Content uploaded by Giovanni Vecchiato
Author content
All content in this area was uploaded by Giovanni Vecchiato on Nov 24, 2014
Content may be subject to copyright.
1 23
Cognitive Computation
ISSN 1866-9956
Cogn Comput
DOI 10.1007/s12559-014-9304-x
How to Measure Cerebral Correlates of
Emotions in Marketing Relevant Tasks
Giovanni Vecchiato, Patrizia Cherubino,
Anton Giulio Maglione, Maria Trinidad
Herrera Ezquierro, Franco Marinozzi, et
al.
1 23
Your article is protected by copyright and all
rights are held exclusively by Springer Science
+Business Media New York. This e-offprint is
for personal use only and shall not be self-
archived in electronic repositories. If you wish
to self-archive your article, please use the
accepted manuscript version for posting on
your own website. You may further deposit
the accepted manuscript version in any
repository, provided it is only made publicly
available 12 months after official publication
or later and provided acknowledgement is
given to the original source of publication
and a link is inserted to the published article
on Springer's website. The link must be
accompanied by the following text: "The final
publication is available at link.springer.com”.
How to Measure Cerebral Correlates of Emotions in Marketing
Relevant Tasks
Giovanni Vecchiato •Patrizia Cherubino •Anton Giulio Maglione •
Maria Trinidad Herrera Ezquierro •Franco Marinozzi •
Fabiano Bini •Arianna Trettel •Fabio Babiloni
Received: 15 December 2013 / Accepted: 18 August 2014
Springer Science+Business Media New York 2014
Abstract Nowadays, there is a growing interest in mea-
suring emotions through the estimation of cerebral vari-
ables. Several techniques and methods are used and
debated in neuroscience. In such a context, the present
paper provides examples of time-varying variables related
to the estimation of emotional valence, arousal and
Approach-Withdrawal behavior in marketing relevant con-
texts. In particular, we recorded electroencephalographic
(EEG), galvanic skin response (GSR) and heart rate (HR)
in a group of healthy subjects while they are watching
different TV commercials. Specifically, results obtained in
the Experiment 1 shows a significant increase of cortical
power spectral density across left frontal areas in the alpha
band and an enhance of cardiac activity during the obser-
vation of TV commercials that have been judged pleasant.
In the Experiment 2, frontal EEG asymmetry, GSR and HR
measurements are used to draw cognitive and emotional
indices in order to track the subject’s internal state frame
by frame of the commercial. A specific case study shows
how the variations of the defined Approach-Withdrawal
and emotional indices can distinguish the reactions of
younger adults from the older ones during the observation
of a funny spot. This technology could be of help for
marketers to overcome some of the drawbacks of the
standard marketing tools (e.g., interviews, focus groups)
usually adopted during the analysis of the emotional per-
ception of advertisements.
Keywords EEG Heart rate Emotions TV
commercials Neuromarketing
G. Vecchiato (&)F. Babiloni
Department Physiology and Pharmacology, ‘‘Sapienza’’
University, Piazzale Aldo Moro 5, 00185 Rome, Italy
e-mail: giovanni.vecchiato@uniroma1.it
G. Vecchiato P. Cherubino A. G. Maglione A. Trettel
F. Babiloni
BrainSigns s.r.l., Via Sesto Celere 7c, 00152 Rome, Italy
P. Cherubino
Department Economics and Marketing, ‘‘IULM’’ University,
Via Carlo Bo 1, 20143 Milan, Italy
A. G. Maglione
Department of Anatomy, Histology, Forensic Medicine
and Orthopedics, ‘‘Sapienza’’ University, Via Borelli 50,
00185 Rome, Italy
M. T. H. Ezquierro
Clinical and Experimental Neuroscience (NiCE-CIBERNED),
Department of Human Anatomy and Psychobiology, School
of Medicine, University of Murcia, Campus Universitario de
Espinardo, 30100 Murcia, Spain
F. Marinozzi F. Bini
Department of Mechanical and Aerospace Engineering,
‘‘Sapienza’’ University, Via Eudossiana 18, 00184 Rome, Italy
F. Babiloni
IRCCS Fondazione Santa Lucia, via Ardeatina 306,
00179 Rome, Italy
123
Cogn Comput
DOI 10.1007/s12559-014-9304-x
Author's personal copy
Introduction
Importance of Emotions in Decision Making
The study of emotions is achieving a growing importance
in the neuroscience field. For decades, researchers have
been approaching decision-making tasks by assuming that
humans use only rational aspects of a situation. The con-
sensus at that time was that decision making followed three
distinct steps: first asking what the best choice is, second
doing it and finally experiencing the feeling that is gener-
ated by the outcome. In the last 10 years, this perspective
has been challenged, mainly by the results obtained by
Damasio [1,2]. According to his theory, ‘‘emotions’’ have
a main role in the whole decision-making process; they are
the humans’ guide for choices, finally executed. Hence, it
becomes clear that the study of emotions, associated to the
decision making, is fundamental for a deep understanding
of the human behavior. Supporting Damasio’s studies,
Bargh and Chartrand [3] affirm that, although humans are
definitely capable of conscious deliberation, many eco-
nomically relevant decisions rely on automatic, fast and
effective processes, which are not under the direct voli-
tional control. Moreover, the human beings are influenced
by unrecognized and finely tuned affective mechanisms,
which often play a decisive role in decision making and
action [4,5]. Many of these processes have been shaped by
evolution in order to serve social purposes [6–8] while
decision making and evaluation in economic contexts are
influenced by mechanisms dedicated to social interaction.
For instance, in man, the observation of different cultural
objects associated with wealth and social dominance (e.g.,
sports cars) elicits activation in reward-related brain areas
[9,10].
Emergence of Neuromarketing
In these last years, it observed an increasing interest in the
use of brain imaging techniques for the analysis of brain
responses to commercial stimuli as well as for the inves-
tigation of purchasing attitudes, using hemodynamic [11,
12] and neuroelectromagnetic measurements [13–16] (see
[17,18] for a review). Such interest is also justified by the
possibility to correlate the observed brain activations with
the proposed commercial stimuli, in order to derive con-
clusions about the adequacy of such stimuli in terms of
interest or emotional engage for the investigated sample.
So far adopted standard marketing techniques often involve
the use of in depth interviews or focus groups during which
customers are exposed to the product before its massive
launch (ad pretest) or afterward (ad posttest). However, it is
now recognized that the verbal advertising pretest is biased
by the respondents’ cognitive processes activating during
the interview [19]. In addition, it was also suggested that
the interviewer may have a great influence on the respon-
dent recalls as well [20,21]. Taking all these consider-
ations in mind, neuroscientists have started to investigate
the brain activity gathered during the watching of TV
commercials by measuring variables linked to the cognitive
and emotional engagement. In fact, there are high hopes
that neuroimaging technology could solve some of the
problems that marketers face, such as streamline marketing
processes and save money by providing more efficient
trade-off between costs and benefits. In fact, neuroimaging
is thought to reveal information about consumer prefer-
ences that are unobtainable through the use of conventional
interviews [22].
Approaching Emotions by Dimensions and Categories
Many scientists have focused their research in order to
achieve a gold standard in the measure of emotion, which
could be approached from both dimensional and discrete
perspectives (see [23,24] for reviews). Scientific evidence
suggests that measuring people’s emotional state is one of
the most vexing problems in affective science. There are a
few fundamental dimensions that organize emotional
response. The most commonly assumed dimensions are
three: valence, arousal and Approach-Withdrawal [25–29].
The valence dimension contrasts states of pleasure (e.g.,
happy) with states of displeasure (e.g., sad), and the arousal
dimension contrasts states of low arousal (e.g., quiet) with
states of high arousal (e.g., surprised). Approach motiva-
tion is characterized by tendencies to approach stimuli
(e.g., as would likely be facilitated by excitement), whereas
avoidance motivation is characterized by tendencies of
avoidance (e.g., as would likely be facilitated by anxiety).
In contrast, the discrete perspective shows that each emo-
tion corresponds to a unique profile in experience, physi-
ology and behavior [30]. As far as the behavioral response
is concerned, early and recent evidence highlights that
facial expression signaling supports the discrimination of
basic categories [31,32]. Although dimensional and dis-
crete perspectives differ in how they conceptualize and
describe emotional states [33], it seems possible to recon-
cile the two theoretical frameworks by proposing that each
discrete emotion represents a combination of several
dimensions [34,35].
Cogn Comput
123
Author's personal copy
Autonomic Nervous System Measurements
for Emotions
As signs of the autonomic nervous system (ANS), the
activity of the sweat glands on the hands and the heart rate
(HR) variation are the neurophysiological variables often
used to describe variations of emotional states according to
the couple of orthogonal axes of valence and arousal [27,
36,37]. The idea that discrete emotions have distinct
autonomic signatures is not fully supported by experi-
mental findings [38]. Instead, relevant studies often point to
relationships among dimensions, particularly those of
valence and arousal, and ANS responses [39,40]. In par-
ticular, the galvanic skin response (GSR) is typically
quantified in terms of skin conductance level (SCL) or
short-duration skin conductance responses (SCRs), while
the most commonly used cardiovascular measure is the
HR. By monitoring autonomic activity using devices able
to record the variation of the skin conductivity and the HR,
it is possible to assess the internal emotional state of the
subject. In fact, the GSR is actually viewed as a sensitive
and convenient measure for indexing changes in sympa-
thetic arousal associated with emotion, cognition and
attention [41]. Studies using functional imaging techniques
have related the generation and the maintenance of the
electrodermal activity level to different specific brain areas.
These specific regions are the ventromedial prefrontal
cortex, the orbitofrontal cortex, the left primary motor
cortex and the anterior and posterior cingulate, which have
been shown to be associated with emotional and motiva-
tional behaviors [41,42]. Such findings also indicate that
the association of peripheral and central measures of
arousal re-emphasize the connections among electrodermal
activity, arousal, attention, cognition and emotion. Also,
several papers reported as the HR correlates with the
emotional valence of a stimulus, e.g., the positive or neg-
ative component of the emotion [43–46]. Moreover, in
experimental psychology, it has been proposed and used
the affect circumplex, in which emotions are mapped in a
two-dimensional space in which horizontal and vertical
axes are related to valence and arousal, respectively [27,
47]. Thus, the joint measurement of HR and GSR and their
positioning on the affect circumplex returns the emotion
perceived by the subject during a specific experimental
task.
Approach-Withdrawal Neural Origins: The Role
of PFC in Emotions
As to Approach-Withdrawal, such parameter can be
obtained by measuring variations of the pre- and frontal
cortex (PFC and FC, respectively; [4]). Despite this is a
region structurally and functionally heterogeneous, its role
in emotion is well recognized [48]. The role of the PFC in
emotion has also addressed by recent fMRI studies, which
highlight the reciprocal relation between amygdala and
ventromedial PFC (vmPFC) in the encoding of emotional
valence [49]. In particular, the connectivity patterns
between amygdala and vmPFC could vary with the role
played by emotional task, being the vmPFC preferentially
engaged to utilitarian and emotional assessments during
moral judgments [50]. In addition, studies with patients
showed minor recruitment of the dlPFC involved in cog-
nitive reappraisal, suggesting focal and aberrant neural
activation during volitional, self-regulation of negative
affective states [51], whereas vmPFC lesions also exhibited
potentiated amygdala responses to aversive images and
elevated resting-state amygdala functional connectivity
[52]. Finally, ventrolateral PFC (vlPFC) deficits in positive
emotions are correlated with social anhedonia and schizo-
phrenia [53]. The relationship between PFC and hedonia is
also discussed in studies performing emotional decision-
making tasks. Specifically, Lin and colleagues [54] suggest
a late participation of the vmPFC in preference decision
making, whereas Li and colleagues [55] showed that the
vmPFC is part of a network coupling both memory and
emotional processes. Finally, Santos and colleagues [56]
affirm that activation of the vmPFC was found when
comparing positive with indifferent or fictitious brands.
Such activation is stronger after the choice than during the
decision process itself. All these results provide evidence
for the critical role of the PFC in regulating emotions in
humans supporting also the notion that vmPFC may be
unimportant in the decision stage concerning brand
preference.
Relationship Between Approach-Withdrawal
and EEG-Based Lateralized Measurements
Electroencephalographic (EEG) spectral power analyses
indicate that the anterior cerebral hemispheres are differ-
entially lateralized for approach and withdrawal motiva-
tional tendencies and emotions. Specifically, findings
suggest that the left PFC is an important brain area in a
widespread circuit that mediates appetitive approach, while
the right PFC appears to form a major component of a
neural circuit that instantiates defensive withdrawal [57,
58]. Sutton and Davidson [59] found that greater left-sided
activation predicted dispositional tendencies toward
approach, whereas greater right-sided asymmetry predicted
dispositional tendencies toward avoidance. In contrast, the
frontal asymmetry measure did not predict dispositional
tendencies toward positive or negative emotions, suggest-
ing an association of frontal asymmetry with Approach-
Withdrawal rather than with valence. Other sources con-
verge on a similar model of frontal asymmetry. Of
Cogn Comput
123
Author's personal copy
particular importance are studies that link anger, an
unpleasant but approach-related emotion, to greater left-
hemispheric activation [60,61]. Also, tendencies toward
worry, thought to be approach-motivated in the sense of
being linked to problem solving, have been linked to rel-
atively greater left frontal EEG activity [62]. Thus, the
emerging consensus appears to be that frontal EEG
asymmetry primarily reflects levels of approach motivation
(left hemisphere) versus avoidance motivation (right
hemisphere). More recently, resting EEG, self-report
measures of Behavioral Activation and Inhibition System
(BAS and BIS) strength, dispositional optimism and a
measure of hedonic tone, were collected and correlated
with alpha asymmetry measures, which yielded significant
frontal and parietal asymmetry correlation patterns [63].
The BAS is conceptualized as a motivational system that is
sensitive to signals of reward, nonpunishment, and that is
important for engaging behavior toward a reward. The BIS,
conversely, inhibits behavior in response to stimuli that are
novel, innately feared, and conditioned to be aversive. In
particular, higher BAS was uniquely related to greater left-
sided activation in the middle frontal gyrus. Optimism was
associated with higher activations in the left-superior
frontal gyrus (BA10) and in the right-posterior cingulate
cortex (BA31). Moreover, Santesso and colleagues [64]
examined the relationship between measures of sensation
seeking and the pattern of resting frontal EEG asymmetry,
thought to reflect a biological predisposition to approach
new experiences. Findings showed that high sensation
seeking was related to a greater relative left frontal activity
at rest. Urry and colleagues [65] also reported that greater
left than right frontal EEG activity was associated with
higher hedonic well-being. Pizzagalli et al. [66] reported
that current density in the left (but not right) prefrontal
cortex was related to reward bias suggesting that these
regions may underlie individual differences in approach-
related behavior.
Emotions Miss Specific Spatial Location
Hence, EEG seems to converge into the dimensional per-
spective of emotions by concluding that relative left-
hemisphere activation is reflective of approach-related
states, whereas relative right-hemisphere activation is
reflective of avoidance-related states, although such acti-
vations are not intended to discriminate emotions com-
pletely. On the other side, other neuroimaging techniques,
such as fMRI, may be better suited than EEG to reveal
emotion specificity in the brain (see [24] for a review). The
debate on this topic (dimensional vs discrete perspective) is
far from the conclusion, and it has to be further investi-
gated since different brain regions can participate in mul-
tiple emotions.
Emotional Signatures in Response to TV
Advertisements
Ioannides and colleagues [14] have employed magnetoen-
cephalography (MEG) to study the neuronal responses in
subjects viewing TV advertisements. Those MEG data
suggest that cognitive advertisements activate predomi-
nately posterior parietal and superior prefrontal cortices,
whereas effective material modulates activity in orbito-
frontal cortices, the amygdala and the brainstem. The
results seem to imply that cognitive rather than affective
advertisements activate cortical centers associated with the
executive control of working memory and maintenance of
higher-order representations of complex visual material.
Other researchers also focus on the EEG frontal imbalance.
In fact, Polish researchers found out that the observation of
two versions of the same TV commercial generated sig-
nificantly different emotional impact in terms of EEG
frontal asymmetry [67]. However, difference in cerebral
activity coding the pleasantness can be also observed by
event-related potentials (ERPs) analysis. In particular,
Handy and colleagues [68] proved that visuocortical pro-
cessing shows an increase of the early positive component
(named P1), at central and parietal sites, along with an
increase of the later negative component (named N2), at
parietal and occipital sites, related to the observation of
disliked logos. Taken together, these mentioned examples
bring evidences that it is possible to link some properties of
the collected EEG rhythms during the watching of some
TV advertisements with the overt preferences of the
observers in terms of emotions. This link could be used to
generate a metric that automatically points to parts of the
advertisings examined that are emotionally OK and part
that are not. This information could be used ‘‘a posteriori’’
to redraw partially the advertising in order to increase the
appearance of the ‘‘like’’ parts while depressing the ‘‘dis-
like’’ parts. Such reduction or modification of the broad-
casted TV advertisings could be then performed using EEG
techniques thanks the high time resolution provided by this
methodology.
Theses and Outline of the Work
In this scenario, the aim of the present paper is to show
how the variation of the EEG frontal cortical activity, GSR
and HR, is differently related to the general appreciation
and the emotional feeling perceived during the observation
of commercial TV advertisements. According to the
aforementioned theoretical framework, we expect that:
1. the EEG frontal imbalance describes appreciation of
TV commercials, showing major (minor) activation on
Cogn Comput
123
Author's personal copy
the left frontal lobe during the observation of pleasant
(unpleasant) scenes…
2. increase (decrease) of GSR and HR variables distin-
guish high (low) arousing and positive (negative)
valence emotions elicited during the observation most
(less) liking advertisements.
Although there could exist a degree of correlation
between GSR (arousal) and HR (valence), we do not
address this issue in the present work but do exploit the
illustrated theory. In addition, we provide an example of
how it is possible to apply neuroscientific tools to real
marketing contexts. For this purpose, we are interested to
analyze the brain activity occurring during the ecologic
observation of commercial advertisings (ads) intermingled
in a random order within a documentary. To measure both
brain activity and autonomic parameters of the investigated
subjects, we used simultaneous EEG, GSR and HR mea-
surements during the whole experiments. We link signifi-
cant variation of EEG and autonomic measurements with
the perceived pleasantness, as declared in the following
subject’s verbal interview. In order to do that, different
indices are used to summarize the cerebral and autonomic
measurements performed, later used in the statistical ana-
lysis. The goal was to recreate, as much as possible, an
ecologic approach to the task, in which the observer is
viewing the TV screen without particular requirements to
accomplish. In fact, subjects were not instructed at all on
the aim of the experiment, and they were not aware that an
interview about the observed TV commercials would be
generated at the end of the video clip.
Introducing Experiment 1 and Experiment 2
The next sections of the paper are articulated in order to
present and comment separately the results of two different
neuromarketing experiments. In particular, Experiment 1
aims to explain the methods and results and discusses the
theoretical basis as well as some practical aspects of the
research; Experiment 2 shows methods and results of a
concrete application, followed by a general discussion.
Specifically, Experiment 2 introduces two neurophysio-
logical indices to describe emotions during the perception
of TV commercials. These indices are derived by a com-
bination of GSR and HR, whereas the second measures the
unbalance of the EEG frontal power exploiting the afore-
mentioned neurophysiological background. The aim of this
section is to provide the reader with an example of the
capabilities of the neuroelectrical measurements in a real
marketing context. The provided results represent a case
study, which could be hopefully of help to describe the
time sequence of a commercial advertisement. For this
reason, we use a representative video clip, which we
believe to highlight clearly the potentiality of the defined
neurophysiological indices to measure emotions in this
kind of experimental tasks.
Experiment 1
Methods
Subjects
Fifteen healthy volunteers (mean age 27.5 ±7.5 years; 7
women) have been recruited for this study. Informed con-
sent was obtained from each subject after explanation of
the study, which was approved by the local institutional
ethics committee. Subjects had no personal history of
neurological or psychiatric disorder. They were free from
medications, alcohol and drugs abuse.
Experimental Paradigm
The procedure of the experimental task consisted in
observing a 30-min-long documentary in which we inserted
three advertising breaks, each composed by two TV com-
mercials, according to the schema illustrated in Fig. 1.
Each interruption was formed by the same number of
commercial video clips, each lasting about thirty seconds.
During the whole documentary, a total of six TV com-
mercials were presented. The clips were related to standard
international brands of commercial products, like cars and
food, and public service announcements (PSA) such as
campaigns against violence. Randomization of the occur-
rence of the commercial videos within the documentary
was made to remove the factor ‘‘sequence’’ as possible
confounding effect.
The experimental subjects enrolled in this experiment
are asked to comfortably seat in front of a computer screen
by means of which we present a documentary intermingled
with several commercial breaks, the target stimuli of the
experiment. During the whole video, the cerebral (EEG)
and autonomic signals (HR, GSR) were collected from the
subjects. The signals gathered during the observation of the
documentary will be used to estimate the personal neuro-
physiological baseline activity. During the experiment,
subjects are not aware that an interview would be held
within a couple of hours from the end of the data recording.
They are simply told to pay attention to what they will
watch. The video consisted in a documentary related to
geography, with the aim to elicit no particular emotional
engage. After data recording, an interview is performed. At
this stage, the experimenter asks the subjects to recall
spontaneously the commercial video clips they memorized.
Then, the experimenter verbally listed the sequence of
Cogn Comput
123
Author's personal copy
advertisements presented within the documentary asking
the subjects to tell which advertisement they remember.
Successively, the experimenter showed on a paper several
frame sequences of each advertisement inserted in the
movie. Analogously, the experimenter also showed several
pictures related to an equal number of advertisements that
were not inserted in the commercial break (distractors).
This was done to provide the subjects the same number of
distractors when compared to the target pictures. Finally,
the experimenter asked subjects to give a score ranging
between 1 and 10 according to the level of pleasantness
they perceived during the observation of each remembered
ad (1, lowly pleasant; 5, indifferent; 10, highly pleasant).
This was performed to divide the gathered neurometric
data in two samples: They comprise the data related to the
clips rated from 7 to 10 (LIKE dataset) and those regarding
the clips rated from 1 to 4 (DISLIKE dataset). In this
report, we do not analyze the data related to a medium
pleasantness score (e.g., from 5 to 6). In such a way, we
divide the EEG, HR and GSR signals of the population
recorded into two different datasets that will be compared
in the following of this study by discarding the data related
to the middle of the pleasantness scale. For each subject, a
2-min EEG segment related to the observation of the
documentary has been further taken into account as base-
line activity. No specific question related to the docu-
mentary was asked during the interview.
The items of the administrated questionnaire are the
following:
1. During the movie, you have watched some advertise-
ments have been showed. Do you remember some of
them? Which one? (if subject does not remember any
advertisement, go to item 3).
2. Are you able to tell me the plot of the advertisements
you remembered?
3. Did you watch any of the following advertisements in
the movie? (the experimenter verbally list a set of
advertisements).
4. Now, I am going to show you some images. You
should tell me whether these images are extracted from
the advertisements you have watched or not.
5. Please, rank the following advertisements according to
the degree of pleasantness you perceived while
watching them. You should give a number between 1
(lowest pleasantness) and 10 (highest pleasantness).
6. Have you watched some of these advertisements
before?
EEG Recordings and Signal Processing
The cerebral activity was recorded by means of a portable
EEG 64-channel system (BE ?and Galileo software,
EBneuro, Italy) according to the 10–20 international sys-
tem configuration. All subjects were comfortably seated on
a reclining chair, in an electrically shielded, dimly lit room.
Electrodes positions were acquired in a 3D space with a
Polhemus device for the successive positioning on the head
model employed for the analysis. Recordings were initially
extra-cerebrally referred and then converted to an average
reference off-line. We collected the EEG activity at a
sampling rate =256 Hz while the impedances kept below
5kX. Each EEG trace was then converted into the Brain
Vision format (BrainAmp, Brainproducts GmbH, Ger-
many) in order to perform signal preprocessing such as
artefacts detection, filtering and segmentation. Raw EEG
traces were first band-pass-filtered (high pass =2 Hz; low
pass =47 Hz), and the independent component analysis
(ICA) was then applied to detect and remove components
due to eye movements and blinks.
These EEG traces were then segmented to extract the
cerebral activity during the observation of the TV com-
mercials and the one associated to the documentary
(baseline period). Then, this dataset has been further seg-
mented into one-second length trials. Later, a semi-auto-
matic procedure has been adopted to reject such EEG trials
that present muscular and other kinds of movement arte-
facts. Only artefacts-free trials have been considered for the
following analysis.
Estimation of Cortical Power Spectral Density
In this work, cortical activity from EEG scalp recordings
was estimated by employing the high-resolution EEG
technologies [69–73] with the use of a realistic head model
known as average head model from McGill University. The
scalp, skull and dura mater compartments were built using
1,200 triangles for each structures, and the Boundary Ele-
ment Model was then employed to solve the forward
electromagnetic model. For each subject, the electrodes
disposition on the scalp surface, through a nonlinear
Fig. 1 Picture presents the schema describing the video stimulation
for the Experiment 1. Gray boxes indicate the time slots related to the
portions of documentary, whereas the white boxes are related to the
time slots of the TV commercials. Durations in minutes and seconds,
for documentary and spots, respectively, are also indicated (Color
figure online)
Cogn Comput
123
Author's personal copy
minimization procedure, has been generated [74]. The
cortical model consists of about 4,000 dipoles uniformly
disposed on the cortical surface, and the estimation of the
current density strength for each dipole was obtained by
solving the electromagnetic linear inverse problem
according to techniques described in previous papers [75–
77]. Each dipole was modeled to be perpendicular to the
cortical surface.
The power spectral density (PSD) of the estimated
cortical signals was calculated using the Welch method
[78] and then mapped onto the average cortex model (MNI
template, 4,096 cortical dipoles) as described above. This
procedure allowed us to obtain a measure of PSD values
for each estimated cortical location and for each trial for all
subject’s datasets, in the frequency range of [1,40] Hz with
a resolution of 1 Hz.
Individual alpha frequency (IAF) has been calculated for
each subject in order to define four bands of interest
according to the method suggested in the literature [79].
Such bands were in the following reported as IAF ?x,
where IAF is the Individual Alpha Frequency, in Hertz, and
x is an integer displacement in the frequency domain,
which is employed to define the band ranges. In particular,
we focused the present analysis in the following frequency
bands: theta (IAF-6, IAF-2), i.e., in the frequency band
between IAF-6 and IAF-2 Hz, alpha (IAF-2, IAF ?2).
To avoid personal baseline, the z-score computation [80]
has been performed for the PSD of each cortical dipole and
subject using the data related to the observation of the
documentary as reference. Then, signals related to the
LIKE and DISLIKE groups have been averaged and sta-
tistically compared by means of z-score subtraction to
obtain the results illustrated in the following sections. In
order to take into account subjects’ personal baseline
activity, we used the neurophysiological signals (mean and
standard deviation) related to the observation of the doc-
umentary to transform into z-score variables the values of
spectral power of the datasets related to the commercials
according to the following formula:
Zspot ¼Xspot ldoc
rdoc ð1Þ
where Z
spot
is the z-score value related to the TV com-
mercials dataset, whereas l
doc
and r
doc
are mean and
standard deviation related to the documentary dataset. Such
procedure has been used to contrast the PSD for each
cortical dipole, and the Bonferroni correction for multiple
comparisons was also adopted [81,82].
Autonomic Recordings and Signal Processing
The GSR and the HR has been recorded by means of the
PSYCHOLAB VD13S system (SATEM, Italy) with a
sampling rate of 10 Hz. Skin conductance was recorded by
the constant voltage method (0.5 V). Ag–AgCl electrodes
(8 mm diameter of active area) were attached to the palmar
side of the middle phalanges of the second and third fingers
of the subject’s nondominant hand by means of a velcro
fastener. SATEM also provided disposable Ag–AgCl
electrodes to acquire the HR signal. Before applying the
sensors, subjects’ skin has been cleaned by following
procedures and suggestions published in the literature [83–
85]. GSR and HR signals have been continuously acquired
for the entire duration of the video and then filtered and
segmented with in-house MATLAB software. As to the
GSR signal processing, we used a band-pass filter with a
low cutoff frequency of 0.2 Hz in order to split the phasic
component of the electrodermal activity from the tonic one,
and a high cutoff frequency of 1 Hz to filter out noise and
suppress artefacts caused by Ebbecke waves [83,86]. As
explained in the previous section, besides the autonomic
activity of the subjects during the observations of the video
clips we used a part of the documentary to estimate the
mean and standard deviation of the electrodermal activity
and the cardiac frequency rate signal in order to compute
their z-score variables. These variables have been com-
puted for each TV spot analyzed and subject recorded.
Specifically, the z-score variables have been computed for
the tonic component of the GSR and for the entire HR
signal. They were used to form the experimental datasets
previously described (LIKE, DISLIKE) to be statistically
compared.
Results
Behavioral Results
The experimental subjects have been divided in two sub-
groups, LIKE and DISLIKE, according to the pleasantness
score they gave during the interview performed after the
recording session. Their median =8 and iqr =2, and
median =3 and iqr =1.75 (Kolmogorov–Smirnov test:
d=1, p\0.001) for LIKE and DISLIKE groups,
respectively. The six TV commercials have been sponta-
neously remembered 32 times across all subjects. The
advertisements have been all remembered after the exper-
imenter verbally and graphically stimulated the subjects.
We did not ask any specific judgement related to the
observation of the documentary.
Neurophysiological Results
A Student’s ttest has been performed between cortical z-
score distributions of the LIKE and DISLIKE groups.
Figure 2presents the related statistical cortical maps
Cogn Comput
123
Author's personal copy
viewed from a frontal perspective. The color scale
employed to color dipoles on the cortex codes the statistical
significance: Gray color is used where the activity of cor-
tical areas does not differ between the two conditions. The
red (blue) color is used when the cortical areas present a
significant increase of power spectral activity for the pop-
ulation that liked (disliked) the commercial videos with
respect to the other. Hence, the picture presents the contrast
between the LIKE and DISLIKE groups in the frequency
bands considered in this analysis. The significant increase
of the frontal activity in the theta band is clearly visible (in
red) in the LIKE group when compared to the DISLIKE
one (upper left part of the Fig. 2). Scattered increased of
cortical activity on the left hemisphere is also present in the
DISLIKE group (in blue). In the alpha frequency band
(upper right of Fig. 2), significant increase of cortical
activity is present on the left hemisphere and on the
orbitofrontal right hemisphere in the LIKE group when
compared to the DISLIKE one.
Figure 3presents the z-score average waveforms of
GSR and Heart Rate (HR) for a representative TV
commercial
From the above picture, it is possible to appreciate the
different time scale of HR signal, during the observation of
the commercial videos. It is possible to appreciate the time
intervals of the HR when exceed the level of statistical
significance (|z|[2, p\0.05). When the signal is within
the range z=[-2, 2], no difference between HR and
documentary appears. Particularly, with respect to the
observation of the documentary, the average values of z-
scores show an increment of cardiac activity at the begin-
ning and at the end of the commercial when the brand is
advertised. Also, similar increase of activity has been
elicited during the observation of scenes showing the actors
in a wheat field. However, the central time interval of the
commercial presents a negative activation when subjects
watch actors in several dark scenes.
As it is possible to see in picture, the average waveform
of the GSR did not exceed the level of statistical
significance.
The analysis of the average values of the autonomic
variables gathered in the experimental group was performed
using a repeated-measures ANOVA with factor AUTO-
NOMIC, with two levels (GSR and HR, including the
Fig. 2 Figure presents two cortical statistical maps, in the theta and
alpha bands. Legend represents cortical areas in which increased
statistically significant activity occurs in the LIKE group when
compared to the DISLIKE group in red, while blue is used otherwise
(p\0.05, Bonferroni corrected). Gray color is used to map cortical
areas where there are no significant differences between the cortical
activity in the LIKE and DISLIKE groups (Color figure online)
Fig. 3 Figure presents the z-score values of GSR (left) and HR (right)
averaged on the entire population during the observation of a
commercial video. The colored rectangles, at the beginning and at the
end of the time scale, depict the interval in which the brand is overtly
presented in the spot. In such case, the brand corresponds to a
particular and well-known kind of biscuits in Italy. The time scale is
in seconds. Dotted lines at z=-2 and z=2 indicate the thresholds
of statistical significance at p\0.05 (Color figure online)
Cogn Comput
123
Author's personal copy
transformed z-score of the GSR and HR recordings,
respectively) and the factors BRAND (levels BRAND and
NO BRAND related to the activity gathered during the
observation of the brands and in the other frame segments,
respectively) and REPORT (LIKE, DISLIKE). Although the
repeated-measures ANOVA did not highlight any significant
result for the main factors BRAND and REPORT, it returned
a statistically significant interaction between all the factors
employed, with a AUTONOMIC 9BRAND 9REPORT
significance of p\0.05. The interactions between the main
factors AUTONOMIC 9REPORT is also statistically sig-
nificant, with a p\0.05 while there are no interactions
between the factors BRAND 9REPORT (p\0.48). The
post hoc analysis performed with the Duncan test returns that
the values of the HR are statistically significantly higher in
the LIKE group versus DISLIKE group in the BRAND
condition (p\0.05) while there is a trend that is not statis-
tically significant in the NO BRAND condition (p\0.08).
There are no significant differences between the values of the
GSR variable between the LIKE and DISLIKE groups, for
both BRAND (p\0.54) and NO BRAND conditions
(p\0.43).
Discussion
The analysis of the statistical cortical maps in the condi-
tions LIKE versus DISLIKE suggested that the left frontal
hemisphere was highly active during the LIKE condition,
especially in the theta and alpha band. The results here
obtained for the LIKE condition are also congruent with
other observations performed with EEG in a group of 20
subjects during the observation of pictures from the Inter-
national Affective Picture System (IAPS, [87,88]). Both
studies indicated an increase of the EEG activity in the
theta and alpha bands for the anterior areas of the left
hemisphere. It is worth to note that there were methodo-
logical differences between the cited studies and the pres-
ent one that are related to the use of different stimuli and
processing algorithms. We could argue that the cerebral
regions involved for processing emotions for static pictures
are also involved for processing emotions during the
observation of moving scenes, such as TV commercials. A
strong involvement of frontal and prefrontal areas has also
been already experienced in a previous study performed
with high-resolution EEG, functional connectivity and
graph theory tools [89–93]. The convergence of these
results, obtained in the ecologic conditions of the obser-
vation of commercial videos within the documentary with
those of more controlled memory and affective tasks,
deserves attention.
The measurements of the HR report a statistically sig-
nificant difference when the population watched commercial
videos that resulted pleasant (LIKE vs. DISLIKE). In par-
ticular, during the observation of the commercials for the
LIKE condition, the z-scored HR is statistically different
when compared to the DISLIKE group. On the contrary, the
z-score levels of the GSR variable during the LIKE/DIS-
LIKE cannot be statistically distinguished. Hence, since
variation of GSR relates to level of arousal [41,42], we could
conclude that the average level of arousal did not change
across the entire set of the commercial videos presented,
irrespective of the experimental conditions.
Moreover, the indications provided by the autonomic
measurements in the analyzed population suggest that HR
is a variable that is useful to track the occurrence of
pleasantness of the commercial videos. In addition, we
observed as the proposed commercials did not elicit par-
ticular changes of GSR in the investigated population. This
is important since it was previously known that participants
react to the viewing of highly aversive films with HR
deceleration and a marked electrodermal increase [43–45,
94]. In this particular case, due to the particular nature of
the video clips presented (commercial advertisements with
a limited arousal content), such orienting and aversive
reaction was not generated.
Experiment 2
Methods
Subjects
In the Experiment 2, we show a concrete application of the
neuroelectrical tools to a real case study. The whole
experimental sample is formed by 24 subjects
(25–54 years, 12 women), which will be in the following
divided in two subgroups (younger adults and older adults)
differing by age. Written informed consent was obtained
from each subject after the explanation of the study. The
present advertisement has been divided into 5 time inter-
vals showing different key scenes of the video clip (as
shown in Fig. 5) in which we computed the AW and EI for
the Like and Dislike groups.
Experimental Paradigm
The procedure of the experimental task consisted in
observing a 20-min-long documentary in which we inserted
two commercial breaks, after 5 and 15 min from the
beginning of the movie, respectively. Each interruption
was formed by seven commercial video clips of different
length (3000,20
00 and 1500). The TV commercial we use here
for case study was unknown to the subjects, and it has been
showed only once during the experiment. Randomization
Cogn Comput
123
Author's personal copy
of the occurrence of the all commercial videos within the
documentary was made to remove the order effect as
possible confounding effect. The specific advertisement is
the one aired by Air Action Vigorsol, a famous brand and
type of chewing gum, which has as protagonists a pair of
lovers who live far away from each other. The TV com-
mercial can be watched at the following link: http://www.
youtube.com/watch?v=o8RCOb3WQOs. The specific fea-
ture of the present advertisement consists in generating an
exhilarating climax with some not-so-nice images.
The present advertisement has been divided into 5 time
intervals showing different key scenes of the video clip (as
shown in Fig. 5) in which we computed the AW and EI for
the Like and Dislike groups.
The procedure of Experiment 2 is the same already
illustrated for Experiment 1, therefore not reported here.
The schema of the video sequence is shown in Fig. 4.
EEG Recordings and Signal Processing
The cerebral activity was recorded by means of a portable
EEG system (BEmicro and Galileo software, EBneuro,
Italy). Informed consent was obtained from each subject
after explanation of the study, which was approved by the
local institutional ethics committee. All subjects were
comfortably seated on a reclining chair, in an electrically
shielded, dimly lit room. Electrodes were arranged
according to an extension of the 10–10 international sys-
tem. Since a clear role of the frontal areas has been
depicted for the investigated phenomena [9–12,95], we
used the following channels: AF7, Fp2, Fpz, Fp1, AF6, F5,
AF3, AFz, AF4 and F6. Recordings were initially extra-
cerebrally referred and then converted to an average ref-
erence off-line. We collected the EEG activity at a sam-
pling rate =256 Hz while the impedances kept below
5kX. Each EEG trace was then converted into the Brain
Vision format (BrainAmp, Brainproducts GmbH, Ger-
many) in order to perform signal pre-processing such as
artefacts detection, filtering and segmentation. The EEG
signals have been band-pass-filtered at 1–45 Hz and de-
purated of ocular artefacts by employing the independent
component analysis (ICA). The EEG data have been re-
referenced by computing the common average reference
(CAR). Individual alpha frequency (IAF) has been
calculated for each subject in order to define the alpha as
alpha =[IAF-2, IAF ?2] [79].
Approach-Withdrawal Index
In order to define an Approach-Withdrawal Index (AW)
according to the theory related to the earlier introduced
EEG frontal asymmetry theory, we computed such
imbalance as difference between the average EEG power
of right and left channels. The formula we used is the
following:
AW ¼1
NPX
i2P
x2
aitðÞ 1
NQX
i2Q
y2
aitðÞ
¼Average Poweraright;frontal Average Poweraleft;frontal
ð2Þ
where xaiand yairepresent the ith EEG channel in the alpha
band that have been recorded from the right and left frontal
lobes, respectively. In addition, P¼Fp2;AF6;AF4;F4fg
and Q¼Fp1;AF7;AF3;F5
fg
,NPand NQrepresent the
cardinality of the two sets of channels. In such a way, an
increase of AW will be related to an increase of interest
and vice versa. The AW signal of each subject has been z-
score-transformed and then averaged to obtain an average
waveform.
Emotional Index
The emotional index is defined by taking into account the
GSR and HR signals. As far as the construction of such
variable concerns, we refer to affect circumplex [27] where
the coordinates of a point in this space are defined by the
HR (horizontal axis) and the GSR (vertical axis). As pre-
sented in the Introduction section, several studies have
highlighted that these two autonomic parameters correlate
with valence and arousal, respectively (see [23] for a
review).
In order to have a mono-dimensional variable, we
describe the emotional state of a subject by defining the
following emotional index (EI):
EI ¼1b
pð3Þ
where.
Fig. 4 Picture presents the schema describing the video stimulation
for the Experiment 2. Gray boxes indicate the time slots related to the
portions of documentary, whereas the white boxes are related to the
time slots of the TV commercials. Durations in minutes and seconds,
for documentary and spots, respectively, are also indicated (Color
figure online)
Cogn Comput
123
Author's personal copy
b¼
3
2pþp#if GSRZ0;HRZ0
p
2#otherwise
8
>
<
>
:ð4Þ
GSR
Z
,HR
Z
represent the z-score variables of GSR and HR,
respectively; #, in radians, is measured as arctang (HR
Z
,
GSR
Z
). Therefore, the angle bis defined in order to transform
the domain of #from [-p,p] to [0, 2 p] and obtain the EI
varying between [-1, 1]. This is why we have two ways to
calculate b. According to Eqs. 2and 3and the affect cir-
cumplex [27], negative (HR
Z
\0) and positive (HR
Z
[0)
values of the EI are related to negative and positive emotions,
respectively, spanning the whole affect circumplex.
Although in Experiment 1 we could not highlight sta-
tistical difference for the GSR, we preferred to maintain
this variable for the emotional index definition due to
evidence published in the literature [41,42].
Results
Behavioral Results
The recording of the neurometric response included the
detection of the EEG signals, HR and GSR parameters on a
sample of 24 subjects (39.56 ±8.11 years; 12 men) that rated
the TV commercial with a pleasantness score distribution with
median =7 and iqr =6. The experimental subjects have
been divided in two subgroups differing by age (10 younger
adults: 31.33 ±3.31; 14 older adults: 44.86 ±5.24; Stu-
dent’s t=7.57, p\0.01). Younger adults resulted rating the
proposed advertisement with higher pleasantness score
(median =8 and iqr =1) with respect to the older adults
(median =3.5, iqr =5; Kolmogorov–Smirnov test:
d=0.78, p\0.01). Later, we also divided the whole
experimental sample in LIKE and DISLIKE groups according
to the pleasantness score they gave during the interview
(LIKE: median =8, iqr =1.5; DISLIKE: median =1,
iqr =1.5; Kolmogorov–Smirnov test: d=1, p\0.01). This
commercial advertisement has been spontaneously recalled
by the 12.5 % of the population. However, all subjects
remembered it when verbally and graphically stimulated. We
did not ask any specific judgement related to the observation
of the documentary. In Fig. 5, the segmentation frame by
frame of the commercial is shown.
Cerebral Indices Results
By analyzing the AW and EI, shown in Fig. 6, it was
evident that the investigated commercial video arouses
Fig. 5 Frame sequence of the Air Action Vigorsol TV commercial
for each second of the video clip. The underlying colors highlight the
different scenes in which it is possible to divide the advertisement, as
the legend on the right shows. In such segments, the average values
for the estimated indices were computed (Color figure online)
Cogn Comput
123
Author's personal copy
different responses with respect to the two used indices. In
particular, as far as concern the investigated sample, the
final frames elicited high value related to the Approach-
Withdrawal index.
The negative effects measured for the emotion on the
opening of the bottle in the video clip has been reported as the
average value for the wholeexperimental sample. However, it
is possible to prove that such effect was sensitive to the age of
the sample subjects. In particular, Fig. 7shows the results in
terms of AW and EI estimators related to the two age-gener-
ated subgroups, younger and older adults. According to the
variations of the AW index, older adults present an increase of
activation in the initial part of the commercial (story telling),
whereas younger adults show a similar increment in the
middle of the spot (funny scenes). The Emotional Index,
except for the initial part,is always higher for younger adults.
In particular, the average EI values for the younger adults if
0.09 ±0.04 while for the older adults shows a level of
-0.06 ±0.08 (Student’s t=10.57, p\0.01). This result is
in agreement with the subjects’ behavioral ratings as it is
possible to appreciate in the previous section.
The whole experimental sample has been also divided
according to the pleasantness score in order to form the
LIKE and DILIKE groups. These two sub-targets have
been further analyzed to highlight difference in terms of
AW and EI in particular time segments of the TV
Fig. 6 Picture presents the average Approach-Withdrawal (left) and
Emotional (right) profiles, across the whole experimental sample,
related to the observation of the Air Action Vigorsol TV commercial.
Both the horizontal axis describes the time evolution of the clip (from
1 to 20 s). The vertical axis of the considered Approach-Withdrawal
Index describes the amplitude of the variable in z-score values.
Vertical axis of the Emotional Index indicates positive emotion from
0 to 0.15 values and negative emotions for values from 0 to -0.2
values. Note that the zero value is the average value of the analyzed
index during the baseline phase. Particular frames of the advertise-
ment are showed on the AW and EI waveforms
Fig. 7 Picture presents the average Approach-Withdrawal (left) and
Emotional (right) profiles across the two analyzed sub-targets, older
adults (continuous line) and younger adults (dotted line), related to
the observation of the Air Action Vigorsol TV commercial. Both
horizontal axes describe the time evolution of the clip (from 1 to
20 s). The vertical axis of the considered Approach-Withdrawal
Index describes the amplitude of the variable in z-score values. Dotted
lines at z=-2 and z=2 indicate the thresholds of statistical
significance (p\0.05) between the single waveforms and the level
related to the documentary. Instead, red circles in the horizontal axis
highlight statistically significant differences between the two groups,
older adults and younger adults (p\0.05, Bonferroni corrected).
Vertical axis of the Emotional Index indicates positive emotion from
0 to 0.15 values and negative emotions for values from 0 to -0.2
values. A zero level of the EI or AW estimators represents the average
value for that estimators during the documentary seen before the TV
commercial analyzed. Particular frames of the advertisement are
showed on the AW and EI waveforms
Cogn Comput
123
Author's personal copy
commercial. Figure 8shows the variation of AW and EI
for the defined time intervals. As to the AW index, data
highlight a statistically significant difference for two seg-
ments (Air Explosion Scene, Brand). Instead, the EI does
not present statistically significant differences between the
Like and Dislike groups across the time segments. How-
ever, it is possible to appreciate how the Like group pre-
sents higher values of EI in all the specific time intervals.
Discussion
The analysis of the Approach-Withdrawal and the Emotional
Index returned, at least apparently, discordant interpretation.
The difference between them is evident in terms of emotional
reaction to the spot (as indicated by the EI), whereas there is a
difference in terms of neuroelectrical reaction to the spot,
just in the central scene, according to the AW index. In
particular, as to the Approach-Withdrawal index, the older
adults prefer the initial part, when the story is told. At the
seconds 1100–1500, when the humor of the TV commercial
comes on in the scenes, they watch the ad with more
detachment, while the younger adults are more interested. In
fact, what we can observe that at first sight the cerebral
Approach-Withdrawal index is characterized by positive
values, whereas the analysis of autonomic signals through
the Emotional Index shows negative emotions. The inter-
pretation of this phenomenon has to be read in the different
nature of the two indices. The former mostly highlights the
cerebral attractiveness or the refusal toward a stimulus; the
latter shows the experienced internal state, through the car-
diac and cutaneous ‘‘body markers.’’ We could speculate
interpreting the described results as a sign of high cerebral
curiosity but an emotional detach of the gathered sample
related to the TV commercial analyzed.
Results suggest that the ugly scenes of the TV com-
mercial elicit negative peaks of emotions but high peaks of
AW index (e.g., approach tendency). A possible interpre-
tation is that such scenes are able to elicit interest from the
subjects although the sudden events depicted in the video
generated negative emotions.
These considerations could be enforced after analyzing
the results of the two different subgroups: younger adults
and older adults. Specifically, the former group not only
shows higher values of the Approach-Withdrawal index,
when compared to the latter, but also highlights a positive
emotion waveform. In this case, it appears that just older
adults have been negatively impressed by some ugly scenes
of the clip. Younger adults, instead, perceive the com-
mercial with a positive emotion along its whole length.
From these last considerations, we can argue that the
analyzed advertisement resulted particularly convincing
and effective for the investigated younger sample. In
addition, the analysis in different time segments, between
the LIKE and DISLIKE groups, highlighted that there are
some particular short periods within the advertisement that
we can assume to do most of the work in actuating
advertising performance measures [96]. However, the
obtained results relate to a single and specific TV com-
mercial and could be performed similar analysis and
description for additional TV commercials. Here, we do
not provide such results because we believe that it could
not add more information to the presented results.
General Discussion
Our results highlight an involvement of frontal and pre-
frontal cerebral regions during the observation of TV
Fig. 8 Picture presents the average Approach-Withdrawal (left)and
Emotional (right) indices calculated within particular segments of
interest of the Air Action Vigorsol TV commercial for the two analyzed
sub-targets, LIKE and DILIKE. On the horizontal axis, the labels of the
particular segments are shown. The vertical axis of the considered
Approach-Withdrawal Index describes the amplitude of the variable in
z-score values. Vertical axis of the Emotional Index indicates positive
emotion from 0 to 0.15 values and negative emotions for values from 0
to -0.2 values. A zero level of the EI or AW estimators represents the
average value for that estimators during the documentary seen before
the TV commercial analyzed. Statistically significant differences in the
time segments are highlighted with the symbol*
Cogn Comput
123
Author's personal copy
commercials that will be judged pleasant. These activations
mostly concern the left frontal lobe showing a de-synchro-
nization of the EEG alpha rhythm. In agreement with the
Approach-Withdrawal theory, the present results affirm that
there are two distinct brain networks processing the pleasant
and unpleasant commercials, which mostly lie in the left and
right frontal areas, respectively. The specific activation of
frontal areas related to the observation of pleasant TV
commercial has been also investigated by several scientists.
Finally, we presented a novel index for the estimation of
the emotions during the observation of TV commercials.
Such neurophysiological variable relies on the measure-
ment of GSR and HR as orthogonal axes of arousal and
valence dimensions, respectively. Our results show that the
increase of such emotional index can discriminate subjects
who liked the advertisements they watched from the one
that did not. Although the increase of HR we found out is
in agreement with previous studies [43,44], our experi-
mental paradigm did not elicit significant enhance of
activity for the GSR. However, since we did not explicit
measure the behavioral arousal, further investigation is
needed to claim results for this autonomic variable.
The neuroimaging methodology illustrated can then
provide additional indications related to the preferences of
people with respect to observation of TV commercials that
are quite different from the standard market research
studies, often based on the generation of written or verbal
questionnaires. This neurometric approach could be useful
to test different ideas or product concepts quite rapidly in
the process of the generation of a new product.
Conclusion
The illustrated results showed an increase of both HR and
cerebral activity, mainly in the theta and alpha bands in the
left hemisphere, when subjects watched pleasant TV com-
mercials. Specifically, the EEG activity highlights a frontal
unbalance between left and right frontal hemispheres in
different frequency bands. These results have been sum-
marized in two neurophysiological indices, which could be
used in neuromarketing research for the evaluation of
already existing TV commercials as well as concepts and
idea of the product, even before their real production. How
far this measurement and interpretation process will bring
the marketing research in the next years is still too early to
state at the light of the present knowledge.
Acknowledgments This work was in part supported by a grant of
Italian Minister of Research and University ‘‘PRIN 2012’’, by EU-
ROCONTROL on behalf of the SESAR Joint Undertaking in the
context of SESAR Work Package E - NINA research project , by the
Italian Minister of Foreign Affairs with a bilateral project between
Italy and China ‘‘Neuropredictor’’ and by a project FILAS ‘‘Brain-
Trained’’ CUP F87I12002500007.
References
1. Dama
´sio AR. Descarte’s error: emotion, reason and the human
brain. New York: Penguin Putnam; 1994.
2. Damasio, The feeling of what happens: body and emotion in the
making of consciousness, Harcourt. 1999.
3. Bargh JA, Chartrand TL. The unbearable automaticity of being.
Am Psychol. 1999;54(7):462–79.
4. Davidson RJ, Irwin W. The functional neuroanatomy of emotion
and affective style. Trends Cogn Sci. 1999;3(1):11–21.
5. Panksepp J. Affective neuroscience: the foundations of human
and animal emotions. 1 ed. Oxford University Press, USA. 2004.
6. Adolphs R. Cognitive neuroscience of human social behaviour.
Nat Rev Neurosci. 2003;4(3):165–78.
7. Astolfi L, Toppi J, De Vico Fallani F, Vecchiato G, Cincotti F,
Wilke CT, Yuan H, Mattia D, Salinari S, He B, Babiloni F.
Imaging the social brain by simultaneous hyperscanning during
subject interaction. IEEE Intell Syst. 2011;26(5):38–45.
8. Cacioppo JT, Berntson GG, Taylor SE, Schacter DL. Foundations
in social neuroscience illustrated edition. The MIT Press. 2002.
9. Erk S, Spitzer M, Wunderlich AP, Galley L, Walter H. Cultural
objects modulate reward circuitry. NeuroReport. 2002;13:2499–503.
10. Braeutigam S. Neuroeconomics–from neural systems to eco-
nomic behavior. Brain Res Bull. 2005;67(5):355–60.
11. Knutson B, Scott R, Wimmer GE, Prelec D, Loewenstein G.
Neural Predictors of Purchases. Neuron. 2007;53:147–56.
12. Bruce AS, Bruce JM, Black WR, Lepping RJ, Henry JM, Cherry
JB et al. Branding and a child’s brain: an fMRI study of neural
responses to logos. SCAN. 2013.
13. Braeutigam S, Rose SP, Swithenby SJ, Ambler T. The distributed
neuronal systems supporting choice-making in real-life situa-
tions: differences between men and women when choosing gro-
ceries detected using magnetoencephalography. Eur J Neurosci.
2004;20(1):293–302.
14. Ioannides AA, Liu L, Theofilou D, Dammers J, Burne T, Ambler
T, et al. Real time processing of affective and cognitive stimuli in
the human brain extracted from MEG signals. Brain Topogr.
2000;13(1):11–9.
15. Silberstein RB, Harris PG, Nield GA, Pipingas A. Frontal steady-
state potential changes predict long-term recognition memory
performance. Int J Psychophys. 2000;39(1):79–85.
16. Astolfi L, Vecchiato G, De Vico Fallani F, Salinari S, Cincotti F,
Aloise F et al. The track of brain activity during the observation
of TV commercials with the high-resolution EEG technology.
Comput Intell Neurosci. 2009:652078.
17. Vecchiato G, Astolfi L, De Vico Fallani F, Toppi J, Aloise F, Bez
F et al. On the Use of EEG or MEG brain imaging tools in
neuromarketing research. Comput Intell Neurosci. 2011. doi:10.
1155/2011/643489.
18. Plassmann H, Ramsøy TZ, Milosavljevicet M. Branding the
brain: a critical review and outlook. J Cons Psych. 2012.
19. Zaltman G. how customers think: essential insights into the mind
of the market 1 ed., Harvard Business Press. 2003.
20. McDonald C. Is your advertising working? WARC. 2003.
21. Franzen G, Bouwman M. The mental world of brands: mind,
memory and brand success. NTC Publications; 2001.
22. Ariely D, Berns G. Neuromarketing: the hope and hype of neu-
roimaging in business. Nat Rev Neurosci. 2010;11(4):284–92.
doi:10.1038/nrn2795.
23. Mauss IB, Robinson MD. Measures of emotion: a review. Cogn
Emot. 2009;23(2):209–37.
Cogn Comput
123
Author's personal copy
24. Hamann S. Mapping discrete and dimensional emotions onto the
brain: controversies and consensus. Trends Cogn Sci. 2012;16(9):
458–66. doi:10.1016/j.tics.2012.07.006.
25. Davidson RJ. Neuropsychological perspectives on affective styles
and their cognitive consequences. In: Power MJ, YorkDalgleish
T, editors. Handbook of cognition and emotion. New. New York:
Wiley; 1999. p. 103–23.
26. Lang PJ, Bradley MM, Cuthbert BN. Motivated attention: affect,
activation, and action. In: Lang PJ, Simons RF, Balaban MT,
editors. Attention and orienting: sensory and motivational pro-
cesses. Mahwah, NJ: Lawrence Erlbaum Associates, Inc. 1997.
p. 97–135.
27. Russell JA, Barrett LF. Core affect, prototypical emotional epi-
sodes, and other things called emotion: dissecting the elephant.
J Pers Soc Psychol. 1999;76(5):805–19.
28. Schneirla TC. An evolutionary and developmental theory of
biphasic processes underlying approach and withdrawal. In: Jones
MR, editor. Nebraska symposium on motivation, 1959. Lincoln,
NE: University of Nebraska Press; 1959. p. 1–42.
29. Watson D, Wiese D, Vaidya J, Tellegen A. The two general
activation systems of affect: structural findings, evolutionary
considerations, and psychobiological evidence. J Pers Soc Psy-
chol. 1999;76(5):820–38.
30. Panksepp J. Neurologizing the psychology of affects: how
appraisal-based constructivism and basic emotion theory can
coexist. Perspect Psychol Sci. 2007;2(3):281–95.
31. Ekman P. Facial expressions of emotion: new findings, new
questions. Psychol Sci. 1992;3:34–8.
32. Rachael JE, Garrod OGB, Schyns PG. Dynamic facial expres-
sions of emotion transmit an evolving hierarchy of signals over
time. Curr Biol. 2014;24(2):187–92.
33. Barrett LF. Solving the emotion paradox: categorization and the
experience of emotion. Pers Soc Psychol Rev. 2006;10(1):20–46.
34. Haidt J, Keltner D. Culture and facial expression: open-ended
methods find more expressions and a gradient of recognition.
Cogn Emot. 1999;13(3):225–66.
35. Carver CS. Self-regulation of action and affect. In: Baumeister RF,
Vohs KD, editors. Handbook of self-regulation: research, theory,
and applications. New York: Guilford Press. 2004. p. 13–39.
36. Baumgartner T, Esslen M, Jancke L. From emotion perception to
emotion experience: emotions evoked by pictures and classical
music. Int J Psychophys. 2006; 60(1):34–4.3.
37. Cacioppo JT, Berntson GG, Larsen JT, Poehlmann KM, Ito TA.
The psychophysiology of emotion. In: Lewis M, Haviland-Jones
JM, editors. The handbook of emotion. New York: Guildford
Press. 2000.
38. Mauss IB, Wilhelm FH, Gross JJ. Is there less to social anxiety
than meets the eye? Emotion experience, expression, and bodily
responding. Cogn Emot. 2004;18(5):631–62.
39. Bradley MM, Lang PJ. Measuring emotion: behavior, feeling,
and physiology. In: Lane RD, NadelL, editors. Cognitive neuro-
science of emotion. New York: Oxford University Press; 2000b.
p. 242–276.
40. Kop WJ, Synowski SJ, Newell ME, Schmidt LA, Waldstein SR,
Fox NA. Autonomic nervous system reactivity to positive and
negative mood induction: the role of acute psychological
responses and frontal electrocortical activity. Biol Psychol.
2011;86(3):230–8. doi:10.1016/j.biopsycho.2010.12.003.
41. Critchley HD. Electrodermal responses: what happens in the
brain. Neuroscientist. 2002;8(2):132–42.
42. Nagai Y, Critchley HD, Featherstone E, Fenwick PB, Trimble MR,
Dolan RJ. Brain activity relating to the contingent negative varia-
tion: an fMRI investigation. NeuroImage. 2004;21(4):1232–41.
43. Baldaro B, Mazzetti M, Codispoti M, Tuozzi G, Bolzani R,
Trombini G. Autonomic reactivity during viewing of an
unpleasant film. Percept Mot Skills. 2001;93:797–805.
44. Palomba D, Sarlo M, Angrilli A, Mini A, Stegagno L. Cardiac
responses associated with affective processing of unpleasant film
stimuli. Int J Psychophysiol. 2000;36:45–57.
45. Steptoe A, Wardle J. Emotional fainting and the psychophysio-
logic response to blood and injury: autonomic mechanisms and
coping strategies. Psychosom Med. 1988;50:402–17.
46. Kreibig SD, Wilhelm FH, Roth WT, Gross JJ. Cardiovascular,
electrodermal, and respiratory response patterns to fear- and
sadness-inducing films. Psychophysiology. 2007;44(5):787–806.
47. Barrett LF, Russell JA. Independence and bipolarity in the
structure of current affect. J Pers Soc Psychol. 1998;74:967–84.
48. Davidson RJ. Anxiety and affective style: role of prefrontal
cortex and amygdala. Biol Psychiatry. 2002;51(1):68–80.
49. Heller AS, Lapate RC, Mayer K, Davidson RJ. The face of
negative affect: trial-by-trial corrugator responses to negative
pictures are positively associated with amygdala and negatively
associated with ventromedial prefrontal cortex activity. J Cogn
Neurosci. 2014.
50. Shenhav A, Greene JD. Integrative moral judgment: dissociating
the roles of the amygdala and ventromedial prefrontal cortex.
J Neurosci. 2014;34(13):4741–9.
51. Rabinak CA, Macnamara A, Kennedy AE, Angstadt M, Stein MB,
Liberzon I, Phan KL. Focal and aberrant prefrontal engagement
during emotion regulation in veterans with posttraumatic stress
disorder. Depress Anxiety. 2014.
52. Motzkin JC, Philippi CL, Wolf RC, Baskaya MK, Koenigs M.
Ventromedial prefrontal cortex is critical for the regulation of
amygdala activity in humans. Biol Psychiatry. 2014.
53. Hooker CI, Benson TL, Gyurak A, Yin H, Tully LM, Lincoln SH.
Neural activity to positive expressions predicts daily experience
of schizophrenia-spectrum symptoms in adults with high social
anhedonia. J Abnorm Psychol. 2014;123(1):190–204.
54. Lin CH, Tuan HP, Chiu YC. Medial frontal activity in brand-
loyal consumers: a behavior and near-infrared ray study. J Neu-
rosci Psychol Econ. 2010;3:59–73.
55. Li X, Lu ZL, D’argembeau A, Ng M, Bechara A. The iowa gam-
bling task in fMRI images. Hum Brain Mapp. 2010;31:410–23.
56. Santos JP, Seixas D, Branda
˜o S, Moutinho L. Investigating the
role of the ventromedial prefrontal cortex (vmPFC) in the
assessment of brands. Front Neurosci. 2011; 5(77).
57. Davidson RJ. What does the prefrontal cortex ‘‘do’’ in affect:
perspectives on frontal EEG asymmetry research. Biol Psychol.
2004;67(1–2):219–33.
58. Davidson RJ. Affective style, psychopathology, and resilience: brain
mechanisms and plasticity. Am Psychologist. 2000;55(11):1196–214.
59. Sutton SK, Davidson RJ. Prefrontal brain asymmetry: a biologi-
cal substrate of the behavioral approach and inhibition systems.
Psychol Sci. 1997;8(3):204–10.
60. Harmon-Jones E, Allen JJB. Anger and frontal brain activity: EEG
asymmetry consistent with approach motivation despite negative
affective valence. J Pers Soc Psychol. 1998;74(5):1310–6.
61. Harmon-Jones E, Lueck L, Fearn M, Harmon-Jones C. The effect
of personal relevance and approach related action expectation on
relative left frontal cortical activity. Psychol Sci. 2006;17(5):
434–40.
62. Heller W, Schmidtke JI, Nitschke JB, Koven NS, Miller GA.
States, traits, and symptoms: Investigating the neural correlates of
emotion, personality, and psycho-pathology. In: Cervone D,
Mischel W, editors. Advances in personality science. New York:
Guilford Press; 2002. p. 106–26.
63. De Pascalis V, Cozzuto G, Caprara GV, Alessandri G. Relations
among EEG-alpha asymmetry, BIS/BAS, and dispositional opti-
mism. Biol Psychol. 2013;94(1):198–209.
64. Santesso DL, Segalowitz SJ, Ashbaugh AR, Antony MM,
McCabe RE, Schmidt LA. Frontal EEG asymmetry and sensation
seeking in young adults. Biol Psychol. 2008;78(2):164–72.
Cogn Comput
123
Author's personal copy
65. Urry HL, Nitschke JB, Dolski I, Jackson DC, Dalton KM, Mu-
eller CJ, Rosenkranz MA, Ryff CD, Singer BH, Davidson RJ.
Making a life worth living: neural correlates of well-being.
Psychol Sci. 2004;15:367–72.
66. Pizzagalli DA, Sherwood RJ, Henriques JB, Davidson RJ. Frontal
brain asymmetry and reward responsiveness. Psychol Sci. 2005;
16:805–13.
67. Ohme R, Reykowska D, Wiener D, Choromanska A. Analysis of
neurophysiological reactions to advertising stimuli by means of
EEG and galvanic skin response measures. J Neurosci Psychol
Econ. 2009;2(1):21–31.
68. Handy TC, Smilek D, Geiger L, Liu C, Schooler JW. ERP evi-
dence for rapid hedonic evaluation of logos. J Cogn Neurosci.
2010;22(1):124–38.
69. Ding L, Lai Y, He B. Low-resolution brain electromagnetic
tomography in a realistic geometry head model: a simulation
study. Phys Med Biol. 2005;50:45–56.
70. He B, Wang Y, Wu D. Estimating cortical potentials from scalp
EEGs in a realistically shaped inhomogeneous head model by
means of the boundary element method. IEEE Trans Biomed
Eng. 1999;46:1264–8.
71. Nunez PL. Neocortical dynamics and human EEG rhythms. New
York: Oxford University Press; 1995.
72. Babiloni C, Babiloni F, Carducci F, Cincotti F, Rosciarelli F,
Rossini PM, et al. Mapping of early and late human somatosen-
sory evoked brain potentials to phasic galvanic painful stimula-
tion. Hum Brain Mapp. 2001;12(3):168–79.
73. De Vico Fallani F, Astolfi L, Cincotti F, Mattia D, Marciani MG,
Salinari S, et al. Cortical functional connectivity networks in
normal and spinal cord injured patients: evaluation by graph
analysis, human brain mapping. 2007; 28(12):1334–46.
74. Astolfi L, De Vico Fallani F, Cincotti F, Mattia D, Bianchi L,
Marciani MG, et al. Neural basis for brain responses to TV
commercials: a high-resolution EEG study. IEEE Trans Neural
Syst Rehabil Eng. 2008; 16(6):522–31.
75. Astolfi L, Cincotti F, Mattia D, Marciani MG, Baccala L, De
Vico Fallani F et al. Comparison of different cortical connectivity
estimators for high-resolution EEG recordings. Hum Brain Mapp.
2007; 28(2):143–57.
76. Astolfi L, De Vico Fallani F, Cincotti F, Mattia D, Marciani MG,
Bufalari S, et al. Imaging functional brain connectivity patterns
from high-resolution EEG And fMRI via graph theory. Psycho-
phys. 2007;44(6):880–93.
77. Babiloni F, Cincotti F, Babiloni C, Carducci F, Basilisco A,
Rossini PM et al. Estimation of the cortical functional connec-
tivity with the multimodal integration of high resolution EEG and
fMRI data by directed transfer function. Neuroimage. 2005; 1,
24(1):118–31.
78. Welch PD. The use of fast fourier transform for the estimation of
power spectra: a method based on time averaging over short,
modified periodograms. IEEE T Audio Electroac. 1967; vol. AU-
15, pp. 70–73.
79. Klimesch W. EEG alpha and theta oscillations reflect cognitive
and memory performance: a review and analysis. Brain Res Rev.
1999;29(2):169–95.
80. Zar JH. Biostatistical analysis. 5th ed. New Jersey: Prentice Hall;
2009.
81. Vecchiato G, De Vico Fallani F, Astolfi L, Toppi J, Cincotti F,
Mattia D, Salinari S, Babiloni F. The issue of multiple univariate
comparisons in the context of neuroelectric brain mapping: an
application in a neuromarketing experiment. J Neurosci Methods.
2010;191(2):283–9.
82. Vecchiato G, Astolfi L, De Vico Fallani F, Cincotti F, Mattia D,
Salinari S, Soranzo R, Babiloni F. Changes in brain activity
during the observation of TV commercials by using EEG, GSR
and HR measurements. Brain Topogr Jun. 2010; 23(2):165–79.
83. Schmidt S, Walach H. Electrodermal Activity. State of the art
measurement and techniques for parapsychological purpose.
J Parapsychol. 2000.
84. Fowles DC, Christie MJ, Edelberg R, Grings WW, Lykken DT,
Venables PH. Publication recommendations for electrodermal
measurements. Psychophysiology. 1981;18:232–9.
85. Venables PH. Autonomic activity. Ann NY Acad Sci. 1991;620:
191–207.
86. Boucsein W. Electrodermal activity 1 ed., Springer. 1992.
87. Aftanas LI, Reva NV, Varlamov AA, Pavlov SV, Makhnev VP.
Analysis of evoked EEG synchronization and desynchronization
in conditions of emotional activation in humans: temporal and
topographic characteristics. Neurosci Behav Physiol. 2004;34(8):
859–67.
88. Balconi M, Brambilla E, Falbo L. BIS/BAS, cortical oscillations
and coherence in response to emotional cues. Brain Res Bull.
2009;80(3):151–7.
89. Astolfi L, Vecchiato G, De Vico Fallani F, Salinari S, Cincotti F,
Aloise F, Mattia D, Marciani MG, Bianchi L, Soranzo R, Babi-
loni F. The track of brain activity during the observation of TV
commercials with the high-resolution EEG technology. Comput
Intell Neurosci. 2009:652078. doi:10.1155/2009/652078. Epub
2009 Jun 22.
90. De Vico Fallani F, da Fontoura Costa L, Aparecido Rodriguez L,
Astolfi L, Vecchiato G, Toppi J, Borghini G, Cincotti F, Mattia D,
Salinari S, Isabella R, Babiloni F. A graph-theoretical approach in
brain functional networks possible implications in EEG studies.
Nonlinear Biomedical Phys. 2010;4(Suppl 1):S8.
91. Astolfi L, De Vico Fallani F, Cincotti F, Mattia D, Bianchi L,
Marciani MG, Salinari S, Colosimo A, Tocci A, Soranzo R,
Babiloni F. Neural basis for brain responses to TV commercials:
a high-resolution EEG study. IEEE Trans Neural Syst Rehabil
Eng. 2008;16(6):522–31. doi:10.1109/TNSRE.2008.2009784.
92. Toppi J, De Vico Fallani F, Vecchiato G, Maglione AG, Cincotti
F, Mattia D, Salinari S, Babiloni F, Astolfi L. How the statistical
validation of functional connectivity patterns can prevent erro-
neous definition of small-world properties of a brain connectivity
network. Comput Math Methods Med. 2012; 2012:130985.
93. Toppi J, Babiloni F, Vecchiato G, Cincotti F, De Vico Fallani F,
Mattia D, Salinari S, Astolfi L. Testing the asymptotic statistic for
the assessment of the significance of partial directed coherence
connectivity patterns. Conf Proc IEEE Eng Med Biol Soc.
2011;2011:5016–9. doi:10.1109/IEMBS.2011.6091243.
94. Oliveri M, Babiloni C, Filippi MM, Caltagirone C, Babiloni F,
Cicinelli P, et al. Influence of the supplementary motor area on
primary motor cortex excitability during movements triggered by
neutral or emotionally unpleasant visual cues. Exp Brain Res.
2003;149(2):214–21.
95. Vecchiato G, Toppi J, Astolfi L, De Vico Fallani F, Cincotti F,
Mattia D et al. Spectral EEG frontal asymmetries correlate with
the experienced pleasantness of TV commercial advertisements.
Med Biol Eng Comput. 2011; 49(5):579–83.
96. Young C. Brain waves, picture sorts
, and branding moments.
J.Advert. Res. 2002, 42–53.
Cogn Comput
123
Author's personal copy
