IN PRESS (ARCHIVES OF SEXUAL BEHAVIOR)
Gender and Visibility of Sexual Cues Influence Eye Movements
While Viewing Faces and Bodies
Jari K. Hietanen4
1Brain Research Unit, O.V. Lounasmaa Laboratory, Aalto University School of Science,
2Department of Biomedical engineering and Computational Science, Aalto University School
of Science, Espoo, Finland
3Turku PET Centre, Turku, Finland
4Department of Psychology, University of Tampere, Tampere, Finland
5Department of Psychology, Åbo Akademi University, Turku, Finland
6Department of Psychology, University of Turku, Turku, Finland
Please address correspondence at
Dr Lauri Nummenmaa
Brain Research Unit, O.V. Lounasmaa Laboratory
Aalto University School of Science
FI-00076 Espoo, Finland
Faces and bodies convey important information for the identification of potential sexual
partners, yet clothing typically covers many of the bodily cues relevant for mating and
reproduction. In this eye tracking study, we assessed how men and women viewed nude and
clothed, same and opposite gender human figures. We found that participants inspected the
nude bodies more thoroughly. First fixations landed almost always on the face, but were
subsequently followed by viewing of the chest and pelvic regions. When viewing nude
images, fixations were biased away from the face towards the chest and pelvic regions.
Fixating these regions was also associated with elevated physiological arousal. Overall, men
spent more time looking at female than male stimuli, whereas women looked equally long at
male and female stimuli. In comparison to women, men spent relatively more time looking at
the chests of nude female stimuli whereas women spent more time looking at the
pelvic/genital region of male stimuli. We propose that the augmented and gender-contingent
visual scanning of nude bodies reflects selective engagement of the visual attention circuits
upon perception of signals relevant to choosing a sexual partner, which supports mating and
Keywords: Eye movements, body, nudity, visual attention, sexuality, mate choice
Human behavior is markedly influenced by the communicative signals conveyed by our
conspecifics. Numerous studies suggest that specialized neurocognitive mechanisms process
the facial and bodily features that guide our social interaction and interpersonal relationships
(Hari & Kujala, 2009; Haxby, Hoffman, & Gobbini, 2000; Minnebusch & Daum, 2009).
Information from faces and bodies is also important for human sexual behavior. Although
identification of mating partners in primates relies extensively on the visual system
(Ghazanfar & Santos, 2004), evidence on the visual and attentional processes involved in
detecting sexual cues from human bodies has remained elusive. In the present eye tracking
study, we demonstrate that observer and stimulus gender, as well as visibility of sexual cues,
influence the visual sampling of human bodies, and discuss how this facilitated processing of
nude figures may support the identification of potential sexual partners.
Numerous studies have shown that both facial and bodily cues are reliable markers of
gender, health, and fertility. Waist-to-hip ratio (WHR) signals health in both genders (Pouliot
et al., 1994); in women it is associated with fertility (Singh, 1993). Facial features, including
symmetry, attractiveness, and sexual dimorphism, are also potential cues to health and fitness
(Rhodes, 2006). Efficient perception of such cues thus enables immediate categorization of
conspecifics as potential mating partners (opposite-gender) or competitors (same-gender),
and subsequently when a potential mating partner is detected, the assessment of his or her
mate value. Humans and other primates display highly selective preferences for viewing the
sexually relevant signals of conspecifics (Deaner, Khera, & Platt, 2005; Grammer, Fink,
Moller, & Thornhill, 2003). When these signals are perceived and evaluated as positive by
the emotion circuit, a physiological arousal response is elicited. This can subsequently trigger
sexual behaviors and ultimately copulation (Walen & Roth, 1987). It could thus be assumed
that detection of sexual cues would be facilitated across various processing stages in the brain
in order to facilitate sexual selection.
Neuroimaging studies have established that a specialized brain network spanning the
occipital and temporal cortices subserves perception of bodies in humans (de Gelder et al.,
2010; Minnebusch & Daum, 2009; Peelen & Downing, 2007). Although this network mainly
codes the configuration of human bodies, functional imaging studies have demonstrated that
responses in subcomponents of this network are further amplified when participants are
presented with nude human figures (Bocher et al., 2001; Schiffer et al., 2008). Interestingly,
such modulation of body processing by sexual cues begins remarkably early, as confirmed by
studies measuring event-related potentials. These studies have found that the body-sensitive
temporocortical regions show enhanced responses towards pictures involving nude versus
clothed humans already at 150-170 ms after stimulus presentation (Costa, Braun, &
Birbaumer, 2003; Hietanen & Nummenmaa, 2011), suggesting very early visual
categorization of potential sexual partners.
Although clothing provides cues to gender, sexual status and rank in many cultures,
Western clothing typically restricts the visibility of the body, especially the primary and
secondary sexual characteristics. Earliest recorded signs of clothing date to 36,000 BCE
(Kvavadze et al., 2009), although genetic and molecular clock estimates of head and body
lice–the latter having little chance of surviving on naked human body–suggest that body lice
have originated already 72,000 ± 42,000 years ago, which could coincide with the beginning
of frequent use of clothing (Kittler, Kayser, & Stoneking, 2003). Considering that the use of
clothing as a cultural habit is relatively recent in the time-scale of evolution, it is likely that
the visual and attentional systems would still be tuned to processing of nude rather than
clothed bodies, and that this tuning would be reflected in the way we move our eyes when
viewing human bodies with and without clothing.
Several lines of evidence support this hypothesis. First, it has been proposed that color
vision might have evolved in primates for discriminating the spectral modulations on the skin
of conspecifics (Changizi, Zhang, & Shimojo, 2006), and in line with this the human visual
system has been found to be particularly sensitive to detecting desaturated reddish targets
resembling human skin tones (Lindsey et al., 2010). Second, when viewing complex,
explicitly sexual scenes with nude figures (e.g., those involving foreplay), observers land
more fixations on bodies rather than on faces but this effect is reversed for scenes involving
clothed figures (Lykins, Meana, & Kambe, 2006). When viewing fully clothed bodies, males
tend to fixate the chest area earlier than women (Hewig, Trippe, Hecht, Straube, & Miltner,
2008); one study even found that nude female chests would typically receive the very first
fixation of male observers, potentially reflecting the tendency to evaluate the attractiveness or
reproductive fitness of the body (Dixson, Grimshaw, Linklater, & Dixson, 2009). Third, nude
bodies have been reported to capture visual attention. Nude but not clothed human bodies
elicit the attentional blink response (Most, Smith, Cooter, Levy, & Zald, 2007) traditionally
assumed to reflect involuntary attention capture; moreover, orienting of attention towards
one’s sexually preferred versus non-preferred gender is facilitated even when the stimuli are
presented outside of visual awareness (Jiang, Costello, Fang, Huang, & He, 2006), suggesting
automated processing of sexual signals.
The Present Study
In sum, there is evidence that the human visual system is biased towards processing of
nude human figures and that clothing might bias the visual processing of the bodily image
embedded in complex sexual scenes. However, controlled eye movement studies exploring
how human observers sample sexual information from same and opposite gender bodies are
practically nonexistent. In the present eye tracking study, we aimed at filling this gap by
answering two critical questions: First, we wanted to address whether covering the cues
relevant to sexual selection by clothes would influence the visual scanning of the body and
the face areas. Second, we wanted to evaluate whether viewing of the bodily image would be
influenced by the gender of the participant as well as of the person shown in the images. To
accomplish these aims, we employed high-resolution eye tracking while participants were
viewing singly presented, nude or clothed bodies of same or opposite-gender individuals.
Free viewing of pictures with no structured social or cognitive task was used to simulate
naturally occurring encounters. We measured where, in which order, and for how long
participants looked while inspecting the figures.
On the basis of evolutionary considerations, we predicted that humans would
automatically focus their attention on the regions that provide reliable information about
conspecifics’ social intentions (i.e., face) as well as reproductive fit (i.e., face, chest, and
pelvic region). We also predicted that observers would show preferential attention towards
the opposite vs. same-sex bodies. Finally, as both bodies and faces relay information relative
to sexual selection but clothing effectively covers all the bodily cues, we predicted that faces
would be predominantly fixated when the stimuli are clothed, whereas chest and pelvic
regions would receive substantially more fixations when the stimuli are presented naked.
Participants A total of 30 male undergraduate psychology students with a mean age of 21
years from the University Turku participated in the experiment and were compensated with
movie tickets. In this and the following experiment, all the participants gave informed
consent and had normal or corrected-to-normal vision. All were heterosexual according to the
Sell Assessment of Sexual Orientation questionnaire scores (Sell, 1996).
Apparatus Stimuli were presented on a 21” monitor (120 Hz refresh rate) with a 3.2 GHz
Pentium IV computer. Participants' eye movements were recorded with an EyeLink 2000
eyetracker (SR Research, Mississauga, Ontario, Canada) connected to a 2.8 GHz Pentium IV
computer. The sampling rate of the eyetracker was 1 kHz, and the spatial accuracy was better
than 0.5°, with a 0.01° resolution in the pupil-tracking mode.
Materials The stimuli (see Fig. 1 for illustrations) were 120 digital photographs of frontal
poses of nude and clothed, attractive and normal-weight adult males and females (30 stimuli
per category) appearing against white background. The stimulus pictures were acquired
mainly from various Internet sites. Clothed stimuli wore sexually non-revealing clothing – at
least a sleeved shirt and long pants/jeans, and some also wore a jacket or a coat. About 10 %
of the clothed stimuli had logos or emblems on their clothing, but these were equiprobable
for male and female stimuli (p > .05 in χ2 test). Nude stimuli clearly showed the chest and
the genitals of the person. The amount of pubic hair varied, although it was typically rather
modest. Penis size and turgidity also varied across nude male stimuli. None wore piercing or
tattoos. Chi-square tests confirmed that across stimulus categories, there were no significant
differences in the frequency of gaze direction (eye contact versus aversion) and facial
expression (smiling versus not smiling) across the stimulus categories (ps > .13 in χ2 test).
Stimulus displays The size of the stimuli was 7 × 10 degrees of visual angle at a viewing
distance of 60 cm. The drift correction target was a black circle with a white center (diameter
1.5 degrees) presented at the center of the screen. Stimuli were presented singly to the left or
right visual field such that the centerpoint of the picture was aligned to an imaginary circle
with a radius of 5.6 degrees. The distance between the stimulus centerpoint and horizontal
axis varied between ±3.7 degrees. This arrangement controlled for biases resulting from, for
example, always performing a horizontal saccade upon picture presentation. As the stimuli
were not initially in the foveal vision we could also address which stimulus regions initially
captured the participants’ overt attention.
Procedure The participants were told that the study concerned eye movements while viewing
pictures of humans. They were explained that on each trial they were going to see a picture of
a nude or clothed male or female, and that their task was to view the pictures similarly as they
were viewing pictures on a computer or while reading a magazine. Next, the eyetracker was
calibrated using a standard nine-point routine. The calibration was accepted if the average
error was less than 0.5°. Each trial (see Fig. 1 for a description of the sequence of trial events)
began with a drift correction. A fixation circle appeared at the center of the screen, and the
participants had to focus their gaze at the center of the circle. When the participant’s eye was
fixated on the circle, the experimenter initiated the trial. A random delay of 0-100 ms was
appended at the beginning of all trials to prevent anticipatory saccades. Next, the stimulus
picture appeared randomly at the left or at the right of the fixation circle for 4 s. After an
inter-trial interval of 1,000 ms, the central fixation point reappeared and the next trial was
initiated. Each participant performed one block of the task with a total of 120 trials, with each
stimulus shown once in a random order. The experiment was preceded by 10 practice trials.
Visual field of the stimuli was counterbalanced. After the experiment, the participants rated
the valence and arousal of the stimulus categories with the self-assessment manikin with
scales ranging from 1 to 9 (Bradley & Lang, 1994). Valence (unpleasantness vs.
pleasantness) reflects the dominant motive system activated (avoidance or approach),
whereas arousal reflects the intensity of motive system activation, from calm to tension
(Lang, 1995). Accordingly, this conceptualization enables an independent assessment of the
likeability of targets, as well as the arousal levels they trigger.
Eye movement analysis Two different analytic strategies were used. The first approach was
based on statistical parametric mapping of fixation heatmaps (see, e.g., Caldara, Zhou, &
Miellet, 2010). Briefly, fixations were first transformed into common xy-originator, as the
actual stimulus position was jiggled from trial to trial. Next, participant-wise fixation
heatmaps for each trial type (nude males, nude females, clothed males, clothed females) were
generated by modeling each fixation as a Gaussian function with mu of fixation’s carthesian
coordinate and sigma of one degree (based on the assumption that the foveal field of view is
roughly 2 degrees; see Wandell, 1995) and multiplied with fixation duration in milliseconds.
Mass univariate t-tests were then used to compare the smoothed fixation distributions across
trial types (male vs. female, nude vs. clothed, nude females vs. nude males and clothed
females vs. clothed males). This resulted in statistical T-maps where pixel intensities reflect
statistical differences in fixation probabilities across conditions. Finally, False Discovery
Rate (FDR) correction with an alpha level of .05 was applied to the statistical maps to control
for false positives due to multiple comparisons.
As our stimulus models varied slightly in posture, the heatmap analyses were also
complemented by classical region-of-interest (ROI) analyses. In this approach, rectangular
ROIs were drawn around the face, chest, and pelvic-genital areas of the stimuli.
Subsequently, we analyzed separately data for fixation events that occurred within or towards
each ROI (see e.g. Calvo & Nummenmaa, 2008; Calvo & Nummenmaa, 2009, for similar
approach using face stimuli). We computed the mean (1) first fixation time, that is, the
latency of the first fixation landing on the area, (2) duration of the first fixation, (3) gaze
duration (i.e., dwell time) for the area, and (4) average pupil size for fixations landing on ROI.
Trialwise total fixation counts were computed to address overall attention allocation to the
stimuli. Finally, latencies of the first saccades with an amplitude exceeding 2 degrees
initiated towards stimuli (irrespective of ROI) were computed to address the speed of
attentional orienting towards different stimulus categories. Two-tailed alpha level of p < .05
was applied in all statistical analyses. Multiple comparisons were corrected using the
Self-report scores The self-reported valence and arousal scores (see Table 1) for the stimuli
were analyzed with a 2 (gender) × 2 (clothing) repeated measures ANOVA. Female pictures
were rated as more pleasant than male pictures, F(1, 29) = 207.02, p < .001, ηp2 = .87, and
nude pictures were rated as more pleasant than clothed pictures, F(1, 29) = 20.01, p < .001,
ηp2 = .40. The interaction of gender and clothing, F(1, 29) = 66.80, p < .001, ηp2 = .69,
showed that whereas both nude and clothed female pictures were rated as equally pleasant,
F(1,29) = 3.53 (ns.), male pictures were rated as more pleasant when they were clothed, F(1,
29) = 58.80, p < .001, ηp2 = .66. Analysis of arousal scores revealed that female pictures were
rated as more arousing than male pictures, F(1, 29) = 45.02, p < .001, ηp2 = .60, and nude
pictures were rated as more arousing than clothed pictures, F(1, 29) = 195.84, p < .001, ηp2 =
.87. The interaction of gender and clothing, F(1, 29) = 14.92, p < .001, ηp2 = .32, showed that
removing clothes from the models resulted in a greater increase of arousal for female than
male models, F(1,29) = 14.92, p < .001, ηp2 = .32.
Fixation heatmaps Fig. 2 shows statistically thresholded fixation density difference images
overlaid on a body stimulus. These maps reveal that female stimuli were inspected in more
detail than male stimuli; the statistical difference in fixation distributions was most profound
in the chest and pelvic regions. Additionally, nude (vs. clothed) stimuli received more
fixations in the chest and pelvic area, whereas the faces of the clothed stimuli were inspected
in more detail.
Global eye movement measures The global eye movement data were analyzed with a 2
(gender) × 2 (clothing) ANOVA. Stimulus clothing, gender or their interaction did not
influence latencies of the first saccade towards the pictures, Fs < 1. However, the number of
fixations was influenced by both gender, F(1, 29) = 17.93, p < .001, ηp2 = .32, clothing, F(1,
29) = 33.71, p < .001, ηp2 = .54, and their interaction, F(1, 29) = 9.89, p < .01, ηp2 = .25.
More fixations were made on female than male pictures (Mfemale = 11.82; Mmale = 11.18) and
on nude than clothed pictures (Mnude = 12.03, Mclothed = 10.97). The interaction resulted from
the fact that fixation frequency for male and female stimuli was equal when the bodies wore
clothes, F < 1, whereas significantly more fixations emerged on female pictures when the
stimuli were presented without clothes, F(1, 29) = 380.39, p < .001, ηp2 = .93.
ROI-based eye movement measures The ROI data were analyzed with a 2 (gender) × 2
(clothing) × 3 (ROI) ANOVA, and the results are summarized in Fig. 3 and in Table 2. For
first fixation time, there were main effects of clothing and ROI. ROIs were looked at earlier
when the stimuli were nude than clothed (Mnude = 1064 ms, Mclothed = 1270 ms), and faces
were looked at earlier than chest, F(1, 29) = 126.89, p < .001, ηp2 = .81, or pelvic region, F(1,
29) = 233.59, p < .001, ηp2 = .89, and chest was looked at earlier than pelvic region, F(1,
29) = 62.19, p < .001, ηp2 = .68. Mean first fixation times were Mface = 499 ms, Mchest = 1229
ms ad Mpelvic = 1774 ms.
There were also interactions of gender and clothing, gender and ROI, clothing and
ROI, as well as gender × clothing × ROI. The gender × clothing interaction reflects the fact
that the latency of first fixation was similar for nude male and female stimuli, F < 1, whereas
it was faster for clothed males vs. females, F(1, 29) = 4.20, p < .05, ηp2 = .13. For the
interaction of gender × ROI, none of the planned comparisons reached significance after
correcting for multiple comparisons. The clothing by ROI interaction resulted from
participants fixating chest, F(1, 29) = 21.64, p < .001, ηp2 = .43, and pelvic region, F(1, 29) =
9.88, p < .01, ηp2 = .25, earlier for nude than clothed figures, whereas faces were looked at
earlier when the stimuli were clothed than nude, F(1, 29) = 55.33, p < .001, ηp2 = .66. The
three-way-interaction reflects the fact that numerically the above effect was larger for male
than female stimuli, but planned comparisons did not reach significance.
For first fixation duration, there were main effects of clothing and ROI. First fixations
were longer for clothed than for nude bodies (Mnude = 292 ms, Mclothed = 323 ms), and for
faces than for chest, F(1, 29) = 59.50, p < .001, ηp2 = .67, or pelvic region F(1, 29) = 40.67, p
< .001, ηp2 = .58, (Mface = 387 ms, Mchest = 250 ms Mpelvic = 285). There were also interactions
of clothing × ROI, and gender × clothing × ROI. First fixations on faces were longer when
bodies were clothed rather than nude, F(1, 29) = 14.76, p < .001, ηp2 = .34, whereas clothing
did not influence first fixation durations for chest or pelvic region, Fs < 1. The three-way
interaction resulted from the fact that faces of clothed versus nude females, F(1, 29) = 16.30,
p < .001, ηp2 = .36, received longer first fixations, whereas the effect was only marginally
significant (p = .09) for males.
For dwell times, the data generally replicated those obtained with the heatmap
analyses. The ANOVA revealed main effects of gender, clothing, and ROI, showing that
participants inspected female pictures longer than male pictures (Mfemale = 878 ms, Mmale =
827 ms), and nude pictures longer than clothed pictures (Mnude= 879 ms , Mclothed = 827 ms).
Furthermore, faces were inspected longest, followed by chest and pelvic ROIs (Mface= 1653
ms, Mchest= 500 ms, Mpelvic= 405 ms). There were also interaction effects of gender × ROI,
clothing × ROI, as well as gender × clothing × ROI. Simple effects tests revealed that
participants looked longer at the female versus male chests, F(1, 29) = 37.10, p < .001, ηp2 =
.56, and pelvic regions, F(1, 29) = 11.31, p < .01, ηp2 = .28, whereas they fixated longer on
male than female faces, F(1, 29) = 6.70, p < .05, ηp2 = .19. The clothing by ROI interaction
reflects the fact that the chest, F(1, 29) = 65.78, p < .001, ηp2 = .69, and pelvic regions, F(1,
29) = 61.84, p < .001, ηp2 = .68, were inspected longer when the bodies were nude, whereas
faces were inspected longer when the bodies were clothed, F(1, 29) = 22.27, p < .001, ηp2 =
.68. The three-way interaction is due to the effect of clothing on viewing times being
different for male than female stimuli. For both genders, chest and genitals were observed
longer when the stimuli were nude, but nudity decreased looking at female faces without
influencing looking times for male faces, Fs > 5.28, ps < .05, ηp2 :s > .15.
For pupil size, there was only a main effect of ROI, with fixations on the pelvic region
resulting in larger pupil sizes than fixations on faces, F(1, 29) = 68.03, p < .001, ηp2 = .70, or
chests, F(1, 29) = 72.00, p < .001, ηp2 = .71.
Experiment 1 confirmed that visibility of sexual cues and the gender of the person
being observed have a strong impact on how the information conveyed by bodies and faces
was sampled by males. First, the data showed that visual scanning of humans begins from the
face, regardless of whether or not the primary and secondary sexual cues were covered by
clothing. The first fixations landed on the faces, with an average latency of 500 ms. Faces
also received longest first fixations and were looked at for the longest duration. This is
compatible with the profound role of faces in social signaling (Calder & Young, 2005), as
well as with studies showing that faces capture attention reflexively (Langton, Law, Burton,
& Schweinberger, 2008; Theeuwes & Van der Stigchel, 2006). We qualify these data by
showing that, even for full-body figures showing a number of cues relevant for social and
sexual perception, the visual information conveyed by the face was almost always addressed
first (see also Janelle, Hausenblas, Ellis, Coombes, & Duley, 2009). Although bodily features
provide reliable cues for women’s reproductive fitness (Pouliot et al., 1994; Singh, 1993), our
data suggest that men put the initial evaluative emphasis on the face, probably because it
conveys information regarding both health and fitness as well as individual’s motivational
and emotional states that may influence the likelihood of successful social interaction and
potential mating (Rhodes, 2006).
The fixations on the face were followed by a gradual downward shift in fixations
towards chest and then finally to pelvic regions (see Fig. 4). Both fixation heatmaps and ROI-
based analyses revealed that the visual scanning of the face, chest, and pelvic regions were
influenced by both stimulus gender and clothing. Participants made more fixations on
opposite than same-gender stimuli, and the spatial distribution of fixations was asymmetrical
for male and female stimuli. Participants looked longer at male rather than female faces,
whereas they looked longer at female vs. male chest and pelvic regions. The bias towards
female chest and pelvic regions probably reflects the fact that these regions signal
reproductive fitness (Jasienska, Ziomkiewicz, Ellison, Lipson, & Thune, 2004; Pouliot et al.,
1994; Singh, 1993), and evaluating these features of opposite-gender humans (i.e., potential
mating partners) would thus be an automatic, biological predisposition. However, such
evaluations would not be necessary for males. Instead, facial information related, for
example, to social dominance or aggressiveness would be more important to acquire.
When sexual characteristics were visible, the stimuli were inspected more thoroughly.
Importantly, our data revealed that the aforementioned primacy in scanning the face was
overshadowed by the stimulus gender and the visibility of sexual characteristics. When the
bodies were shown without clothes, first fixations on faces occurred later and were much
shorter. On the contrary, first fixations on both chest and pelvic regions were longer and
occurred much earlier on the nude stimuli. Moreover, the total time spent observing these
sexually relevant regions was significantly longer when the stimuli were shown nude rather
than with clothing. This suggests that clothing indeed covers important information related to
sexual processing that male observers nevertheless strive to acquire. The inspection of the
chest and pelvic regions was also associated with elevated physiological arousal as evidenced
by pupillometric responses, confirming that viewing these regions was probably related to
sexual interest. Importantly, all these effects were observed in a free viewing condition rather
than under specific instructions; thus, they reflect the observers’ natural, biological
predisposition to scanning the bodily image. Finally, it must be stressed that although both
nude and clothed opposite-sex stimuli were considered pleasant, only nude stimuli were rated
highly arousing. Accordingly, it is likely that the arousal level rather than the activation of the
approach motivation system is associated with the enhanced scanning of the sexual features
of the opposite-sex nudes.
Although Experiment 1 convincingly demonstrated that male human observers
inspect nude bodies more thoroughly than clothed bodies and that they are biased towards
viewing the opposite-gender bodily regions that are relevant for identifying potential sexual
partners, it could be argued that there simply is more significant information in this region of
female rather than male bodies (especially WHR). Hence, the scan patterns may simply
reflect the amount of information conveyed by the male and female stimuli, rather than
sexual interest. Additionally, it is possible that the gaze patterns observed in Experiment 1
might not generalize to woman observers. It has actually been established that men prefer
physically attractive partners more than women (Buss & Barnes, 1986), which suggests that
men and women could indeed view same and opposite gender bodies differently. To
generalize our results to both sexes, we conducted Experiment 2 in which we evaluated the
gaze patterns of men and women who viewed pictures of nude and clothed male and female
stimuli. If biases towards viewing chest and pelvic regions truly reflect viewing strategies
specific for sexual interest, we expected to observe an interaction of stimulus gender,
observer gender and region of interest for the dwell times.
Participants and Procedure Experiment 2 essentially replicated Experiment 1 with the
exception that eye movements of both men and women were studied. Thirty-eight
undergraduate students (22 women, 16 men) participated for course credit. All were
heterosexual on the basis of the Sell Assessment of Sexual Orientation scores (Sell, 1996).
ROI-based data analyses were conducted similarly to Experiment 1 with the exception that
participant gender was introduced as a between-subjects factor in the ANOVA.
Global eye movement measures Saccade latencies were not influenced by any experimental
factor, Fs < 1.5. For fixation count, there was a main effect of stimulus gender, F(1, 36) =
7.43, p < .01, ηp2 = .17, and clothing, F(1, 36) = 37.21, p < .001, ηp2 = .51. In general, more
fixations were made on female than male stimuli (Mfemale = 10.44, Mmale = 10.20) and on nude
than clothed stimuli (Mnude = 10.84, Mclothed = 9.80). These main effects were qualified by an
interaction of stimulus gender × clothing × participant gender, F(1, 36) = 6.31, p < .05, ηp2 =
.15. Simple effects tests revealed that men made more fixations on nude than clothed stimuli,
F(1, 15) = 50.52, p < .001, ηp2 = .77, whereas for women the difference between nude and
clothed stimuli was larger for female than for male stimuli F(1, 21) = 10.74, p < .01, ηp2 =
ROI-based eye movement measures: Interactions with participant gender The overall
pattern of first fixation time, first fixation duration and dwell time essentially replicated that
observed in Experiment 1. Only the analysis of dwell times resulted in interactions involving
subject gender; thus, the other main effects and interactions not involving subject gender are
not presented here for the sake of conciseness (see Table 3 for the full ANOVA results for
these measures). For dwell time, there were interactions of participant gender and stimulus
gender F(1, 36) = 7.76, p < .01, ηp2 = .18, and participant gender × stimulus gender ×
clothing × ROI, F(2, 72) = 3.75, p < .05, ηp2 = .10. The two-way interaction reflects the fact
that women observed the male and female stimuli equally long, whereas men showed a clear
preference towards female stimuli, F(1, 15) = 10.54, p < .01, ηp2 = .41. The four-way
interaction was decomposed by analyzing data from each ROI separately with 2 (participant
gender) × 2 (stimulus gender) × 2 (clothing) mixed ANOVAs and corresponding simple
For face region, there was a three-way interaction of participant gender, stimulus
gender and clothing, F(1, 36) = 16.67, p < .001, ηp2 = .32. Simple effects tests showed that
women looked longer at male than female faces, F(1, 21) = 7.75, p < .05, ηp2 = .18, and at
faces of clothed bodies than faces of nude bodies, F(1, 21) = 9.76, p < .01, ηp2 = .32 whereas
men looked more at faces of nude male than nude female figures, F(1, 15) = 11.44, p < .01,
ηp2 = .43, and more at faces of clothed female than nude female figures, F(1, 15) = 9.02, p <
.01, ηp2 = .38.
For chest region, there was a main effect of participant gender, F(1, 36) = 8.94, p <
.01, ηp2 = .20, showing that men spent overall more time looking at the chest region than did
women. Furthermore, the bias towards nude vs. clothed chests was larger in magnitude
among men than women, as evidenced by a clothing × participant gender interaction, F(1, 36)
= 5.93, p < .05, ηp2 = .14.
For pelvic region, there was a three-way interaction of participant gender × stimulus
gender × clothing, F(1, 36) = 7.38, p < .01, ηp2 = .17. This interaction resulted from the fact
that men looked equally long at nude male and female pelvic region, as well as clothed male
and female pelvic region, whereas women looked more at nude male vs. female pelvic region
F(1, 21) = 17.51, p < .001, ηp2 = .46, with no significant differences in looking times for
clothed female and male pelvic regions. Additionally, men looked longer at nude female
pelvic regions than women, F(1, 37) = 5.23, p < .05, ηp2 = .12.
For pupil size, there was a main effect of ROI, F(2, 72) = 41.59, p < .001, ηp2 = .53,
as well as an interaction of stimulus gender and ROI, F(2, 72) = 3.75, p < .05, ηp2 = .10. Pupil
size was larger when participants were inspecting pelvic rather than chest F(1, 36) = 43.73, p
< .001, ηp2 = .55, or face, F(1,36) = 45.83, p < .001, ηp2 = .56, region, and larger when they
were looking at chest rather than face region, F(2, 72) = 25.66, p < .001, ηp2 = .42. None of
the planned comparisons for the two-way interaction reached significance. There was also an
interaction of participant gender and stimulus gender, F(1, 36) = 8.02, p < .01, ηp2 = .18.
Women showed a greater pupillary response towards male than female stimuli, F(1, 21) =
5.38, p < .05, ηp2 = .20, whereas the opposite was true for men although the effect was only
marginally significant, F(1, 15) = 3.01, p = .10, ηp2 = .17.
Experiment 2 confirmed that women showed a similar spatiotemporal pattern of
fixations on bodies as seen with male observers in Experiment 1. Moreover, clothing had a
strong effect on both men’s and women’s gaze patterns. For both men and women, nude vs.
clothed stimuli received more fixations, and removal of the clothing biased fixations away
from the face towards the chest and pelvic regions. This confirms that, for both genders,
nudity is an important attentional cue, which leads to more detailed inspection of the human
However, participant and stimulus gender as well as stimulus clothing influenced
interactively the viewing patterns: Whereas men showed a clear preference for viewing the
opposite-gender stimuli, women did not show a preference towards either gender. Only when
fixations on the face region were considered, more fixations on opposite-gender faces were
found for women. These data accord with findings showing that men pay more attention to
visual qualities in mate choice than females (Buss & Barnes, 1986), and are also compatible
with the prevailing view of sexual responsiveness, suggesting a greater discrimination of
physiological responses to sexually arousing opposite-gender vs. same-sex stimuli among
men than women (Alexander & Charles, 2009; Costa et al., 2003; Costell, Lunde, Kopell, &
Wittner, 1972; Hietanen & Nummenmaa, 2011; Lykins, Meana, & Strauss, 2008; Quinsey,
Ketsetzis, Earls, & Karamanoukian, 1996). On the other hand, our pupillometric measures
suggest that viewing opposite vs. same-sex stimuli elicit larger arousal responses in both
genders, indicative of arousal contingent on sexual interest while viewing the bodies. This
extends prior studies showing an elevated pupillary response to auditorily presented, sexually
arousing versus non-arousing cues (Dabbs, 1997) by demonstrating that similar effects are
also observed in the visual domain.
Both men and women looked longer at female versus male chests, even with clothed
stimuli. Furthermore, men spent more time looking at the chest region than women, and
clearly preferred the chests of nude females to nude males. This is in line with the findings
showing that, in women, mere breast size is a reliable marker of reproductive fitness
(Jasienska et al., 2004) and thus probably attracts more attention from male viewers.
Consistent with this, it has been found that judging attractiveness from headless semi-nude
stimuli biases eye movements towards the chest region (Cornelissen, Hancock, Kiviniemi,
George, & Tovee, 2009). On the other hand, women showed a clear preference for opposite
vs. same-gender pelvic region in the nude figures, whereas a similar bias (i.e., enhanced
attention to nude female vs. male pelvic region) was not found for men. It has been proposed
that the human penis size has evolved particularly due to female sexual selection (Miller,
1998), and in line with this, a considerable number of women value the size of partner’s penis
(Francken, van de Wiel, van Driel, & Weijmar Schultz, 2002; Stulhofer, 2006), which may
explain women’s selective scanning of the pelvic-genital regions of the nude males.
In two eye movement experiments, we investigated how the visibility of sexual cues
of male and female bodies influenced visual processing of the bodily image by men and
women. The experiments yielded three important conclusions. First, we were able to
characterize a spatiotemporal, top-to-bottom viewing pattern that human observers follow
when inspecting conspecifics’ bodies. Second, we demonstrated how the clothing of the
bodies modulated this overall pattern of fixations, with visible sexual characteristics leading
to more detailed overall inspection of the image and to a particular focus on the features
relevant to identifying potential sexual partners and their mate value. Third, we demonstrated
that participants’ gender (and, simultaneously, their sexual interest, as we only included
heterosexual participants) had a large impact on how nude bodies were inspected.
How Do Humans View Bodies And Faces?
The first fixation typically landed on the face; moreover, faces were inspected for the
longest duration, although participants also spent considerable time viewing the lower chest
and pelvic regions of the body. As the face conveys information regarding both typical and
situational behavior (Calder & Young, 2005; Haxby et al., 2000), the initial processing of the
facial information (such as gender, facial expression, and so forth) supports subsequent
interpretation of information acquired from the bodily image. Although attention capture by
faces against objects and animals is a robust phenomenon (Langton et al., 2008; Theeuwes &
Van der Stigchel, 2006), prior studies have provided contradictory evidence regarding the
primacy of examining the face versus other body regions when viewing full-body images
(Dixson et al., 2009; Hewig et al., 2008). Our high-resolution eye tracking data confirmed
that even when the chest and genital regions of figures were made sexually salient by
removal of the clothes, the face was still fixated first. All the other previous studies have
shown the stimuli at fixation; thus, even the very first fixation was forced to land on a
predefined region of the body, which obviously confounds with the obtained results. In
contrast, we presented stimuli outside of the participants’ initial foveal field of view and
jittered the vertical stimulus position to ensure that stereotypical viewing strategies, such as
making always a horizontal saccade upon stimulus presentation, cannot contaminate the
results. Taken together, these data suggest that the face is indeed the most relevant signal for
human interactions, and human observers strive to grasp the information conveyed by the
Enhanced Attention to Nude Bodies
As predicted, scanning of all three ROIs (face, chest, and pelvic regions) was
contingent on the visibility of sexual cues. Most importantly, nude stimuli received more
fixations than clothed ones. Increased number of fixations on scene regions and objects is
indicative of how much diagnostic information they contain and how much observers prefer
them (Henderson, 2003; Shimojo, Simion, Shimojo, & Scheier, 2003). Thus, the automatic
tendency to pay more attention to nude bodies indicates that the visual system is biased to
process this type of biologically salient information. Although faces were, in general,
inspected first and longest, this tendency was dramatically reduced when the bodies were
presented without clothing. Viewing nude bodies was associated with earlier fixations on the
chest and pelvic regions, and enhanced attention paid (as indexed by dwell time) to these
Prior evidence from eye movement studies suggests that the chest and pelvic regions
are important for the assessment of features relevant to sexual selection. When judging
gender from point-light walker stimuli, participants focus on the hip region of the figures,
although a substantial number of fixations also land on the shoulder region (Saunders,
Williamson, & Troje, 2010). Rating WHR from headless, semi-nude bodies biases fixations
towards the pelvic region, whereas rating attractiveness biases fixations towards the chest
region (Cornelissen et al., 2009). Our study did not involve an explicit judgment task so we
cannot disentangle from the obtained results the relative contribution of attractiveness and
reproductive fitness. However, it is plausible to assume that both features play a role in
guiding eye movements to the chest and pelvic regions in the nude bodies.
The finding that looking at chest and pelvic regions was associated with elevated
physiological arousal and that the latencies of the first fixations to these regions were shorter
when the bodies were shown without clothes accords with prior eye movement studies
demonstrating an automatic bias in directing gaze towards pleasant, highly arousing visual
content (Calvo & Lang, 2004; Nummenmaa, Hyönä, & Calvo, 2006, 2009). However, on the
basis of the present data, we cannot conclude whether or not the nude chest and genital
regions would have captured visual attention automatically. Nevertheless, as prior studies
have demonstrated automatic attention capture by sexual cues (Jiang et al., 2006; Most et al.,
2007), it seems plausible to assume that the bias towards nude chests and pelvic regions
observed in the present study would reflect automatic engagement of the visual attention
circuits upon perception of sexual cues.
Men and Women View Bodies Differently
Pupillometric analyses confirmed that the stimuli triggered an arousal response
contingent on sexual interest: opposite-gender stimuli were associated with elevated arousal
in both heterosexual men and women. The gaze patterns while viewing bodily images were
also strongly contingent on both the observer and stimulus gender. Men made more fixations
on the opposite rather than same-gender stimuli, which accords with the data showing that
men value partners’ physical qualities in mate choice more than women (Buss & Barnes,
1986). Specifically, men prefer the female chest and pelvic regions at the expense of the face,
whereas an opposite pattern emerged when they viewed male stimuli (cf. Fig. 2). Men also
preferred viewing the nude versus clothed chests more than women. As the female chest and
pelvic regions are indicative of both attractiveness and physical and reproductive fitness
(Jasienska et al., 2004; Pouliot et al., 1994; Singh, 1993), fixations on these regions probably
reflect an automatic tendency to evaluate these features.
Women also showed selective and strong biases towards specific features of opposite-
gender figures, but, unlike men, they paid more attention to the opposite- versus same-gender
faces. Although bodily cues, such as muscularity, are also important for women’s appraisals
of men’s attractiveness (Frederick & Haselton, 2007), certain facial characteristics might be
even more important for mate valuation. The sexual strategies theory (Buss & Schmitt, 1993)
posits that, in human sexual selection, men emphasize more youth and good looks, whereas
women are more attentive to cues signaling characteristics related to ambition and status.
Findings demonstrating that facial characteristics, such as eye gaze and maturity, rather than
specific bodily features, provide cues that signal social status in humans (Allison, Puce, &
McCarthy, 2000; Rule & Ambady, 2008; Todorov, Said, Engell, & Oosterhof, 2008) may
explain women’s bias towards viewing male faces. On the other hand, although faces are a
reliable cue for age, bodily features, such as form and breast development, are more
important markers of age and sexual development of females, which would explain the men’s
bias towards viewing the chest regions in female stimuli.
Limitations and Future Directions
One obvious limitation of the study was that our stimulus figures were not fully
standardized across categories. Even though we were careful to match stimulus categories
with respect to gaze contact and facial expression, the postures varied slightly across models.
Although such variability increases the ecological validity of the study by introducing natural
variability and unpredictability to the stimuli, it is possible that it may have affected the eye
movement patterns. Furthermore, we did not parametrically vary the amount of clothing the
stimulus persons were wearing. We have recently established that body-sensitive event-
related potentials measured from the occipitotemporal cortex are parametrically modulated
by the degree of clothing (nude-swimsuits–full clothing) worn by the stimulus persons
(Hietanen & Nummenmaa, 2011); thus, it would be interesting to use a similar approach in
eye movement studies. Also, studies on subpopulations with low sexual desire, such as
children or neurological patient groups with hypo- and hypersexuality, would provide
important insight regarding the role of sexual drive in guiding attentional deployment during
perception of nude and clothed bodies.
It is also possible that the nude stimuli could simply have been more novel and hence
captured attention more readily. However, this explanation seems unlikely, given that all the
stimuli were photos of unfamiliar models (e.g., pictures of famous actors were not used)
acquired from Internet sites. Thus, there is no reason to expect that the participants would
have been more familiar with any of the clothed figures. Alternatively, it could be argued that
we see nude stimuli more infrequently; hence, they would be more novel and more attention-
grabbing. However, it must be noted that the gaze patterns to nude versus clothed figures
were contingent on the observer’s gender; hence, it is unlikely that mere novelty of nude
bodies could explain the differential gaze patterns to nude versus clothed figures. Finally, it
must be stressed that we employed static photograph stimuli, which fail to capture the
intrinsic dynamic nature of human bodies. Given that eye movements can be easily recorded
and analyzed during dynamic body perception as well (see, e.g., Nummenmaa, Hyönä, &
Hietanen, 2009), future studies need to explore gender differences in more naturalistic tasks
involving dynamic clothed and nude persons.
Presence of sexual cues biases the human visual system in extracting information
from the human bodily image. Nude bodies attract more attention, particularly to the regions
relevant for the identification of sexual partners. We propose that the augmented and gender-
contingent visual scanning of nude bodies reflects selective engagement of the visual
attention circuits upon perception signals relevant to mate value, which supports mating and
reproduction. When this sexually relevant information is not available, fixations shift towards
the face, which also conveys socially and sexually relevant information, but is rarely covered
by clothing. It is thus intriguing to ask whether the often reported strong biases towards
viewing human faces could at least partially reflect the (learned) fact that faces are usually
the most reliably available source of information relevant in forming sexual and interpersonal
relationships in societies where clothes are worn regularly.
This research was supported by the AivoAalto grant from the Aalto University, and Academy
of Finland (grant # 251125 to LN). We thank Sanni Aalto, Jenny Wahlström and Anna
Backström for their help with data acquisition.
Alexander, G. M., & Charles, N. (2009). Sex differences in adults’ relative visual interest in
female and male faces, toys, and play styles. Archives of Sexual Behavior, 38, 434-
Allison, T., Puce, A., & McCarthy, G. (2000). Social perception from visual cues: Role of the
STS region. Trends in Cognitive Sciences, 4, 267-278.
Bocher, M., Chisin, R., Parag, Y., Freedman, N., Weil, Y. M., Lester, H., et al. (2001).
Cerebral activation associated with sexual arousal in response to a pornographic clip:
A O-15-H2O, PET study in heterosexual men. NeuroImage, 14, 105-117.
Bradley, M. M., & Lang, P. J. (1994). Measuring emotion. The Self-Assessment Mannequin
and the semantic differential. Journal of Behavior Therapy and Experimental
Psychiatry, 25, 49-59.
Buss, D. M., & Barnes, M. (1986). Preferences in human mate selection. Journal of
Personality and Social Psychology, 50, 559-570.
Buss, D. M., & Schmitt, D. P. (1993). Sexual strategies theory - An evolutionary perspective
on human mating. Psychological Review, 100, 204-232.
Caldara, R., Zhou, X., & Miellet, S. B. (2010). Putting culture under the Spotlight reveals
universal information use for face recognition. PLoS ONE, 5, e9708.
Calder, A. J., & Young, A. W. (2005). Understanding facial identity and facial expression
recognition. Nature Reviews Neuroscience, 6, 641-651.
Calvo, M. G., & Lang, P. J. (2004). Gaze patterns when looking at emotional pictures:
Motivationally biased attention. Motivation and Emotion, 28, 221-243.
Calvo, M. G., & Nummenmaa, L. (2008). Detection of emotional faces: salient physical
features guide effective visual search. Journal of Experimental Psychology: General,
Calvo, M. G., & Nummenmaa, L. (2009). Eye-movement assessment of the time course in
facial expression recognition: Neurophysiological implications. Cognitive, Affective,
& Behavioral Neuroscience, 9, 398-411.
Changizi, M. A., Zhang, Q., & Shimojo, S. (2006). Bare skin, blood and the evolution of
primate colour vision. Biology Letters, 2, 217-221.
Cornelissen, P. L., Hancock, P. J. B., Kiviniemi, V., George, H. R., & Tovee, M. J. (2009).
Patterns of eye movements when male and female observers judge female
attractiveness, body fat and waist-to-hip ratio. Evolution and Human Behavior, 30,
Costa, M., Braun, C., & Birbaumer, N. (2003). Gender differences in response to pictures of
nudes: A magnetoencephalographic study. Biological Psychology, 63, 129-147.
Costell, R. M., Lunde, D. T., Kopell, B. S., & Wittner, W. K. (1972). Contingent negative
variation as an indicator of sexual object preference. Science, 177, 718-720.
Dabbs, J. M. (1997). Testosterone and pupillary response to auditory sexual stimuli.
Physiology and Behavior, 62, 909-912.
de Gelder, B., Van den Stock, J., Meeren, H. K. M., Sinke, C. B. A., Kret, M. E., & Tamietto,
M. (2010). Standing up for the body. Recent progress in uncovering the networks
involved in the perception of bodies and bodily expressions. Neuroscience and
Biobehavioral Reviews, 34, 513-527.
Deaner, R. O., Khera, A. V., & Platt, M. L. (2005). Monkeys pay per view: Adaptive
valuation of social images by rhesus macaques. Current Biology, 15, 543-548.
Dixson, B., Grimshaw, G., Linklater, W., & Dixson, A. (2009). Eye-tracking of men’s
preferences for waist-to-hip ratio and breast size of women. Archives of Sexual
Behavior, 40, 43-50.
Francken, A. B., van de Wiel, H. B. M., van Driel, M. F., & Weijmar Schultz, W. C. M.
(2002). What importance do women attribute to the size of the penis? European
Urology, 42, 426-431.
Frederick, D. A., & Haselton, M. G. (2007). Why is muscularity sexy? Tests of the fitness
indicator hypothesis. Personality and Social Psychology Bulletin, 33, 1167-1183.
Ghazanfar, A. A., & Santos, L. R. (2004). Primate brains in the wild: The sensory bases for
social interactions. Nature Reviews Neuroscience, 5, 603-616.
Grammer, K., Fink, B., Moller, A. P., & Thornhill, R. (2003). Darwinian aesthetics: Sexual
selection and the biology of beauty. Biological Reviews, 78, 385-407.
Hari, R., & Kujala, M. V. (2009). Brain basis of human social interaction: From concepts to
brain imaging. Physiological Reviews, 89, 453-479.
Haxby, J. V., Hoffman, E. A., & Gobbini, M. I. (2000). The distributed human neural system
for face perception. Trends in Cognitive Sciences, 4, 223-233.
Henderson, J. M. (2003). Human gaze control during real-world scene perception. Trends in
Cognitive Sciences, 7, 498-504.
Hewig, J., Trippe, R. H., Hecht, H., Straube, T., & Miltner, W. H. R. (2008). Gender
differences for specific body regions when looking at men and women. Journal of
Nonverbal Behavior, 32, 67-78.
Hietanen, J. K., & Nummenmaa, L. (2011). The naked truth: The face and body sensitive
N170 response is enhanced for nude bodies. PLoS One, 6(11), e24408.
Janelle, C. M., Hausenblas, H. A., Ellis, R., Coombes, S. A., & Duley, A. R. (2009). The time
course of attentional allocation while women high and low in body dissatisfaction
view self and model physiques. Psychology & Health, 24, 351-366.
Jasienska, G., Ziomkiewicz, A., Ellison, P. T., Lipson, S. F., & Thune, I. (2004). Large
breasts and narrow waists indicate high reproductive potential in women. Proceedings
of the Royal Society B: Biological Sciences, 271, 1213-1217.
Jiang, Y., Costello, P., Fang, F., Huang, M., & He, S. (2006). A gender- and sexual
orientation-dependent spatial attentional effect of invisible images. Proceedings of the
National Academy of Sciences, 103, 17048-17052.
Kittler, R., Kayser, M., & Stoneking, M. (2003). Molecular evolution of Pediculds Humanus
and the origin of clothing. Current Biology, 13, 1414-1417.
Kvavadze, E., Bar-Yosef, O., Belfer-Cohen, A., Boaretto, E., Jakeli, N., Matskevich, Z., et al.
(2009). 30,000-year-old wild flax fibers. Science, 325, 1359-1359.
Lang, P. J. (1995). The emotion probe: Studies of motivation and attention. American
Psychologist, 50, 372-385.
Langton, S. R. H., Law, A. S., Burton, A. M., & Schweinberger, S. R. (2008). Attention
capture by faces. Cognition, 107, 330-342.
Lindsey, D. T., Brown, A. M., Reijnen, E., Rich, A. N., Kuzmova, Y. I., & Wolfe, J. M.
(2010). Color channels, not color appearance or color categories, guide visual search
for desaturated color targets. Psychological Science, 21, 1208-1214.
Lykins, A. D., Meana, M., & Kambe, G. (2006). Detection of differential viewing patterns to
erotic and non-erotic stimuli using eye-tracking methodology. Archives of Sexual
Behavior, 35, 569-575.
Lykins, A. D., Meana, M., & Strauss, G. P. (2008). Sex differences in visual attention to
erotic and non-erotic stimuli. Archives of Sexual Behavior, 37, 219-228.
Miller, G. F. (1998). How mate choice shaped human nature: A review of sexual selection
and human evolution In C. Crawford & D. Krebs (Eds.), Evolution and human
behavior: Ideas, issues, and applications. (pp. 87-93). Hillsdale, NJ: Lawrence
Minnebusch, D. A., & Daum, I. (2009). Neuropsychological mechanisms of visual face and
body perception. Neuroscience and Biobehavioral Reviews, 33, 1133-1144.
Most, S. B., Smith, S. D., Cooter, A. B., Levy, B. N., & Zald, D. H. (2007). The naked truth:
Positive, arousing distractors impair rapid target perception. Cognition & Emotion,
Nummenmaa, L., Hyönä, J., & Calvo, M. G. (2006). Eye movement assessment of selective
attentional capture by emotional pictures. Emotion, 6, 257-268.
Nummenmaa, L., Hyönä, J., & Calvo, M. G. (2009). Emotional scene content drives the
saccade generation system reflexively. Journal of Experimental Psychology: Human
Perception and Performance, 35, 305-323.
Nummenmaa, L., Hyönä, J., & Hietanen, J. K. (2009). I'll walk this way: Eyes reveal the
direction of locomotion and make passersby look and go the other way. Psychological
Science, 20, 1454-1458.
Peelen, M. V., & Downing, P. E. (2007). The neural basis of visual body perception. Nature
Reviews Neuroscience, 8, 636-648.
Pouliot, M. C., Despres, J. P., Lemieux, S., Moorjani, S., Bouchard, C., Tremblay, A., et al.
(1994). Waist circumference and abdominal sagittal diameter–Best simple
anthropometric indexes of abdominal visceral adipose-tissue accumulation and related
cardiovascular risk in men and women. American Journal of Cardiology, 73, 460-
Quinsey, V. L., Ketsetzis, M., Earls, C., & Karamanoukian, A. (1996). Viewing time as a
measure of sexual interest Ethology and Sociobiology, 17, 341-354.
Rhodes, G. (2006). The evolutionary psychology of facial beauty. Annual Review of
Psychology, 57, 199-226.
Rule, N. O., & Ambady, N. (2008). The face of success: Inferences from chief executive
officers' appearance predict company profits. Psychological Science, 19, 109-111.
Saunders, D. R., Williamson, D. K., & Troje, N. F. (2010). Gaze patterns during perception
of direction and gender from biological motion. Journal of Vision, 10.
Schiffer, B., Krueger, T., Paul, T., de Greiff, A., Forsting, M., Leygraf, N., et al. (2008).
Brain response to visual sexual stimuli in homosexual pedophiles. Journal of
Psychiatry & Neuroscience, 33, 23-33.
Sell, R. L. (1996). The Sell Assessment of Sexual Orientation: Background and scoring.
Journal of Lesbian, Gay and Bisexual Identity 1, 295-310.
Shimojo, S., Simion, C., Shimojo, E., & Scheier, C. (2003). Gaze bias both reflects and
influences preference. Nature Neuroscience, 6, 1317-1322.
Singh, D. (1993). Adaptive significance of female physical attractiveness: Role of waist-to-
hip ratio. Journal of Personality and Social Psychology, 65, 293-307.
Stulhofer, A. (2006). How (un)important is penis size for women with heterosexual
experience [Letter to the Editor]? Archives of Sexual Behavior, 35, 5-6.
Theeuwes, J., & Van der Stigchel, S. (2006). Faces capture attention: Evidence from
inhibition of return. Visual Cognition, 13, 657-665.
Todorov, A., Said, C. P., Engell, A. D., & Oosterhof, N. N. (2008). Understanding evaluation
of faces on social dimensions. Trends in Cognitive Sciences, 12, 455-460.
Walen, S. R., & Roth, D. (1987). A cognitive approach. In J. H. Geer & W. T. O'Donohue
(Eds.), Theories of human sexuality (pp. 335-362). New York: Plenum Press.
Wandell, B. A. (1995). Foundations of vision. Sunderland, MA: Sinauer.
Table 1. Mean valence and arousal scores for the clothed and nude male and female stimuli
used in Experiments 1 and 2. For both ratings, the scales range between 1 (unpleasant/calm)
and 9 (pleasant/aroused).
Valence M SD M SD
Male 5.35 1.05 3.29 1.57
Female 7.26 0.89 7.65 0.91
Male 1.42 0.72 2.00 1.24
Female 4.35 1.54 6.35 1.43
Table 2. Summary of the results of the ANOVAs for the ROI data in Experiment 1.
Source df F
First Fixation Time
Gender 1,29 1.77
Clothing 1,29 65.62 0.69
ROI 2,58 154.08 0.84
Gender × Clothing 1,29 4.44
Gender × ROI 2,58 4.45
Clothing × ROI 2,58 27.64 0.49
Gender × Clothing × ROI 2,58 4.01
First Fixation Duration
Gender 1,29 0.23
Clothing 1,29 16.77 0.37
ROI 2,58 49.15 0.63
Gender × Clothing 1,29 0.17
Gender × ROI 2,58 0.63
Clothing × ROI 2,58 7.11
Gender × Clothing × ROI 2,58 3.74
Gender 1,29 8.23
Clothing 1,29 6.12
ROI 2,58 166.70 0.85
Gender × Clothing 1,29 0.25
Gender × ROI 2,58 17.31 0.37
Clothing × ROI 2,58 38.32 0.57
Gender × Clothing × ROI 2,58 24.47 0.46
Gender 1,29 1.98
Clothing 1,29 0.23
ROI 2,58 57.02 0.66
Gender × Clothing 1,29 0.24
Gender × ROI 2,58 1.00
Clothing × ROI 2,58 0.76
Gender × Clothing × ROI 2,58 0.22
η ηp2 p-value
Table 3. Summary of the results of the ANOVAs for the ROI data in Experiment 2.
Gender × Participant Gender
Clothing × Participant Gender
ROI × Participant Gender
Gender × Clothing
Gender × Clothing × Participant Gender
Gender × ROI
Gender × ROI × Participant Gender
Clothing × ROI
Clothing × ROI × Participant Gender
Gender × Clothing × ROI
Gender × Clothing × ROI × Participant Gender 2,72
Gender × Participant Gender
Clothing × Participant Gender
ROI × Participant Gender
Gender × Clothing
Gender × Clothing × Participant Gender
Gender × ROI
Gender × ROI × Participant Gender
Clothing × ROI
Clothing × ROI × Participant Gender
Gender × Clothing × ROI
Gender × Clothing × ROI × Participant Gender 2,72
η ηp2 p-value
2,72 115.76 0.76
2,72 2.24 0.06
1,36 17.26 0.32
1,36 2.49 0.07
2,72 13.41 0.27
2,72 0.21 0.01
2,72 45.24 0.56
2,72 1.55 0.04
2,72 13.21 0.27
1,36 0.81 0.02
1,36 0.01 0.00
1,36 8.02 0.18
1,36 1.93 0.05
1,36 1.32 0.04
2,72 41.16 0.53
2,72 0.56 0.02
1,36 2.36 0.06
1,36 1.82 0.05
2,72 3.75 0.09
2,72 0.96 0.03
2,72 0.02 0.00
2,72 0.31 0.01
2,72 0.11 0.00
Figure 1. Illustration of the trial events (a) and experimental stimuli (b) in Experiments 1-2.
Figure 2. Statistical T-maps displaying differences in fixation patterns across experimental
conditions overlaid on a sample body stimulus in Experiment 1. Yellow to red codes a bias
towards females (a, c, and d) or nude bodies (b), turquoise to blue codes a bias towards males
(a, c, and d) or clothed bodies (b). The data are thresholded at p < .05 (FDR corrected for
Figure 3 Means and standard errors of the ROI-based measures of first fixation times (a),
first fixation duration (b), dwell time, (c) and pupil size (d) in Experiment 1.
Figure 4 Time course of allocating attention to the face, chest and pelvic regions of clothed
and nude male (a) and female (b) stimuli in Experiment 1. Y-axis shows the proportion of
fixations in each 200-ms time bin.
Figure 5 Means and SD of the dwell times for face (a), chest, (b) and pelvic (c) region, as a
function of stimulus and participant gender in Experiment 2.
41 Download full-text