevic et al. (2003) found that the

4Greebles are a class of novel stimuli that have been utilized toinvestigate the role of expertise and configural processing in faceperception studies because, like faces, they have a number of smallcomponents arranged in a commom configuration.

P1 component was also modulated by Thatcherization.Both P1 and N170 amplitudes were significantly largerfor Thatcherized faces than normal faces. Boutsenet al. (2006) compared the effects of Thatcherizationon a range of stimulus classes (faces, houses, andchairs). Although an inversion effect occurred for P1amplitude (face stimuli only), there was no Thatchereffect. The amplitude and latency of the N170 wereboth modulated by Thatcherization. The exact natureof the relationship between Thatcherization and P1modulation remains to be determined.

Because of its sensitivity to configural processing,the N170 may provide insight into the neural under-pinnings of the other-race effect. While behavioralstudies have yielded relatively consistent findings withrespect to the other-race effect, ERP studies of theN170 have provided conflicting results. Halit et al.(2000) demonstrated that the N170 component waslarger (in amplitude) for atypical faces than typicalfaces (faces were made to look “atypical” by verti-cally stretching the face and distorting its appearance).Since other-race faces are more atypical to observersthan faces of their own race, it seems the N170 shouldbe larger in amplitude for other-race faces. In support,several studies have demonstrated an increased N170amplitude for other-race faces (Herrmann et al., 2007;James, Johnstone, & Hayward, 2001; Stahl, Wiese, &Schweinberger, 2008; Walker et al., 2008). However,the N170 has also been demonstrated to increase asa result of expertise (Rossion et al., 2000), suggestingthat the N170 should be larger for same-race faces thanother-race faces since we are “experts” with our ownrace. A number of studies have, in fact, demonstratedlarger N170 amplitudes in response to same-race facesas compared to other-race faces (Caldara et al., 2003;Ito & Urland, 2005, Tanaka & Curran, 2001). Thereis yet a third class of ERP studies that have demon-strated that the N170 is not race-sensitive (Caldara etal., 2002; Caldara, Rossion, Bovet, & Hauert, 2004).To date, the mechanisms by which race may modulatethe N170 response remain unclear.

In the present study, we aimed to demonstrate thatconfigural processing strategies are utilized to a greaterextent for own-race faces than other-race faces, bymeasuring the effect of Thatcherization electrophysio-logically. Because the N170 response is modulated byexpertise, we hypothesized that the change in the N170response following Thatcherization (i.e., configuraldisruption) would be larger for same-race faces thanother-race faces, thus supporting the expertise theoryof the other-race effect. Previous studies have shownthe N170 amplitude both to increase (Milivojevic etal., 2003) and decrease (Boutsen et al., 2006) follow-ing Thatcherization. We predicted that the magnitude

Dow

nloa

ded

by [

Uni

vers

ity o

f St

And

rew

s], [

Am

anda

Hah

n] a

t 08:

12 2

0 Ju

ly 2

011

4 HAHN, JANTZEN, SYMONS

of change in terms of N170 peak amplitude would belarger for same-race than other-race faces, regardlessof the direction of this change. The P1 response waspredicted to be affected by low-level stimulus char-acteristics of our displays (e.g., brightness), but wasunlikely to be affected by expertise or Thatcherization(Stahl et al., 2008). Halit et al. (2000) found differ-ences in the P250 response when comparing typicalto atypical faces. Their atypical faces were created byaltering the spatial configuration of “normal” faces,suggesting that Thatcherization could have an impacton P250 amplitudes. However, we did not anticipate aninteraction between race and Thatcherization for theP250.

METHOD

Participants

Nineteen students from Western WashingtonUniversity (12 female) were recruited from anintroductory psychology subject pool to participate inthe present study. Volunteers received course creditand were recruited on the basis that they were ofCaucasian descent and had been raised by Caucasianparents. All participants were between 18 and 25years of age and reported right-handedness as wellas normal or corrected-to-normal vision. Prior totaking part in the study, all participants gave informed,written consent.

Measures

A racial contact survey, the Social ExperienceQuestionnaire (SEQ), developed by Brigham and col-leagues (J. Brigham, personal communication, 23January 2009), was used as a measure of other-racecontact. The SEQ contains 56 questions pertaining tolevel of contact with African-Americans across vari-ous stages of the life span (including business setting,personal settings, and intimate settings). The SEQassesses both the extent and quality of interactionswith members of another race. Participants completedthe entire three-part questionnaire; however, only thegeneral information section of the SEQ was used fordata analysis because this section deals only withamount of contact. Scores for this section of the sur-vey can range from 0 (no contact) to 9 (very high levelsof contact) and are calculated by the average responseacross all questions.

Stimuli

Stimuli consisted of 28 digitally manipulated colorphotographs of human faces. Photographs of 14Caucasian (same-race) faces (7 female) and 14African-American (other-race) faces (7 female) wereobtained from the CAL/PAL Face Database (Minear& Park, 2004). All faces obtained were classifiedas emotionally neutral (visible teeth are not nec-essary to produce the Thatcher Illusion; Lewis &Johnston, 1997). The 28 original photographs wereedited with Adobe Photoshop. Each image wascropped to 349×437 pixels and included the per-son’s face and neck. The backgrounds were allreplaced with a solid white background. Each of thesenewly manipulated photographs was used to createfour different test conditions: normal upright, nor-mal inverted, Thatcherized upright, and Thatcherizedinverted (Figure 1). Thatcherizing the digital imagesconsisted of rotating the eyes and the mouth by180◦ and then blending the manipulated regions toensure that they appeared to be a natural part ofthe face. Because the Thatcher illusion is orientation-dependent, the inverted condition was included asa control to verify the illusion. In total, 112 dif-ferent test faces were used. These consisted of 28faces of each race. Each of these faces was pre-sented both as Thatcherized and as normal, andeach of these was presented both upright andinverted.

Normal

Same-Race

Other-Race

Thatcher

Figure 1. Example stimuli in normal and thatcherized versions.

Dow

nloa

ded

by [

Uni

vers

ity o

f St

And

rew

s], [

Am

anda

Hah

n] a

t 08:

12 2

0 Ju

ly 2

011

THATCHERIZATION IMPACTS THE PROCESSING 5

EEG apparatus, recording, dataprocessing, and analysis

Stimuli were presented in full color on a 19-inchDell LCD monitor. Stimulus presentation and responserecording were controlled by in-house software writ-ten in Visual Basic. All responses were made with aCedrus 8-button box (Cedrus Corporation, San Pedro,CA, USA). EEG was continuously recorded from64 scalp sites, using BioSemi ActiveTwo Ag/AgClelectrodes and hardware (Biosemi, Amsterdam, TheNetherlands). The electrodes were placed according tothe 10-5 electrode system (Oostenveld & Praamstra,2001), using a nylon electrode cap. EEG signals wereamplified with a bandpass of DC-104 Hz by BioSemiActive-Two amplifiers, sampled at 512 Hz. Off-linesegmentation and averaging of EEG signals was per-formed with EEGlab v6.01b, running on Matlab 7.3.0(Mathworks, Inc., Natick, MA, USA). In a small num-ber of cases, a single channel demonstrated excessivenoise and was replaced by a new channel derived byspherical interpolation of the surrounding channels.Data were then filtered by a high-pass of 1 Hz and alow-pass of 12 Hz. The continuous EEG recording wassegmented into epochs extending from −100 to 500ms around the onset of each stimulus. Trials containingartifacts were detected by the EEGlab automatic epoch

rejection tool and visually confirmed by us beforeremoval. Trials with values outside of the range −40to +40 μV were rejected.

Visual inspection of the grand-averaged visuallyevoked responses indicated that electrodes PO7, PO8,P8, P6, P9, and P10 best demonstrated the clas-sic face-related potentials (including the P1, N170,P2, and LPC) and showed maximum peak N170amplitudes (Figure 2). The occipital temporal loca-tion of face-specific ERPs is in keeping with exten-sive previous literature (Bentin et al., 1996; Caldaraet al., 2002, 2004; Carbon et al., 2005; Eimer, 2000a,2000b; Stahl et al., 2008), and, as such, all furtheranalysis was restricted to these six electrode sites.Dependent measures of the neural response were theaverage N170 amplitude and latency for each of the 19participants across each of the six electrode sites.

Procedure

Participants first completed the SEQ survey. Prior tocompleting the experimental trials, participants weregiven a practice task consisting of five Thatcherizedand five normal faces in order to familiarize them tothe experimental task. Participants were instructed to

Figure 2. Exemplary set of electrodes showing the classic face-related waveform as well as maximum N170 amplitude.

Dow

nloa

ded

by [

Uni

vers

ity o

f St

And

rew

s], [

Am

anda

Hah

n] a

t 08:

12 2

0 Ju

ly 2

011

6 HAHN, JANTZEN, SYMONS

rate each face in terms of the bizarreness or grotesque-ness of its appearance. Because Thatcherization causesa face to appear bizarre or grotesque, these grotesque-ness ratings can be interpreted as an index ofthe configural disruption caused by Thatcherization.Participants were also asked to refrain from blinkingor making any extraneous movements while viewingphotographs.

EEG data were collected over three blocks. Withina block, each of the 112 faces was presented twice(for a total of 224 trials per block). Each trial beganwith the presentation of a black fixation cross at thecenter of the screen on a gray background for 1500ms. A face then appeared in the center of the screenfor 1000 ms. The face was then removed from thescreen and replaced by the rating task. Participantswere given 5000 ms to rate its “grotesqueness” on a 5-point Likert scale, where 1 corresponded to “not at allgrotesque” and 5 corresponded to “very grotesque,” bypressing the corresponding key on a response box. Atrial ended when a rating was made. Trials were sepa-rated by a 1000-ms interstimulus interval. Participantswere given a short (1–2 min) break between blocks.

RESULTS

Other-race expertise

Although the possible range of scores on the generalinformation section of the SEQ ranged from 0 to 9,participants’ contact scores ranged only from 0.12 to5.76 (M = 2.28, SD = 1.43), with only one partici-pant scoring above the midpoint (i.e., 5); this indicatesthat all participants had little or no experience withAfrican-Americans.

A positive relationship was anticipated betweenaverage contact score and the grotesqueness ratingscores for the other-race, Thatcherized, upright faces.However, the correlation between average contactscore on the SEQ and grotesqueness rating was notsignificant, r(19) = .202, p = .204. This finding is con-sistent with Ng and Lindsay (1994) and is likely dueto the limited range of contact scores obtained in thepresent data set.

Grotesqueness ratings

A within-subject ANOVA was carried out onthe grotesqueness ratings with factors of Race,Thatcherization, and Orientation. Grotesqueness rat-ings were affected by Race, F(1, 18) = 6.03, p= .024, MSE = 0.072, ηp

2 = .25; Thatcherization,

Figure 3. Mean grotesqueness rating for each face category.

F(1, 18) = 192, p < .001, MSE = 0.307, ηp2 =

.91; and Orientation, F(1, 18) = 10.3, p = .005,MSE = .233, ηp

2 = .36. Orientation interacted withboth Race, F(1, 18) = 10.8, p = .004, MSE =0.025, ηp

2 = .38, and Thatcherization, F(1, 18) =26.9, p < .001, MSE = 0.225, ηp

2 = .60. Inversionresulted in a larger decrease in overall grotesque-ness rating for other-race faces than it did for same-race faces. Similarly, inversion effects were onlyapparent in the Thatcherized condition, while non-Thatcherized faces received similar grotesqueness rat-ings regardless of Orientation. There was no inter-action between Race and Thatcherization, F(1, 18)< 1, p = .718, MSE = 0.023, ηp

2 = .01, sug-gesting that Thatcherization had an equal impact,overall, on same- and other-race faces. As predicted,a three-way interaction occurred, F(1, 18) = 11.0,p = .004, MSE = 0.019, ηp

2 = .38. However, asFigure 3 demonstrates, the direction of this inter-action was counter to our original prediction;Thatcherization, in the upright condition, had a greatereffect on other-race faces, rather than same-race faces.

The P1 component

P1 peak amplitudes were evaluated in a 100–130-mspost-stimulus search window. As seen in Figure 4,peak amplitudes were larger for other-race faces thansame-race faces, F(1, 18) = 58.3, p < .001, MSE =1.02, ηp

2 = .76, as well as for inverted faces com-pared to upright faces, F(1, 18) = 11.0, p = .004,

Dow

nloa

ded

by [

Uni

vers

ity o

f St

And

rew

s], [

Am

anda

Hah

n] a

t 08:

12 2

0 Ju

ly 2

011

THATCHERIZATION IMPACTS THE PROCESSING 7

Figure 4. Mean P1 peak amplitude values for each face category.

MSE = 8.56, ηp2 = .38. Amplitudes showed no effect

of Thatcherization, F(1, 18) = 1.90, p = .185, MSE= 2.00, ηp

2 = .10. Other-race faces elicited signifi-cantly larger peak P1 amplitudes than same-race faces,a result that may reflect differences in low-level visualproperties between Caucasian and African-Americanfaces. There was no interaction between Race andThatcherization, F(1, 18) < 1, p = .691, MSE =2.19, ηp

2 = .01, nor between Race and Orientation,F(1, 18) = 1.77, p = .200, MSE = 2.62, ηp

2 = .09.Orientation interacted with Thatcherization, F(1, 18)= 13.1, p = .002, MSE = 1.26, ηp

2 = .42, suchthat Thatcherization resulted in increased P1 ampli-tudes for the upright condition only, mirroring thebehavioral aspects of this illusion. Importantly, therewas a significant three-way interaction between Race,Thatcherization, and Orientation, F(1, 18) = 9.06,p = .008, MSE = 1.48, ηp

2 = .34, because theThatcherization by Orientation interaction was onlyapparent for the same-race faces. Thus, the P1 was sen-sitive to configural manipulations of the same-race butnot the other-race faces.

P1 Latencies were not affected by Race, F(1, 18)< 1, p = .646, MSE = 15.9, ηp

2 = .01; Thatcherization(F (1, 18) < 1, p = .588, MSE = 4.96, ηp

2 =.02; or Orientation, F(1, 18) = 2.56, p = .127, MSE= 31.5, ηp

2 =.13. Neither Thatcherization, F(1, 18)< 1, p = .756, MSE = 8.30, ηp

2 =.01, nor Orientation,F(1, 18) < 1, p = .879, MSE = 11.1, ηp

2 < .01,interacted significantly with Race; however, there wasa significant interaction between Thatcherization andOrientation, F(1, 18) = 7.01, p = .016, MSE = 5.66,ηp

2 = .28, such that inversion delayed the latencyfor Thatcherized faces more so than non-Thatcherized

Figure 5. P1 peak latencies for each face category.

faces. There was no three-way interaction, F(1, 18)< 1, p = .399, MSE = 11.8, ηp

2 = .04 (Figure 5).

The N170 component

Peak N170 amplitude was measured within a post-stimulus window of 140–210 ms. N170 amplitude wassignificantly increased following inversion, F(1, 18) =47.6, p < .001, MSE = 8.53, ηp

2 = .73. Other-racefaces tended to elicit larger N170 amplitudes, overall,than did same-race faces, F(1, 18) = 3.54, p = .076,MSE = 2.19, ηp

2 = .16, while Thatcherization resultedin slightly decreased N170 amplitudes, F(1, 18) =3.99, p = .061, MSE = 2.17, ηp

2 = .18. However,the latter effects were only marginally significant.Post-hoc paired comparison t-tests of same-race andother-race faces in the upright normal condition only,indicated that peak N170 amplitudes were not affectedby Race alone, t(18) = 0.028, p = .978, suggestingthat the marginally significant main effect of Racewas due to differences across experimental condi-tions rather than differential responses to same- versusother-race faces. Race did not interact with eitherThatcherization, F(1, 18) = 0.82, p = .378, MSE =1.27, ηp

2 = .04, or Orientation, F(1, 18) = 0.63, p =.44, MSE = 4.95, ηp

2 = .03. The interaction betweenThatcherization and Orientation failed to reach sig-nificance, F(1, 18) = 2.34, p = .144, MSE = 3.07,ηp

2 = .12. The expected three-way interaction wasonly marginally significant, F(1, 18) = 3.32, p =.085, MSE = 3.63, ηp

2 = .16; however, as predicted,Thatcherization modulated N170 amplitude for same-race faces to a greater extent than for other-race facesin the upright condition (Figure 6).

Dow

nloa

ded

by [

Uni

vers

ity o

f St

And

rew

s], [

Am

anda

Hah

n] a

t 08:

12 2

0 Ju

ly 2

011

8 HAHN, JANTZEN, SYMONS

Figure 6. Mean N170 peak amplitude values for each face cate-gory. Note that an increase along the vertical axis indicates greaternegativity of the N170 response. Error bars represent standard error.

N170 peak latency was longer for other-race facesthan same-race faces, F(1, 18) = 7.44, p = .014, MSE= 33.3, ηp

2 = .29. Inversion resulted in a marginallysignificant delay to peak N170 latency, F(1, 18) =4.23, p = .055, MSE = 128.9, ηp

2 = .19. Surprisingly,N170 latencies were not affected by Thatcherization,F(1, 18) = 2.11, p = .164, MSE = 25.2, ηp

2 = .11.There was no two-way interaction between Race andThatcherization, F(1, 18) < 1, p = .721, MSE = 6.85,ηp

2 = .01; Race and Orientation, F(1, 18) = 2.41, p =.138, MSE = 41.8, ηp

2 = .12; or Thatcherization andOrientation, F(1, 18) = 1.21, p = .285, MSE = 30.9,ηp

2 = .06. There was also no three-way interaction,F(1, 18) = 1.65, p = .216, MSE = 62.5, ηp

2 = .08(Figure 7).

Figure 7. Latency of the N170 response at peak amplitude. Errorbars represent SE.

The P250 component

Peak P250 amplitudes were measured during a searchwindow of 220–260 ms post-stimulus onset. NeitherRace, F(1, 18) = 2.58, p = .125, MSE = 2.30, ηp

2 =.13, nor Thatcherization, F(1, 18) = 1.43, p = .247,MSE = 1.40, ηp

2 = .07, impacted P250 amplitudes.Inverted faces had a significantly larger peak amplitudethan did upright faces, F(1, 18) = 16.8, p < .001, MSE= 6.45, ηp

2 = .48, and Orientation interacted withThatcherization, F(1, 18) = 9.57, p = .006, MSE =1.60, ηp

2 = .35, such that inversion increased the peakP250 amplitude more so for Thatcherized faces thannon-Thatcherized faces. Race did not interact withOrientation, F(1, 18) = 0.90, p = .354, MSE = 1.70,ηp

2 = .05; however, a marginally significant inter-action occurred between Race and Thatcherization,F(1, 18) = 3.81, p = .067, MSE = 1.37, ηp

2 = .18(Figure 8), such that Thatcherization modulated P250amplitude to a greater extent for same-race faces thanfor other-race faces. The three-way interaction, how-ever, failed to reach significance, F(1, 18) = 1.44, p =.246, MSE = 1.35, ηp

2 = .07.Neither Race, F(1, 18) = 2.52, p = .151, MSE =

28.4, ηp2 = .11, nor Thatcherization, F(1, 18) = .647,

p = .432, MSE = 21.4, ηp2 = .04, affected P250 laten-

cies. Orientation had a marginally significant effect,F(1, 18) = 3.93, p = .063, MSE = 259, ηp

2 = .18,such that inverted faces had a later peak latency thanupright faces. Thatcherization did not interact withRace, F(1, 18) < 1, p = .589, MSE = 45.7, ηp

2 =.02, or Orientation, F(1, 18) = 1.31, p = .268, MSE =

Figure 8. Peak P250 Amplitudes depicting the marginal interac-tion between race and Thatcherization (independent of orientation).Error bars represent standard error.

Dow

nloa

ded

by [

Uni

vers

ity o

f St

And

rew

s], [

Am

anda

Hah

n] a

t 08:

12 2

0 Ju

ly 2

011

THATCHERIZATION IMPACTS THE PROCESSING 9

38.08, ηp2 = .07. A marginally significant interac-

tion existed between Race and Orientation, F(1, 18) =3.29, p = .086, MSE = 81.6, ηp

2 = .16, such that inver-sion affected same-race faces more so than other-racefaces. The three-way interaction also failed to reachsignificance, F(1, 18) < 1, p = .968, MSE = 34.0,ηp

2 < .01.

DISCUSSION

In the present study, we sought to provide both behav-ioral and electrophysiological evidence for the exper-tise theory of the other-race effect. Expertise withown-race faces was expected to affect configural pro-cessing (Carey & Diamond, 1977), assessed using theThatcher effect. Behaviorally, Thatcherized faces wereperceived as more grotesque in appearance than werenormal faces. This grotesque appearance was expectedfor Thatcherized faces because the spatial-relationalinformation is in disagreement with the upright facialorientation. As such, it is obvious that there is some-thing “wrong” with the face.

With expertise leading to the use of configural pro-cessing strategies, we had predicted that the effectof Thatcherization on grotesqueness ratings wouldbe larger for same-race faces than other-race faces.However, the effect of Thatcherization was actu-ally less pronounced for same-race faces than forother-race faces. Upon further inspection of the data,we think this pattern may be due to the fact thatother-race faces were rated as slightly less grotesquewhen un-Thatcherized and slightly more grotesquewhen Thatcherized. The pattern of results seen in thegrotesqueness-rating data set suggests that the per-ception of other-race faces is actually just as relianton configural processing as, if not more so than, theperception of same-race faces. This finding does notreplicate that of Murray et al. (2003). In their study,the effects of component and configural disruptionsacross races were examined. Using Caucasian partic-ipants and Caucasian and Asian faces, they found nodifferences in susceptibility to component (i.e., featu-ral) distortions based on race. However, configurallydisrupted faces were rated as more bizarre in appear-ance if they were of the same race of the observer thanif they were of a different race.

Interestingly, but perhaps not surprisingly,the electrophysiological data do not parallel thegrotesqueness-rating data. When dealing with sen-sitive social issues, such as race, participants mayseek to respond in a socially or personally acceptablemanner, and this may require suppression of implicitbias and result in reluctance to rate faces of other races

as grotesque (as opposed to bizarre). In contrast, theevoked responses may reflect implicit categorizationoccurring prior to proposed bias detection and controlmechanisms (e.g., Stanley, Phelps, & Banaji, 2008).Indeed, this has been demonstrated behaviorally withimplicit measures of racism (Greenwald, McGhee, &Schwarz, 1998) and in face-processing studies show-ing that race-related alteration in the N170 precedesmore frontal cortical responses that correlate withimplicit bias (He, Johnson, Dovidio, & McCarthy,2009). Thus, because explicit behavioral responsescan be consciously edited, it is possible that the EEGdata provide a more accurate insight into differencesin facial processing strategies (although, notably, somedifferences in ERP amplitude and latencies may reflectsocial bias associated with racial categorization).

Effects on P1

The P1 component may reflect processing of low-level visual properties of stimuli such as shape, con-trast, and luminance (Itier & Taylor, 2004). Thus,the greater P1 response in viewing African-American(other-race) faces than Caucasian (same-race) facesmay be due in part to race-related differences in con-trast and color. We also found that, in the uprightcondition, Thatcherized faces yielded larger ampli-tudes and longer latencies. Milivojevic et al. (2003)found a similar Thatcherization effect on P1, whichthey attributed to an increase in captured attentiondue to the bizarreness or grotesque appearance ofThatcherized faces. Conversely, Boutsen et al. (2006)failed to show P1 modulation with Thatcherization.They attributed this lack of modulation to the lackof emotional expression in the facial stimuli becausethey used faces with neutral expressions rather than thesmiling faces of Milivojevic et al. (2003).

Surprisingly, Thatcherization affected P1 ampli-tudes only for own-race faces. This suggests that P1represents a response that is sensitive to more thanjust low-level features and attention. In particular, theeffect of expertise appears to be present in this veryearly component. Both same- and other-race faceswere rated as equally grotesque when Thatcherized(in the upright condition), yet only same-race facesshowed modulation of the P1 component. The currentdata suggest that same-race faces are being processedmore configurally than other-race faces and that whileP1 activity may reflect the processing of low-level fea-tures and attention, it appears to be also modulated byexperience and configural processing. Greater modula-tion for same-race faces is likely the result of increaseduse of configural processing strategies (as compared to

Dow

nloa

ded

by [

Uni

vers

ity o

f St

And

rew

s], [

Am

anda

Hah

n] a

t 08:

12 2

0 Ju

ly 2

011

10 HAHN, JANTZEN, SYMONS

other-race faces) and thus reflects a greater disruptionof the structural encoding process. A second inter-esting outcome of the P1 analyses was the apparentdisconnect between the P1 responses and the behav-ioral data. For our behavioral results, Thatcherizationdid not interact with race, but it did for the P1. Thiswould imply that the P1 response is occurring beforethe point of bias detection and control (e.g., Stanleyet al., 2008).

Effects on N170

N170 amplitudes were marginally reduced in mag-nitude following Thatcherization, suggesting that theN170 is sensitive to configural processing. Previousresearch appears mixed with respect to the influenceof Thatcherization on the N170 amplitude. A num-ber of studies report that Thatcherization increasesN170 amplitude (Carbon et al., 2005; Milivojevicet al., 2003), whereas more recent work and thepresent study report that Thatcherization decreasesN170 amplitude (Boutsen et al., 2006). This differ-ence in direction may be the result of differencesin the nature of the Thatcherization manipulation. Inmany Thatcherization studies to date, the Thatcherizedface versions have been created with choppily cutsections and minimal blending once these sectionshave been inverted. The edges created by this manip-ulation create extra information. Thus, the observedincreases in N170 amplitude may be due to someunspecified extra effort required to analyze this newinformation. Conversely, when the inverted featureshave been smoothed and blended, as in the presentstudy, the N170 is affected by only the configuraldisruption. The switch to featural analysis presum-ably requires less processing effort, thus resulting indecreased amplitude.

Thatcherization did not have an orientation-dependent effect on N170 amplitudes, althoughslightly greater changes in amplitude followingThatcherization were seen for upright as comparedto inverted faces. A number of previous ERP stud-ies have found similar Thatcherization effects in bothupright and inverted faces (Boutsen et al., 2006;Carbon et al., 2005). Boutsen et al. (2006) suggest thatThatcherization affects the encoding of both uprightand inverted faces, resulting in modulation of the face-sensitive N170 component. Again, greater modulationin the upright condition is likely due to the increasedavailability of configural information in upright versusinverted faces.

Thatcherization reduced N170 amplitude more for aface of one’s own race than for another race. Changes

in the N170 peak amplitude probably indicate differ-ences in the processing effort required to extract con-figural information from within faces (Zion-Golumbic& Bentin, 2007) or a switch in processing strategy(from configural to featural) when the configural infor-mation has been disrupted to an extent that extractionbecomes too difficult (Rossion & Gauthier, 2002).That Thatcherization modulated the N170 amplitudeto a greater extent for same-race faces suggests thatthe processing of same-race faces relies on configu-ral processing strategies more so than the processingof other-race faces, as in our original hypothesis.Previous work utilizing the face inversion effect (FIE)as a configural disruption in same- and other-race faceshas yielded complementary evidence (Rhodes, Brake,Taylor, & Tan, 1989; Vizioli, Foreman, Rousselet, &Caldara, 2010). Vizioli and colleagues (2010) havedemonstrated that the magnitude of N170 ampli-tude change due to inversion is larger for same-racefaces than other-race faces. While both inversion andThatcherization are considered configural distortions,Thatcherization may be a stronger test of configuralprocessing strategies, as it involves an explicit manip-ulation of the configural information of a given face.Because all of our participants had low contact scores,it can be concluded that none of them had sufficientexperience with African-Americans to have developedthe expertise necessary to utilize configural processingto the same extent as they would with Caucasian faces.

Effects on P250

The effects of inversion and Thatcherization on theP250 component in the present study are in line witha number of previous studies. We found the P250to be sensitive to face inversion, with inverted facesresulting in larger amplitudes and delayed peak laten-cies; similar inversion effects on P250 have beendescribed by Boutsen et al. (2006). Thatcherizationinteracted with orientation in the present data suchthat Thatcherization resulted in decreased P250 ampli-tude for the upright condition only. While Boutsenet al. found Thatcherization to lower P250 ampli-tude independently of orientation, Milivojevic et al.(2003) also demonstrated an orientation-dependenteffect of Thatcherization on P250 amplitudes. Theyargue that negativity of the P250 component likelyreflects the processing of configural information.Similarly, Halit et al. (2000) found that stretchingfaces to appear “atypical” (i.e., distorting the spa-tial relations among features) decreased P2 ampli-tudes (analogous to our P250 component). Theyargued that differences in observed P2 amplitudes

Dow

nloa

ded

by [

Uni

vers

ity o

f St

And

rew

s], [

Am

anda

Hah

n] a

t 08:

12 2

0 Ju

ly 2

011

THATCHERIZATION IMPACTS THE PROCESSING 11

for “typical” versus “atypical” faces were the resultof configural information being used to recognizeindividual faces. The Thatcherization effects on theP250 are, therefore, indicative of disruptions to localconfigural processing. Most interestingly, an interac-tion was observed between race and Thatcherization,with Thatcherization resulting in significantly greateramplitude reduction for own-race faces than other-racefaces. Following Milivojevic’s argument that modula-tion of the P250 component is indicative of disruptionsto local configural processing, the present findingssuggest that configural processing is utilized for theprocessing of own-race faces more so than for other-race faces.

Conclusions

When considered together, our EEG data appearto agree with previously observed behavioral data(Murray et al., 2003). Taken together, the findings ofthese studies provide evidence for the expertise the-ory of the other-race effect. The configural disruptionused in the present study impacted the processing ofsame-race faces more so, overall, than other-race faces.According to Farah, Wilson, Drain, and Tanaka (1998),configural and featural processing strategies for objectrecognition form a continuum, and the processing offaces lies on the configural end. The present study sug-gests that while faces hold a position on the configuralend of this processing strategy, it is possible to parseout subtle differences in the levels of configural andfeatural processing utilized for different types of faces,and that these differences may be occurring very earlyin the structural encoding stage of face processing,rather than later in processing when social judgmentsare thought to occur.

Original manuscript received 27 August 2010Revised manuscript accepted 14 April 2011

First published online day/month/year

REFERENCES

Ashley,V., Vuilleumier, P., & Swick, D. (2004). Time courseand specificity of event-related potentials to emotionalexpressions. Neuroreport, 15(1), 211–216.

Bartlett, J. C., & Searcy, J. (1993). Inversion and configu-ration of faces. Cognitive Psychology, 25, 281–316. doi:10.1006/cogp.1993.1007

Bentin, S., Allison, T., Puce, A., Perez, E., & McCarthy, G.(1996). Electrophysiological studies of face perception inhumans. Journal of Cognitive Neuroscience, 8, 551–565.

Blais, C., Jack, R. E., Scheepers, C., Fiset, D., & Caldara, R.(2008). Culture shapes how we look at faces. PloS ONE,3, e3022.

Bothwell, R. K., Brigham, J. C., & Malpass, R. S.(1989). Cross-racial identification. Personality and SocialPsychology Bulletin, 15, 19–25.

Bötzel, K., Schulze, S., & Stodieck, R. G. (1995). Scalptopography and analysis of intracranial sources of face-evoked potentials. Experimental Brain Research, 104,135–143.

Boutsen, L., & Humphreys, G. W. (2003). The effect ofinversion on encoding of normal faces and the Thatcherillusion. Quarterly Journal of Experimental Psychology,56A, 955–975.

Boutsen, L., Humphreys, G. W., Praamstra, P., & Warbrick,T. (2006). Comparing neural correlates of configuralprocessing of faces and objects: An ERP study of theThatcher illusion. NeuroImage, 32, 352–367.

Brigham, J. C., & Malpass, R. S. (1985). The role of experi-ence and context in the recognition of faces of own- andother-race persons. Journal of Social Issues, 4,139–155.

Caldara, R., Rossion, B., Bovet, P., & Hauert, C.-A. (2004).Event-related potentials and time course of the ‘other-race’ face classification. Cognitive Neuroscience andNeuropsychology, 15(5), 905–910.

Caldara, R., Thut, G., Servior, P., Michel, C., Bovet, P., &Renault, B. (2002). Face versus non-face object percep-tion and the ‘other-race’ effect: An event-related poten-tials study. Clinical Neurophysiology, 114(3), 515–528.

Caldara, R., Thut, G., Servoir, P., Michel, C., Bovet, P., &Renault, B. (2003). Face versus non-face object percep-tion and the “other-race” effect: A spatio-temporal ERPstudy. Clinical Neurophysiology, 114(3), 515–528.

Carbon, C.-C., Schweinberger, S. R., Kaufmann, J. M., &Leder, H. (2005). The Thatcher illusion seen by thebrain: An event-related potentials study. Cognitive BrainResearch, 24, 544–555.

Carey, S., & Diamond, R. (1977). From piecemeal toconfigurational representation of faces. Science, 195,312–314.

Chance, J. E., Turner, A. L., & Goldstein, A. G. (1982).Development of differential recognition for own- andother-race faces. Journal of Psychology, 112, 29–37.

Chiroro, P., & Valentine, T. (1995). An investigation of thecontact hypothesis of the own-race bias in face recogni-tion. Quarterly Journal of Experimental Psychology, 48,879–894.

Cunningham, W. A., Johnson, M. K., Gatenby, J. C., Gore,J. C., & Banaji, M. R. (2003). Neural componentsof social evaluation. Journal of Personality and SocialPsychology, 85, 639–649.

Cunningham, W. A., Johnson, M. K., Raye, C. L., Gatenby,J. C., Gore, J. C., & Banaji, M. R. (2004). Separableneural components in the processing of black and whitefaces. Psychological Science, 15, 806–813.

Diamond, R., & Carey, S. (1986). Why faces are and are notspecial: An effect of expertise. Journal of ExperimentalPsychology: General, 115, 107–117.

Eimer, M, (1998). Does the face-specific N170 componentreflect the activity of a specialized eye processor. NeuroReport, 9, 2945–2948.

Eimer, M. (2000a). Effects of face inversion on the struc-tural encoding and recognition of faces: Evidence fromevent-related brain potentials. Cognitive Brain Research,10, 145–158.

Eimer, M. (2000b). Event-related brain potentials distin-guish processing stages involved in face perception andrecognition. Clinical Neurophysiology, 111, 694–705.

Dow

nloa

ded

by [

Uni

vers

ity o

f St

And

rew

s], [

Am

anda

Hah

n] a

t 08:

12 2

0 Ju

ly 2

011

12 HAHN, JANTZEN, SYMONS

Eimer, M., & Holmes, A. (2002). An ERP study on thetime course of emotional face processing. Neuroreport,13, 427–431.

Farah, M. J., Tanaka, J. W., & Drain, H. M. (1995). Whatcauses the face inversion effect? Journal of ExperimentalPsychology: Human Perception and Performance, 21,628–634.

Farah, M. J., Wilson, K. D., Drain, M., & Tanaka, J.W. (1998). What is “special” about face perception?Psychological Review, 105, 482–498.

Freire, A., Lee, K., & Symons, L. A. (2000). The face-inversion effect as a deficit in the encoding of config-ural information: Direct evidence. Perception, 29(2),159–170.

Gauthier, I., & Tarr, M. J. (1997). Becoming a ‘greeble’expert: Exploring the face recognition mechanism. VisionResearch, 37, 1673–1682.

Goffaux, V., & Rossion, B. (2007). Face inversion dis-proportionately impairs the perception of vertical butnot horizontal relations between features. Journalof Experimental Psychology: Human Perception andPerformance, 33, 995–1002.

Golby, A. J., Gabrieli, J. D. E., Chiao, J. Y., & Eberhardt, J.L. (2001). Differential responses in the fusiform region tosame-race and other-race faces. Nature Neuroscience, 4,845–850.

Goodman, G. S., Sayfan, L., Lee, J. S., Sandhei, M., Walle-Olsen, A., Magnussen, S., et al. (2007). The developmentof memory for own- and other-race faces. Journal ofExperimental Child Psychology, 98, 233–242.

Greenwald, A. G., McGhee, D. E., & Schwarz, J. L.K. (1998). Measuring individual differences in implicitcognition: The Implicit Association Test. Journal ofPersonality and Social Psychology, 74(6), 1464–1480.

Halgren, E., Raij, T., Marinkovic, K., Jousmaki, V., &Hari, R. (2000). Cognitive response profile of the humanfusiform face area as determined by MEG. CerebralCortex, 10, 69–81.

Halit, H., de Haan, M., & Johnson, M. H. (2000). Modulationof event-related potentials by prototypical and atypicalfaces. Neuroreport, 11(9), 1871–1875.

Hart, A. J., Whalen, P. J., Shin, L. M., McInerney, S. C.,Fischer, H., & Rauch, S. L. (2000). Differential responsein the human amygdala to racial outgroup vs. ingroupface stimuli. Neuroreport, 11, 2351–2355.

He, Y., Johnson, M. K., Dovidio, J. F., & McCarthy, G.(2009). The relation between race-related implicit associ-ations and scalp-recorded neural activity evoked by facesfrom different races. Social Neuroscience, 4(5), 426–442.

Herrmann, M. J., Schreppel, T., Jager, D., Koehler, S., Ehlis,A. C., & Fallgatter, A. J. (2007). The other-race effect forface perception: An event-related potential study. Journalof Neural Transmission, 114(7), 951–957.

Höfel, L., & Jacobsen, T. (2007). Electrophysiologicalindices of processing aesthetics: Spontaneous orintentional processes? International Journal ofPsychophysiology, 65(1), 20–31.

Itier, R. J., Herdman, A. T., George, N., Cheyne, D., &Taylor, M. J. (2006). Inversion and contrast – rever-sal effects on face processing assessed by MEG. BrainResearch, 1115, 108–120.

Itier, R. J., & Taylor, M. J. (2002). Inversion and contrastpolarity reversal affect both encoding and recognitionprocesses of unfamiliar faces: A repetition study usingERPs. NeuroImage 15, 353–372.

Itier, R. J., & Taylor, M. J. (2004). Effects of rep-etition learning on upright, inverted and contrast-reversed face processing using ERPs. NeuroImage, 21,1518–1532.

Ito, T. A., & Urland, G. R. (2003). Race and genderon the brain: Electrocortical measures of attention torace and gender of multiply categorizable individuals.Journal of Personality and Social Psychology, 85, 616–626.

Ito, T. A., & Urland, G. R. (2005). The influence of process-ing objectives on the perception of faces: An ERP studyof race and gender perception. Cognitive, Affective, andBehavioral Neuroscience, 5, 21–36.

Jacques, C., & Rossion, B. (2007). Early electrophysio-logical responses to multiple face orientations correlatewith individual discrimination performance in humans.NeuroImage, 36(3), 863–876.

James, M. S., Johnstone, S. J., & Hayward, W. G.(2001). Event-related potentials, configural encoding,and feature-based encoding in face recognition. Journalof Psychophysiology, 15, 275–285.

Johnston, V. S., & Oliver-Rodriguez, J. C. (1997). Facialbeauty and the late positive component of event-related potentials. Journal of Sex Research, 24(2),188–198.

Levin, D. T. (2000). Race as a visual feature: Using visualsearch and perceptual discrimination tasks to understandface categories and the cross-race recognition deficit.Journal of Experimental Psychology: General, 129,559–574.

Lewis, M. B., & Johnston, R. A. (1997). The Thatcher illu-sion as a test of configural disruption. Perception, 26,225–227.

Lindsay, S., Jack, P. C., & Christian, M. A. (1991). Other-race face perception. Journal of Applied Psychology,76(4), 587–589.

Linkenkaer-Hansen, K., Palva, J. M., Sams, M., Hietanen,J. K., Aronen, H. J., & Ilmoniemi, R. J. (1998). Face-selective processing in human extrastriate cortex around120 ms after stimulus onset revealed by magneto-and electroencephalography. Neuroscience Letters, 253,147–150.

Malpass, R. S., & Kravitz, J. (1969). Recognition for facesof own and other race. Journal of Personality and SocialPsychology, 13, 330–334.

Marzi, T., & Viggiano, M. P. (2007). Interplay betweenfamiliarity and orientation in face processing: An ERPstudy. International Journal of Psychophysiology, 65(3),182–192.

McCarthy, G., Puce, A., Belger, A., & Allison, T. (1999).Electrophysiological studies of human face perception.II. Response properties of face-specific potentials gen-erated in occipitotemporal cortex. Cerebral Cortex, 9,431–444.

Meissner, C. A., & Brigham, J. C. (2001). Thirty yearsof investigating the own-race bias in memory for faces:A meta-analytic review. Psychology, Public Policy, andLaw, 7, 3–35.

Michel, C., Caldara, R., & Rossion, B. (2006a). Same-race faces are perceived more holistically than other-racefaces. Visual Cognition, 14, 55–73.

Michel, C., Rossion, B., Han, J., Chung, C.-S., & Caldara,R. (2006b). Holistic processing is finely tuned forfaces of one’s own race. Psychological Science, 17(7),608–615.

Dow

nloa

ded

by [

Uni

vers

ity o

f St

And

rew

s], [

Am

anda

Hah

n] a

t 08:

12 2

0 Ju

ly 2

011

THATCHERIZATION IMPACTS THE PROCESSING 13

Milivojevic, B., Clapp, W. C., Johnson, B. W., &Corballis, M. C. (2003). Turn that frown upside down:ERP effects of Thatcherization of misorientated faces.Psychophysiology, 40, 967–978.

Minear, M., & Park, D. C. (2004). A lifespan databaseof adult facial stimuli. Behavior Research Methods,Instruments, & Computers, 36, 630–633.

Murray, J. E., Rhodes, G., & Schuchinsky, M. (2003). Whenis a face not a face? The effects of misorientation on acti-vation of the face recognition module. In M. Peterson& G. Rhodes (Eds.), Perception of faces, objects, andscenes: Analytic and holistic processes (pp. 75–91).New York, NY: Oxford University Press.

Ng, W. J., & Lindsay, R. C. L. (1994). Cross-racefacial recognition. Journal of Cross-Cultural Psychology,25(2), 217–232.

Oliver-Rodríguez, J. C., Guan, Z., & Johnston, V. S. (1999).Gender differences in late positive components evoked byhuman faces. Psychophysiology, 36, 176–185.

Oostenveld, R., & Praamstra, P. (2001). The five per-cent electrode system for high-resolution EEG andERP measurements. Clinical Neurophysiology, 112,713–719.

Oram, M. W., & Perrett, D. I. (1992). Time course of neuralresponses discriminating different views of the face andhead. Journal of Neurophysiology, 68(1), 70–84.

Perrett, D. I., Heitanen, J. K., Oram, M. W., & Benson, P. J.(1992). Organization and functions of cells responsive tofaces in the temporal cortex. Philosophical Transactionsof the Royal Society of London. Series B, BiologicalSciences, 335, 23–30.

Pizzagalli, D. A., Lehmann, D., Hendrick, A. M., Regard,M., Pascual-Marqui, R. D., & Davidson, R. J. (2002).Affective judgments of faces modulate early activ-ity (approximately 160 ms) within the fusiform gyri.NeuroImage 16, 663–677.

Pizzagalli, D. A., Lehmann, D., Koenig, T., Regard, M., &Pascual-Marqui, R. D. (2000). Face-elicited ERPs andaffective attitude: Brain electric microstate and tomogra-phy. Clinical Neurophysiology, 111, 521–531.

Pizzagalli, D. A., Regard, M., & Lehmann, D. (1999).Rapid emotional face processing in the human right andleft brain hemispheres: An ERP study. Neuroreport, 10,2691–2698.

Rebai, M., Poiroux, S., Bernard, C., & Lalonde, R. (2001)Event-related potentials for category-specific informationduring passive viewing of faces and objects. InternationalJournal of Neuroscience, 106, 209–226.

Rhodes, G., Brake, S., Taylor, K., & Tan, S. (1989).Expertise and configural coding in face recognition.British Journal of Psychology, 80, 313–331.

Rossion, B., & Gauthier, I. (2002). How does the brainprocess upright and inverted faces? Behavioral andCognitive Neuroscience Reviews, 1, 62–74.

Rossion, B., Gauthier, I., Tarr, M. J., Despland, P.-A.,Linotte, S., Bruyer, R., et al. (2000). The N170 occipito-temporal component is enhanced and delayed to invertedfaces but not to inverted objects: An electrophyiological

account of face-specific processes in the human brain.Neuroreport, 11, 1–6.

Rossion, B., Joyce, C. A., Cottrell, G. W., & Tarr, M. J.(2003). Early lateralization and orientation tuning forface, word, and object processing in the visual cortex.NeuroImage, 20, 1609–1624.

Rossion, B., Kung, C. C., & Tarr, M. J. (2004). Visual exper-tise with nonface objects leads to competition with theearly perceptual processing of faces in the human occipi-totemporal cortex. Proceedings of the National Academyof Sciences of the United States of America, 101,14521–14526.

Schutter, D. J. L. G., de Haan, E. H. F., & van Honk,J. (2004). Functionally dissociated aspects in ante-rior and posterior electrocortical processing of facialthreat. International Journal of Psychophysiology, 53,29–36.

Stahl, J., Wiese, H., & Schweinberger, S. R. (2008).Expertise and own-race bias in face processing: An event-related potential study. Neuroreport, 19, 583–587.

Stanley, D., Phelps, E., & Banaji, M. (2008). The neu-ral basis of implicit attitudes. Current Directions inPsychological Science, 17, 164–170.

Tanaka, J. W., & Curran, T. (2001). A neural basis for expertobject recognition. Psychological Science, 12, 43–47.

Tanaka, J. W., Kiefer, M., & Bukach, C. M. (2004). Aholistic account of the own-race effect in face recogni-tion: Evidence from a cross-cultural study. Cognition, 93,B1–B9.

Thompson, P. (1980). Margaret Thatcher: A new illusion.Perception, 9, 483–484.

Vizioli, L., Foreman, K., Rousselet, G. A., & Caldara, R.(2010). Inverting faces elicits sensitivity to race on theN170 component: A cross-cultural study. Journal ofVision, 10(1), 1–23.

Walker, P. M., & Hewstone, M. (2006). A perceptual dis-crimination investigation of the own-race effect andintergroup experience. Applied Cognitive Psychology, 20,41–75.

Walker, P. M., Silvert, L., Hewstone, M., & Nobre, A. C.(2008). Social contact and other-race face processing inthe human brain. SCAN, 3, 16–25.

Walker, P. M., & Tanaka, J. W. (2003). An encoding advan-tage for own-race versus other-race faces. Perception,32(9), 1117–1125.

Werheid, K., Alpay, G., Jentzsch, I., & Sommer, W. (2005).Priming emotional facial expressions as evidenced byeventrelated brain potentials. International Journal ofPsychophysiology, 55, 209–219.

Yin, R. K. (1969). Looking at upside-down faces. Journal ofExperimental Psychology, 81, 141–145.

Young, A. W., Hellawell, D., & Hay, D. C. (1987).Configurational information in face perception.Perception, 16, 747–759.

Zion-Golumbic, E., & Bentin, S. (2007). Dissociated neuralmechanisms for face detection and configural encoding:Evidence from N170 and induced gamma-band oscilla-tion effects. Cerebral Cortex, 17(8), 1741–1749.

Dow

nloa

ded

by [

Uni

vers

ity o

f St

And

rew

s], [

Am

anda

Hah

n] a

t 08:

12 2

0 Ju

ly 2

011