AD ALTA
JOURNAL OF INTERDISCIPLINARY RESEARCH
attractiveness of the face may be based on attributed personality
traits of the evaluated faces and that people tend to prefer and
consider more attractive those faces that resemble their own
personality traits. Little, Burt, and Perrett (2006) found that male
faces, attributed with a higher degree of extraversion, are
preferred by women who are also extraverted.
2 Objective
From the data mentioned above, we may ask the question, based
on the preferred personality traits present in the evaluated faces,
whether people tend to consider those faces that resemble their
own characteristics to be more attractive. It is also questionable,
whether this preference (if valid) applies generally (extends to a
non-mating context), or it is only valid in the evaluation of
potential sexual partners. To find an answer, as an example, we
have chosen extraversion/introversion as a personality trait and
intelligence.
3 Methods
The subjects (N=1,903) were Slovak men (N=754; 39.6%) and
women (N=1149; 60.4%) between the ages of 15 and 67 (mean
23.96; st.dev. 9.128) who were asked to fill in a battery of tests.
The intelligence score of each participant was measured using
two subtests (measuring verbal and visuospatial IQ) from the
standardized “Test of the Level of Mental Abilities” (Vonkomer,
1992) and the score of self-reported extraversion from the
“Personality Inventory KUD” (Miglierini & Vonkomer, 1986).
For the assessment of the attractiveness of faces, eight computer-
modified facial composites were used that represented high and
low levels of intelligence and high and low levels of extraversion
in both sexes. The technique of computer-modified facial
composites is based on the process, where initially a large
number of individuals complete the self–report extraversion/
introversion questionnaire (or intelligence tests). Further,
a smaller number of participants (e.g. 10%) who score the
highest and the 10% who scored the lowest are selected as
representative extraverted/introverted (or high/low intelligent)
individuals. These representatives are photographed and to
construct composites, hundreds (e.g., in case of extraversion/
introversion there were 219) of standard feature points are
marked as facial landmarks on each face. The mean coordinates
of each delineated feature point are calculated to generate
average shape information (Penton-Voak, Pound, Little, and
Perrett, 2006). Information about the typical shape is applied to
the facial average picture (an artificially created face computed
from the many photographs) and a face typical of an
extravert/introvert (or a person high/low in intelligence) is
created.
Within the face assessment task, participants were asked (along
with other tasks not listed here) to choose the most attractive
face from a set of two nearly identical male and female faces.
The participants were blind to the fact, that the two faces within
each set differed in the specific features known to indicate
different levels of intelligence (see Figure 1 – Kleisner,
Charvatova, & Flegr 2014) and different levels of
extraversion/introversion (see Figure 2 – Penton-Voak, Pound,
Little, and Perrett, 2006). A statistical analysis was carried out
using the SPSS program, version 16.
(a)
(b)
(a)
(b)
Figure 1. Face composites manipulated to show different levels
of intelligence. Faces marked a) refer to faces indicating low
intelligence, b) faces indicating high intelligence (Source:
Source: Kleisner, Charvatova, & Flegr 2014; http://journals.plos.
org/plosone/article?id=10.1371/journal.pone.0081237)
(a)
(b)
(a)
(b)
Figure 2. Face composites manipulated for facial extraversion
and introversion. Faces marked a) refer to faces indicating
introversion, b) to faces indicating extraversion (Source: male
faces: Penton-Voak, Pound, Little, & Perrett, 2006, p. 622;
female faces: http://www.bbc.co.uk/wales/radiowales/sites/scie
nce cafe/updates/ 20111016.shtml [cit. 2016-06-28])
4 Results
To answer the question of whether people tend to consider those
faces, that resemble their own characteristics based on their
preferred personality traits present in the evaluated faces to be
more attractive, we measured the intelligence and extraversion of
the participants and their attractiveness preferences for low/high
intelligence male/female faces and extravert/introvert
male/female faces. The intelligence and extraversion of the
participants were measured and found to have a normal
distribution (values for intelligence: mean=13.79; st. dev. 2.63;
skewness=-.286; kurtosis=.160, values for extraversion:
mean=10.85; st. dev.=3.59; skewness -.697; kurtosis=.051),
therefore parametric tests were used in further computations.
Table 1
T-test for differences in intelligence between 2 groups according
to the preference for low/high intelligence facial composites
Preferred
male facial
composite
N
Mean
St. dev.
T
Sig.
(2-
tailed)
Low
intelligent
905
13.85
2.747
.877
.381
High
intelligent
998
13.74
2.524
Preferred
female facial
composite
N
Mean
St. dev.
T
Sig.
(2-
tailed)
Low
intelligent
732
13.67
2.931
-1.581
.114
High
intelligent
1171
13.87
2.425
Table 1 shows there is no statistically significant difference in
the intelligence level of the participants which preferred the face
indicating low intelligence to those who considered the face
indicating high intelligence to be more attractive, neither in the
case of male facial composite nor female facial composites. It,
therefore, seems that the intelligence level of the evaluator does
not significantly influence the preference or assessed
attractiveness of facial features typical of the presence of
intelligence. Similarly (see table 2), we did not find a statistically
significant difference in the self-reported extraversion levels of
the participants which preferred the introverted facial composite
to those who considered the extraverted facial composite to be
more attractive, neither in the case of the male facial composite
nor the female facial composite. It is likely that extraversion
levels do not have an impact on the attractiveness ratings of
faces having typical facial features of introversion or
extraversion.
Identical conclusions can be drawn from the t-tests results
computed separately for the analysis of the facial preferences of
potential sexual partners (male choices for female facial
composites and vice versa). Tables 3 and 4 do not show any
significant differences in intelligence (Table 3) and extraversion
- 42 -