The current issue and full text archive of this journal is available on Emerald Insight at:

www.emeraldinsight.com/1355-5855.htm

Consumer
Which wine do you prefer? An behaviour and
analysis on consumer behaviour brain activity

and brain activity during a wine

tasting experience 1149

Letizia Alvino Received 8 April 2019

Revised 17 July 2019
International Business and Language, Accepted 31 July 2019
Rotterdam University of Applied Sciences, Rotterdam, The Netherlands and
TiSEM, Tilburg University, Tilburg, The Netherlands
Rob van der Lubbe
Cognitive Psychology and Ergonomics,
University of Twente, Enschede, The Netherlands
Reinoud A.M. Joosten
Industrial Engineering and Business Information Systems,
University of Twente, Enschede, The Netherlands, and
Efthymios Constantinides
Department of NIKOS, University of Twente, Enschede, The Netherlands

Abstract
Purpose – The purpose of this paper is to assess whether or not electroencephalography (EEG) provides a
valuable and substantial contribution to the prediction of consumer behaviour and their preferences during
product consumption. In this study, the authors especially focus on individual preferences during a wine
tasting experience.
Design/methodology/approach – A consumer neuroscience experiment was carried out with 26 participants
that evaluated different red wines while their brain activity was recorded with EEG. A within-subjects design
was employed and the experiment was carried out in two sessions. All participants took part in a blind taste
session (no label session), in which information about the wine was not disclosed, and a normal taste session
(label session), during which the bottle and its label were visible.
Findings – The findings suggest that EEG is a useful tool to study brain activity during product experience.
EEG has high temporal resolution, low costs, small dimensions and superior manoeuvrability compared to
other consumer neuroscience tools. However, it is noticed that there is a lack of solid theoretical background
regarding brain areas (e.g. frontal cortex) and brain activity (e.g. brain waves) related to consumer preferences
during product experience. This lack of knowledge causes several difficulties in replicating and validating the
findings of other consumer neuroscience experiments for studying consumer behaviour.
Originality/value – The experiment presented in this paper is an exploratory study. It provides insights into
the possible contribution of EEG data to the prediction of consumer behaviour during product experience.
Keywords Wine tasting, EEG, Neuromarketing, Consumer neuroscience, Beta band, Wine preferences
Paper type Research paper

1. Introduction
Recent years have seen an explosion in the use of neuroimaging techniques in marketing
research. The use of neuroimaging and psychophysiological tools for marketing purposes is
usually labelled as consumer neuroscience (mostly used in academia) or neuromarketing.
Consumer neuroscience uses both psychological and neuroscience methods to investigate Asia Pacific Journal of Marketing
human neural mechanisms and cognitive functions. Researchers in this field aim at and Logistics
Vol. 32 No. 5, 2020
pp. 1149-1170
The authors would like to thank all of those who took part in our study. The authors also would thank © Emerald Publishing Limited
1355-5855
Prof. Dr Petra C. de Weerd-Nederhof for her support. DOI 10.1108/APJML-04-2019-0240
APJML understanding consumer behaviour by focusing on the underlying neuronal mechanisms
32,5 that might play an important role in the perception and processing of marketing stimuli
(Alvino et al., 2018; Fortunato et al., 2014). Consumer neuroscience also studies how
consumers make decisions with regard to these stimuli (Alvino et al., 2018; Fortunato et al.,
2014; Lee et al., 2006; Sung and Phau, 2018).
The use of both psychological and neuroscience methods enables to study the
1150 decision-making process, which ultimately may lead to buying a product, at different levels.
Psychology helps to decompose the various factors playing a role in human decision making
as it studies human cognitive functions (e.g. verbal communication, attention, memory) and
emotions (e.g. happiness, sadness, positive vs negative, intense vs moderate), which
influence the decision-making process and buying behaviour. Neuroscientific methods can
help to link cognitive and emotional aspects with physiological and underlying neuronal
processes; to study how cognitive functions are accomplished in the brain and which brain
areas play a crucial role in these processes. Thus, neuroscience can support psychological
research investigating how cognitive processes (e.g. perception, memory, attention,
consciousness) originate in the brain and identify the brain areas involved in the explication
of cognitive functions.
Overall, consumer neuroscience contributes to a systematic understanding of consumer
behaviour and decision-making processes thereby improving existing marketing theories.
However, studies in this field come with some limitations. Consumer neuroscience
experiments seem to extremely simplify the complexity of the decision process (Heit, 2015;
Koschate-Fischer and Schandelmeier, 2014; Lee and Chamberlain, 2018) because researchers
do not consider the interaction between the participants and the stimuli. In fact, studies in
this field rarely investigate the neural mechanisms and responses involved in product
experiences. In real-life purchase experiences (except for on-line purchase), consumers can
interact with the product. For instance, consumers can touch, smell or taste the products.
Too frequently, in consumer neuroscience experiments, stimuli (e.g. potential products) are
only digitally presented (e.g. digital pictures, videos, websites). Thus, there is no real
interaction between the consumer and the product. This seems to be a major reduction in the
factors that influence consumer behaviour. The question may be raised whether it is
possible to determine the neural mechanisms and responses involved in product experience.
To resolve these issues, we need to investigate whether it is possible to measure changes
in brain activity during product experience but also how these changes can be related to
individual preferences for a product. Studying neural mechanisms involved in product
experience can help researchers in different ways. First, it can help researchers to achieve a
better understanding of the neuronal processes underlying consumer behaviour, without
explicitly asking participants to reflect on such internal mental states. Second, it can provide
added value to the prediction of consumer preferences in relations to different intrinsic
(e.g. taste) and extrinsic product characteristics (e.g. label, price). Finally, investigating
changes in brain activity related to product experience is important in order to achieve
maximal generalisability of consumer neuroscience research. In fact, it can help researchers
to create accurate models of consumer behaviour that can be applied to real-world situation,
for instance in designing and implementing marketing strategies in the Fast Moving
Consumer Good (FMCG) sector.
This paper also aims to investigate whether electroencephalography (EEG) may be a
suitable tool to study neural processes during product experiences.
First, EEG is relatively inexpensive compared to other neuroimaging tools, for instance
MRI or fMRI (Boksem and Smidts, 2015; Freeman and Quiroga, 2012). The cost of a
complete EEG setup is approximately 5 per cent of that of an fMRI machine ($1.3 to $2.6m)
(Hammou et al., 2013). The costs per-hour of EEG are also much lower compared to other
techniques (Miljkovic and Alcakovic, 2010).
 Second, EEG equipment is nowadays small and easily manoeuvrable. Usually, an EEG Consumer
system consists of only two computers (control and acquisition), two monitor screens and a behaviour and
cap (Cheryl and Kieffaber, 2014). Hence, EEG equipment can be moved or used more easily brain activity
to recreate real-life conditions (e.g. shop, restaurant) than scan machines such as fMRI.
An important benefit of using EEG is that it allows for relatively natural testing conditions.
In fMRI studies, participants usually lie in a narrow tube. In EEG studies, participants can
sit in a comfortable chair and perform small movements (Boksem and Smidts, 2015). Unlike 1151
fMRI, participants can really interact with the product, for instance touching the product or
independently hold a glass. Finally, EEG has more favourable testing conditions due to its
low noise level compared to other tools. Using fMRI implies very noisy conditions that can
have a distort impact on measuring and analysing the brain activity. Overall, EEG might
be a more suitable tool to study neural processes during product experiences. No study to
date has related EEG measures to individual preferences for wines during product
experience. Hence, the question may be raised whether EEG is a valid instrument to assess
individual preference during product experience.
Based on the aforementioned considerations, we used a literature review and qualitative/
quantitative research methods to study the neural mechanisms underlying individual
preference during a real product experience. The literature review provides an overview of
consumer neuroscience studies on product experience different tools used to measure brain
activity during product experience.
We also used qualitative research methods to study changes in brain activity related to
individual preferences during a real product experience. An experiment was carried out in
order to measure individual preferences for wines and changes in the brain activity of
participants during the product experience (wine tasting) using EEG. We performed a
conceptual replication of Boksem and Smidts (2015), in order to examine the replicability of
consumer neuroscience studies on product experience.
The paper is organised as follows: Section 2 describes the literature review and the
research questions. In Section 3, the employed methodology is described and in the fourth
section the results are presented. In the final section, the conclusions and limitations of the
employed study are discussed.

2. Background and research questions

2.1 Literature review
Consumer neuroscience research examines human neural mechanisms and cognitive
functions to verify and improve existing marketing theories on consumer decision-making
and behaviour. In particular, consumer neuroscience tools could be useful to identify and
analyse the neural mechanisms involved in product experiences. Product experience relies
on a set of brain mechanisms, psychological processes (e.g. expectation, valence) and
subjective factors (e.g. culture, economic status) (Plassman et al., 2008; Ramsøy, 2014).
Several methods have been used in marketing to study consumers’ preferences during
product experience (e.g. surveys, questionnaires, simulated choice experiments). However,
the combination of neural data and traditional methods can provide unique added value to
subjective factors, psychological and cognitive processes that drive consumer behaviour
during product experience. Relating brain activity to choice and preferences can also be
useful to predict consumer behaviour (Boksem and Smidts, 2015; Ohme et al., 2010).
Hitherto, the number of consumer neuroscience studies that examine consumer behaviour
and product experience is still limited.
We employed a literature review to provide an overview of consumer neuroscience
studies that investigate brain and body responses during product experience. We
considered only studies which focus on beverage products (e.g. wine, soft drinks, coffee) for
two reasons. First, in this paper, individual brain responses were measured during a wine
APJML tasting experience. As for any other drinking products (e.g. water, beer), wine quality and
32,5 preferences can be assessed only during consumption, hence during product experience.
Thus, it is useful to understand whether brain responses and preferences are assessed in
the same way for different beverage products. Second, it is important to identify studies
that use different approaches (e.g. fMRI, EEG, Emotiv) in order to measure brain and
psychophysiological responses during product experience. Based on the findings, it is
1152 possible to determine which tool is more reliable to measure consumer preference during
product experience.
For the literature review, we selected only peer-reviewed academic papers and books.
Papers published in journals were selected from two different platforms: Scopus and Science
Direct. We used the following keywords to select the papers: neuromarketing, consumer
neuroscience, decision-making process, product experience, food, beverage and wine. We
identified six most relevant studies that use neuroimaging tools to study brain and body
responses to product experience for beverage products.
fMRI, is currently one of the most popular brain imaging technique used in consumer
neuroscience (Ramsøy, 2014). fMRI detects changes in brain activity through the increasing
of the blood oxygenation level-dependent (BOLD), while a subject is lying a scanner. Due to
its high spatial resolution, fMRI allows to identify which brain areas appear to be activated
(Buxton, 2013). Thus, fMRI provides images of specific regions of the brain that “light up”
during a task or stimulus presentation, which implies that these areas are more active
(Gore, 2003). For instance, McClure et al. (2004) used fMRI to investigate how brand
knowledge during soft drinks delivery (Pepsi and Coca Cola) can influence an individual’s
preference, and brain activity in the hippocampus and the dorsolateral prefrontal cortex.
These are structures that are usually related to working memory as well as in the
acquisition and recall of declarative memories.
Similarly, Plassman et al. (2008) examined how increasing the price of a wine modifies
consumers’ preferences as well as brain activity in medial orbitofrontal cortex, an area that
is usually related to experienced pleasantness during experiential tasks. Results show that
increasing the price of a wine increases subjective reports of flavour pleasantness as well as
BOLD in the medial orbitofrontal cortex.
EEG measures the electrophysiological signals resulting from brain activity (Zhang et al.,
2014). There are several advantages in the use of EEG as a tool for studying neurocognitive
processes. First, EEG has a high temporal resolution (Abhang et al., 2016; Cohen, 2014;
Lakshmi et al., 2014; Michel and Murray, 2012). It means that EEG can capture the dynamics
of brain processes in the time frame in which they occur (Cohen, 2014; Freeman and Quiroga,
2012). Moreover, cognitive, perceptual, linguistic, emotional and motor processes are fast
(Cohen, 2014). In fact, most cognitive processes occur within tens to hundreds of milliseconds
(Cohen, 2014; Freeman and Quiroga, 2012). EEG is also well suited to capture these fast,
dynamic, time sequenced cognitive events (Cohen, 2014). Unlike fMRI, EEG has a low spatial
resolution (Michel and Murray, 2012; Ramsøy, 2014). Hence, EEG is not well suited for studies
in which precise functional localisation is important (Cohen, 2014). The EEG signal can be
described of varies oscillations that differ in their frequencies (Abhang et al., 2016). The main
frequencies of the human EEG have been classified in five frequencies bands: delta (0.5–4 Hz),
theta (4–8 Hz), alpha (8–13 Hz), beta (13–30 Hz) and gamma (31–100 Hz) (Abhang et al., 2016;
Rahman et al., 2015). Oscillations of different frequencies are thought to reflect global state
changes of the brain (Engel and Fries, 2010). In particular, theta and beta oscillations in frontal
cortical areas have been related to the reward processing system. For instance, Lucchiari and
Pravettoni (2012) used EEG to study the influence of mineral water brands on brain activity in
the theta and beta bands, and their examined preferences. Their study suggests that power in
the beta band increases for known and preferred brands. Instead, tasting unknown water
brands (considered low-quality and subjective bad tasting) was associated to an increase of
theta activity, similar to the one observed during processing of negative feedback. Yucel et al. Consumer
(2015) also studied the influence of coffee brands on consumer preference for taste and brain behaviour and
activity in theta and alpha bands. In contrast to the results of Lucchiari and Pravettoni (2012), brain activity
theta band activity was related to subjects’ preferences for a coffee brand. In fact, the intake of
the preferred coffee brand had a relaxing effect on subjects. Instead, tasting the least preferred
coffee generated stress and an increase in alpha activity, as according to the authors alpha
band has been used as a neurophysiological measure of activation. 1153
Emotiv cap is a wearable headset that allows to record emotions and brain activity.
Emotiv performance metrics report real-time changes in the subjective emotions
experienced by the user (engagement, frustration, meditation, excitement (instantaneous
and long term) (Russo, 2015). Russo (2015) used Emotiv cap to measure differences in the
level of frustration between wine experts and non-experts, during a wine tasting experience.
Likewise, a more recent study conducted by Horska et al. (2016) measured participants’
emotional responses, using Emotiv cap (e.g. frustration, engagement, meditation) and facial
expression recognition (e.g. happiness, sadness, disgust), during the tasting of eight white
wines. However, both studies did not link the participants’ emotional responses (frustration,
engagement and meditation) with specific brain waves. It is not possible to determine
whether the participants’ emotional responses can be associated with activity in the alpha,
beta or theta bands like in experiments that use a traditional EEG cap.

2.2 Research questions

We present an experiment that was designed in an attempt to recreate a real wine tasting
experience. Using EEG, the participants’ brain activities in the beta band were recorded
during wine tasting, and behavioural responses were measured during the process. Beta band
(12–30 Hz) is usually linked to cortical excitability during sensory and motor tasks (Hasler
et al., 2016; Engel and Fries, 2010). Beta band is also associated with reward processing and
consumer preferences for a product (Boksem and Smidts, 2015; HajiHosseini et al., 2012). For
instance, beta-band activity has been observed to increase 200–400 ms after positive feedback
informing about monetary gains using EEG (HajiHosseini et al., 2012). Beta-band activity
seems to be modulated by the experience of pleasure associated with a favourite brand
(Lucchiari and Pravettoni, 2012; Boksem and Smidts, 2015). In particular, beta-band
oscillations originating from the frontal cortex (FC) were examined. Several studies indicate
that the FC is anatomically and functionally connected to structures that are related to
positive and negative emotions (Davidson and Irwin, 1999; Maglione et al., 2017). In particular,
beta-band oscillations (12–30 Hz) originating from frontal regions have been associated with
reward and pleasure (Boksem and Smidts, 2015; HajiHosseini et al., 2012; Khushaba et al.,
2013; Lucchiari and Pravettoni, 2012; Rakshit and Lahiri, 2016; Vecchiato et al., 2013). Based
on these findings, a conceptual replication of Boksem and Smidts (2015) was performed to
examine whether or not frontal beta-band oscillations are related to individual wine
preferences, and whether activity in the beta band adds predictive power to self-reported
preferences. Thus, the present study tries to answer the following research questions:
RQ1. Do wines influence participants’ preferences and brain activities (beta band)
differently?
RQ2. Is preference for a wine related to increased beta-band activity?
Research in the field of consumer neuroscience also examines the effects of extrinsic cues on
quality assessment and individual preferences for a product. These studies have focused on
the positive or negative effect of these attributes on buying behaviour. In particular, studies
show that extrinsic cues can influence people’s product experiences (e.g. tasting, smelling) and
preferences (Balconi et al., 2014; Park et al., 2015; Plassman et al., 2008; Plassmann and Weber,
APJML 2015; da Rocha et al., 2013). The controversial nature of wine quality makes it difficult to
32,5 assess, especially for no expert wine drinkers. Thus, it is important to study the influence of
external attributes on quality perception during the wine tasting or the selection process.
Marketing studies analysed how price (Chen and McCluskey, 2016; Mueller and Szolnoki,
2010b; Goldstein et al., 2008) and label (e.g. colour, design) (Barwich, 2017; Mueller and
Szolnoki, 2010a; Szolnoki et al., 2008) influence consumer’s choice and preferences. Similarly,
1154 we tested whether extrinsic cues such as label and price can have an influence on consumer
preferences and beta band oscillations associated with product experience. In the current
study, all participants took part in a “blind taste session”, in which information about the wine
was not disclosed, and a normal taste session, in which the information about the wine was
disclosed before the test. The following research questions are tested in this study:
RQ3. Do participants have strong preferences for more expensive wines?
RQ4. Do labels influence participants’ preferences and brain activity (beta band)?
Finally, marketing studies show that people’s preferences can influence their price
perception. (Chen and McCluskey, 2016; Goldstein et al., 2008; Hollebeek et al., 2007; Mueller
and Szolnoki, 2010a). The term usually refers to customer’s subjective perception of what is
given up or sacrificed to acquire the product (Zeithaml, 1988; Shintaputri and Wuisan, 2017).
Since, customers do not always remember or know the actual price of a product, they
subjectively encode the prices as “expensive” or “cheap”. Thus, it suggests that customer’s
perception of price may be different from the objective price (Zeithaml, 1988; Shintaputri
and Wuisan, 2017). The perceived price of a product is also influenced by consumer
preferences. In fact, a positive preference towards a product is known to be positively
correlated with perceived price (Rao and Monroe, 1988; Plassman et al., 2008). When the
information about price is not disclosed, the individual is more likely to believe that a
product that he/she prefers is more expensive than a product that he/she does not like.
Hence, individual preferences for a wine can influence the participants’ price perception of
wines. The final research question is tested in this study:
RQ5. Does the preference for a wine influence the participant’s perception of price?
Participants selected were not expert wine drinkers. Hence, this paper investigates brain
responses and preferences of non-expert wine consumers.

3. Methods
3.1 Participants
In total, 31 participants, all non-paid volunteers, were recruited among students or
employees of the university. All volunteers were asked to participate in two sessions. The
local ethics committee of the Faculty of Behaviour, Management and Social sciences
approved the procedures employed, which were all in line with the declaration of Helsinki.
Participants had no history of neurological illness or damage, were not using drugs or
psychiatric medication, and had normal or corrected-to-normal vision and had no
colour-blindness. Non-smoker volunteers were preferred; smokers were required not to
smoke 12 hours before the experiment. Participants were also instructed to abstain from
alcohol and caffeine-containing substances 12 hours before the experiment. A questionnaire
and Alcohol Use Disorders Identification Test (AUDIT) (sent by e-mail) were used to decide
whether volunteers should be allowed to participate in the experiment. According to
the World Health Organization (WHO) (2001) the AUDIT is a simple method of screening for
excessive drinking behaviours. Participants with a score higher than 19 in the AUDIT were
excluded as they can be considered to display hazardous (or risky) drinking behaviour,
harmful drinking or alcohol dependence.
 Five participants were excluded for different reasons. For two participants, a different Consumer
amplifier was used in the first and the second session, due to EEG equipment failure, which behaviour and
makes a comparison between sessions difficult. One participant was not able to take part in brain activity
the second session. Two other participants were excluded because of excessive artifacts in
their EEG recordings. The final sample consisted of 26 participants (16 male and 10 female)
between 18 and 40 years of age ( female: mage ¼ 27.3, SD ¼ 4.6, ranging from 23 to 39 years;
male: mage ¼ 26.2, SD ¼ 3.4, ranging from 19 to 33 years) that participated in both sessions. 1155
Participants originated from 17 different countries (Austria, Belarus, China, Cuba, France,
Germany, Great Britain, Honduras, India, Italy, Lithuania, Mexico, the Netherlands,
Pakistan, Poland, Spain and Turkey).
In the questionnaire, four categories were used to classify wine knowledge of the
participants: amateur, basic knowledge, expert or professional. Participants knowledge of
wine was based on self-reported measure, as only participants with no knowledge or little
knowledge of wine were selected. For the remaining 26 participants who took place in
the experiment, the results showed that 12 participants could be considered as amateur,
while 14 participants displayed basic knowledge (Figure 1).

3.2 Procedure
Once participants arrived at the laboratory, they were asked to sign an informed consent form.
Subsequently, participants received both written and verbal detailed instructions on the tasks
they were going to perform in the experiment and a small guideline on wine tasting.
Participants were invited to sit in a comfortable chair, at a distance of approximately
100 cm at the eye level in front of a 24-inch AOC G2460P LED computer screen. The room
was sound-attenuated and illuminated. EEG electrodes were applied on the participants’
scalps. Participants were instructed to relax and reduce blinking. Participants were also
asked to minimise sudden movements while executing the tasks and follow the instructions
displayed on the computer screen.

3.3 Task
During the experiment, participants tasted and judged four different wines. The procedure
consisted of four steps that were repeated for each wine[1]:
(1) volunteers started with rinsing their mouth with water after which they had to wait
for 10 seconds;
(2) a glass containing red wine was presented in front of the participant for 20 seconds;
(3) volunteers had to smell the wine twice; first with a stationary glass and second after
swirling the wine in the glass for three seconds; and

Step 1 Step 2 Step 3 Step 4

Water Wine Wine Wine
Rinse your mouth Appreciate the color Smell the wine Taste it

Figure 1.
5 seconds 20 seconds 15 seconds 5 seconds Steps of the wine
tasting procedure
Note: Participants repeated these steps for each wine
APJML (4) volunteers subsequently tasted the wine by taking a small sip, and swirling the wine
32,5 in their mouth, to appreciate the full taste.
After these four steps, participants were asked to rate each wine. The wine tasting task was
performed in two sessions, which were separated by two weeks (see Subsection 3.8).
A time frame of two weeks was chosen in order to reduce the possibility that volunteers
would remember the wines. In both sessions, the participants were also asked to give an
1156 overall rating of the wine (wine preferences), and to indicate in which price category the
wine should be located. In the session with the labels, the volunteers were additionally asked
to rate the labels.
The participants were randomly assigned to one of the two conditions for the first
session (order of sessions). As shown in Figure 2, the order of sessions (no label/label;
label/no label) was counterbalanced between participants.

3.4 Materials and stimuli

Wines were selected based on the type of grape, price and country of origin. In
order to reduce the number of variables, the wines selected had the same type of grape
(Cabernet Sauvignon, 100 per cent).
The wines were selected from two different price ranges, two wines were high-priced
wines (price category: €24–€27), while the two other wines were low-priced wines (price
category: €3–€5). In order to have a realistic evaluation of the wines, the prices were
compared on the same website[2]. All the wines can easily be bought online.
In order to have a realistic wine testing experience, the wines were selected from two
very popular countries for wines: Italy and Chile. In total, two wines, referred to as Chilean
Expensive (CE) and the Italian Expensive (IE) were high-priced wines, while the two other
wines, named Chilean Cheap (CC) and the Italian Cheap (IC) were low-priced wines. The
order in which wines were presented was counterbalanced across participants. The order
was changed according to the nationality (Italian, Chilean) and the price (Cheap, Expensive).

3.5 Behavioural responses

During the task, self-reported drinking behaviours, wine and label preferences and
perceived price of each participant were collected:
• Drinking behaviours: participants were asked to complete a questionnaire and to
indicate their age, nationality, gender, drinking habits, and wine knowledge.
• Wine preferences: the volunteers were asked to give an overall rating of the wine,
according to their preferences (six-points Likert scale, the greater the preference the
higher the value).
• Perceived price: the volunteer chose the corresponding price category of each wine
(€3–€5, €6–€8, €9–€11, €12–€15, €16–€21, €22–€27).

Group 1 Group 2

No Label Label

Figure 2. 2 weeks
The figure shows
the groups and
order of sessions Label No Label
 • Label preferences: for the label condition only, the volunteers rated the labels of the Consumer
wines (six-points Likert scale, the greater the preference the higher the value). behaviour and
brain activity
3.6 Electroencephalographic (EEG) measures
The EEG was recorded continuously from 32 active Ag/AgCl electrode sites using an
EasyCap-62 channel cap (standard international 10–20 system layout) connected to an
ActiChamp amplifier and BrainVision Recorder software (version 1.21.0102). 1157
The electrodes were located at the following sites: AFz, AF3, AF4, AF7, AF8, F1, F2, F5,
F6, FCz, FC3, FC4, FT7, FT8, C3, C4, C5, C6, CPz, CP3, CP4, TP7, TP8, P1, P2, P5, P6, POz,
PO3, PO4, PO7 and PO8 (see Figure 3). The horizontal and vertical electro-oculogram (hEOG
and vEOG) were recorded. Two electrodes were placed at the outer canthi of both eyes to
measure the electrical activity generated by horizontal eye movements. Electrodes located
on the infraorbital and supraorbital regions of the left eye placed in line with the pupil
enabled to measure vertical eye movements and blinks. The resistance of the electrodes was
kept below 10 kΩ by using electrode gel and standard procedures to improve conductivity.
Presentation® software (Neurobehavioral Systems, 2012, www.neurobs.com/menu_
presentation/menu_features/features_overview) installed on a separate computer (DELL),
was used to present instructions to the participants and send appropriate markers
signalling relevant events to be picked up by BrainVision Recorder software (Version 1.21).
A QWERTY keyboard was used where the keys 1, 2, 3, 4, 5, 6, on the top
left, and the space bar registered answers and controlled the sequence of events during
the experiment.

Fpz
Fp1 Fp2

AF7 AF8
AF3 AFz AF4

F7 F8
F5 F3 F4 F6
F1 Fz F2
FT
10
FT7 FT8
FC5 FC3 FC1 FC2 FC4 FC6
FC2

T7 C5 C3 C1 Cz C2 C4 C6 T8

CP3 CP1 CP2 CP2 CP4

CP5 CP6
TP7 TP8
TP
TPg 10
P3 P1 Pz P2 P4
P5 P6
P7 P8

PO3 POz PO4

PO7 PO8

O1 O2
Oz
Figure 3.
The figure shows
the position of the
electrodes employed
Note: The electrodes were located at the red encircled positions, as in this study
described in the 10-20 system for electrodes placement
APJML 3.7 EEG data analysis
32,5 EEG data were processed using the BrainVision Analyzer software (Version 2.1.1).
First, an interpolation method was used to check the sufficient quality of the signal
acquired from the electrodes. This method allows to identify noisy channels and to replace
them by two or three existing channels (located close to the noisy channel). Noisy channels
were identified and corrected for 12 participants.
1158 Second, the continuous data were divided into four segments (one for each wine), each
segment (0–40,000 ms) started at the beginning of each rinsing condition (water) and lasted
the duration of the whole procedure.
Each segment (wine) was subsequently separated into four segments of 19,500 ms
(2,500–22,000 ms) for the colour (visual), smell (smelling), taste (tasting) and rinsing
condition (water tasting). Each segment (colour, smell, taste and water tasting) was in turn
partitioned in sized segments of 5,000 ms. Then, the baseline of every segment was adjusted
(0 to 100 ms). Artifacts were detected and removed for each segment. Next, Fast Fourier
transformations (FFTs) analysis was performed on tasting condition (wine) and rinsing
condition (water) data for each wine, using a standard Hanning window. A log10
transformation was applied[3] on the preprocessed data for wine. The resulting spectral
EEG data per wine and water were averaged for all participants individually.
In order to obtain a baseline, a subtraction was performed between tasting (wine) and
rinsing (water) log10 transformed data. Finally, data were imported to be used with SPSS.

3.8 Statistical analysis

Three different statistical analyses were used to test our hypotheses, which included
repeated measures ANOVA, Friedman tests and Correlational analyses.
A repeated measures ANOVA[4] was used to analyse changes in the beta bands (log10
transformed data) between different conditions, such as different types of wine, electrodes,
frequency bands and sessions (label, no label). The repeated measures ANOVA was used to
establish the relation between the dependent variable (beta-band power 12–30 Hz) and the
multiple independent variables (wine, electrodes, frequencies bands and sessions). Associated
degree of freedom, F-values, p-values, means and Partial η2 were reported (see Tables III and IV).
Repeated measures ANOVA are particularly susceptible to the violation of the assumption of
sphericity[5], thus associated Mauchly’s Test of sphericity was also used to test if the
assumption of sphericity was violated (Field, 2013). Corrected results (Greenhouse–Geisser or
Huynh–Feldt corrections)[6] were reported only when the sphericity assumption was violated.
In order to perform the statistical analysis:
• Six different electrodes located over the frontal and central cortex were selected: AFz,
F1, F2, FCz, FC3, FC4. The electrodes were chosen on the basis of data analysis used
in other studies (Boksem and Smidts, 2015; Khushaba et al., 2013; Lucchiari and
Pravettoni, 2012; Rakshit and Lahiri, 2016).
• The beta band (12–30 Hz) was separated in three different frequency ranges: low-beta
band (12–16 Hz), mid-beta band (16–20 Hz) and high-beta band (20–30 Hz). For
comparable procedures see Spironelli et al. (2013), Engel and Fries (2010), Abhang
et al. (2016) and Boksem and Smidts (2015).
• Two different sessions were analysed: label and no label. In the label session, the
wines were presented with the label. In the no label session, the wines were presented
without any label.
• Participants were divided in two groups according to the order of sessions attended
(group 1 ¼ no label-label; Group 2 ¼ label-no label). The order of sessions was used as
between-subject variable in the repeated measures ANOVA.
For the behavioural data, a Friedman test[7] was used to analyse changes in participants Consumer
responses for wines rating (wine preferences), price perception (perceived price) and behaviour and
label rating (label preferences) for the different wines and sessions (label/no label). The brain activity
Friedman-test was used to test whether the different wines and sessions (label/no label)
have an effect on the dependent variables wine preferences, perceived price label
preferences. A Wilcoxon signed-rank test[8] was used as post hoc test to analyse differences
in the different combinations of related groups (wines and sessions). Associated degree of 1159
freedom, χ2[9], Z-scores and p-values, were reported.
Finally, a correlation analysis[10] was performed on the behavioural data in order to
determine the relationship between wine preferences and perceived price for all the wines in
the two different sessions (label/no label). A correlation analysis was used to test whether the
direction of the relationship is positive (wine preferences and perceived price are positively
correlated) or negative (wine preferences and perceived price are negatively correlated).

4. Results
This section describes the results obtained for both Behavioural and EEG data.

4.1 Behavioural results

The results show that there was a statistically significant difference for Wine preferences
(six-points Likert scale, the greater the preferences the higher the value) among wines and
sessions (w~ 2 (7) ¼ 73.444, p o0.0001). Next, post hoc analysis with Wilcoxon signed-rank
tests was conducted with a Bonferroni correction applied ( p o0.0031) in order to determine
where there were significant differences among the wines and conditions. As presented in
Table I, the results show that there was a significant difference in participants’ preferences
(Wine preferences) for the IE wine between the two sessions (Z ¼ −4.492, p o0.0001).
Similarly, there was a significant difference between IE and CE (Z ¼ −4.527, p o0.0001), IE
and CC (Z ¼ −4.508, p o0.0001), CE and IC (Z ¼ −4.527, p o0.0001) in the no label session.
As shown in Table I, no significant difference was observed between participants’
preferences for the other wines in the different sessions.
Next, the results show that there was no significant difference in perceived price
responses (six-points Likert scale, the greater the preferences the higher the value)

Wine 1 Session Wine 2 Session Z p-value

Chilean Cheap Label Italian Expensive Label −0.336 0.737

Chilean Expensive Label Chilean Cheap Label −0.769 0.442
Chilean Expensive Label Italian Cheap Label −1.536 0.125
Italian Cheap Label Chilean Cheap Label −0.926 0.355
Italian Cheap Label Italian Expensive Label −0.580 0.562
Italian Expensive Label Chilean Expensive Label −1.211 0.226
Chilean Cheap No Label Chilean Expensive No Label −0.538 0.59
Chilean Cheap No Label Italian Expensive No Label −4.508 0.0001
Chilean Expensive No Label Italian Cheap No Label −1.186 0.235
Italian Cheap No Label Chilean Cheap No Label −0.532 0.594
Italian Expensive No Label Italian Cheap No Label −4.527 0.0001
Italian Expensive No Label Chilean Expensive No Label −4.527 0.0001
Table I.
Chilean Cheap No Label Chilean Cheap Label −0.613 0.54 Comparison between
Chilean Expensive Label Chilean Expensive No Label −1.292 0.196 the Wine preferences
Italian Cheap No Label Italian Cheap Label −0.371 0.71 for the four wines in
Italian Expensive Label Italian Expensive No Label −4.492 0.0001 the two sessions
Note: Significant differences between the wines are displayed in italic (Label/No Label)
APJML depending on wines and sessions (w~ 2 (6) ¼ 11.189, p o0.083). Thus, post hoc analysis with
32,5 Wilcoxon signed-rank test was not performed.
There was a significant difference in label preferences (six-points Likert scale, the greater
the preferences the higher the value) depending on wines and sessions (w~ 2 (3) ¼ 13.396,
p o0.004). Post hoc analysis with Wilcoxon signed-rank tests was conducted with a
Bonferroni correction applied ( p o0.0083) in order to determine if there was significant
1160 difference among the wines and the sessions for the Label preferences. The results show
that there was a significant difference in the participants’ responses between the IE label
and the IC label (Z ¼ −3.128, p o0.002). No significant difference was observed comparing
the other labels: IE and CE (Z ¼ −2.034, p o0.042); IE and IC (Z ¼ −0.202, p o0.840); CE
and CC (Z ¼ −1.479, p o0.139); CE and IC (Z ¼ −2.137, p o0.033); CC and IC (Z ¼ −1.263,
p o0.207). Although the difference between CE and IC shows the expected pattern, but it
did not survive the correction for different testing.
Finally, behavioural data were also used to determine whether there was a positive or
negative relationship between Wine preference and Perceived price. The Pearson correlation
coefficient was obtained for different data conditions: all the participants, different wines
and Label/No Label conditions. As shown in Table II, the analysis shows that there was a
positive correlation between the participants’ Wine preferences and Perceived price. In fact,
Wine preference can be explained for 55 per cent by the assessment of price.
The correlation coefficient suggests also a moderate uphill relationship between the
preference and the price for both conditions. However, the correlation coefficient was
slightly higher for the label session (0.572) than the no label session (0.54). As shown in
Table II, the data also show a positive relationship between preferences and price for all the
wines (CE:0.417; IE:0.616; CC:0.621; IC:0.49) in both sessions, no label session (CE:0.347;
IE:0.634; CC:0.694; IC:0.348) and Label session (CE:0.466; IE:0.59; CC:0.573; IC:0.615).

4.2 EEG results

A repeated measures ANOVA was performed to examine changes in the beta-band activity
(log10 transformed data) on the base of different conditions such as electrodes, frequency
ranges, sessions, wines as within-subject factors.
First, as described in Table III, the results show that the different electrodes
(F(2.3, 3.39) W0.66, p o0.536), frequency ranges (F(2, 1.24) W0.54, p o0.585) and sessions
(F(1, 1.44) W0.22, p o0.641) did not have an effect on beta-band activity of the participants.
However, the results suggest that tasting different wines affect beta-band activity

Data condition Variable 1 Variable 2 Correlation coefficient

All data Wine preference Perceived price 0.552

No Label Wine preference Perceived price 0.54
Label Wine preference Perceived price 0.572
Chilean Expensive Wine preference Perceived price 0.417
Italian Expensive Wine preference Perceived price 0.616
Chilean Cheap Wine preference Perceived price 0.621
ItalianCheap Wine preference Perceived price 0.49
No Label-Chilean Expensive Wine preference Perceived price 0.347
Table II. Label-Chilean Expensive Wine preference Perceived price 0.466
Correlation between No Label-Italian Expensive Wine preference Perceived price 0.634
wine preferences Label-Italian Expensive Wine preference Perceived price 0.59
for each wine and No Label-Chilean Cheap Wine preference Perceived price 0.694
perceived price in Label-Chilean Cheap Wine preference Perceived price 0.573
both sessions No Label-Italian Cheap Wine preference Perceived price 0.348
(label/no label) Label-Italian Cheap Wine preference Perceived price 0.615
(F(2.5, 13.18) W2.9, po0.049). In particular, it was examined if tasting different wines is Consumer
related to an increase or decrease of beta-band activity. behaviour and
A comparison between the different wines shows that there was a substantial decrease in brain activity
the beta-band activity for CE (see Tables IV and V ). Specifically, there was a significant
effect of the CE compared to the IE (F(1, 58.04) W 5.635, p o0.026),the CC (F(1, 12.17) W5.61,
p o0.026) and the IC (F(1, 37.43) W 4.532, p o0.044). However, no significant differences
were found between the IE and the CC (F(1, 17.05) W1.99, p o0.171) or IC (F(1, 2.24) W0.29, 1161
p o0.592). Similarly, no difference was observed between CC and IC (F(1, 6.91) W0.79,
p o0.592).
Finally, the interaction effect between all the conditions was tested. The results show
that there was no significant interaction effect between all the factors.

4.3 Discussion
This subsection summarises the results presented in subsections 4.1 and 4.2. The present
experiment aims at investigating whether neural measures can contribute to the assessment
of consumers’ preferences during product experience (wine tasting). In particular, the
purpose of this experiment was to examine whether beta-band activity generated by the FC
(often associated with reward and pleasure) can be related to participants’ preferences for
wines. Beta-band activity in the FC was obtained from a sample of 26 volunteers, during the
tasting of four different wines. During the experiment, also self-reported preferences
(behavioural data) for wines, labels and perceived price were measured.

Conditions Correction df Mean square F p-value η2 Table III.

Results of the
Electrodes Greenhouse–Geisser 2.3 3.39 0.669 0.536 0.027 repeated measures
Frequency Sphericity assumed 2 1.24 0.542 0.585 0.022 ANOVA analysis for
Wine Huynh–Feldt 2.5 13.18 2.93 0.049 0.109 the different
Label Huynh–Feldt 1 1.44 0.22 0.641 0.009 conditions

Wine Wine df Mean Square F p-value η2

Chilean Expensive Chilean Cheap 1 12.17 5.61 0.026 0.189

Chilean Expensive Italian Cheap 1 37.43 4.532 0.044 0.159 Table IV.
Chilean Expensive Italian Expensive 1 58.04 5.635 0.026 0.19 The comparison
Italian Expensive Chilean Cheap 1 17.05 1.99 0.171 0.077 between the different
Italian Expensive Italian Cheap 1 2.24 0.295 0.592 0.012 wines with respect to
Italian Cheap Chilean Cheap 1 6.915 0.791 0.383 0.032 beta-band activity

Wine Mean SE

Chilean Expensive −0.247 0.072

Italian Expensive −0.133 0.065
Chilean Cheap 0.002 0.085
Table V.
Italian Cheap −0.047 0.08 Mean of beta band
Notes: The Chilean Expensive wine and the two Italian wines have negatives means (decrease in beta band for each of the four
activity). For the Chilean Cheap wine the mean is positive (increase in beta band activity). Please notice that wines that the
these are log transformed data participants tasted
APJML We investigated whether tasting different wines influences participants’ preferences and
32,5 brain activity (beta bands) (RQ1). The results show that there was a significant difference in
the participants’ brain activity after the tasting of the four different wines. In particular,
frontal beta-band oscillations in three different frequency ranges (12–16, 16–20, 20–30 Hz)
and for six different electrodes (AFz, F1, F2, FCz, FC3, FC4) were different among wines.
Participants self-reported data also confirmed that there was a significant difference in
1162 participants’ preferences among wines. Hence, it is possible to confirm that tasting different
wines influenced both preferences and brain activity. Furthermore, participants were able to
identify the differences among the four wines, tested as well as to report their preferences
for the wines. In addition, EEG data showed that beta-band activity for the CE wine
decreased compared to the other wines. Thus, these findings also suggest that differences in
the participants’ preferences for the wines could be related to changes in the participants’
brain activities.
We also performed a conceptual replication of Boksem and Smidts (2015). Hence, we
tested whether preference for a specific wine is related to increase in the beta-band activity
(RQ2). Boksem and Smidts (2015) show that the higher participants ranked a movie, the
higher the amplitude of beta-band oscillations. In the present study, results show that there
was an opposite trend, higher preferences for a wine corresponded to a stronger decrease in
beta-band oscillations. In fact, a decrease in beta-band oscillation was observed for the most
preferred wine (CE), while a higher beta-band oscillation was observed for the least
preferred wines (IE, CC and IC). These results suggest that it is not possible to relate an
increase in beta-band activity to preferences for the wine product. However, these findings
might also suggest that the cognitive and neural mechanisms involved in movie evaluation
(Boksem and Smidts, 2015) are different from neural processes involved in product
experiences. For instance, watching a movie might involve cognitive and neuronal
processes, like visual attention that are different from those that are triggered during wine
tasting, which implies processing of olfactory and gustatory signals. The increased
activation of beta-band activity in the study of Boksem and Smidts (2015) might be related
to an arousal of the visual system during increased visual attention. Instead, wine tasting
does not imply a visual attentional process but processing of olfactory and gustatory
signals that might have a different impact on beta-bands oscillations. We conclude that the
findings of Boksem and Smidts (2015) do not apply for preferences for wine product it may
not be possible to validate the findings of for the wine product.
In this study, we also intended to test whether participants would prefer more expensive
wines (RQ3). Behavioural data show that participants preferred the expensive wines (CE
and IE) rather than cheaper wines (CC and IT). It is noticed that even though participants
were amateur or inexpert wine drinking, they were able to identify the quality difference
among the four wines. In fact, the wines selected had the same type of grape (Cabernet
Sauvignon), however, there was a wide difference in price range between the wines
(expensive: €24–€27, cheap: €3–€5), hence different quality. Participants were able to
indicate their preferences for the more expensive wines in a meaningful and consistent way
in both the blind taste session (no label session) and the normal taste session (label session).
This study also tested whether the visibility of the wine label might influence beta-band
oscillations associated with product experience (RQ4). Moreover, it was tested whether
individual preferences for a wine and a label influence participants’ price perception of
wines (RQ5). Regarding tRQ4 results show that there was no significant difference in the
beta-band oscillations between the label and no label sessions. Hence, no important effect of
the extrinsic cue label was observed for the neural measures (beta-band). Results suggest
that there were no changes in the beta-band activity when the label was presented.
It was also partially confirmed by the participants’ self-reported preferences. No significant
differences were observed in participants’ preferences among wines for the two
sessions, except for the Italian expensive wine. Regarding RQ5 behavioural results show Consumer
that the relationship between participants’ preferences and perceived price is positive. behaviour and
Some differences could be observed for CE and IE wines in both sessions. These differences brain activity
might indicate that there was a strong relationship between participants’ preferences for the
wines and the price that they attributed to wines (high price).

5. Conclusions and limitations 1163

The present study aims at examining whether it is possible to determine the neural
mechanisms and responses involved in product experience, EEG is a valid instrument to
assess consumer preferences during product experience. We used a literature review and
qualitative/quantitative research methods to define the current state of affairs in consumer
neuroscience research.
The literature review highlights that it is possible to measure consumer behaviour
during product experience using three different tools: fMRI, EEG and Emotiv cap. The most
cited and well-known studies were conducted using fMRI. Both studies found that the
medial orbitofrontal cortex and the orbitofrontal cortex (two adjacent brain areas) were
active for increased consumer preferences during beverage tasting.
We also found two studies (Lucchiari and Pravettoni, 2012; Yucel et al., 2015) that
analyse individual preferences for beverage during product experiences using EEG that
gave contrasting results. Both experiments try to determine whether individual preferences
for a brand could influence the brain activity during the tasting process. However, changes
in different brain waves (alpha and theta, beta and theta) were associated with the same
process, such as preferences for a brand during product tasting. We can conclude that
unlike studies conducted using fMRI, the study of brain activity using EEG seems based on
isolated attempts to relate EEG components to the study of individual preferences. Usually,
it is not clear why researchers link some brain wave frequency (e.g. theta bands) to a specific
cognitive process (e.g. reward mechanisms). However, EEG can still be considered a tool to
study brain activity during product experience. In fact, tools such as fMRI might limit the
product experience, due to significant duration of the test, unnatural position (horizontal),
limited movements (participants lie in a narrow tube) and noisy level (typically more than
90 dB). The adverse environmental characteristics (like in fMRI studies) might also have a
negative impact on cognition and choice. Thus, researchers might face serious distortions in
measuring and analysing and brain activity and behavioural data. On the other hand, EEG
provides a high temporal resolution, low costs ( for both the instrument and experiments),
small dimensions and it is more manoeuvrable compared to other tools. EEG, if compared to
fMRI, does not restrict the participant in a small space as it allows the participant to
accomplish small movements. Thus, EEG allows to recreate a more realistic and natural
experience for the consumer during the experiment.
For this reason, we performed a conceptual replication of the Boksem and Smidts (2015)
study. We investigated if beta-band activity can be related to consumers’ preferences during
a wine tasting experience. Based on our findings, we draw the conclusion that it is possible
to link consumer preference and brain activity during wine testing. However, the findings of
Boksem and Smidts (2015) study do not apply for the wine product. Watching a movie
might involve different cognitive and neuronal processes (e.g. visual attention) that are
different from those that arouse from wine tasting (e.g. processing of olfactory and
gustatory signals). It is also possible that dissimilar products (e.g. wine, movie) influence
consumers’ experience and neural processes in different ways. Overall, our experiment was
the first attempt to study consumer preferences for wine using EEG. Due to the limited
number of studies that use EEG to investigate consumer preferences during product
experience, it is difficult for both academicians and practitioners to understand which brain
oscillations could be related to individual preferences during product experience. Thus, the
APJML results of studies using EEG to examine consumer behaviour during product experience
32,5 cannot always be validated. For this reason, researchers can use the findings discussed in
this study to better understand brain mechanisms involved during product experience for
different consumer neuroscience tools. This study can also help researchers to determine
whether measuring brain activity in the beta-band could be useful to investigate consumer
preferences for a specific product (e.g. wine, coffee) or task (e.g. drinking). Finally,
1164 practitioners can use the finding presented in this study as guideline to design realistic and
natural conditions to study individual preferences during product experience.
In conclusion, the study of consumer behaviour during product experiences is still in a
fledgling stage, and current investigations have been mostly targeted to basic research.
The lack of a solid theoretical background makes it difficult to identify specific brain areas
or brain activity involved in product experience. One reason for the current state of affairs
may be that studies have utilised experimental designs, which, although useful for
establishing cause and effect relationships, tend to drastically limit the number of
variables that can be examined in combination. For that reason, to date, the findings
derived from the use of EEG tools in product evaluation must be carefully examined as
well as the use of EEG to predict consumer preferences. However, we do believe that
academicians should pay particular attention to the replication of consumer neuroscience
studies. Testing the conceptual replication of these studies can help academicians
investigating consumer preference during product experience for different product
(e.g. food, clothes) and sensory modalities (e.g. visual, olfactory), creating valid models
that can help the prediction of consumer behaviour.

5.1 Limitations and future work

Taken together, these results confirm certain previous findings and challenge others in
consumer neuroscience literature. However, the present study has a few research
constraints that can be addressed in future studies. For instance, it would be useful to
examine the effect of multiple external cues, such as price and brand, on consumers’
preferences. Future studies can also inquire into gender differences, regarding wine taste
and the reactions to product external cues, or differences between professional and no
expert wine consumers. Finally, it would be useful to combine several consumer
neuroscience tools to investigate consumers’ preference during product experience. For
instance, researchers could use eye-tracker, galvanic-skin conductance sensors or heart rate
monitor in combination with EEG. It would help researchers to have more accurate and
reliable measurements of the neural and physiological mechanisms underlying consumer
behaviour during product experience.

Notes
1. The procedure used refers to the FISAR procedure for wine tasting (Italian Federation of
Sommeliers Hoteliers Restaurateurs).
2. www.uvinum.co.uk
3. Log10 transformation allows to normalise the EEG data. For more details see Figures A1 and A2
in Appendix 1.
4. The repeated measures ANOVA are a parametric test that allows to compare the means across
one or more variables that are based on repeated observations. The repeated measures ANOVA
“allows to determine whether the means of three or more measures from the same person are
similar or different” (Plichta and Garzon, 2009). Repeated measures ANOVA can be used to
investigate either changes in mean scores over three or more time points or differences in mean
scores under three or more different conditions (Statistics.laerd.com, 2019). The independent
variable has categories called levels or related groups (Statistics.laerd.com, 2019).
 5. The effect of violating sphericity is “a loss of power (e.g. an increased probability of a Type II Consumer
error) and a test statistics (F-ratio) that simply cannot be compared to tabulated values of the behaviour and
F-distribution” (Field, 2013).
brain activity
6. In general, the recommendation is to use the Greenhouse–Geisser correction if εo0.75 and the
Huynd–Feldt correction if εW0.75.
7. The Friedman test is the non-parametric alternative to the repeated measures ANOVA (Statistics.
laerd.com, 2019). It is used to test differences between groups when the dependent variable is 1165
ordinal (Statistics.laerd.com, 2019).
8. The Wilcoxon signed-rank test is the nonparametric test equivalent to the dependent t-test
(Statistics.laerd.com, 2019).
9. w~ 2 allows to test the relationship between two categorical variables (Bluman, 2007).
10. Correlation determine the strength of a linear relationship between two variables. When there is
no correlation between two variables, then there is no tendency for the values of the variables to
increase or decrease in tandem (Bluman, 2007).

References
Abhang, P.A., Gawali, B.W. and Mehrotra, S.C. (2016), Introduction to EEG-and Speech-Based Emotion
Recognition, Academic Press, London.
Alvino, L., Constantinides, E. and Franco, M. (2018), “Towards a better understanding of consumer
behavior: marginal utility as a parameter in neuromarketing research”, International Journal of
Marketing Studies, Vol. 10 No. 1, pp. 90-106.
Balconi, M., Stumpo, B. and Leanza, F. (2014), “Advertising, brand and neuromarketing or how
consumer brain works”, Neuropsychological Trends, Vol. 16 No. 16, pp. 15-21.
Barwich, A.S. (2017), “Up the nose of the beholder? Aesthetic perception in olfaction as a
decision-making process”, New Ideas in Psychology, Vol. 47, pp. 157-165.
Bluman, A.G. (2007), Elementary Statistics (A Step by Step Approach), McGraw-Hill Companies,
New York, NY.
Boksem, M.A.S. and Smidts, A. (2015), “Brain responses to movie trailers predict individual preferences
for movies and their population-wide commercial success”, Journal of Marketing Research,
Vol. 52 No. 4, pp. 482-492.
Buxton, R.B. (2013), “The physics of functional magnetic resonance imaging (fMRI)”, Reports on
Progress in Physics, Vol. 76 No. 9, pp. 1-61.
Chen, K.J. and McCluskey, J.J. (2016), “Price impacts of tasting notes across wine segments”, 2016
Annual Meeting, Agricultural and Applied Economics Association, Boston, MA, 31 July–2
August 2016.
Cheryl, L.D. and Kieffaber, P.D. (2014), “EEG methods for the psychological sciences”, British Journal of
Psychology, Vol. 105 No. 3, pp. 438-439.
Cohen, M.X. (2014), Analyzing Neural Time Series Data: Theory and Practice, MIT Press, MA.
da Rocha, A., Rocha, F. and Arruda, L. (2013), “A neuromarketing study of consumer satisfaction”,
SSRN 2321787.
Davidson, R.J. and Irwin, W. (1999), “The functional neuroanatomy of emotion and affective style”,
Trends in Cognitive Sciences, Vol. 3 No. 1, pp. 11-21.
Engel, A.K. and Fries, P. (2010), “Beta-band oscillations-signalling the status quo?”, Current Opinion in
Neurobiology, Vol. 20 No. 2, pp. 156-165.
Field, A. (2013), Discovering Statistics Using IBM SPSS Statistics, Sage, London.
Fortunato, V.C.R., Giraldi, J.d.M.E. and Oliveira, J.H.C.d. (2014), “A review of studies on
neuromarketing: practical results, techniques, contributions and limitations”, Journal of
Management Research, Vol. 6 No. 2, pp. 201-220.
APJML Freeman, W. and Quiroga, R.Q. (2012), Imaging Brain Function with EEG: Advanced Temporal and
32,5 Spatial Analysis of Electroencephalographic Signals, Springer Science & Business Media,
New York, NY.
Goldstein, R., Almenberg, J., Dreber, A., Emerson, J.W., Herschkowitsch, A. and Katz, J. (2008), “Do
more expensive wines taste better? Evidence from a large sample of blind tastings”, Journal of
Wine Economics, Vol. 3 No. 1, pp. 1-9.
1166 Gore, J.C. (2003), “Principles and practice of functional MRI of the human brain”, Journal of Clinical
Investigation, Vol. 112, pp. 4-9.
HajiHosseini, A., Fornells, A. and Marco, J. (2012), “The role of beta-gamma oscillations in unexpected
rewards processing”, Neuroimage, Vol. 60 No. 3, pp. 1678-1685.
Hammou, K.A., Galib, M.H. and Melloul, J. (2013), “The contributions of neuromarketing in marketing
research”, Journal of Management Research, Vol. 5 No. 4, pp. 20-33.
Hasler, R., Perroud, N., Meziane, H.B., Herrmann, F., Prada, P., Giannakopoulos, P. and Deiber, M.
(2016), “Attention-related EEG markers in adult ADHD”, Neuropsychologia, Vol. 87, pp. 120-133.
Heit, E. (2015), “Brain imaging, forward inference, and theories of reasoning”, Frontiers in Human
Neuroscience, Vol. 8 No. 1056, pp. 1-5.
Hollebeek, L.D., Jaeger, S.R., Brodie, R.J. and Balemi, A. (2007), “The inuence of involvement on
purchase intention for new world wine”, Food Quality and Preference, Vol. 18 No. 8,
pp. 1033-1049.
Horska, E., Bercik, J., Krasnodebski, A. and Matysik-Pejas, R. (2016), “Innovative approaches to
examining consumer preferences when choosing wines”, Agricultural Economics, Vol. 62 No. 3,
pp. 124-133.
Khushaba, R.N., Wise, C., Kodagoda, S., Louviere, J., Kahn, B.E. and Townsend, C. (2013), “Consumer
neuroscience: assessing the brain response to marketing stimuli using electroencephalogram
(EEG) and eye tracking”, Expert Systems with Applications, Vol. 40 No. 9, pp. 3803-3812.
Koschate-Fischer, N. and Schandelmeier, S. (2014), “A guideline for designing experimental studies in
marketing research and a critical discussion of selected problem areas”, Journal of Business
Economics, Vol. 84 No. 6, pp. 793-826.
Lakshmi, M.R., Prasad, T. and Prakash, C. (2014), “Survey on EEG signal processing methods”,
International Journal of Advanced Research in Computer Science and Software Engineering,
Vol. 4 No. 1, pp. 84-91.
Lee, N. and Chamberlain, L. (2018), “Welcome to the jungle! The neuromarketing literature through the
eyes of a newcomer”, European Journal of Marketing, Vol. 52 Nos 1-2, pp. 4-38.
Lee, N., Broderick, A.J. and Chamberlain, L. (2006), “What is neuromarketing? A discussion and agenda
for future research”, International Journal of Psychophysiology, Vol. 63 No. 2, pp. 199-204.
Lucchiari, C. and Pravettoni, G. (2012), “The effect of brand on EEG modulation: a study on mineral
water”, Swiss Journal of Psychology, Vol. 71 No. 4, pp. 199-204.
McClure, S., Li, J., Tomlin, D., Cypert, K., Montague, L. and Montague, R. (2004), “Neural correlates of
behavioral preference for culturally familiar drinks”, Neuron, Vol. 44, pp. 379-387.
Maglione, A.G., Brizi, A., Vecchiato, G., Rossi, D., Trettel, A., Modica, E. and Babiloni, F. (2017),
“A neuroelectrical brain imaging study on the perception of figurative paintings against only
their color or shape contents”, Frontiers in Human Neuroscience, Vol. 11, No. 378, pp. 1-14.
Michel, C.M. and Murray, M.M. (2012), “Towards the utilization of EEG as a brain imaging
tool”, NeuroImage, Vol. 61 No. 2, pp. 371-385, available at: http://dx.doi.org/10.1016/
j.neuroimage.2011.12.039
Miljkovic, M. and Alcakovic, S. (2010), “Neuromarketing: marketing research future?”, Singidunum
revija, Vol. 7 No. 2, pp. 273-283.
Mueller, S. and Szolnoki, G. (2010a), “The relative inuence of packaging, labelling, branding and
sensory attributes on liking and purchase intent: consumers differ in their responsiveness”,
Food Quality and Preference, Vol. 21 No. 7, pp. 774-783.
Mueller, S. and Szolnoki, G. (2010b), “Wine packaging and labelling-do they impact market price? A Consumer
hedonic price analysis of US scanner data”, University of Auckland Business School, PhD thesis. behaviour and
Ohme, R., Reykowska, D., Wiener, D. and Choromanska, A. (2010), “Application of frontal EEG brain activity
asymmetry to advertising research”, Journal of Economic Psychology, Vol. 31 No. 5, pp. 785-793.
Park, W., Ki, D., Kim, D., Kwon, G., Kim, S. and Kim, L. (2015), “EEG correlates of user satisfaction of
haptic sensation”, IEEE International Conference on Consumer Electronics, pp. 569-570.
Plassman, H., O’Doherty, J., Shiv, B. and Rangel, A. (2008), “Marketing actions can modulate neural 1167
representations of experienced pleasantness”, Proceedings of the National Academy of Sciences
of the United States of America, Vol. 105 No. 3, pp. 1050-1054.
Plassmann, H. and Weber, B. (2015), “Individual differences in marketing placebo effects: evidence
from brain imaging and behavioral experiments”, Journal of Marketing Research, Vol. 52 No. 4,
pp. 493-510.
Plichta, S.B. and Garzon, L.S. (2009), Statistics for Nursing and Allied Health, Lippincott Williams & Wilkins.
Rahman, F.A., Othman, M.F. and Shaharuddin, N.A. (2015), “Analysis method of EEG signals: a review
in EEG application for brain disorder”, Jurnal Teknologi, Vol. 72 No. 2, pp. 67-72.
Rakshit, A. and Lahiri, R. (2016), “Discriminating different color from EEG signals (a neuromarketing
study on the effect of color to cognitive state)”, 1st IEEE International Conference on Power
Electronics; Intelligent Control and Energy Systems, pp. 1-6.
Ramsøy, T.Z. (2014), Introduction to Neuromarketing and Consumer Neuroscience, Neurons.
Rao, A.R. and Monroe, K.B. (1988), “The moderating effect of prior knowledge on cue utilization in
product evaluations”, Journal of Consumer Research, Vol. 15 No. 2, pp. 253-264.
Russo, V. (2015), Neuromarketing, comunicazione e comportamenti di consumo. Principi, strumenti e
applicazioni nel food and wine, Franco Angeli, Milan.
Shintaputri, I. and Wuisan, A.J. (2017), “The impact of perceived price towards perceived value through
the mediation of perceived quality: a case of brand X Smartphone in Indonesian middle-class
customers”, iBuss Management, Vol. 5 No. 1, pp. 29-42.
Spironelli, C., Manfredi, M. and Angrilli, A. (2013), “Beta EEG band: a measure of functional brain
damage and language reorganization in aphasic patients after recovery”, Cortex, Vol. 49 No. 10,
pp. 2650-2660, available at: http://dx.doi.org/10.1016/j.cortex.2013.05.003
Statistics.laerd.com (2019), Repeated Measures ANOVA - Understanding a Repeated Measures
ANOVA | Laerd Statistics, available at: https://statistics.laerd.com/statistical-guides/repeated-
measures-anova-statistical-guide.php (accessed 2 February 2019).
Sung, B. and Phau, I. (2018), “Applying biometric methods to understand luxury consumers’ emotional
responses”, Global Marketing Conference, Tokyo, 26-29 July.
Szolnoki, G., Hoffmann, D. and Herrmann, R. (2008), “The inuence of verbal and non-verbal information
on the consumer decision-analysis using the example of white wine”, 4th International
Conference of the Academy of Wine Business Research, Siena, pp. 17-19.
Vecchiato, G., Cherubino, P., Trettel, A. and Babiloni, F. (2013), Neuroelectrical Brain Imaging Tools
for the Study of the Efficacy of TV Advertising Stimuli and their Application to
Neuromarketing, Springer, Berlin.
World Health Organization (WHO) (2001), “World health report 2001: mental health”, No. EM/RC48/
INF. DOC. 1.
Yucel, N., Yucel, A., Yimaz, A.S., Cubuk, F., Orhan, E.B. and Simsek, A.I. (2015), “Coffee tasting
experiment from the neuromarketing perspective”, The 2015 WEI International Academic
Conference Proceedings, Harvard, pp. 29-35.
Zeithaml, V. (1988), “Consumer perceptions of price, quality and value: a meansend model and
synthesis of evidence”, Journal of Marketing, Vol. 52 No. 2, pp. 2-22.
Zhang, X., Lei, X., Wu, T. and Jiang, T. (2014), “A review of EEG and MEG for brainnetome research”,
Cognitive Neurodynamics, Vol. 8 No. 2, pp. 87-98.
APJML Appendix 1. Log10 transformation to normalise the EEG data
32,5

1.0

1168
0.8
Frequency

0.6

Figure A1. 0.4

The figure shows that
the output of the EEG
0.2
data for the electrode
FC4 (12–16 frequency)
is normally 0.0
distributed after the
–1.5 –1.0 –0.5 0.0 0.5 1.0 1.5 2.0
log10 transformation
Log10 of EEG power spectrum with baseline correction

1.0

0.8
Frequency

0.6

0.4
Figure A2.
The figure shows that
the output of the EEG 0.2
data for the electrode
FC4 (16–20 frequency)
is normally 0.0
distributed after the
–1.5 –1.0 –0.5 0.0 0.5 1.0 1.5
log10 transformation
Log10 of EEG power spectrum with baseline correction
Appendix 2. Graphs Consumer
behaviour and
brain activity
Label
0.10
Label Session
No-Label Session
1169
0.00
Estimated Marginal Means

–0.10

Figure A3.
The graph shows the
–0.20 changes in beta-band
activity for each
wine (Chilean
Expensive, Italian
Expensive, Chilean
–0.30 Cheap and Italian
Cheap) in the
two sessions
CE IE CC IC
(label/no label)
Wine

0.10 Wine
CE
IE
CC
IC

0.00
Estimated Marginal Means

–0.10

Figure A4.
The graph shows the
–0.20 changes in beta-band
activity for each
wine (Chilean
Expensive,
Italian Expensive,
–0.30 Chilean Cheap
and Italian Cheap) in
the two sessions
Label session No Label session
(label/no label)
Label
APJML About the authors
32,5 Letizia Alvino obtained the Bachelor Degree in Business Administration for SMEs and followed post
graduate studies in Management and Business Control in Naples. In 2018, she received her doctoral
degree from the University of Molise, Campobasso, Italy. Letizia was also Researcher at the University
of Twente, where she conducted research and gave lectures of marketing and consumer neuroscience
during and after her doctoral degree. Letizia is Lecturer of Marketing and Research at Rotterdam
Business School, in the Netherlands. She is also external supervisor for Master thesis at Tilburg
1170 University. Her main research and teaching interests fall into areas such as Consumer Behaviour,
Branding, Marketing Strategies and Consumer Neuroscience. Letizia Alvino is the corresponding
author and can be contacted at: letizia.alvino@hotmail.it
Rob Van der Lubbe studied psychology at the University of Amsterdam, and got his PhD Degree at
the Free University in Amsterdam. Subsequently, he worked as a postdoc at the Medical University of
Lübeck and the University of Utrecht. In 2005 he got a position of Assistant professor at the
department Cognitive Psychology and Ergonomics at the University of Twente, The Netherlands, and
in 2010 he was promoted to the position of Associate Professor. In 2011, he also got a position as
Associate Professor at the University of Finance and Management at the Department of Cognitive
Psychology in Warsaw (Poland). In 2013 and 2014, Van der Lubbe also held a temporary position as
Visiting Professor at the Adam Mickiewicz University at the Physics Department. His research interest
is in lateralized power spectra of the EEG to index visuospatial attention, visual short term memory,
nociceptive spatial attention, changes in spatial attention due to a mindfulness based stress reduction
training, and variations in spatial attention due to different cue validities.
Reinoud A.M. Joosten studied econometrics in Tilburg and specialized in Mathematical Economics.
He obtained his PhD Degree from Maastricht University sponsored by MERIT, the Maastricht
Economic Research institute on Innovation and Technology, under the supervision of Koos Vrieze,
Hans Peters and Frank Thuijsman. For his post doc period, he obtained two scholarships from
METEOR, the graduate school of Maastricht University, and three scholarships from the Max Planck
Society to research at the Max Planck Institute on Economic Systems in Jena (director Ulrich Witt).
Since 2001, he has been employed by the University of Twente as Assistant Professor. His research
interest is in Economics (micro, resource and environmental), Game theory (stochastic games,
cooperative games, evolutionary games), Finance (real options (non-expected) utility, value) and Others
(internet security, social dilemmas, dynamic systems, mathematical biology, Markov chain).
Efthymios Constantinides, PhD, studied Economics in the Athens University of Economics and
Business and followed post graduate studies in Economics of European Integration in the University of
Amsterdam. He received his PhD at the University of Twente. During a corporate career of about ten
years he worked for corporations like Ericsson, Northern Telecom and KLM. He is currently Associate
Professor of Digital Marketing at the Faculty of Behavioural, Social and Management Sciences at the
University of Twente. His research is focused on the Digital Marketing and the impact of digital
technologies on the Marketing practice. He also is interested in the role of social media as marketing
tools and the role of emerging technologies and fields like Internet of Things, consumer neuroscience,
Artificial Intelligence and Big Data on marketing management activities.

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com