An explanation for enhanced perceptions of attractiveness after alcohol consumption
Acute alcohol consumption increases ratings of attractiveness to faces. This may help to explain increased frequencies of sexual encounters during periods of alcohol intoxication. At least in part, such increased attraction may be the result of alcohol consumption decreasing ability to detect bilateral asymmetry, presumably because of the reductions in the levels of visual function. We tested the hypotheses that acute alcohol consumption decreases ability to detect asymmetry in faces and reduces preference for symmetrical faces over asymmetrical faces. Twenty images of a pair of faces and then 20 images of a single face were displayed on a computer, one at a time. Participants were instructed to state which face of each of the face pairs displayed was most attractive and then whether the single face being displayed was symmetrical or not. Data were collected near campus bars at Roehampton University. Sixty-four self-selecting students who undertook the study were classified as either sober (control) or intoxicated with alcohol. For each face pair or single face displayed, participant response was recorded and details of the alcohol consumption of participants that day were also obtained. Sober participants had a greater preference for symmetrical faces and were better at detecting whether a face was symmetrical or otherwise, supporting the hypotheses. A further, unexpected finding was that males made fewer mistakes than did females when determining whether individual faces were asymmetrical. The reduced ability of inebriated people to perceive asymmetry may be an important mechanism underlying the higher ratings of facial attractiveness they give for members of the opposite sex and hence their increased frequency of mate choice.
This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
Author's personal copy
An explanation for enhanced perceptions of attractiveness after
, J.W. Huber
, R.D.J. Bufton
, A.C. Little
School of Human and Life Sciences, Roehampton University, London SW15 4JD, UK
Department of Psychology, University of Stirling, Stirling FK9 4LA, UK
Received 9 June 2009; received in revised form 27 November 2009; accepted 10 February 2010
Acute alcohol consumption increases ratings of attractiveness to faces. This may help to explain increased frequencies of sexual encoun-
ters during periods of alcohol intoxication. At least in part, such increased attraction may be the result of alcohol consumption decreasing
ability to detect bilateral asymmetry, presumably because of the reductions in the levels of visual function. We tested the hypotheses that
acute alcohol consumption decreases ability to detect asymmetry in faces and reduces preference for symmetrical faces over asymmetrical
faces. Twenty images of a pair of faces and then 20 images of a single face were displayed on a computer, one at a time. Participants were
instructed to state which face of each of the face pairs displayed was most attractive and then whether the single face being displayed was
symmetrical or not. Data were collected near campus bars at Roehampton University. Sixty-four self-selecting students who undertook the
study were classiﬁed as either sober (control) or intoxicated with alcohol. For each face pair or single face displayed, participant response
was recorded and details of the alcohol consumption of participants that day were also obtained. Sober participants had a greater preference
for symmetrical faces and were better at detecting whether a face was symmetrical or otherwise, supporting the hypotheses. A further, unex-
pected ﬁnding was that males made fewer mistakes than did females when determining whether individual faces were asymmetrical. The
reduced ability of inebriated people to perceive asymmetry may be an important mechanism underlying the higher ratings of facial attrac-
tiveness they give for members of the opposite sex and hence their increased frequency of mate choice. Ó 2010 Elsevier Inc. All rights
Keywords: Attraction; Faces; Symmetry perception; Acute alcohol consumption
Throughout the animal kingdom, individuals discriminate
between potential mates with regard to their physical appear-
ance. Evolutionary psychologists typically consider that such
discrimination represents a sexually selected adaptation to
enhance mate choice (Thornhill and Gan gestad, 1999 ). For
instance, the degree of symmetry in bilatera l eleme nts of
the body provides a veritable indication of phenotypic quality
(Watson and Thornhill, 1994). Comparative data for a range
of organisms demonstrates that individuals who are more
exposed to environmental or gene stressors during develop-
ment exhibit a greater divergence from structural symmetry
(Jones et al., 2001; Leary and Allendorf, 1989; Møller,
1992). Therefore, symmetry may well play an important role
in mate selection. Subtleties of body morphology are also an
important factor in human mate choice with preferences for
symmetry in breasts (Singh, 1995) and in general body shape
e et al., 2000). As might be expected, ratings of attrac-
tiveness by people about adult members of the opposite sex
are also inﬂuenced by degree of facial bilateral symmetry
(Little et al., 2007; Perrett et al., 1999).
Studies conducted in both a ‘‘ﬁeld’’ setting (Jones et al.,
2003) and a laboratory setting (Parker et al., 2008) have
demonstrated that acute alcohol consumption increases
ratings of attractiveness to faces of the opposite sex, which
may help to explain increased frequencies of mate choice,
that is, sexual practice, during periods of alcohol intoxica-
tion (hereafter ‘‘intoxication’’). A further study (Egan and
Cordan, 2008) found only alcohol-induced increases in
ratings of facial attractiveness for sexually mature (female)
faces wearing make-up. However, this ﬁnding agrees with
those of the aforementioned studies, which presented
stimuli of mature individuals (i.e., older than 18 years). A
recent paper suggests that, in part, this increase in attrac-
tiveness ratings during intoxication may be the result of
alcohol consumption decreasing ability to detect bilateral
asymmetry (Souto et al., 2008), presumably because of
* Corresponding author. Tel.: þ44(0)208-392-8120; fax: þ44(0)208-
E-mail address: email@example.com (L.G. Halsey).
0741-8329/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved.
Alcohol 44 (2010) 307e313
Author's personal copy
reductions in level of visual function (Andre, 1996; Colzato
et al., 2004; Watten and Lie, 1996). However, Souto et al.
(2008) assessed perception of asymmetry using simple
geometric images and it is presently unclear whether percep-
tion of symmetry in faces involves the same or different brain
mechanisms as perception of symmetry in patterns and
shapes (Oinonen, 2003; Rhodes et al., 2005; Wilkinson
et al., 2000; Wilson and Wilkinson, 2002). Furthermore,
Souto et al. (2008) did not test the effects of alcohol
consumption in combination with symmetry/asymmetry on
ratings of attractiveness.
The present study progresses the work of Souto et al.
(2008) by testing whether acute alcohol consumption
decreases ability to detect asym metry in faces and, further-
more, whether such consumpt ion reduces preference for
symmetrical face s. We hypothesize that participants intoxi-
cated with alcohol are less able to detect asymmetry in
faces and have a reduced preference for symmetrical faces
over asymmetrical faces. If supported, this would suggest
that during periods of acute alcohol consumption the reason
that people ﬁnd faces more attractive may be because of
their reduced ability to detect asymmetry in faces, and this
may also explain their reduced preference for symmetrical
Materials and methods
A between-groups design was used in the present study,
which was quasi-experimental as participants were self-
assigned to a group. Participants, classed as either sober or
intoxicated, undertook two tests. Test ‘‘a’’: they were shown
a series of paired faces where the two faces of a pair were of
the same person and had been manipulated such that one was
symmetrical and the other asymmetrical, and were asked to
choose the more attractive face. Test ‘‘b’’: they were shown
a series of single faces to which they responded on whether
they perceived the image to be symmetrical or asymmetrical.
For test ‘‘a,’’ the dependent variable was the numb er of paired
faces of which the participants preferred the symmetrical
face, and for test ‘‘b,’’ it was the number of correctly identi-
ﬁed asymmetrical faces.
Participants and location
In total 69 participants (33 males and 36 females, 66
were white , 3 were of various Asian ethnicities, combined
mean age 6 one standard deviation: 22 6 5.0 years) under-
took the test between March and June of 2008, and the data
from 64 of them were included in analyses. Data for ﬁve
participants were not included because they were not clas-
siﬁed as either sober or intoxicated. Data for the present
study were collected from students and bar staff around
the bars at Roehampton University, London. Thus, similarly
to the study by Egan and Cordan (2008) and Jones et al.
(2003), the present study has strong ecological validity. In
other words, the study ﬁndings are more likely to be repre-
sentative of the effects that acute alcohol consumption has
on people in typical life situations in terms of ability to
perceive facial asymmetry and preference for facial
symmetry. In turn, ﬁndings supporting the ﬁrst stud y
hypothesis are more likely to be true explanations, at least
in part, for the increases in mate choice by people who are
intoxicated. Most of the data were collected between the
hours of 20:00 and 23:00, and all between 19:15 and
01:30. Sober people and intoxicated people were tested
during both the earlier hours and the later hours of the
evening. The tests were conducted at locations near to the
bar where each particular participant had been present.
However, the locations chosen ensured that those partici-
pants received minimal distraction from their surrounding
environment while undertaking the tests, that is, outside
the bar or in an adjacent room. Requests were made to
potential participants to undertake the test on the evening
of the test and, in the case of intoxicated participants, after
they had been drinking. Therefore, participants were never
encouraged to drink in preparation for doing the test. All
the participants were tested voluntarily. The procedures fol-
lowed were in accordance with the ethical standards of
Roehampton University and with the Helsinki Declaration
of 1975, as revised in 1983. None of the participants who
undertook the test sober reported that they consistently
abstained from drinking alcohol.
Before undertak ing the test, participants were asked to
wear glasses if they required them and they were also asked
if their glasses corrected their eyesight to normal in terms
of viewing the computer screen. After the test, all the
participants reported that they could view the faces on the
computer screen clearly.
All images were photographs of white individuals (aged
between 18 and 25 years) without spectacles or obvious
facial hair. Photographs were taken under standardized
lighting conditions and with the participants posed with
a neutral expression, following previous studies (Jones
et al., 2003; Parker et al., 2008). To equate size, all images
were aligned to standardize the position of the pupils in the
image. Composite images, composed of multiple images of
different individuals, were used as base faces (10 male and
10 female composite images each made of 5 individual
images). The composite faces were created using specially
designed software. Details of the software and the compos-
iting and transformation techniques can be found in Benson
and Perrett (1993) and Tiddleman et al. (2001). This tech-
nique has been used to create composite images in many
previous studies (Benson and Perrett, 1993; Little and
Hancock, 2002; Tiddleman et al., 2001). Images were made
perfectly symmetrical and then a transform applied. The
transform was different for each image representing the
difference between an original image and its symmetrical
308 L.G. Halsey et al. / Alcohol 44 (2010) 307e313
Author's personal copy
counterpart. In this way, the transform then applied the
symmetry apparent in an original individual image. A
similar technique, although not using composites, has been
used in previous studies ( Little et al., 2007 ). The transform
then created two images, one symmetrical and one asym-
metrical, for each base face. Images were then masked on
the outline of the face so that hair and clothing cues were
not visible in the image. Fig. 1 shows an example of trans-
formed faces made using these methods.
The size of each face image as it appeared on the
computer screen (IBM Thinkpad model 2682, length of
diagonal: 36 cm) was approximately 15 10 cm wide.
Participants sat with the screen directly in front of them
and viewed it freely from a distance of approximately
50 cm. Twenty pairs of faces were presented to participants
in test ‘‘a’’ (Fig. 1), and 20 single faces were presented to
participants in test ‘‘b.’’ All of the faces presented in test
‘‘b’’ had been shown in test ‘‘a.’’
An explanation of the tas k was read to the participant
from an instruction sheet, which fully informed them about
the nature of the study. The participants were then asked to
provide information including their age, weight, how much
they estimated they had drunk so far that day, at what time
they had started drinking, and at what time they had last
eaten a meal.
Firstly, 20 paired faces were shown to participants; one
face was symmetrical and the other asymmetrical. They
were displayed using PowerPoint (Microsoft Corp.). The
pairs of faces were shown in the same order to all the partic-
ipants, which alternated between female and male image
pairs. In 10 pairs, the symmetrica l face was displayed on
the left. Each pair of faces appeared on the screen for 5 s
after which the screen became blank. After this point, the
participant told the experimenter which face they found
more attractive. When the participant was ready, the
experimenter pressed a key on the keyboard to view the
next image. The duration of the test itself was approxi-
mately 2 min.
Secondly, 20 single faces were shown, displayed in the
same pseudo-random order to all the participants. Each face
was either the symmetrical or asymmetrical face of the
pairs shown in test ‘‘a.’’ Seven of the faces were symmet-
rical. Each face was shown for 5 s after which the screen
became blank. The participant then told the experimenter
whether they perceived the image to be symmetrical or
asymmetrical. This test also took about 2 min to complete.
All the participants were asked to abstain from drinking
alcoholic beverages during the test. Directly after the test,
participants were asked to blow into a portable breath
alcohol analyzer (Alcosign CA-1000; Industrial Computer
Components Limited, Auckland, NZ), which measures the
amount of alcohol in the deep lung air. This provides an
estimate of blood alcohol levels: blood alcohol content
(BAC, %, g alcohol/100 g blood).
Participants were allowed to stop their participation at
any time during the test.
Determining sober and intoxicated groups
Of the 69 participants who undertook the test, 64 were
classiﬁed into either the sober or intoxicated group. A
considerable majority of the people present at the bars
during the periods that the tests were conducted had drunk
alcohol to some extent that day. To ensure satisfactory
group-level sample sizes, the sober group of participants
necessarily included some individuals who had drunk
a small amount of alcohol. Based on the experience in con-
ducting previous, related research (e.g., Souto et al., 2008),
it was anticipated that for participants to experience
a reduced ability to perceive asymmetry in a face they
would often need to be fairly heavily intoxicated. The
breath alcohol analyzer only gave a maximum reading of
0.08%, which is the legal limit for driving in the UK and
is a value which does not typically indicate being heavily
under the inﬂuence of alcohol. Therefore, classiﬁcation of
participants was undertaken based on several criteria
including BAC values. Participants were deemed sober if
their BAC value was 0.05% or less and they had either
drunk less than four units of alcohol that day or had drunk
less than six units that day and at a rate lower than two units
per hour. The average amount having been drunk that day
by sober participants was 1.2 units. Participants were
deemed as intoxicated if their BAC value was 0.08%, indi-
cating that their blood alcohol levels were 0.08% or higher,
and either they had drunk a total of 10 or more units of
alcohol that day or had drunk more than 6 units at a rate
higher than 2 units per hour just before taking the tests.
Participants were not included in statistical analysis if
excluded from the two groups. This exclusion of ﬁve partic-
ipants served as a ‘‘buffer’’ with the intention of creating
a more marked difference in the mean amount of alcohol
Fig. 1. A pair of f aces shown to participants in test ‘‘a.’’ The l eft-hand
face is symmetrical, whereas the right-hand face is asymmetrical. Partici-
pants were asked which face they found more attractive.
309L.G. Halsey et al. / Alcohol 44 (2010) 307e313
Author's personal copy
ingested and concentration of alcohol in the blood of the in-
toxicated group compared with the sober group.
Values are provided as means 6 standard error of the
mean. To test for a preference in sober and intoxicated
participants for symmetrical faces, single-sample t-tests
were run to compare the number of times in test ‘‘a’’ that
the symmetrical face was preferred compared with the
number of times such a face should be preferred if
symmetry does not inﬂuence attraction (i.e., by chance
and hence 50% of the time; 10 times). Paired t-tests were
used to test for differences between the sober and intoxi-
cated groups. Firstly, a difference was tested for in the
percentage of sober and intoxicated partic ipants who
preferred the symmetrical face over the asymmetrical face
across a set of face pairs (test ‘‘a’’). Secondly, a difference
was tested for in the ability of sober and intoxicated partic-
ipants to detect symmetry and asymmetry across a set of
single faces (test ‘‘b’’). Across all participants, correlations
were conducted on preference for the symmetrical face of
a face pair against ability to detect symmetry or asymmetry
in faces. Finally, paired t-tests were used to test for differ-
ences in responses by males and females to the two tests,
within the sober group and within the intoxicated group.
Several statistical tests were undertaken to check that
certain possible confounds and methodological limitations
were not inﬂuencing the results. Pearsons’s correlations
were run to test for relationships indicating a learning effect
by participants during the test ‘‘b,’’ that is, between ability
to perceive whether a face is symmetrical or asymmetrical
and the number of faces already viewed, for both groups
separately. Pearson’s correl ations were also run to test for
relationships between the responses of participants in both
test ‘‘a’’ and test ‘‘b’’ against time of day, again for both
groups separately, to check that the time the test was taken
did not affect responses to the tests. Statistics were run in
Excel (Microsoft Corp.) and SPSS v.15 (SPSS Inc.).
For the 69 participants, 36 were classiﬁed as sober, 28 as
intoxicated, and 5 as neither. For the sober and the intoxi-
cated groups, both the frequency distribution of the data
for test ‘‘a’’ and for test ‘‘b’’ did not differ signiﬁcantly
from normal (one-sample KolmogoroveSmirnov tests,
P O .05 in all the cases).
Effects of intoxicat ion on preference for and detection of
Both sober participants and intoxicated participants
preferred symmetrical to asymmetrical faces (test ‘‘a,’’
single-sample t-test, preference for symmetrical faces:
66.9 6 3.25%; t
5 5.21, P # .001 and 59.3 6 2.91%;
5 3.19, P # .01, respectively). However, sober participants
preferred the symmetrical face over the asymmetrical face of
a pair signiﬁcantly more often than did intoxicated partici-
pants (paired t-test, t
5 2.95, P # .01; Fig. 2).
When participants wer e asked to decide whether a face
was symmetrical or asymmetrical (test ‘‘b’’) across the
set of faces shown, sober participants more frequently gave
the correct answer than did intoxicated participa nts. This
was the case both when considering the symmetrical and
asymmetrical faces (paired t-test, number of wrong
answers: 3.5 6 0.36, 4.7 6 0.51; t
5 2.47, P # .05) and
when considering only the asymmetrical faces (paired
t-test, number of wrong answers: 2.9 6 0.31, 3.9 6 0.48;
5 2.39, P # .05).
A Pearson’s correlation found no signiﬁcant correlation
between preference for symmetry and ability to perceive
symmetry or asymmetry across all the participants
5 0.21, P 5 .11); however, a Spearman’s Rho correla-
tion, which tests for any form of monotonic relationship,
reported a signiﬁcant, positive correlation (r
P # .05).
Comparison of preference and detection between genders
There was no signiﬁcant difference in the responses of
male and female participants to test ‘‘a’’ (paired t-test,
5 0.326, P 5 .748; intoxicated, t
P 5 .110). There were signiﬁcant differences in the
responses of the genders to test ‘‘b‘‘ in sober participants
when considering the symmetrical and asymmetrical faces
(paired t-test, number of wrong answers: 2.8 6 0.42,
4.3 6 0.54; t
5 2.358, P # .05) and when considering on-
ly the asymmetrical faces (paired t-test, number of wrong
answers: 2.2 6 0.35, 3.5 6 0.49, t
5 2.143, P # .05), with
females giving wrong answers more often in both cases.
There was also a signiﬁcant difference in the responses of
males and females to test ‘‘b’’ within intoxicated partici-
pants when only the asymmetrical faces were considered
(paired t-test, number of wrong answers: 3.6 6 0.43,
Fig. 2. The percentage of sober (open squares) and intoxicated (closed
triangles) participants who preferred the symmetrical face over the asym-
metrical face of each face pair. The face pairs are numbered in the order
that they were shown to participants. The values on the far right are means
of all the 20 face pairs, and the associated bars represent 6one standard
310 L.G. Halsey et al. / Alcohol 44 (2010) 307e313
Author's personal copy
4.2 6 0.96, t
5 3.313, P # .01), again with females giving
more wrong answers. However, there was no signiﬁcant
difference within intoxicated participants when considering
symmetrical and asymmetrical faces together (paired t-test,
number of wrong answers: 4.5 6 0.66, 4.9 6 0.95,
5 2.000, P 5 .060).
In test ‘‘b,’’ there was no improvement in the ability of
participants to perceive whe ther a face was symmetrical
or asymmetrical during the course of viewing the 20 faces
(Pearson’s correlation, sober: r 5 0.305, P 5 .191; intoxi-
cated: r 5 0.099, P 5 .677). There was no relat ionship
between the responses of participants and the time of day
(Pearson’s correlation, test ‘‘a,’’ sober: n 5 36, r 5 0.115,
P 5 .505; test ‘‘a,’’ intoxicated: n 5 27, r 5 0.097,
P 5 .630; test ‘‘b,’’ sober: n 5 36, r 5 0.282, P 5 .096;
test ‘‘b,’’ intoxicated: n 5 27, r 5 0.118, P 5 .559).
Conﬁrming previous results, sober participants had
a preference for symmetrical versus asymmetrical faces
(Little et al., 2007; Perrett et al., 1999). Intoxicated partic-
ipants also had a preference for symmetrical faces.
However, the preference of intoxicated participants for
symmetrical faces was less than that of sober participants.
Intoxicated participants were also less able to detect when
faces were asym metrical than were sober participants.
Furthermore, there was some evidence for a correlation
between symmetry preference and detection, which was
also reported by Oinonen and Mazmanian (2007).
Together, these ﬁndings provide evidence for the
hypotheses of the present study that intoxication reduces
ability to detect asymmetry in faces (supp orting the claims
of Souto et al., 2008) and reduces preference for symmet-
rical faces. In reality, faces are always asymmetrical and
hence the effect of intoxication in reducing ability to
perceive that asymmetry may be an important mechanism
underlying higher ratings of facial attractiveness for
members of the opposite sex when inebriated (Jones
et al., 2003), at least in sexually mature females wearing
make-up (Egan and Cordan, 2008), and, in turn, increased
frequency of mate choice. This mechanism may also at
least in part explain the otherwise unexpected ﬁnding that
intoxicated heterosexual participants rate same sex faces
more highly compared with sober heterosexual participants
(Parker et al., 2008). Th e reduced preference for symmet-
rical faces by intoxicated participants may be explained
by the fact that those participants exhibit a greater increase
in perceived attra ctiveness of less symmetrical faces than
more symmetrical faces. However, it is far from certain that
the reduction in ability to perceive asymmetry by intoxi-
cated participants explains their reduction in facial
symmetry preference. The correlation between symmetry
preference and detection does not, of course, infer causality
and the correlation coefﬁcient, although signiﬁcant, was not
strong. This in fact suggests that there is a limit on the
impact that reduction in detection ability has on symmetry
There was also an effect of gender on the ability of
participants to discern whether a face was symmetrical or
asymmetrical. To our knowledge, this difference has not
been reported before. Males made fewer mistakes than
did females when asked to decide whether individual faces
were symmetrical or not, in both the sober and intoxicated
conditions. This might be at least partly explained by the
fact that physical appearance is considered to play a larger
role in the mate choices of males than females (Buss and
Schmitt, 1993). For example, Rhodes et al. (1998) reported
males ﬁnd the appeal of symmetrical opposite sex faces
stronger than do females. However, there was no gender
difference in preference for symmetrical versus asymmet-
rical faces in the present study.
Similarly to the studies of Egan and Cordan (2008),
Jones et al. (2003), and Souto et al. (2008), the ‘‘ﬁeld’’
element of the present study, while providing ecological
validity (Sheridan, 1979), meant that it did not include
speciﬁc controls for potential personality differences
between the two groups of participants. The present study
cannot rule out the possibility that intoxicated participants
generally tend to drink more than do sober participants
and that this either resulted in different personality types
in the two participant groups and/or that chronic alcohol
consumption affects asymmetry detection (Oinonen and
Sterniczuk, 2007). However, none of the sober participants
claimed that they consistently abstained from alcohol and
indeed a number of them were seen regularly drinking on
the campus while data were being collected for the present
study. Furthermore, the relatively young age of most of the
study participants meant that they were unlikely to have had
the opportunity to develop chronic alcoholism.
Intoxicated participants might experience changes
alongside deterioration in visual perception that would
affect performance on the test, for example effects of
fatigue, particularly if they have been drinking and active
for a large proportion of the day. Yet importantly, none of
the intoxicated participants was observed to be sufﬁciently
inebriated, or indeed tired, that they undertook the test
without due care, concentration, and composure from start
to ﬁnish. Indeed, it was noteworthy how well intoxicated
participants concentrated on the tests, presumably partly
because peopl e are usually interested in faces. These obser-
vations are supported by the symmetry preference shown by
intoxicated participants indi cating that they were not per-
forming at random. However, longer, more arduous tests
might have been confounded by a difference in mental
focus between sober and intoxicated groups. An important
and arguably neces sary aspect of the tests used in the
present study was that they were concise in nature, with
participants exposed to a relatively small number of stimuli
311L.G. Halsey et al. / Alcohol 44 (2010) 307e313
Author's personal copy
such that they had to concentrate for only a couple of
minutes at a time.
However, using a small number of stimuli has the disad-
vantage of providing a smaller opportunity to uncover
a signiﬁcant difference (i.e., the design had a lower power),
which was problem atic in the present study because, for
example, testing for a difference in prefere nce for symmet-
rical faces is subtle. This is because even with no symmetry
preference whatsoever, participants would on average be
expected to report the symmetrical face of a pair to be more
attractive for 10 out of the 20 pairs shown to them, that is,
50% of the time by chance, whereas the preference of sober
participants for symmetrical faces is not always particularly
high despite being signiﬁcant (e.g., 57% by European
participants; Little et al., 2007). Therefore, the present
study was searching for a small signal when testing for
a difference in symmetry preference between sober and
intoxicated participants with only a small number of stimuli
used to reveal that signal.
Furthermore, the stim uli varied considerably in terms of
how differently sober and intoxicated participants reacted
to them, for example with certain face pairs eliciting a much
higher symmetry preference in sober partic ipants compared
with intoxicated participants than other face pairs (Fig. 2).
The reasons for this variation are unclear but may indicate
that stimuli could in general be improved, enhancing the
ability to uncover differences between sober and intoxi-
cated people. Use of a breath alcohol monitor that provided
a greater range of readings at a higher resolution, coupled
with a greater subject sample size, would have provided
the opportunity to test for correlations between levels of
intoxication and perception of faces as well as differences
between groups, which arguably would give more compel-
ling evidence of a true association with alcohol consump-
tion. However, this assumes a strong relationship between
blood alcohol levels and its effect of that alcohol on rele-
vant perceptual capacities.
The present study’s design is validated despite the afore-
mentioned methodological constraints and the subtlety of
the effect being tested for, because it provides evidence
for intoxication reducing the ability to detect facial asym-
metry within the naturalistic setting of university bars. This
ﬁnding could represent one of the mechanisms to explain
the increased perceived attractiveness of faces after acute
alcohol consumption. In turn, this mechanism could help
explain the increased frequency of sexual encounters by
inebriated people. Although this mechanism is almost
certainly not the only one explaining increased frequency
of sexual encounters, because, for example, alcohol is
known to reduce behavioral inhibitions, the effects found
in the present study could operate in an additive fashion
with other effects of intoxication.
Although our study does not address the issue of the
underlying mechanisms that may explain the effects of
alcohol on symmetry perception, there are a number of
papers that report on physiological effects of acute alcohol
consumption on visual ability. For example, Andre (1996)
suggests that ethanol reduces contrast sensitivity, particu-
larly for details (i.e., higher frequencies of visual informa-
tion). Face perception and the detection of symmetry are
likely to rely on the perception of high frequencies. Thus,
alcohol may reduce perception of detail and therefore
reduce the ability to detect symmetry. Given that inebria-
tion appears to reduce judiciousness of mate choice, future
studies should test whether mate choices made during
alcohol intoxication result in reduced beneﬁts in terms of
potential offspring ﬁtness, using indirect measures such as
compatibility of genes, relationship length, and mea sures
of relationship satisfact ion (Roberts and Little, 2008).
Furthermore, investigations should be undertaken into
whether other changes that occur during intoxication, such
as reduced behavioral inhibitions, serve as the main expla-
nations to decrease mate choosiness or if reduced percep-
tual capacity is a dominant mechanism. Such studies may
be best served with a controlled, experimental design
including randomization and a placebo group (Parker
et al., 2008 ).
We thank the bar and security staff at Roehampton
University who facilitated data collection. Financial
support for this study by the School of Human and Life
Sciences, Roehampton University is gratefully acknowl-
edged. Drs. Ben Jones and Gina Pauli kindly provided
comments on a draft of this article.
Andre, J. (1996). Visual functioning in challenging conditions: effects of
alcohol consumption, luminance, stimulus motion, and glare on
contrast sensitivity. J. Exp. Psychol. Appl. 2, 250–269.
Benson, P., and Perrett, D. I. (1993). Extracting prototypical facial images
from exemplars. Perception 22, 257–262.
Buss, D., and Schmitt, D. (1993). Sexual strategies theory: an evolutionary
perspective on human mating. Psychol. Rev. 100, 204–232.
Colzato, L. S., Erasmus, V., and Hommel, B. (2004). Moderate alcohol
consumption in humans impairs feature binding in visual perception
but not across perception and action. Neurosci. Lett. 360, 103–105.
Egan, V., and Cordan, G. (2008). Barely legal: is attraction and estimated
age of young female faces disrupted by alcohol use, make up, and the
sex of the observer? Br. J. Psychol. 13, 415–427.
Jones, B. C., Little, A. C., Penton-Voak, I. S., Tiddeman, B. P., Burt, D. M.,
and Perrett, D. I. (2001). Facial symmetry and judgements of apparent
health. Support for a ‘‘good genes’’ explanation of the attractivenessd
symmetry relationship. Evol. Hum. Behav. 22, 417–429.
Jones, B. T., Jones, B. C., Thomas, A. P., and Piper, J. (2003). Alcohol
consumption increases attractiveness ratings of opposite-sex faces:
a possible third route to risky sex. Addiction 98, 1069–1075.
Leary, R., and Allendorf, F. (1989). Fluctuating asymmetry as an indicator
of stress: implications for conservation biology. Trends Ecol. Evol. 4,
Little, A. C., Apicella, C., and Marlowe, F. (2007). Preferences for symmetry
in human faces in two cultures: data from the UK and the Hadza, an iso-
lated group of hunter-gatherers. Proc. R. Soc. Lond. B 274, 3113–3117.
312 L.G. Halsey et al. / Alcohol 44 (2010) 307e313
Author's personal copy
Little, A. C., and Hancock, P. (2002). The role of masculinity and distinc-
tiveness on the perception of attractiveness in human male faces. Br. J.
Psychol. 93, 451–464.
Møller, A. (1992). Parasites differentially increase the degree of ﬂuctu-
ating asymmetry in secondary sexual characters. J. Evol. Biol. 5,
Oinonen, K. (2003). The Effects of Hormones on Symmetry Detection and
Perceptions of Facial Attractiveness. Ontario: Lakehead University.
Oinonen, K., and Mazmanian, D. (2007). Facial symmetry detection ability
changes across the menstrual cycle. Biol. Psychol. 75, 136–145.
Oinonen, K., and Sterniczuk, R. (2007). An inverse relationship between
typical alcohol consumption and facial symmetry detection ability in
young women. J. Psychopharmacol. 21, 507–518.
Parker, L., Penton-Voak, I. S., Attwood, A., and Munafo, M. (2008). Effects
of acute alcohol consumption on ratings of attractiveness of facial
stimuli: evidence of long-term encoding. Alcohol Alcohol. 43, 636–640.
Perrett, D., Burt, D., Penton-Voak, I., Lee, K., Rowland, D., and
Edwards, R. (1999). Symmetry and human facial attractiveness. Evol.
Hum. Behav. 20, 295–307.
Rhodes, G., Peters, M., Lee, K., Concetta Morrone, M., and Burr, D.
(2005). Higher-level mechanisms detect facial symmetry. Proc. R.
Soc. Lond. B 272, 1379–1384.
Rhodes, G., Profﬁtt, F., Grady, J., and Sumich, A. (1998). Facial symmetry
and the perception of beauty. Psychon. Bull. Rev. 5, 659–669.
Roberts, S., and Little, A. C. (2008). Good genes, complementary genes
and human mate preferences. Genetica 132, 309–321.
Sheridan, C. (1979). Methods in Experimental Psychology. Kansas City:
Holt, Rinehart and Winston.
Singh, D. (1995). Female health, attractiveness, and desirability for rela-
tionships: role of breast asymmetry and waist-to-hip ratio. Ethol. Soci-
obiol. 16, 465–481.
Souto, A., Bezerra, B. M., and Halsey, L. G. (2008). Alcohol intoxication
reduces perception of asymmetry: an explanation for increased percep-
tions of facial attractiveness after alcohol consumption? Perception 37,
Thornhill, R., and Gangestad, S. (1999). Facial attractiveness. Trends
Cogn. Sci. 3, 452–460.
Tiddleman, B. P., Burt, D. M., and Perrett, D. I. (2001). Prototyping and
transforming facial texture for perception research. IEEE Comput.
Graph. Appl. 21, 42–50.
e, M., Tasker, K., and Benson, P. (2000). Is symmetry a visual cue to
attractiveness in the human female body? Evol. Hum. Behav. 21, 191–200.
Watson, P., and Thornhill, R. (1994). Fluctuating asymmetry and sexual
selection. Trends Ecol. Evol. 9, 21–25.
Watten, R., and Lie, I. (1996). Visual functions and acute ingestion of
alcohol. Opthalmol. Physiol. Opt. 16, 460–466.
Wilkinson, F., James, T., Wilson, H., Gati, J., Menon, R., and Goodale, M.
(2000). An fMRI study of the selective activation of human extrastriate
form vision areas by radial and concentric gratings. Curr. Biol. 10,
Wilson, H., and Wilkinson, F. (2002). Symmetry perception: a novel
approach for biological shapes. Vision Res. 42, 589–597.
313L.G. Halsey et al. / Alcohol 44 (2010) 307e313