《社会心理学》课程教学资源（文献资料）Close relationship——Genetic Factors That Increase Male Facial Masculinity Decrease Facial Attractiveness of Female Relatives

Psychological Science OnlineFirst,published on December 30,2013 as doi:10.1177/0956797613510724 aos Research Article PSYCHOLOGICAL SCIENCE Genetic Factors That Increase Male Facial ints and per Masculinity Decrease Facial Attractiveness of Female Relatives SAGE aret J.Wright Nich Martin Brendan P.ZietschMatthew C.Keller,and nsland:De for Be Abs en,choosing a facially masculine man as a mate is thought to confer genetic benefits to offspring.Crucial assumptions of this hypothesis have not been adequately tested.It has been assumed that variation in facial masculinity is due to genetic variation and that genetic factors that increase male facial masculinity do not increase facial masculinity cral mas linity in photos ntical 1)an0 nonidenti and female facial n also found that n ale faces is ated to thei attractiveness and that facially masculine men tend to have facially masculine,less-attractive sisters.These findings challenge the idea that facially masculine men provide net genetic benefits to offspring and call into question this popular theoretical framework Keywords sexual dimorphism,intralocus sexual conflict,evolution,immunocompetence-handicap principle,good genes. pathogen,sexually antagonistic selection Received 4/22/13:Revision accepted 9/29/13 Little,DeBruine,&Jones The studies just cited other features such as shading or texture The widely culine mates in circumstances thought to increase the accepted interpretation of these findings is that male relative importance of indirect benefits of mate choice facial masculinity is a signal of genetic quality ("good (.e.,genetic bene toopring)as opposed to direct )and women ave accord ngly evolved to en od&si for facially masculine men when considering a short-term 2011:Roberts Little 2008) Facial masculinity is thought to be an honest signal of 2002),dun nng the fe ile phase of the menstrual Penton-Voaketal 1999 whe Jones Depruine.200),when ness is high (Little,Burt,Penton-Voak,Perrett,2001), Co

Psychological Science 201X, Vol XX(X) 1–9 © The Author(s) 2013 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/0956797613510724 pss.sagepub.com Research Article A large body of research has shown that women attend to facial masculinity when assessing potential mates. Women tend to show greater preference for facially mas￾culine mates in circumstances thought to increase the relative importance of indirect benefits of mate choice (i.e., genetic benefits to offspring) as opposed to direct benefits of mate choice (e.g., resource provision, protec￾tion). For example, women show increased preference for facially masculine men when considering a short-term or extrapair partner (Little, Jones, Penton-Voak, Burt, & Perrett, 2002), during the fertile phase of the menstrual cycle (Gangestad, Thornhill, & Garver-Apgar, 2010; Penton-Voak et al., 1999), when sex drive is high (Welling, Jones, & DeBruine, 2008), when self-perceived attractive￾ness is high (Little, Burt, Penton-Voak, & Perrett, 2001), and when pathogens are prevalent or health is threat￾ened (DeBruine, Jones, Crawford, Welling, & Little, 2010; Little, DeBruine, & Jones, 2011). The studies just cited focused largely on masculine face shape as opposed to other features, such as shading or texture. The widely accepted interpretation of these findings is that male facial masculinity is a signal of genetic quality (“good genes”) and that women have accordingly evolved to attend to facial masculinity when choosing mates (Gangestad & Simpson, 2000; Little, Jones, & DeBruine, 2011; Roberts & Little, 2008). Facial masculinity is thought to be an honest signal of genetic quality because of the immunosuppressive effects of testosterone (Folstad & Karter, 1992). The idea is that 510724PSSXXX10.1177/0956797613510724Lee et al.Genetics of Facial Masculinity research-article2013 Corresponding Author: Brendan P. Zietsch, School of Psychology, McElwain Building, University of Queensland, St Lucia, Queensland 4029, Australia E-mail: zietsch@psy.uq.edu.au Genetic Factors That Increase Male Facial Masculinity Decrease Facial Attractiveness of Female Relatives Anthony J. Lee1 , Dorian G. Mitchem2,3, Margaret J. Wright4 , Nicholas G. Martin4 , Matthew C. Keller2,3, and Brendan P. Zietsch1,4 1 School of Psychology, University of Queensland; 2 Department of Psychology and Neuroscience, University of Colorado Boulder; 3 Institute for Behavioral Genetics, University of Colorado Boulder; and 4 QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia Abstract For women, choosing a facially masculine man as a mate is thought to confer genetic benefits to offspring. Crucial assumptions of this hypothesis have not been adequately tested. It has been assumed that variation in facial masculinity is due to genetic variation and that genetic factors that increase male facial masculinity do not increase facial masculinity in female relatives. We objectively quantified the facial masculinity in photos of identical (n = 411) and nonidentical (n = 782) twins and their siblings (n = 106). Using biometrical modeling, we found that much of the variation in male and female facial masculinity is genetic. However, we also found that masculinity of male faces is unrelated to their attractiveness and that facially masculine men tend to have facially masculine, less-attractive sisters. These findings challenge the idea that facially masculine men provide net genetic benefits to offspring and call into question this popular theoretical framework. Keywords sexual dimorphism, intralocus sexual conflict, evolution, immunocompetence-handicap principle, good genes, pathogen, sexually antagonistic selection Received 4/22/13; Revision accepted 9/29/13 Psychological Science OnlineFirst, published on December 30, 2013 as doi:10.1177/0956797613510724 Downloaded from pss.sagepub.com by Cai Xing on January 7, 2014

Lee et al. only men with good innate immune functioning can changes the ass ound. n th tephen,Clark.Penton 1975)Supporting this im handicap hypothesis.research shows that facial mascu- quantify the masculinity of facial shape in photographs linity is positively associated with circulating testosterone of a large sample of identical and nonidentical (same-sey nton-Voak Chen,2004),and ma and opposit metric. Chan.Simmons,2003 Thorhl remale facial mas Fin sexual conflict by assessing the correlation in facial mas. Gangestad,2006).An alternative (or additional)explana culinity betw in the for greater attractiveness of male offspring.This situation can create a self-reinforcing runaway effect that exaggerates ce and the preferred trait (Fisher,1915; Method 20 The iden that ulinity signals heritabl Participants genetic quality,manifested as immunoco petence.sexv Participants were 1.193 individual twins and 106 of their sons,or both,has gained broad acceptance(Ga gestad Scheyd,2005;Gangest 2010 1 Clark.Boothrovd.Penton-voak 2012)However.this day (mean age =16.03 years,SD=0.47 years),and their idea depends on two key assumptions that have not been siblings were tested and photographed as close as pos adequately tested.First, is assume that male facia of thu participant tion):otherwise it could not be inherited by offs rom the Human Research Ethics Committee at OIMR Berghofer. genes that increase male facia mascu lnity are by any genetic benefits to male offspring would be counter acted by the detriment to female offspring(this is termed n2011 ak digi waves wer Comwell and 2008)em pirically addr under standard indoor lighting these assumptions by analyzing ratings of masculinity measures of masculinity and subjective ratings of mascu- sof the faces linity and attractiveness were obtained from these wer ndent raters identified a total of 18 land members of a standard nuclear family qually share both marks on each face.Raters we ined fo eral week genes and family environment,which are therefore com in hour-long sessions in which landmarks were defined pletely confounded In another study,Mitchem et a anatomically.Figure I shows the location of each land al ph or mon and mark.I were randoml n Ior eac of and family s to dis as them pixel location chosen by these and attractiveness again,however,no objective mea Photographs of participants were not originally taken sures were used.It has been show previously that sub for shape analysis. re,the photographs varie on addit ways tha than morphologica masculinity, icipant's capt

2 Lee et al. only men with good innate immune functioning can afford to support the levels of testosterone required to develop masculine facial features (Folstad & Karter, 1992; Zahavi, 1975). Supporting this immunocompetence￾handicap hypothesis, research shows that facial mascu￾linity is positively associated with circulating testosterone levels (Penton-Voak & Chen, 2004), and male facial mas￾culinity has been found to correlate positively with both perceived and actual health (Rantala et al., 2012; Rhodes, Chan, Zebrowitz, & Simmons, 2003; Thornhill & Gangestad, 2006). An alternative (or additional) explana￾tion of the relevance of male facial masculinity to genetic quality is the sexy-son hypothesis, according to which the genetic benefits to offspring come in the form of greater attractiveness of male offspring. This situation can create a self-reinforcing runaway effect that exaggerates both the preference and the preferred trait (Fisher, 1915; Huk & Winkel, 2008). The idea that male facial masculinity signals heritable genetic quality, manifested as immunocompetence, sexy sons, or both, has gained broad acceptance (Gangestad & Scheyd, 2005; Gangestad & Simpson, 2000; Little, Jones, et al., 2011; Perrett et al., 1998; Rantala et al., 2012; Roberts & Little, 2008; although see Puts, 2010; Scott, Clark, Boothroyd, & Penton-Voak, 2012). However, this idea depends on two key assumptions that have not been adequately tested. First, it is assumed that male facial masculinity is substantially heritable (i.e., a substantial proportion of the variation is due to additive genetic vari￾ation); otherwise, it could not be inherited by offspring and could not signal good genes. Second, it has been assumed that the genes that increase male facial mascu￾linity are not detrimental to female offspring (e.g., by increasing their facial masculinity, which has been shown previously to decrease female attractiveness); otherwise, any genetic benefits to male offspring would be counter￾acted by the detriment to female offspring (this is termed intralocus sexual conflict; see Bonduriansky & Chenoweth, 2009; Garver-Apgar, Eaton, Tybur, & Thompson, 2011). Cornwell and Perrett (2008) empirically addressed these assumptions by analyzing ratings of masculinity and attractiveness of the faces in family photographs. However, no objective measures of masculinity were used, and heritability could not be estimated because members of a standard nuclear family equally share both genes and family environment, which are therefore com￾pletely confounded. In another study, Mitchem et al. (2013) used facial photos of monozygotic (identical) and dizygotic (nonidentical) twins to distinguish the influence of genes and family environment on facial masculinity and attractiveness; again, however, no objective mea￾sures were used. It has been shown previously that sub￾jective ratings of masculinity are based on additional factors other than morphological masculinity, which changes the association with traits such as attractiveness (Scott, Pound, Stephen, Clark, & Penton-Voak, 2010). In the research reported here, we used geometric mor￾phometrics, the statistical analysis of shape, to objectively quantify the masculinity of facial shape in photographs of a large sample of identical and nonidentical (same-sex and opposite-sex) twins and siblings. Using biometrical modeling, we estimated the heritability of male and female facial masculinity. Finally, we tested for intralocus sexual conflict by assessing the correlation in facial mas￾culinity between opposite-sex twins and siblings, and we investigated the relationship, for each sex, between the objective masculinity and rated attractiveness of the photographs. Method Participants Participants were 1,193 individual twins and 106 of their siblings from 575 families who took part in the Genes for Cognition study and were part of the Brisbane Adolescent Twin Studies (Wright & Martin, 2004). Twins were tested and photographed as close as possible to their 16th birth￾day (mean age = 16.03 years, SD = 0.47 years), and their siblings were tested and photographed as close as pos￾sible to their 18th birthday (mean age = 17.80, SD = 0.46). All participants gave informed written consent, and approval to code and analyze these data was obtained from the Human Research Ethics Committee at QIMR Berghofer. Photographs Photographs of participants were taken between 1996 and 2010. In the earliest waves of data collection, photo￾graphs were taken using film cameras and later scanned to digital format. Photographs from later waves were taken with digital cameras. Each photograph was taken under standard indoor lighting conditions. Objective measures of masculinity and subjective ratings of mascu￾linity and attractiveness were obtained from these photographs. Ten independent raters identified a total of 18 land￾marks on each face. Raters were trained for several weeks in hour-long sessions in which landmarks were defined anatomically. Figure 1 shows the location of each land￾mark. Two raters were randomly chosen for each land￾mark, and the coordinate for that landmark was calculated as the mean pixel location chosen by these two raters. Photographs of participants were not originally taken for shape analysis. Therefore, the photographs varied in ways that could alter the shape information captured by the landmarks (e.g., the participant’s head angle facing Downloaded from pss.sagepub.com by Cai Xing on January 7, 2014

Genetics of Facial Masculinity 3 common shape space,removes size effects by standard- izing centroid size to 1,and removes rotational effects by minimizing the summed,squared distances between ates that pure information.The coordinates are then tansformed int shape variables via a principal component analysis Shape variable are a decomposition of the ongir have tage of being compatible with conventional statistica techniques without the need for adjustments.For full ysisviarrustes analysis andldncha 1 et a To compute a data-driven single measure of facial masculinity,we conducted a discriminant-function anal ysis (DFA)with sex as the grouping varable (femal 1).DFA unction tha discriminated between male and female landmark con figurations.Thus,the discriminant function from this dimorphisn scores ously to co pute data-driven scores of facial masculinit (Gangestad et al 2010:Scott et al.2010).The DEA sex the c or the We ween pa assumed that mos reported in Gang tad et al (2010) cratic and would therefore simply add error variance The discriminant function correctly classified the sex of cdPoshay 81%of participants,which is lower s wou d th ue rep in scott e 10 -alidation makeit for rated degree of smiling did not affect the results (data difficult to interpret their very high rate of correct not reported here). classification 1o cro a te our Facial masculinity scores fun We used geometric morphometrics,the statistical analysis correct-classification rate of 80%,which indicates that the of shape through land ark coordinates,to analyze the masculinity iderski. eets and it therefore encapsulates all other information such The discriminant scores were standardized by sex to as distances and angles between different landmark: produce a facial masculinity score for each participant in extract s shape information from raw facial nd elation to other participants of the same sex.Five outl a ge ed fro procedure removes translation effects (position of the these outliers vielded results virtually identical results to object in the shape space)by standardizing all faces to a those reported here. 20

Genetics of Facial Masculinity 3 the camera or the participant’s facial expression). We assumed that most of this type of variation was idiosyn￾cratic and would therefore simply add error variance rather than biasing the results in any particular direction. However, to avoid the possibility that smiles would bias the measures, we did not use landmarks around the mouth, and we subsequently confirmed that controlling for rated degree of smiling did not affect the results (data not reported here). Facial masculinity scores We used geometric morphometrics, the statistical analysis of shape through landmark coordinates, to analyze the faces (Bookstein, 1991; Zelditch, Swiderski, Sheets, & Fink, 2004). Shape is defined as the differences between objects that are not due to translation, size, or rotation, and it therefore encapsulates all other information, such as distances and angles between different landmarks. To extract shape information from raw facial land￾marks, we conducted a generalized Procrustes analysis (Zelditch et al., 2004) on raw x- and y-coordinates. This procedure removes translation effects (position of the object in the shape space) by standardizing all faces to a common shape space, removes size effects by standard￾izing centroid size to 1, and removes rotational effects by minimizing the summed, squared distances between homologous landmarks across a range of faces. This pro￾duces Procrustes coordinates that purely represent shape information. The coordinates are then transformed into shape variables via a principal component analysis. Shape variables are a decomposition of the original Procrustes coordinates and completely maintain the shape information. Shape variables also have the advan￾tage of being compatible with conventional statistical techniques without the need for adjustments. For full details of generalized Procrustes analysis and shape anal￾ysis via geometric morphometrics, see Zelditch et al. (2004). To compute a data-driven single measure of facial masculinity, we conducted a discriminant-function anal￾ysis (DFA) with sex as the grouping variable (female = 0, male = 1). DFA produced a discriminant function that was the linear combination of shape variables that best discriminated between male and female landmark con￾figurations. Thus, the discriminant function from this analysis represents the sexual-dimorphism dimension (see Fig. 2 for the distribution of scores on the discrimi￾nant function). Related analyses have been used previ￾ously to compute data-driven scores of facial masculinity (Gangestad et al., 2010; Scott et al., 2010). The DFA performed on the twins’ data yielded a point-biserial correlation of .66 between participant’s sex and the dis￾criminant score, which was slightly higher than the cor￾responding value reported in Gangestad et al. (2010). The discriminant function correctly classified the sex of 81% of participants, which is lower than the correspond￾ing value reported in Scott et al. (2010), but their high ratio of predictors to participants (which can cause model overfitting) and lack of cross-validation make it difficult to interpret their very high rate of correct classification. To cross-validate our measure, we applied this same function to the siblings’ data; this yielded a point-biserial correlation between sex and masculinity of .65 and a correct-classification rate of 80%, which indicates that the masculinity measure discriminated between the sexes equally well in the approximately 18-year-old siblings and the approximately 16-year-old twins, further validat￾ing our measure. The discriminant scores were standardized by sex to produce a facial masculinity score for each participant in relation to other participants of the same sex. Five outli￾ers on facial masculinity (≥ ±3 SD from the mean) were omitted from all analyses; however, an analysis retaining these outliers yielded results virtually identical results to those reported here. Fig. 1. Facial landmarks (green crosses) used to compute facial masculinity. Downloaded from pss.sagepub.com by Cai Xing on January 7, 2014

Lee et al. Female ☐Male 60 50 20 0 Masculinity Score Observer ratings of facial uctions on ho attractiveness and masculinity features that are considered to be sexually dimorphic in trait For this stud whe photograph on humans ere p the 1.mode masculinity rat to check whether facial ma ofemale rater scores calculated from landmark coordinates correlated correlated very highly with the averaged score from all raters (r=.94 for male raters and r=.92 for female rat so the l er c sed for all analy who were not involyed in identifving the facial land- eror than the separate scor for male and female marks)and asked them to rate all faces on attractiveness and facial masculinity.Ratings were given on 7-poin Interrater agreement was low for masculinity (intra D4 19a=.6 )Neverthe

4 Lee et al. Observer ratings of facial attractiveness and masculinity Observers also rated the photographs on a number of traits. For this study, we were primarily interested in the attractiveness ratings, but we also analyzed the facial masculinity ratings to check whether facial masculinity scores calculated from landmark coordinates correlated with subjective perceptions of facial masculinity. We presented the photos in a random order to 8 undergraduate research assistants (4 men and 4 women who were not involved in identifying the facial land￾marks) and asked them to rate all faces on attractiveness and facial masculinity. Ratings were given on 7-point scales (for attractiveness, 1 = low attractiveness, 7 = high attractiveness; for masculinity, 1 = very feminine, 7 = very masculine). We did not give raters instructions on how to judge attractiveness, but we did inform them of facial features that are considered to be sexually dimorphic in humans. Interrater agreement for attractiveness was moderate (intraclass correlation coefficient = .44, p < .001; α = .87). Averaged scores from male raters and from female raters correlated very highly with the averaged score from all raters (r = .94 for male raters and r = .92 for female rat￾ers), so the latter composite score was used for all analy￾ses because it contained substantially less measurement error than the separate scores for male and female raters. Interrater agreement was low for masculinity (intra￾class correlation coefficient = .19; α = .66). Nevertheless, there was still a significant (though modest) correlation 0 10 20 –4.00 –3.50 30 40 50 60 70 80 Frequency Masculinity Score Female Male –2.50 –2.00 –1.50 –1.00 1.00 2.00 –0.50 0.00 0.50 1.50 2.50 3.00 3.50 4.00 –3.00 Fig. 2. Frequency of objective facial masculinity scores from the discriminant-function analysis for males (M = .92, SD = .94) and females (M = −.80, SD = .97), before standard￾ization separately by sex. The purple portions of the bars represent overlapping distribu￾tions for males and females. Downloaded from pss.sagepub.com by Cai Xing on January 7, 2014

Genetics of Facial Masculinity 5 between obiective and rated masculinity (male faces:r= the correlations between nonidentical twin pairs (ie. .23.p<001;female faces:r=.25,p<001).Objective male-male,female-female,and male-female)did not differ sed only on shape an nd was not associ formnthecohtio2bcoecnorpo0die图 of 29)and acne (female faces:r=.29.p<001;male faces: dnm r=21.)and were presumably influenced by in 3.Int =0.04,P= 85,or ents measure correlated much more strongly with the compo- shown in Table 1.Correlations between identical twins were markedly greater than correlations between same ed shape variables (male faces: sex nonidentical twins and siblings for both male ces:r 1.92 sis that th and em Z(D =4.93.p=.05 available online for details of the analysis.For more nent of facial masculinity in both sexes.The estimated proportions of variation in facial masculinity due to genetic and reported and lem Statistical anabsis genetic factors.whereas virtually no variation was attrib Identical twins share all their genes,whereas nonidenti- uted to shared environmental influences.This finding is cal twins of their comple y environ for good genes environmental (C).and residual (sources.As is stan- One of our main goals was to determine the degree to liked for twin-family designsi we conducted maximum- The factt and twin or sibling pair c and sibl :sce Table D suggests that For further details on the type of twin analysis that we heritable factors that increase male facial masculinity also (2003) cond 2 ces bet means and correlations of different zygosity groups were Zygosity group r195%CIl thehangen mde罗 and test as Pa山 which All identical twins 50L305g Nonidentical female twins (n=113 pairs) 30[11,45 Results 16-.04.35 mean facial masculinit Male siblings (n=39 pairs) -09上382 of th ,221=248=20 230.3 Means of female or male members of same-sex pairs did iblings 231.09.36 not differ significantly from means of female or male mem- pairs.() 85,which Opposite-sex siblings (n=120 pairs) of nins Opposite-sex twins and siblings it Me did not differ significantly from means of other siblings ()=3.60,p=.17,which suggests that there was nothing unusual about the facial masculinity of twins.Furthermore

Genetics of Facial Masculinity 5 between objective and rated masculinity (male faces: r = .23, p < .001; female faces: r = .25, p < .001). Objective masculinity was based only on shape and was not associ￾ated with ratings of grooming or acne, whereas masculin￾ity ratings were associated with ratings of grooming (female faces: r = −.44, p < .001; male faces: r = −.05, p = .29) and acne (female faces: r = .29, p < .001; male faces: r = .21, p < .001) and were presumably influenced by cues other than shape, such as skin color and tone, heavi￾ness of brow, and facial hair. Our objective masculinity measure correlated much more strongly with the compo￾nent of the masculinity ratings that is captured by the landmark-based shape variables (male faces: r = .53, p < .001; female faces: r = .57, p < .001) than with the raw masculinity measure. (See the Supplemental Material available online for details of the analysis.) For more detail on the rating process and genetic analyses of observer ratings, see Mitchem et al. (2013). Statistical analysis Identical twins share all their genes, whereas nonidenti￾cal twins share, on average, half of their segregating genes, and all twins completely share the family environ￾ment. Therefore, we were able to partition the variation in scores into three sources: additive genetic (A), shared environmental (C), and residual (E) sources. As is stan￾dard for twin-family designs, we conducted maximum￾likelihood modeling, which determines the combination of A, C, and E that best matches the observed data (i.e., means, variances, and twin or sibling pair correlations). For further details on the type of twin analysis that we used, see Neale & Cardon (1992) and Posthuma et al. (2003). All analyses were conducted in the Mx software package, Version 1.54a (Neale, Boker, Xie, & Maes, 2006). As is standard in twin modeling, differences between the means and correlations of different zygosity groups were tested by equating the relevant parameters in the model and testing the change in model fit (distributed as χ2 ) against the change in degrees of freedom (which equals the change in the number of parameters estimated). Results Preliminary testing found that mean facial masculinity scores did not significantly differ between identical and nonidentical twins of the same sex, χ2 (2) = 2.48, p = .29. Means of female or male members of same-sex pairs did not differ significantly from means of female or male mem￾bers of opposite-sex pairs, χ2 (2) = 0.31, p = .85, which suggests that prenatal hormone transfer from one twin to the other had no influence on this trait. Means of twins did not differ significantly from means of other siblings, χ2 (2) = 3.60, p = .17, which suggests that there was nothing unusual about the facial masculinity of twins. Furthermore, the correlations between nonidentical twin pairs (i.e., male-male, female-female, and male-female) did not differ significantly from the correlations between corresponding nontwin sibling pairs, χ2 (3) = 2.18, p = .54, as expected given the equivalent genetic and environmental similarity of nonidentical-twin and sibling pairs; these correlations were equated in subsequent modeling. There was no significant effect of age on facial mascu￾linity scores in males, χ2 (1) = 0.04, p = .85, or females, χ2 (1) = 0.63, p = .43. Intraclass correlation coefficients are shown in Table 1. Correlations between identical twins were markedly greater than correlations between same￾sex nonidentical twins and siblings for both males, χ2 (1) = 11.92, p < .001, and females, χ2 (1) = 4.93, p = .03, which suggests that there is an important genetic compo￾nent of facial masculinity in both sexes. The estimated proportions of variation in facial masculinity due to genetic and environmental sources are reported in Table 2. For both males and females, approximately half of the variation in facial masculinity was attributed to additive genetic factors, whereas virtually no variation was attrib￾uted to shared environmental influences. This finding is consistent with the assumption that variation in facial masculinity is substantially heritable, which is a neces￾sary condition for facial masculinity to serve as a signal for good genes. One of our main goals was to determine the degree to which genes that affect masculinity in men have the same effect in women. The fact that facial masculinity scores were significantly correlated between opposite-sex twins and siblings (r = .23, p < .001; see Table 1) suggests that heritable factors that increase male facial masculinity also Table 1. Intraclass Correlation Coefficients for Objective Facial Masculinity of Twin and Sibling Pairs Zygosity group r [95% CI] Identical female twins (n = 110 pairs) .50 [.36, .61] Identical male twins (n = 88 pairs) .50 [.34, .62] All identical twins .50 [.39, .59] Nonidentical female twins (n = 113 pairs) .30 [.11, .45] Female siblings (n = 55 pairs) .20 [–.16, .46] All nonidentical female twins and siblings .28 [.11, .42] Nonidentical male twins (n = 93 pairs) .16 [–.04, .35] Male siblings (n = 39 pairs) –.09 [–.38, .22] All nonidentical male twins and siblings .09 [–.08, .26] All nonidentical same-sex twins and siblings .23 [.10, .35] Nonidentical opposite-sex twins (n = 171 pairs) .23 [.09, .36] Opposite-sex siblings (n = 120 pairs) .23 [.04, .39] Opposite-sex twins and siblings .23 [.12, .33] Note: Means and variances were equated across zygosity (within sex). Sibling pairs are not independent (i.e., a nontwin sibling can have a sibling relationship with each member of a twin pair). CI = confidence interval. Downloaded from pss.sagepub.com by Cai Xing on January 7, 2014