How facial lesions impact attractiveness and perception: Differential effects of size and location

The LaryngoscopeVC 2011 The American Laryngological,Rhinological and Otological Society, Inc.

How Facial Lesions Impact Attractiveness and Perception:Differential Effects of Size and Location

Andres Godoy, MD; Masaru Ishii, MD, PhD; Patrick J. Byrne, MD; Kofi D. O. Boahene, MD;

Carlos O. Encarnacion, BS; Lisa E. Ishii, MD, MHS

Objectives/Hypothesis: To determine the effect of facial lesion size and location on perceptions of attractiveness andimportance for repair. We hypothesized that attractiveness scores and importance for repair would be dependent on lesionsize and location.

Study Design: Randomized controlled experiment.Methods: Forty-five subjects viewed 35 photographs of normal faces and faces with lesions of different sizes and loca-

tions. They rated attractiveness, how disfiguring, how bothered, and how important they considered repair.Results: Iterated factor analysis showed ‘‘bothered, disfigured, and important to repair’’ addressed the same domain, so

a disfigured/bothersome/repair factor score (DBRFS) was used. A mixed-effects regression model for attractiveness showedsmall-central and small-peripheral coefficients were not significantly different, v2(1) ¼ 0.03, P ¼ 1.000; but large-central andlarge-peripheral differences and small-central and large-peripheral differences were significantly different, v2(1) ¼ 10.34, P ¼0.004; and v2(1) ¼ 50.55, P < .001, respectively. DBRFS and attractiveness were poorly correlated (v ¼ �0.29). A mixed-effects regression for DBRFS showed small-central to large-central and the small-central to large-peripheral coefficients weresignificantly different, v2(1) ¼ 129.20, P < .001; and v2(1) ¼ 115.25, P < .001; but large-central to large-peripheral coeffi-cients were not, v2(1) ¼ 0.14, P ¼ 1.000.

Conclusions: The attractiveness penalty caused by a lesion was correlated with size but not location. Importance torepair was correlated with how disturbing and bothersome it was but not with how the lesion diminished attractiveness. Alllarge lesions and small central lesions were considered important to repair by observers. These results will help us predictthe true impact of lesions and support evidence-based treatment plans.

Key Words: Facial attention, facial deformity, facial perception.Laryngoscope, 121:2542–2547, 2011

INTRODUCTIONFor humans, faces are among the most captivating

visual stimuli we encounter. In our own unique way, weuse our faces to communicate ideas and emotion and toreceive information from the world we encounter aroundus. As such, facial deformities can cause profound socialhandicaps, including inhibition of normal social habits,increased emotional vulnerability, depression, and, ulti-mately, social alienation.1,2 This is a real concern as it isestimated that 10% of the population has some type offacial deformity.3 Given the impact of facial deformities,it is surprising that there is a paucity of data on the spe-cific characteristics of facial lesions that influence social

judgment and the extent to which they do so. Althoughit is probably safe to say that facial deformities affectfacial appearance, we should not assume all lesionsrender the same penalty. However, there is a gap in ourknowledge on how specific lesion features, such as sizeand location, contribute to their impact.

Facial lesions may be the result of a number ofcauses, including trauma, congenital abnormalities, orcancer. For example, basal cell carcinoma is the most com-mon malignancy in humans, affecting an estimated 2million people each year,4 with 70% occurring on theface.5 With an aging population we can expect more facialcutaneous malignancies and will be called on for anincreasing number of facial reconstructive procedures.This will occur concomitantly with an era of shrinkinghealth-care dollars. It is incumbent on us to responsiblyallocate health-care dollars by prioritizing resourceswhere they will have the greatest impact. To that end, abetter understanding of the potential value of our recon-structive techniques is necessary. The objective of thecurrent study was to define how specific lesion character-istics, size and location, impact social judgment by casualobservers as measured through their effect on attractive-ness and how bothersome, disturbing, and important torepair they were considered. First, we hypothesized thatfaces with lesions would be regarded as less attractive

From the Division of Rhinology (M.I.) and Division of Facial Plasticand Reconstructive Surgery (K.D.O.B., P.J.B., L.E.I.).; and Department ofOtolaryngology–Head and Neck Surgery (A.G.), Johns Hopkins School ofMedicine, Baltimore, Maryland, U.S.A.; and San Juan Bautista School ofMedicine (C.O.E.), San Juan, Puerto Rico.

Editor’s Note: This Manuscript was accepted for publication July25, 2011.

The authors have no funding, financial relationships, or conflictsof interest to disclose.

Send correspondence to Lisa E. Ishii, MD, MHS, Assistant Profes-sor, Department of Otolaryngology–Head and Neck Surgery, Johns Hop-kins School of Medicine, 601 North Caroline Street, Baltimore, MD 21287.E-mail: [email protected]

DOI: 10.1002/lary.22334

Laryngoscope 121: December 2011 Godoy et al.: Facial Lesion Size and Location

2542

than those without and that this effect would be depend-ent on the size and location of the lesion. Our furthersupposition was that facial lesions would be consideredbothersome, disturbing, and important to repair in a man-ner dependent on their size and location. We used Mohssurgery defects as our model for studying lesions.

MATERIALS AND METHODS

ParticipantsInstitutional review board approval was received for this

study. A group of 45 randomly selected subjects (25 females and20 males) participated in the study from March to July 2010.The investigator stood at the entrance to the Johns HopkinsOutpatient Center and approached every 10th person enteringthe building. The entrance is one used by faculty, staff, patients,and their families and was thus felt to be representative of thepopulation of the Baltimore-Washington catchment area. Sub-jects ranged in age from 18 to 56 years and were raciallyheterogeneous. Individuals were excluded as subjects if theyexperienced affective psychiatric conditions (schizophrenia, au-tism and related spectrum diseases) owing to establisheddifferences in the way individuals with those disorders viewand direct attention toward a face, as well as how they conveyand are able to infer emotional states from a person’s face.6 Thesubjects were naive with respect to the purpose of the study.

InstrumentThirty-five photographs were randomly selected from our

clinical archive of facial images using search terms to locate allnormal images and all images with lesions contained in our

archive. Images from patients who did not give informed con-sent to have their pictures used in research were excluded; theremaining images were then randomly selected for inclusionusing a random sampling without replacement approach. These35 faces were composed of seven normal faces (no facial defor-mity/lesion), seven faces with small central lesions, seven faceswith large central lesions, seven faces with small peripherallesions, and seven faces with large peripheral lesions.

Lesions were considered small if the surface area of thelesion was smaller than the surface area of the patient’s iris(approximately 8 mm). The surface area of an iris at conversa-tional distance corresponds roughly to the area of a visualfixation—meaning that the entire lesion can be assessed duringa single glance. Large lesions in our series require multiple fix-ations to fully sample the lesion (Fig. 1). These photographswere randomly assembled into a photographic survey.

For each picture presented, subjects were asked to ratethe attractiveness of the face on a scale of 1 to 10, with 10 beingthe most attractive. They were then asked if the presented facehad a lesion (yes or no) and, if so, if the wound was small orlarge. If a subject identified a lesion, they were then asked torank how disfiguring he/she thought the lesion was, how both-ered he/she was by the lesion, and how important it was torepair the lesion. These three metrics were also scored on a 10-point scale, with 10 being the most disturbing, bothersome, orimportant to repair and 1 being the least disturbing, bother-some, or important to repair.

ProcedureAfter recruitment, informed consent was obtained. Sub-

jects then answered a brief demographic questionnaire toensure they met the inclusion criteria. Study instructions weregiven to the subject, and then he or she was given the photo-graphic survey to fill out.

Data analysisData were maintained in an Excel spreadsheet. We per-

formed our statistical analysis using Stata 11 SE (StataCorporation, College Station, TX). Survey results were tabu-lated and are listed by lesion type (size and location) in Table I.First, histograms for the attractiveness scores for the normalfaces and faces with lesions were examined (Fig. 2). TheKruskal-Wallis equality of populations rank test was used toshow that the normal faces were viewed differently than faceswith lesions. Next, the correlation between the disfiguring,bothersome, and need to repair questions was measured andfound to be high, suggesting they all measured the same do-main, so iterated principal factor analysis was used to confirmthis conjecture. We used Kaiser’s criterion to determine thenumber of retained factors. Thompson scoring was then used tocombine the disfiguring, bothersome, and need to repair

Fig. 1. A large peripheral lesion that required multiple fixations tofully sample the lesion. [Color figure can be viewed in the onlineissue, which is available at wileyonlinelibrary.com.]

TABLE I.Rating Scores by Group Type.

Group Mean Attractiveness1,2 Bothered1,2 Disfiguring1,2 Repair1,2 DBRFS

Normal 6.04 (1.89) N/A N/A N/A N/A

Small central 4.97 (1.84) 6.11 (2.60) 4.77 (2.32) 6.30 (2.57) �0.170 (0.94)

Small peripheral 5.00 (1.97) 4.63 (2.54) 3.33 (1.90) 4.69 (2.75) �0.739 (0.91)

Large central 3.73 (1.72) 7.70 (2.28) 6.65 (2.39) 7.83 (2.17) 0.473 (0.75)

Large peripheral 4.13 (1.79) 7.64 (2.34) 6.66 (2.47) 7.84 (2.34) 0.440 (0.803)

Standard deviations are shown in parentheses.DBRFS ¼ disfigured/bothersome/repair factor score; N/A ¼ not applicable.


2543

variables into a single normalized factor score. The factor scoreand the attractiveness score were used for mixed-effects regres-sion analysis to determine how lesions affect casual observers’perception of faces. For the attractiveness regression we usedfour fixed covariates, allowing us to study the effects that smallcentral, large central, small peripheral, and large peripherallesions have on attractiveness. Because the faces and observerswere selected at random, we assumed that there were two rea-sons for the spread in the attractiveness at each level: 1) thedifferences in attractiveness naturally encountered in a randompopulation and 2) intrinsic differences in the way individualtest subjects judge attractiveness. We account for these twosources of variation using a random intercept term to treatsources of variation introduced by subjects and a random errorterm to account for the intrinsic variability in attractiveness ofthe randomly sampled faces, in addition to traditional sourcesof error encountered in regression modeling.

The factor scores are only valid for faces with lesions, sowe chose small peripheral lesions as a reference and includedthree dummy variables indicating the presence of small central,large central, and large peripheral lesions as our covariates forthe fixed effects. The same sources of variation exist for the fac-tor scores as in the attractiveness score, so we used the samerandom terms for the factor score regression.

Post hoc testing was used to further understand modelresults. Bonferroni’s method was used to adjust the P values formultiple comparisons. We set the experiment-wide alpha at0.05. Because Stata does not offer a multivariate mixed-effectsregression option, we regressed each term separately and usedthe Bonferroni method to adjust significance levels for eachregression accordingly.

RESULTS

DemographicsThe subjects were heterogeneous in terms of sex,

income status, and racial distribution.

Analysis of Rating ScoresHistograms of the attractiveness scores for the nor-

mal faces and faces with lesions are shown in Figure 2.

The histograms appear skewed and display distinct cent-roids. The mean attractiveness score for the normalfaces was 6.04 (standard error [SE]: 0.124), and the ranksum was 184,680 with 231 observations; the meanattractiveness score for the faces with lesions was 4.48(SE: 0.061), and the rank sum was 499,185 with 938observations. The difference in rank sums was statisti-cally significant, v2ties(1) ¼ 118.86, P < .001, suggestingthat faces with lesions are less attractive than normalfaces. To understand this effect, we studied the datamore carefully.

There are no good objective measures for determin-ing how disfiguring or bothersome a facial lesion is, sowe queried observers about how bothered they were byeach lesion, how disfiguring they found each lesion, andhow important they thought it was to repair each lesion.Average scores and standard deviations by lesion typeare shown in Table I. The results were highly correlated(product momentum correlation coefficients [q]: botheredto disfiguring 0.906; bothered to repair 0.957; disfiguringto repair 0.882; all P values < .001). We used factoranalysis to determine how many independent pieces ofinformation these three questions contained. Iteratedprincipal factor analysis showed that nearly 100% oftotal variance was extracted by the first factor; thisbehavior suggests that all three questions probed thesame domain, and only one factor was retained for fur-ther analysis. We assumed that the retained factor givesa measure of how disfiguring/bothersome a lesionappears to a casual observer because of the nature of thequestions. Next we used Thompson (regression) scoringto develop an index of how disfigured/bothersome/impor-tant to repair a lesion appeared based on the observer’sbothersome, disfiguring, and need to repair scores. (Thefinal factor score was found to be a weighted sum of thestandardized bothersome, disfiguring, and repair scoreswith weights 0.747, 0.072, and 0.191, respectively.)The resulting disfigured/bothersome/repair factor score(DBRFS) is tabulated in Table I and was used for

Fig. 2. Histogram of attractiveness scores for normal faces and faces with lesions. [Color figure can be viewed in the online issue, which isavailable at wileyonlinelibrary.com.]


2544

further analysis. The factor scores ranged from �2.081to 1.273, with larger negative numbers indicating a lessbothersome or disturbing lesion with a low need forrepair and larger positive numbers indicating the oppo-site. The DBRFS and the attractiveness scores arepoorly correlated (q ¼ �0.29), confirming our conjecturethat they measure different aspects of how casualobservers view faces.

To understand how lesions affect perception, weturn to regression analysis. The regression results forattractiveness scores are displayed in Table II and TableIII. The results show that normal faces have an averageattractiveness score of 6.015 (equal to the constantterm). The fixed portion of the model shows that smalllesions decrease attractiveness by approximately 1 point,and large lesions decrease attractiveness by roughly 2points, with central lesions decreasing slightly morethan 2 points and peripheral lesions slightly less. Allcoefficients were statistically significant. Post hoc testingshowed that the small-central and small-peripheral coef-ficients were not statistically significantly different, v2(1)¼ 0.03, P ¼ 1.000; that is, there was no difference in thedecrease in attractiveness caused by small peripheraland small central lesions, and the large-central andlarge-peripheral differences and the small-central andlarge-peripheral differences were statistically signifi-cantly different, v2(1) ¼ 10.34, P ¼ .004; and v2(1) ¼50.55, P < .001, respectively; Bonferroni adjusted P val-ues given for all post hoc tests.

The DBRFS regression results are shown in TableIV and Table V. As with the attractiveness scores, thevariability in scores between subjects and within individ-ual subject’s scores is similar. Unlike the attractivenessscores, there is a dramatic difference between how sub-jects scored a small peripheral lesion compared to asmall central lesion, with small-peripheral lesions beingnot particularly disturbing or important to repair andsmall central lesions being disturbing and important to

repair. Large lesions are more disturbing/bothersomeand important to repair than small lesions, with largecentral lesions being more so than large peripherallesions. Post hoc testing shows that the small-central tolarge-central and the small-central to large-peripheralcoefficients are statistically significantly different, v2(1)¼ 129.20, P < .001; and v2(1) ¼ 115.25, P < .001; thelarge-central to large-peripheral coefficients are not,v2(1) ¼ 0.14, P ¼ 1.000. The Bonferroni method wasused to adjust P values for multiple comparisons.

DISCUSSIONThese novel findings provide the first evidence that

there are differences in the way faces with lesions areregarded that are dependent on the size and location ofthe lesion. Faces with lesions were considered lessattractive than those without, and this effect variedbased on lesion size but not location. Interestingly, theimportance the subjects in our study placed on repairinga facial lesion was highly correlated with how disturbedor bothered they were by the lesion but not correlatedwith the attractiveness penalty caused by the lesion. Inother words, the more bothersome or disturbing theyfound the lesion, the more important they considered itto repair; but it was not more important to repair by vir-tue of its ability to diminish attractiveness alone. Alsonotable from this study is the finding that observersdeem large lesions and lesions located within the centraltriangle as important to repair. As a whole, these resultshave implications for evidence-based treatment plansand may justify the cost of more complex repairs andreferrals to specialists for repair of these lesions, scar re-vision, and multistage repairs. We expand on thesefindings herein.

It is estimated that 10% of the population has sometype of facial deformity, and this number is expected torise with our aging population.2 Given the significance

TABLE II.Attractiveness Fixed Effects Results.

CovariateRegressionCoefficient

StandardError P Value

97.5%ConfidenceInterval

Small central �1.027 0.124 <.001 (�1.306, �0.749)

Small peripheral �1.001 0.125 <.001 (�1.285, �0.725)

Large central �2.327 0.124 <.001 (�2.605, �2.048)

Large peripheral �1.925 0.126 <.001 (�2.206, �1.644)

Constant 6.015 0.202 <.001 (5.56- 6.467)

TABLE III.Attractiveness Random Effects Results.

RandomEffectsParameters Estimate

StandardError

97.5%ConfidenceInterval

SD: subject term 1.215 0.135 (0.920-1.604)

SD: residual 1.341 0.028 (1.239-1.450)

SD ¼ standard deviation.

TABLE IV.Disfigured/Bothersome/Repair Factor Scores

Fixed Effects Results.

CovariateRegressionCoefficient

StandardError P Value

95%ConfidenceInterval

Small central 0.557 0.059 <.001 (0.425-0.689)

Large central 1.211 0.058 <.001 (1.082-1.342)

Large peripheral 1.189 0.058 <.001 (1.059-1.319)

Constant �0.786 0.0936 <.001 (�0.996, �0.577)

TABLE V.Disfigured/Bothersome/Repair Factor Score

Random Effects Results.

RandomEffectsParameters Estimate

StandardError

95%ConfidenceInterval

SD: subject term 0.562 0.063 (0.436-0.724)

SD: residual 0.623 0.015 (0.591-0.657)

SD ¼ standard deviation.


2545

of facial deformities as lesions that cannot be easily con-cealed from view, it is remarkable that there are limiteddata showing which characteristics of facial deformitiesaffect salient social perceptions, such as attractive-ness.1,7 Ishii et al.8 were the first to provide objectivedata demonstrating that observers allocate attention dif-ferently when gazing on faces with deformities ascompared to normal faces. However, until now there hasbeen no specific exploration into the penalty caused bylesions based on features of the lesion such as size andlocation. This is important because we cannot assumeall lesions have the same effect on perception, and weshould be able to more reliably predict the impact of dif-ferent types of lesions.

In this study, the faces with lesions were perceivedas markedly less attractive than normal faces; these find-ings were consistent with other data in the literature.7

Lesion size had a substantial effect on attractiveness,with large lesions decreasing attractiveness by twostandard deviations, and small lesions decreasing attrac-tiveness by one standard deviation. Surprisingly, lesionlocation in and of itself, central versus peripheral, did notaffect attractiveness. Conversely, all large lesions, locatedcentrally and peripherally, were considered bothersome,disturbing, and important to repair, but only small centrallesions were similarly considered bothersome, disturbing,and important to repair. Interestingly, small peripherallesions were not considered particularly bothersome, dis-turbing, and important to repair. These findings meritparticular consideration in light of the allocation of lim-ited health-care resources and the importance ofpracticing evidence-based medicine. For those lesions con-sidered important to repair by society, it is incumbent onus to ensure satisfactory reconstruction. These lesionswarrant attention by specialists trained in soft-tissuereconstruction for optimal results. However, for that classof lesions that society regards as not bothersome, disturb-ing, or important to repair, perhaps less complexreconstructive techniques are warranted, with resourceinvestment reserved for the former class of defects. Fromthe approach of evidence-based medicine, these data pro-vide evidence on societal expectations about appearancethat should factor into our management algorithm. How-ever, although society may deem small peripheral lesionsunimportant to repair, surgeons should remain cognizantof an intrinsic bias they may have against repairingthem. Repair will be particularly important for individualpatients who place a high premium on attractiveness, asthe attractiveness penalty for small peripheral lesions isstill significant.

Given the differential impact of lesion characteristicson attractiveness ratings as compared to feelings of dis-turbance/bothered/need to repair, perhaps there is adifferent mechanism for the perception of these two qual-ities. One would speculate that small peripheral lesionswould be gazed on for shorter periods of time than centrallesions of equal size, but they result in the same penaltyto attractiveness scores. This suggests the perception ofattractiveness is not mitigated through accumulatedattentional distraction, that is, time spent gazing on thelesion, but rather through a single glance mechanism.

There is a large body of literature showing thatpatients who undergo changes that diminish theirattractiveness are vulnerable to depression.1,2,7,9,10 Ourdata showed that large lesions caused dramatic changesin attractiveness ratings, with ratings plummeting twostandard deviations from the normal attractiveness dis-tribution; that is, an individual of average attractivenessdrops to the bottom 15% on the attractiveness scale witha large lesion. The impact caused by small lesions wasless dramatic although still substantial, with theseattractiveness ratings dropping by one standard devia-tion. These data are important to consider whencounseling patients because they validate the patientexperiences described in quality-of-life surveys.11–13

Lesions diminish attractiveness, and individuals withdiminished attractiveness are vulnerable to depression.For all individuals, but particularly those for whomattractiveness is of great value and tied to self worth,the impact of facial lesions, large and small, cannot beminimized. Our reconstructive procedures have thepotential to markedly improve the quality of life of ourpatients. The data are also important to consider foreducating our colleagues in other specialties where moreattention may be focused on systemic illness manage-ment as compared to facial lesions, although both clearlyhave great potential to impact overall patient well-being.

It is important to appreciate that subjects displaysignificant variability in how they rank attractivenessand how disfiguring/bothersome/important to repair alesion is, with in between subject variability being ofsimilar order to within subject variability (see Tables IIIand V). This implies that there is a lot of variability inthe way individuals rate faces, and two people may ratethe same face differently. This variability must beaccounted for in the design and analysis of experiments,and physicians must be sensitive to this matter whendiscussing lesions with patients, because the patient’sperception of how disfiguring a lesion is may be quitedifferent from the physician’s.

All of the photographs were randomly selectedfrom our clinical archive of photographs and consistedprimarily of patients with lesions resulting from theresection of skin cancers. The normal photographswere randomly selected from our clinical archive. Weare not aware of data that show skin cancers occur onpeople that are as a group less attractive, and thuswe assume that the only difference in the facesbetween the groups is the presence of the lesion itself.Thus, the concerns of the patients regarding fear ofnegative judgment are based in reality and should berecognized as such.

The study does have limitations that will beaddressed in further investigation. For this pilot study, alimited number of faces were included. There were sevenfaces included in each of the six groups, for a total of 42faces. Although this number was adequate for this pilotproof-of-principle study, we recognize that it limits thegeneralizability of the results. We will expand this num-ber in future studies to make it more generalizable. Wewill also expand the number of subjects. Also, weincluded lesions that differed by size and location, and


2546

then grouped them dichotomously (small/large, central/peripheral). It will be interesting to expand on thesedata to try to determine if there are more discretethresholds for changes in perception.

CONCLUSIONIn the current study we showed there was vari-

ability in how lesions affected facial perception based onlesion size and location. Attractiveness was correlatedwith lesion size but not location. Importance to repairwas directly correlated with how bothersome or dis-turbing the lesion was considered but not withattractiveness. The results provide insight into howspecific types of lesions affect our patients. This mayenable us to reliably predict their impact and devise anevidence-based treatment approach.

BIBLIOGRAPHY

1. Valente SM. Visual disfigurement and depression. Plast Surg Nurs 2009;29:10–6; quiz 7–8.

2. Valente SM. Visual disfigurement and depression. Plast Surg Nurs 2004;24:140–146; quiz 7–8.

3. Robinson E, Partridge J. The Psychology of Facial Disfigurement: A Guidefor Health Professionals. London: Changing Faces; 1996.

4. Society AC. Cancer facts and figures 2010. Available at: http://www.cancer.org/AboutUs/DrLensBlog/post/2008/08/27/Can-Skin-Cancer-Be-Prevented-If-You-Are-At-Risk.aspx. Accessed July 5, 2011.

5. Erba P, Farhadi J, Wettstein R, Arnold A, Harr T, Pierer G. Morphoeic ba-sal cell carcinoma of the face. Scand J Plast Reconstr Surg Hand Surg2007;41:184–188.

6. Nakano T, Tanaka K, Endo Y, et al. Atypical gaze patterns in children andadults with autism spectrum disorders dissociated from developmentalchanges in gaze behaviour. Proc Biol Sci 2010;277:2935–2943.

7. Levine E, Degutis L, Pruzinsky T, Shin J, Persing JA. Quality of life andfacial trauma: psychological and body image effects. Ann Plast Surg2005;54:502–510.

8. Ishii L, Carey J, Byrne P, Zee DS, Ishii M. Measuring attentional bias toperipheral facial deformities. Laryngoscope 2009;119:459–465.

9. Orr DA, Reznikoff M, Smith GM. Body image, self-esteem, and depressionin burn-injured adolescents and young adults. J Burn Care Rehabil1989;10:454–461.

10. Ogden J, Lindridge L. The impact of breast scarring on perceptions ofattractiveness: an experimental study. J Health Psychol 2008;13:303–310.

11. Patrick H, Neighbors C, Knee CR. Appearance-related social comparisons:the role of contingent self-esteem and self-perceptions of attractiveness.Pers Soc Psychol Bull 2004;30:501–514.

12. Noles SW, Cash TF, Winstead BA. Body image, physical attractiveness,and depression. J Consult Clin Psychol 1985;53:88–94.

13. Connors J, Casey P. Sex, body-esteem and self-esteem. Psychol Rep 2006;98:699–704.


2547

Documents

How facial lesions impact attractiveness and perception: Differential effects of size and location