Content uploaded by Laura Jenkins
Author content
All content in this area was uploaded by Laura Jenkins on Nov 26, 2017
Content may be subject to copyright.
38
Volume 1 • Issue 1 • 1000107Madridge J Neuro Sci.
Madridge
Journal of Neuroscience
Research Article Open Access
Does personality effect facial emotion recognition? A
comparison between the older Ekman Emotion Hexagon
Test and a newly created measure
Laura Jenkins
Northumbria University, Department of Psychology, Newcastle-upon-Tyne, UK
Article Info
*Corresponding author:
Laura Jenkins
Northumbria University
Department of Psychology
Newcastle-upon-Tyne
UK
E-mail: laurajenkins840@gmail.com
Received: October 24, 2017
Accepted: November 10, 2017
Published: November 16, 2017
Citation: Laura J*. Does personality effect
facial emotion recognition? A comparison
between the older Ekman Emotion Hexagon
Test and a newly created measure. Madridge
J Neuro Sci. 2017; 1(1): 38-46.
Copyright: © 2017 Laura J. This work is
licensed under a Creative Commons
Attribution 4.0 International License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the
original work is properly cited.
Published by Madridge Publishers
Abstract
Recent research has highlighted one possible problem faced when it comes to
assessing the recognition of emotion in the human face. Previous research has suggested
that the images used in the methods of assessment are becoming too familiar within the
psychological research domain. They therefore suggest new ways of creating images to
look at facial emotion recognition.
To investigate this issue, the current study created a new emotion recognition task
(Laura’s Emotion Hexagon Task) and compared this to the older Ekman’s Emotion
Hexagon Test. 60 participants completed both tasks along with the IPIP-NEO. This
measure looked at whether personality could predict the scores for the two facial
emotion recognition tasks. Results showed that the agreeableness and extraversion
personality were significant predictors of Laura’s Emotion Hexagon Task. Discussions of
the findings are in relation to previous research about methodological issues surrounding
facial emotion recognition, and results are discussed in relation to previous findings of
how personality can effect facial emotion recognition.
Introduction
Emotion recognition has been researched for many years, dating back to the late
1800s. One of the first theorists to conduct research about the expression of emotion
was Darwin [6]. Darwin wrote a book describing how emotions can be in voluntary in all
animals and humans. He focused more upon the biological reasons for displays of
emotion, relating it to both animal and human species. This was the first real topic for
literature critiques to look at within the field of emotion recognition.
In recent years, Cacioppo and Gardener [3] have given their accounts of emotion
and looked at the issue with a more psychological perspective. The study looked at how
human emotion is measured. They reviewed the past literature that had aimed to
discover more about human emotions from different perspectives. Cacioppo and
Gardener [3] described different methodologies of measuring emotion, such as fMRI
techniques. A review of the literature, which looked at how situational variables influence
emotion. One perspective involved looking at literature that investigated emotion using
PET scans [12]. Researchers also looked at how intelligence links in some way to emotion
recognition [15].
When discussing emotion recognition, it is very important to look at the research
conducted by theorists such as Ekman. Ekman, Friesen and Tomkins [11] had also looked
at past research relating to Darwin [6]. They noted that Darwin had mainly focused on the
biological reasons of how emotional expressions can occur in the human face (for example,
Madridge Journal of Neuroscience
39
Madridge J Neuro Sci. Volume 1 • Issue 1 • 1000107
which muscles and tissues are in use during expression).
Conclusions suggest that research should have a slight change
in direction. Ekman, Friesen and Tomkins decided to investigate
the emotions that observers could see in the human face (for
example, happiness or disgust) instead of focusing on how the
emotions themselves are displayed and what muscles are used.
This gave several opportunities to question how researchers
could actually measure emotion recognition in the human face,
therefore inspiring Ekman to create several measures and
methodologies to help measure emotion recognition in faces.
One emotion recognition test is named as Ekman’s 60
Faces Test. In this test, participants have to decide which
emotion is being displayed in a series of photographs. The
photographs are taken from ‘Pictures of Facial Affect’ [7].
Ekman’s 60 Faces Test was created with 10 participants
(models) who had six different facial expression photographs
taken. The expressions were happiness, anger, sadness, fear,
disgust and surprise, and Ekman had identified all as the six
basic human emotions in his earlier work. As the name of the
test suggests, Ekman’s 60 Faces Test contains 60 facial
photographs so therefore participants can obtain a total score
of 60. A score of 60 would mean that a participant had an
excellent ability to recognise the emotions in human faces.
Other researchers investigated the uses of Ekman’s 60
Faces Test by looking to see whether the test could be used as
a diagnostic tool to highlight any problems with facial emotion
recognition in patients who had frontotemporal dementia [7].
Results demonstrated that Ekman’s 60 Faces Test discriminates
between patients with frontotemporal dementia and healthy
patients when it came to recognise emotions in the human
face.
The research conducted by Ekman [9] was not the only
line of investigation that was taking place. Matsumoto,
LeRoux, Wilson-Cohn, Raroque, Kooken, Ekman, Yrizzary,
Loewinger, Uchida, Yee, Amo and Goh [17] had attempted to
create a new way of measuring facial emotion recognition.
They designed the Japanese and Caucasian Brief Affect
Recognition Test (JACBART) to help researchers understand
and measure people’s individual differences in emotion
recognition ability. Matsumoto had worked with Ekman in
previous years to create measures such as the Japanese and
Caucasian Facial Expressions of Emotion (JACFEE) and Neutral
Faces (JACNeuF). Literature surrounding these measures had
shown differing results lacking in validity and reliability.
Matsumoto and colleagues therefore decided to create a new
measure (JACBART) and included improvements to the
criticism made earlier in their research. They used photographs
that had good validity and reliability, and included
photographs of people who were of a variety of ethnic origins.
One notable criticism of facial emotion recognition
methods was that the same series of photographs were being
used on multiple occasions. The photographs were taken
from ‘Pictures of Facial Affect’, which was created by Ekman
and Friesen [11].
Research (Suzuki, Hoshino and Shigemasu) [20] had
suggested that the non-morphed images used form ‘Pictures
of Facial Affect’ [11] were becoming too easy to recognise.
They therefore proposed a morphing technique which could
increase the difficultly level of the emotion recognition
photographs.
Ekman’s Emotion Hexagon Test has been created so that
individuals would have to choose one emotion out of a
photograph that would have contained two different
emotions. The photographs, used in ‘Pictures of Facial Affect’
were morphed to form images that contained two emotions
out of the six basic emotions (happiness, sadness, anger, fear,
surprise and disgust). It was suggested that an individual
would find Ekman’s Emotion Hexagon Task more challenging
than his 60 Faces Task because the different features that
could be recognised in each emotion were less apparent.
Ekman’s Emotion Hexagon Test was shown that it could
be used successfully [4]. Researchers were investigating
human facial emotion recognition across the adult life span so
therefore created two mini studies to do this. Study 1 used the
original photos created for Ekman’s 60 Faces Task. Study 2
used the morphed images created for Ekman’s Emotion
Hexagon Test instead of the non-morphed images as
described above. Results showed that with increasing age,
participants found it more difficult to identify five out of the
six basic emotions. The emotion of disgust was the opposite
of this. With increasing age, participants found it easier to
interpret the emotion of disgust in faces.
Ekman’s Emotion Hexagon Test has very good reliability
and validity. The FEEST Manual [22] gives details of both
Ekman’s 60 Faces Task and more importantly for this study; it
gives details about Ekman’s Emotion Hexagon Test.
Split-half reliability scores demonstrated reliability. 40
participants were shown the images used in the Emotion
Hexagon Test to ensure that each emotion could be
successfully recognised. Only 50 percent of the images were
reliable in allowing a participant to identify emotions in the
human face. Only this 50 percent of photographs were used
in Ekman’s Emotion Hexagon Test and the photographs that
were unsuccessful were taken out of the procedure. Reliability
techniques (split-half reliabilities) gave statistical significance
for the emotion of sadness, anger, fear, disgust, and surprise.
The validity of Ekman’s Emotion Hexagon Test is successful
due to the use of it on many occasions. Many empirical studies
[4] [2] [11], have managed to use the series of images to create
Ekman’s Emotion Hexagon Test. This gives the test a very
good rate of validity as each individual study used the pictures
to show the correct displays of emotion to different
participants.
Research has also looked at Ekman’s Emotion Hexagon
Test [5]. This investigation looked at strategies used when
participants have to choose a certain emotion in the human
face. They concluded that it was a very advantageous to
include some forced-choice answers. For example, in Ekman’s
Emotion Hexagon Test, participants have to choose from a
variety of six basic emotions instead of just writing down the
ne displayed emotion. Researchers noted that this was a very
good way of assessing emotion recognition in faces as it
Madridge Journal of Neuroscience
40
Madridge J Neuro Sci. Volume 1 • Issue 1 • 1000107
eliminates the choice of simply guessing the answer.
With the issue of universality of previous photographs
becoming more apparent, new measures help researchers to
identify new relationships within the field of emotion
recognition in faces. This is, therefore, one of the main aims of
the current study, to create a new and updated measure of
emotion recognition.
The present researchers had noted that for a successful
comparison to take place there had to be a specific variable to
measure. In recent years, these variables have included aspects
such as emotional intelligence, personality and age.
Austin [2] used pictures taken from Ekman and Friesen
[11] to show that emotional intelligence correlated with
emotion facial recognition. Researchers had used Ekman’s 60
Faces Task and the Trait Emotion Intelligence Questionnaire
to show these relationships.
Terracciano, Merritt, Zonderman and Evans [21] had
demonstrated that the personality trait of openness can also
affect emotion facial recognition and concluded that the
more open an individual is, then the easier they can successfully
identify the basic emotions in the human face. Researchers
aimed to replicate a past study conducted by Matsuomo et al.
[17] who had conducted research using Caucasian and Asian
students. Terracciano et al. [21] used the NEO Five Factor
Inventory to assess each individual’s personality traits, and
used two sub-tasks of the Perception of Affect Task (PAT). The
two sub tasks included images, whichwas taken from the
research of Ekman and Friesen. Results did demonstrate that
openness could affect an individual’s facial emotion
recognition and that conscientiousness had a similar (but
smaller) effect.
Mill, Allik, Realo and Valk [18] looked at how openness
could affect an individual’s facial emotion recognition. They
were initially looking at age related differences in emotion
recognition but came across the finding that both openness
and conscientiousness could influence facial emotion
recognition. They used the measures Japanese and Caucasian
Facial Expressions of Emotion (JACFEE) and Neutral Faces
(JACNeuF), created by Matsumoto and Ekman [16], which
included the pictures used in Ekman’s work. Researchers also
used the NEO-FFI [1] which assessed the Big Five Factors of
Personality in each participant. Both openness and
conscientiousness were found to have positive correlations
with emotion recognition. This indicates that the more open
and conscientious a person is then the better they will do on
the facial emotion recognition tasks.
Aims and Predictions
The overall aim of this investigation is to create a new and
modern version of Ekman’s Emotion Hexagon Test and name
this Laura’s Emotion Hexagon Task. Laura’s Emotion Hexagon
Task will have the same layout as Ekman’s Emotion Hexagon
Test, using one male volunteer who agreed to have facial
emotive photographs taken. The male volunteer had six facial
emotion photographs taken, similar to Ekman’s - expression
of happiness, sadness, anger, fear, disgust and surprise.
Due to the most recent literature focussing on how
personality can relate to facial emotion recognition, the
variable of personality is going to be used to help compare
Ekman and Laura’s tasks. To do this, researchers will use the
IPIP-NEO (Goldberg) [13] to measure the Big Five Factors of
Personality. Goldberg et al. [14] discussed the development of
the IPIP-NEO (Goldberg) [13]. The IPIP-NEO allows people in
the public domain to have access to a personality test that
was simple to administer. This is one of the reasons as to why
the current investigation uses the IPIP-NEO. No extensive
administer training is needed to administer the test so
therefore it is a simple but effective way of assessing an
individual’s personality type. The variable of personality will
be measured because this area of psychology also has an
increasing number of research studies to support the
relationship between personality and emotion facial
recognition. Examples include research such as Terracciano et
al. [21].
From this study, it is predicted that there will be a positive
relationship between openness and both Laura’s task score
and Ekman’s task score. The more open an individual is then
the higher their scores will be on both Ekman’s Emotion
Hexagon Test and Laura’s Emotion Hexagon Task. It is also
predictions that openness will be a predictor of both Laura’s
and Ekman’s task scores. A final hypothesis is that Laura’s
Emotion Hexagon Task will be significantly correlated with
Ekman’s Emotion Hexagon Test.
Method
Design and Rationale
A 5 x 2 repeated measures design is used for the purposes
of this experiment. There are two factors for the experiment.
The first factor is ‘personality’ and this has five levels being
‘openness’, ‘conscientiousness’, ‘extraversion’, ‘agreeableness’
and ‘neuroticism’. The level of ‘openness’ will be split into two
categories of ‘low openness’ and ‘high openness’. This is so
that researchers can have a more detailed look at whether
facial emotion recognition is predicted by a person’s openness.
The second factor is ‘task type’ and this has two levels. The
first level is ‘Laura’s Emotion Hexagon Task’ and the second
level is ‘Ekman’s Emotion Hexagon Test’. There is only one
dependent variable in the experiment and this is the ‘score of
the tasks’. The scores of each task will be noted separately
(and not added together) so that a comparison can take place.
The second factor of ‘task type’ will use a repeated
measures design as participants are required to complete
both Laura’s Emotion Hexagon Task and Ekman’s Emotion
Hexagon Task. This will allow a comparison to between both
tasks. A between subjects’ design is being used on the first
factor of ‘openness score’. This is because participants can
only fall into one of the two categories of ‘high openness’ or
‘low openness’.
Participants
Before investigations could take places, two participants
Madridge Journal of Neuroscience
41
Madridge J Neuro Sci. Volume 1 • Issue 1 • 1000107
(one male and one female volunteer) had their photographs
taken for the creation of Laura’s Emotion Hexagon Task. These
two participants did not take part in any other phase of the
investigation. Researchers decided not to use the photographs
taken from the female volunteer as Ekman’s Test only used
males.
Sixty experimental participants (5 males and 55 females)
were recruited from a North East University so that testing
could begin. Participants had to be students of the university
and ideally aged between 18 and 65 years of age.
Materials
The IPIP NEO [13] was used to assess participants’
personality types. This comprised of a 120 item multiple-
choice questionnaire that was given to participants as an
online version.
Participants were given Ekman’s Emotion Hexagon Task
(2002) to assess their emotion recognition abilities within
facial photographs. This was a computer-based task, given to
participants by the use of a CD-ROM. The task comprised of
150 facial photographs of which participants had to decide
which basic emotion was present. The photographs consisted
of six different facial emotion expressions that had been
morphed to create a series of new images (photographs) for a
male model. Each new photograph (morphed) contained a
total of two basic emotions (ranging from 10-90 percent
ratios between the displays of emotions). Participants had to
note down which emotion was present. Participants’ maximum
score was 120 (20 points from each emotion expression).
Images displaying expressions with a 50:50 emotion ratio
were not counted in the points score. Participants had one
practice trial consisting of 30 photographs (points were not
counted at this stage); then had five test trials (points were
counted at this stage).
To create Laura’s Emotion Hexagon Task, photographs
were taken of a male volunteer expressing the five basic
emotions of happiness, sadness, anger, surprise and disgust.
As this task was looking at morphed facial images, the software
Phanta Morph was used to combine the emotions in the facial
images. Each emotion facial photograph contained two facial
expressions that could be identified and there were a total of
sixty photographs created (thirty for the male participant and
thirty for the female participant). After the computer
programme had been created, researchers decided to omit
the photographs taken from the female volunteer so that a
more appropriate comparison could take place.
Laura’s Emotion Hexagon Task (2010) was given to
participants to assess their emotion recognition abilities
within facial photographs. Please see figure 1 for an example
of the male emotive faces. This task was a computer based
programme consisting of 150 photographs from one male
volunteer. Each photograph contained two emotion
expressions which had been morphed into one photograph.
There were six possibilities of emotion expression – happiness,
sadness, anger, fear, surprise and disgust. The scoring was the
same as Ekman’s Emotion Hexagon Task, where participants
would not be scored for a photograph that had been created
with an emotion expression ratio of 50:50. The total score for
Laura’s Emotion Hexagon Task was 120 (20 points maximum
coming from each one of the six basic emotion expressions).
Figure1. Example of the images used in Laura’s Emotion Hexagon
Task.
Bottom right is the expression of sadness, bottom left is
the expression of happiness. Top left is the emotion of surprise
and top right is the emotion of disgust
Procedure
Participants were contacted to arrange an appropriate
meeting time and place for testing. This was normally in the
seating area of the 3rd Floor Northumberland Building at
Northumbria University as it was a quiet area used for
independent learning.
Participants were asked to read the participant information
sheet so that they could discover the aims behind the
investigation. Participants were informed that they could
withdraw from the investigation at any point up until the data
was sent for analysis.
When participants were satisfied, they signed a consent
form so that researchers had a clarification that participants
were willing to take part.
When this was completed, testing could begin. Participants
completed the IPIP-NEO (Goldberg) [13]. This was an online
multiple-choice questionnaire, which assessed each
participant’s personality.
Participants completed Ekman’s Emotion Hexagon Task
(2002). Please see list of materials for explanation of the task.
This task comprised of a practice phase (where scores were
not noted) and then five testing phases where scored were
noted. This took approximately ten minutes.
Participants completed Laura’s Emotion Hexagon Task
(2010). This was the second emotion recognition task using
150 morphed photographs of facial expressions. Please see
materials for explanation of the task. This again took
approximately 10 minutes and the practice trials did not
receive any score.
After all testing had been completed, participants were
Madridge Journal of Neuroscience
42
Madridge J Neuro Sci. Volume 1 • Issue 1 • 1000107
given a full debrief using the participant debrief sheet. This
was a two page debrief which detailed the full aims of the
investigation and gave a chance for participants to ask
questions. Once again, participants were reminded that they
could withdraw from the investigation within the allocated
time limit (before results were analysed).
If participants had ticked the appropriate box on the
consent form, they were then informed that they would be
notified of the overall results of the investigation, and that
they could contact the researchers at any point with queries.
Results
Sixty participants provided data from completing both
Laura’s Emotion Hexagon Task and Ekman’s Emotion Test.
Each test could have a maximum score of 120. Participants’
personality trait scores of openness, conscientiousness,
agreeableness, extraversion and neuroticism were also
collected. Each personality trait could have a maximum score
of 100.
Evaluation of Overall Task
Table 1. Mean scores (and Standard Deviations) of all
variables, including personality traits and both emotion
hexagon tests.
Laura’s
Task
Ekman’s
Test
Open-
ness
Conscientious-
ness
Extraver-
sion
Agreeable-
ness
Neuroti-
cism
81 (12.1) 99 (19.2) 21
(19.0)
46
(24.9)
50
(25.3)
55
(26.0)
47
(24.4)
Table Notes
The first section of the analysis was a correlation between
Laura’s Emotion Hexagon Test and Ekman’s Emotion Hexagon
Test. This correlation showed a positive (and significant)
correlation of r =.422, p=.001. Because there was a positive
correlation, results suggested that if the higher the score was
achieved on Ekman’s Emotion Hexagon Test, then the higher the
score would also be achieved on Laura’s Emotion Hexagon Test.
Please see table 2 for the correlational analysis statistics.
Table 2. Correlations between all personality and emotion task
variables.
123456
1Laura’s Test
2Ekman’s Test .422***
3Openness .114 .248
4Consciousness .015 .144 .246
5Agreeableness .226 .147 .243 .443**
6Extraversion -.268* .028 .400** .341** .260*
7Neuroticism .105 .119 -.155 -.195 -.399** -.393*
Table Notes: *p<.05, ** p<.01, *** p<.001
Sensitivity of Overall Task to Personality
Variables
Two multiple regression analyses were then ran. The first
was between Laura’s Emotion Hexagon Task and all five-
personality traits (openness, conscientiousness, agreeableness,
extraversion and neuroticism). This analysis presented a
significant model, R²=.459, F(5,54)=2.876, p=.023. The two
personality traits that were found to be significant predictors
of Laura’s Emotion Hexagon Task score were extraversion
t(54)= -2.695, p=.009 and agreeableness , t(54)=2.242, p=.029.
As extraversion has a negative Beta value of -.393, this
demonstrates the negative relationship between the score of
Laura’s Emotion Hexagon Task and extraversion. In contrast,
agreeableness has a positive Beta value of .319, demonstrating
the positive relationship between the score of Laura’s Emotion
Hexagon Task and agreeableness. Please see Table 3 for the
regression statistics for this model.
The second regression analysis was between Ekman’s
Emotion Hexagon Test and all five-personality traits. In this
case, the regression model was not significant suggesting that
not one of the five personality traits could predict an
individual’s score on Ekman’s Emotion Hexagon Test. Please
see table 4 for the regression statistics for this model although
it did not reach significance.
Table 3. Regression statistics the regression of the five personality
predicting Laura’s Emotion Hexagon Test
BTΒeta RR² ΔR²
Model .459 .210 .137
Openness .136 1.593 .214
Consciousness -.014 -.199 -.028
Agreeableness .149 2.242* .319
Extraversion -.188 -2.695** -.393
Neuroticism .043 .632 .086
Table 4. Regression statistics the regression of the five personality
predicting Ekman’s Emotion Hexagon Test
BTΒeta RR² ΔR²
Model .339 .115 .033
Openness .251 1.758 .250
Consciousness .063 .550 .082
Agreeableness .098 .877 .132
Extraversion -.043 -.371 -.057
Neuroticism .155 1.353 .196
Openness Trait
One of the aims of this investigation was to look in more
detail at the personality trait of openness. Despite the
regression analysis showing little detail about the openness
trait, a t-test analysis provided further details.
Scores for the trait of openness were categorised into
‘high’ and ‘low’ scores. High scores were those over 52 and
low scores were those under 48.
The t-test analysis demonstrated that openness does have
a significant effect upon an individual’s facial emotion
recognition in Laura’s Emotion Hexagon Test, t(58)= -2.740,
p=.008. The people who were seen to have low openness had
a mean task score of 79.8(11.46) in comparison to the high
openness people who scored 92.57(11.95).
In contrast to this, openness does not have a significant
effect on an individual’s facial emotion recognition in Ekman’s
Emotion Hexagon Test, t(58)= 1.414, p=.133. This result both
supported and did not support the first prediction made by
Madridge Journal of Neuroscience
43
Madridge J Neuro Sci. Volume 1 • Issue 1 • 1000107
researchers. As the t-test only proved a significant effect of
openness upon Laura’s Emotion Hexagon Task, the researchers’
prediction cannot be said to be truly correct and supported.
Individual Emotions
Regressions looked at whether the Big Five Personality
Traits (openness, conscientiousness, extraversion,
agreeableness and neuroticism) could predict the scores for
the individual emotions within Laura’s Emotion Hexagon Test.
These emotions were happiness, sadness, disgust, fear, anger
and surprise.
Regressions demonstrated that personality predicted an
individual’s performance on one of the six basic emotions
within Laura’s Emotion Hexagon Task. This emotion was
sadness, R²=.239, F(5,54)= 3.390, p=.010.Significant predictors
of disgust were both agreeableness, t(58)= 2.613, p=.012,
with a positive Beta value of .362 and extraversion, t(58)=
-2.683, p=.010, with a negative Beta value of -.384.
Another set of regressions looked to see if the five
personality traits (openness, conscientiousness, extraversion,
agreeableness and neuroticism) could predict the scores for
the individual emotions within Ekman’s Emotion Hexagon
Test. Again, regressions demonstrated that personality can be
used to predict an individual’s performance on one of the six
basic emotions within Ekman’s Emotion Hexagon Test. This
emotion was disgust, R²=.498, F(5,54)=3.566, p=.007.The only
significant predictor of disgust was neuroticism, t(58)= 3.444,
p=.001, with a positive Beta value of .459.
Discussion
The aim of this investigation was to see if personality
could predict the scores on the two facial emotion recognition
tasks. Because there had been previous issues regarding the
different emotion recognition methods, a new method
(Laura’s Emotion Recognition Task) was created and compared
an older method of Ekman’s Emotion Hexagon Test.
Findings from this study demonstrated that both Laura’s
Emotion Hexagon Task and Ekman’s Emotion Hexagon Test
were appropriate measures of facial emotion recognition.
Results showed that the personality traits of agreeableness
and extraversion were predictors of facial emotion recognition
concerning Laura’s Emotion Hexagon Task. The analysis
relating Ekman’s Emotion Hexagon Test had no significant
predictions. The result of a further analysis indicated that
openness did have an effect on an individual’s facial emotion
recognition only when participants had completed Laura’s
Emotion Recognition Task. The participant who has low
openness also had lower task scores in this case.
The current researchers made the prediction that
openness could predict how well an individual would score on
both Laura’s Emotion Hexagon Task and Ekman’s Emotion
Hexagon Test. This prediction had no support. Researchers
also predicted that the other four personality variables
(extraversion, agreeableness, conscientiousness and
neuroticism) would not have any significant effect on an
individual’s facial emotion recognition score on both tasks.
This again had no support from the results. Finally, it was
predicted that there would be some form of correlation
between Laura’s Emotion Hexagon Task and Ekman’s Emotion
Hexagon Test. This prediction was supported.
The findings from this investigation did not show full
support for previous research [21] [18].
Terracciano, Merritt, Zonderman and Evans [21] suggested
that facial emotion recognition was predicted by a person’s
personality trait of openness, and slightly by the trait of
conscientiousness. They found no significant results with
agreeableness, extraversion or neuroticism. This study, on the
other hand, found that openness, agreeableness and
extraversion is to predict an individual’s facial emotion
recognition when participants completed Laura’s Emotion
Hexagon Task.
Mill, Allik, Realo and Valk [18] found that both openness
and conscientiousness correlated with their facial emotion
recognition task. This is slightly different to the current study
who found correlations with agreeableness, extraversion and
a significant effect of openness when being concerned with
participants’ facial emotion recognition on Laura’s Emotion
Hexagon Task. There are many reasons why the results of
both studies could have differed.
One possible difference could have been because there
were different methods used in each study. Terracciano,
Merritt, Zonderman and Evans [21] had used the Perception
of Affect Task (PAT) which has a different structure to Ekman’s
Emotion Hexagon Test. Participants are not asked to look at
morphed images in the PAT so therefore the methods used by
Terracciano could be seen as an easier test to take.
Mill, Allik, Realo and Valk [18] had also used a different
method when assessing facial emotion recognition. In this
case, the Japanese and Caucasian Facial Expressions of
Emotion (JACFEE) was used. Despite using images taken from
‘Picutres of Facial Affect’ (Ekman and Friesen), not all the
photographs were placed in the same order and the
photographs were not morphed giving the test a slightly
lower level of difficulty. Mill, Allik, Realo and Valk [18] showed
participants only 32 images with the use of the Japanese and
Caucasian Facial Expressions of Emotion. This could have been
one of the main reasons for the difference in results compared
to the current research. Both Laura’s Emotion Hexagon Task
and Ekman’s Emotion Hexagon Test gave participants the
opportunity to look at 150 images of emotion expressions in
the human face. Suggestions could propose that the larger
amount of photographs allowed for participants to gain more
experience in identifying the emotions in the images. This in
turn would mean that participants’ scores could be higher
when recognising the six basic emotions.
One other possible reason for the difference in results
could have been due to the samples in each study. The current
study used 60 Northumbria University students whereas
Terracciano et al. [21] use 152 participants (African American
and Hispanic). There are many differences with these samples,
which could mean that one sample is better than the other is
Madridge Journal of Neuroscience
44
Madridge J Neuro Sci. Volume 1 • Issue 1 • 1000107
at recognising emotions in the human face. Factors include
things such as culture, education and socioeconomic status.
Because there has been little research done in this area, these
reasons can only be a possibility (and can only be suggested)
because they do not have empirical support. We cannot be
sure whether individuals from different cultures would
perform the same on Laura’s Emotion Hexagon Task, Ekman’s
Emotion Hexagon Test, The Japanese and Caucasian Facial
Expressions of Emotion, and the Perception of Affect Task.
Cross-cultural investigations could be an option to help
discuss this issue.
Another possible reason (relating to the participants)
could have been the differences for exposure individuals had
previously had to each emotion recognition task. The current
study used Psychology students who will have more than
likely covered the topic of emotion recognition somewhere
within the Psychology Degree, including the tasks used. Both
Terracciano et al. [21] and Mill et al. [18] did not use students
so therefore their exposure to both emotion recognition tasks
could be more limited. Any differences in results could have
been down to the fact of more exposure and more practice
from the participants of the current study. In order to control
this, researchers could have noted down whether participants
had encountered Ekman’s Emotion Hexagon Test prior to the
testing phase.
When analysing results, a positive correlation between
both Laura’s Emotion Hexagon Task and Ekman’s Emotion
Hexagon Test was presented. This gives support to suggest
that Laura’s Emotion Hexagon Task is a very good measure of
facial emotion recognition. If participants did well on Laura’s
Emotion Hexagon Task then they would also do well on
Ekman’s Emotion Hexagon Test.
From looking at the statistical averages calculated from
the data, results demonstrated that participants had
performed better on Ekman’s Emotion Hexagon Test
compared to Laura’s Emotion Hexagon Task. There are again
many possible reasons for this difference.
The first reason can relate to the reliability and validity of
each facial emotion recognition test. Ekman’s Emotion
Hexagon Test has been shown to have very good split-half
reliability (FEEST Manual) whereas Laura’s Emotion Hexagon
Task does has not have any previous research conducted
using the task itself. In order to compare the two facial
emotion recognition tasks further, investigations about the
reliability of Laura’s Emotion Hexagon Task needs to be
considered.
Calder et al. [4] had given support for the validity of
Ekman’s Emotion Hexagon Test by using it to look at emotion
recognition within the adult life span. Because the test was
used successfully on a number of occasions, this gave
researchers a good basis to say that Ekman’s Emotion Hexagon
Test was a valid method of assessing the recognition of
emotion in the human face.
Because the current study is the first piece of research to
use Laura’s Emotion Hexagon Task, researchers cannot
conclude that this method is valid. Even though Laura’s
Emotion Hexagon Task was created in the same way as
Ekman’s Emotion Hexagon Test was created (using the same
number of photographs with similar ratios of emotions), it
cannot be seen as valid until further research is conducted
using the test in different situations.
The final reason for the difference in results could simply
be down to the involvement of order effects. All participants
completed Laura’s Emotion Hexagon Task before they
completed Ekman’s Emotion Hexagon Test. Results could
possibly have been different if participants completed the
tasks in a different order, therefore reducing the involvement
of order effects.
One of the main aims of this investigation was to look at
the argument surrounding the photographs in Ekman’s
Emotion Hexagon Test (and the earlier methods). The
argument put forward by Suzuki et al. [20] suggested that
participants were finding it very easy to recognise the
emotions within the photographs. This was simply because
the photographs are becoming universal and are being
displayed on many occasions. Suzuki et al. [20] suggested that
recent methods are still using the photographs taken from
‘Pictures of Facial Affect’ (Ekman and Friesen) so therefore
anyone with background knowledge of these images would
find the task very easy.
Laura’s Emotion Hexagon Task helps to provide evidence
for this argument. Because participants performed better in
Ekman’s Emotion Hexagon Test (with a higher average than in
Laura’s Emotion Hexagon Task), it can be suggested that
Suzuki et al. [20] were right in some ways.
The current study also looked to see if any individual basic
emotion scores (happiness, sadness, fear, disgust, surprise and
anger) could be predicted by an individual’s personality. This
was not one of the main aims but results gave some good
points to discuss. Results (from the current study) demonstrated
that in Laura’s Emotion Hexagon Task, only the detection of
the sadness emotion could be predicted by the five personality
traits. In the case of Ekman’s Emotion Hexagon Test, the only
emotion that could be predicted by the five personality traits
was the emotion of disgust.
Very little research uses the ‘normal population’ so
therefore results will be discussed in relating to individuals
with personality disorders.
Domes, Czieschnek, Weidler, Berger, Fast and Herpertz [8]
looked at individuals who have a diagnosis of Borderline
Personality Disorder to see if there were any links between the
disorder and facial emotion recognition. Researchers asked 25
females to look at a set of stimuli that involved pictures created
by Ekman and Friesen. It was demonstrated that the individuals
who had Borderline Personality Disorder had no deficits in
recognising emotions, but were able to identify the emotions of
anger better than the other five basic emotions. Participants with
Borderline Personality Disorder found it difficult to identify the
emotion of surprise and fear amongst the pictures. Researchers
suggested that these basic emotions could have had some sort
of sensitivity to the Borderline Personality Disorder, therefore
creating a difficulty in recognising the emotions.
Madridge Journal of Neuroscience
45
Madridge J Neuro Sci. Volume 1 • Issue 1 • 1000107
Results from Domes, Czieschnek, Weidler, Berger, Fast
and Herpertz [8] differ to the results demonstrated by the
present study. From the present study, researchers had
indicated a sensitivity towards to emotion of sadness (relating
to Laura’s Emotion Hexagon Task) and disgust (in Ekman’s
Emotion Hexagon Test).
Differences may have occurred because of the ‘normal
population’ used in the current study. Because there were no
specific personality disorders present in the current sample of
participants, it could mean that the sensitivity to different
emotions is also different. This study therefore gives an insight
(and a direction) into the possibility of conducting further
research using the normal population instead of conducting
research-using individuals with specific disorders.
There are some areas of improvement if this investigation
was repeated. Firstly, using live models to assess an individual’s
facial emotion recognition. Instead of having still images,
participants could be asked to look at a model who is
displaying one of the six basic emotions of anger, sadness,
happiness, fear, disgust or surprise. This would increase the
ecological validity of the investigation as you could put each
participant in a real-life scenario where human interactions
can take place.
One other improvement could be with the design of the
experiment itself. The current study used a repeated measures
design but there are also other options for the choice of
design. To eliminate any order effects, a matched pairs design
could have been used. Participants would have to match
participants based on their personality types and other
demographic factors (like age). This would mean that
participants would not have to do both Ekman’s Emotion
Hexagon Test and Laura’s Emotion Hexagon Task so it could
be more difficult to make a comparison. One notable criticism
of this study is that participants seemed to become very tired
after completing the first of the two emotion recognition
tasks. Eliminating fatigue effects would be done by
counterbalancing the areas of the investigation.
With the current study in mind, there could be several
areas for future research. The main area of investigation is the
reliability and validity of Laura’s Emotion Hexagon Task. Just
because the task has worked once does not mean that the
same results will occur again so the only way to look at this is
to put the test in different situations.
By using the previous research findings related age [4]
and emotional intelligence (Austin) [2], researchers could
discover whether or not Laura’s Emotion Hexagon Task could
be used in these two situations just like Ekman’s Emotion
Hexagon was.
One other possible area for future research could be to
look at individuals diagnosed with specific disorders.
Sprengelmeyer et al. [19] looked at instances when individuals
can lose the ability to recognise emotions correctly. Research
was conducted using individuals who had been diagnosed
with Huntington’s disease - a disorder that can affect muscle
control and can cause the de-generation of the brain.
Participants recalled different facial photographs. These
photographs were of both familiar faces and unfamiliar faces,
all displaying different emotions. Results showed that
participants with Huntington’s disease were impaired in
discriminating the emotion of anger and fear. Participants,
however, showed the ability to discriminate between
photographs displaying the happiness and sadness emotions,
suggesting that there could be some problems with only
negative emotion recognition. If the task has good reliability
and validity then it could be used in situations with the non-
normal population.
Conclusion
In conclusion, the current study has demonstrated that
Laura’s Emotion Hexagon Task is a more up-dated facial
emotion recognition task by initially presenting good
reliability for this test. The task was used to show that
personality does have a relation to facial emotion recognition
in humans. Despite this finding, the validity of the new
measure was not tested; therefore further investigations are
needed to look specifically at the validity of this task. This,
however, does not stop the opportunity for future research
into the creation of new methodologies for assessing facial
emotion recognition alongside the different variables that can
affect their result.
Conflict of interest
The authors confirm that there is no conflict of interest
regarding this manuscript.
References
1. Allik J, Laidra K, Realo A, Pullmann H. Personality development from 12 to
18 years of age: Changes in mean levels and structure of traits. European
Journal of Personality. 2004; 18: 445-462. doi: 10.1002/per.524
2. Austin EJ. An investigation of the relationship between trait emotional
intelligence and emotional task performance. Personality and Individual
Differences. 2004; 36: 1855-1864. doi: 10.1016/j.paid.2003.07.006
3. Cacioppo JT, Gardner WL. Emotion. Annual Review of Psychology. 1999;
50: 191-214. doi: 10.1146/annurev.psych.50.1.191
4. Calder AJ, Keane J, Manly T, Sprengelmeyer R, Scott S, et al. Facial
Expression Recognition Across The Adult Life Span. Neuropsychologia.
2003; 41: 195-202. doi: 10.1016/S0028-3932(02)00149-5
5. Calder AJ, Young AW, Rowland D, Perrett DI, Hodges JR, et al. Facial
Emotion Recognition after Bilateral Amygdala Damage: Differentially
Severe Impairment of Fear. Cognitive Neuropsychology. 1996; 13: 699-745.
doi: 10.1080/026432996381890
6. Darwin C. The Expression of the Emotions in Man and Animals. London:
Penguin Books Ltd; 1872-2009.
7. Diehl-Schmid J, Pohl C, Ruprecht C, Wagenpfeil S, Foerstl H, et al. The
Ekman 60 Faces Test As A Diagnostic Instrument In Frontotemporal
Dementia. Archives of Clinical Neuropsychology. 2007; 22: 459-464. doi:
10.1016/j.acn.2007.01.024
8. Domes G, Czieschnek D, Weidler F, Berger C, Fast K, et al. Recognition of
Facial Affect in Borderline Personality Disorder. Journal of Personality
Disorders. 2008; 22(2): 135-147. doi: 10.1521/pedi.2008.22.2.135
9. Ekman P. Universal Facial Expressions of Emotion. California Mental Health
Readers Digest. 1970; 8(4): 151-158.
Madridge Journal of Neuroscience
46
Madridge J Neuro Sci. Volume 1 • Issue 1 • 1000107
10. Ekman P. Darwin, Deception, and Facial Expression. Annual New York
Academy of Sciences. 2003; 1000: 205-221. doi: 10.1196/annals.1280.010
11. Ekman P, Friesen WV, Tomkins SS. Facial Affect Scoring Technique: A First
Validity Study. 1971.
12. George MS, Ketter TA, Parekh PL, Horwitz B, Herscovitch P, et al. Brain
activity during transient sadness and happiness in healthy women. Annual
Journal of Psychiatry. 1995; 152: 341-351. doi: 10.1176/ajp.152.3.341
13. Goldberg LR. The IPIP-NEO: International Personality Item Pool
Representation of the NEO PI-R. 1999.
14. Goldberg LR, Johnson JA, Eber HW, Hogan R, Ashton MC, et al. The
International Personality Item Pool and the Future of Public-Domain
Personality Measures. Journal of Research in Personality. 2006; 40: 84-96.
doi: 10.1016/j.jrp.2005.08.007
15. Goleman D. Emotional Intelligence. New York: Bantam Publishing
Company. 1995.
16. Matsumoto D, Ekman P. The relationship among expressions, labels, and
descriptions of contempt. Journal of Personality and Social Psychology.
2004; 87: 529-540.
17. Matsumoto D, Le Roux J, Wilson-Cohn C, Raroque J, Kooken K, et al. A
New Test To Measure Emotion Recognition Ability: Matsumoto and
Ekman’s Japanese and Caucasian Brief Affect Recognition Test (JACBART).
Journal of Nonverbal Behaviour. 2000; 24(3): 179-209.
18. Mill A, Allik J, Realo A, Valk R. Age-Related Differences In Emotion
Recognition Ability: A Cross-Sectional Study. Emotion. 2009; 5: 619-630.
doi: 10.1037/a0016562
19. Sprengelmeyer R, Young AW, Calder AJ, Karnat A, Lange H, et al. Loss of
Disgust Perception of Faces and Emotions in Huntington’s Disease. Brain.
1996; 119: 1647-1665.
20. Suzuki A, Hoshino T, Shigemasu K. Measuing Individual Differences in
Sensitivities to Basic Emotions in Faces. Cognition. 2006; 99: 327-353. doi:
10.1016/j.cognition.2005.04.003
21. Terracciano A, Merritt M, Zoderman AB, Evans MK. Personality Traits and
Sex Differences in Emotions Recognitions among African Americans and
Caucasians. Ann N Y Acad Sci. 2003; 1000: 309-312.
22. Young A, Perrett D, Calder A, Sprengelmeyer R, Ekman P. Facial Expressions
of Emotion – Stimuli and Tests (FEEST). London: Thames Valley Test; 2002.
