(PDF) Neuroscience in branding: A functional magnetic resonance imaging study on brands' implicit and explicit impressions

AUTHOR COPY

www.palgrave-journals.com/bm/

Correspondence:

Jos é Paulo Santos

ISMAI – Superior Institute

of Maia, Av. Carlos Oliveira

Campos – Cast ê lo da Maia,

4475-690 Avioso S. Pedro,

Portugal

E-mail: jpsantos@ismai.pt

Original Article

Neuroscience in branding:

A functional magnetic resonance

imaging study on brands ’

implicit and explicit impressions

Received (in revised form): 21 st May 2012

Jos é Paulo Santos

recently completed the PhD programme from the Institute of Economy and Management, Technical University of Lisbon,

Portugal, with the theme Neuroscience in Branding. He has 15 years of professional experience in industrial management,

focusing on business-to-business brands, and, more recently, end-customer brands. Also, Santos researches at Socius,

focusing on consumer behaviour. His research interests include Social Neuroscience and associated neuroscientiﬁ c

techniques as fMRI and fNIRS, sociology of brands, Symbolic Interactionism, and self-concept drive by brands.

Daniela Seixas

has a primary degree in medicine, and is currently a neuroradiologist at S ã o Jo ã o Hospital, Portugal. She is also a PhD

student at the Faculty of Medicine of the University of Porto, in collaboration with the Pain Imaging Neuroscience

Group, Oxford Centre for Functional Magnetic Resonance Imaging of the Brain (FMRIB), University of Oxford, United

Kingdom. Seixas ’ s main research interests are neuroimaging techniques applied to neuroscience research: she is at

present studying pain in neurological diseases using DTI and fMRI, and also collaborating in neuromarketing research.

She has published a recent review on ethics in fMRI studies.

Soﬁ a Brand ã o

completed her BD in Radiology from the Superior School of Health Technologies of Porto, Porto, Portugal and is

currently pursuing an MD on Biomedical Engineering, and another MD on Medical Informatics, at the Faculty of Medicine

of the University of Porto. Her thesis focuses on segmentation of deep grey nuclei using SPM5 software tools. Brand ã o

has been working since 1999 at the S ã o Jo ã o Hospital, in the Resonance Magnetic Unit. She is also a lecturer in the

Imagiology area. She has several published papers in medical and technical conferences. Her research interests include

Resonance Magnetic techniques.

Luiz Moutinho

is Professor of Marketing and holds the Foundation Chair of Marketing at the School of Business and Management,

University of Glasgow. He has held posts at the Cardiff Business School, University of Wales, Cleveland State University,

Northern Arizona University and California State University, as well as Visiting Professorship positions in several countries.

Moutinho has published 22 books, and extensively in academic journals. He is the Editor of the Journal of Modelling in

Marketing and Management . His research interests include mathematical and computer modelling in marketing, consumer

behaviour and marketing of services.

ABSTRACT Although the use of neuroscientiﬁ c knowledge to investigate marketing

issues has been widely discussed, to date, few empirical studies have been published.

This study is a ﬁ rst approach in the development of a theory of the perception of

brands, which is based on neuroscience. In a Functional Magnetic Resonance Imaging

experiment, we stimulated participants with commercial brands ’ logos, with and

without explicit instructions on how to assess them, in an attempt to capture the

AUTHOR COPY

Santos et al

2

overt verbalisations and which directly tackle

the brain structures and networks that sup-

port human behaviour. Despite this, few

empirical articles have been published in

peer-reviewed journals. As our aim is to

con tribute to cutting-edge research, we

chose commercial brands, represented by

their logos, to target a neuroscientiﬁ c app-

roach to the perception of brands.

We selected Functional Magnetic Reso-

nance Imaging (fMRI) owing to certain

advantages. For example, it is ethically accep-

table to apply this technique to healthy parti-

cipants. Other advantages include spatial

resolution and the long established wealth

of knowledge based on cognitive studies

using fMRI. This is crucial in guaranteeing

the nomological validity of the eventual

ﬁ ndings. However, it is important to indi-

cate certain disadvantages to fully under-

stand the results. The most important is the

association of magnetic resonance ima ging

(MRI) scanners with hospitals, which brings

an unusual context to the expe ri ments. The

extent of the contribution of such an envi-

ronmental inﬂ uence is unknown, notwith-

standing the adoption of practices that aim

to reduce it, such as the wearing of informal

clothes by team members and, before scan-

ning, 15 min of conversation in a relaxing

atmosphere to reduce anxiety. Other major

INTRODUCTION

Neuroscientiﬁ c knowledge has been used

by several researchers and practitioners to

investigate marketing issues ( Zaltman, 2003 ;

Lee et al , 2007 ; Plassmann et al , 2007 ;

Ambler, 2008 ; Hubert and Kenning, 2008 ;

Penn, 2008 ), and a theory of brand building

has already been proposed ( Walvis, 2007 ).

Neuroscience can help in overcoming

known hurdles in the realm of social

science. Direct measures of marketing para-

meters, for example, may be considerably

biased if collected outside important con-

text attributes, such as engaging subjects in

unusual cognitive processes, which produces

spurious measures and reports ( Mueller et al ,

2010 ). In another example, Bernard et al

(1984) studied the problem of informant

accuracy and the validity of retrospective

data and concluded that ‘ on average, about

half of what informants report is probably

incorrect in some way ’ ( Bernard et al ,

1984 ). They also add that ‘ a weak relation-

ship between a concept and the accurate

measurement of the concept is unaccep-

table ’ . Such deviations are important mainly

when subjects are required to self-report on

their own emotional states ( Chamberlain

and Broderick, 2007 ). Neuroscience, in parti-

cular neuroimaging techniques, may pro-

vide methods that overcome the need for

real-life experience of evaluating brands. We found common activations in both

situations in the medial frontal pole, the paracingulate gyrus, the frontal orbital

cortex, the frontal medial cortex, and in the hippocampus. In a general scheme of

brands ’ perception, we hypothesised a relationship between Theory of Mind and

meta-representations, in particular self-reﬂ exive ones: ‘ I think about what others

are thinking about me ’ . We suggest that brands have an important social dimension.

Brands may function like a social currency, which every individual uses to assess

others, and which others are expected to use in their assessments of the individual.

Brands are most probably social tools.

Journal of Brand Management advance online publication, 13 July 2012;

doi: 10.1057/bm.2012.32

Keywords: brand names ; impression management ; social neuroscience ; symbolism ;

self-reference

AUTHOR COPY

Neuroscience in branding

3

disadvantages of fMRI are the noise (which

is intrinsic to MRI scanner operations) and

the extremely limited interaction with parti-

cipants. The Appendix at the end of this

article includes some introductory notes on

fMRI that describe the basics of this tech-

nique and which also help in paving the

way to the correct interpretation of the

results. The newcomer to Neuromarketing

is invited to ﬁ rst read the Appendix.

In the present study, we aimed to cap-

ture the same strategy that individuals use

in everyday life when they face brands ’

logos in the social environment. We believe

that when individuals make social inter-

actions within social groups, they gene rally

do not use rational, fully conscious and

expli cit strategies. On the contrary, they

employ implicit and automated, ready-

made short cuts as a matter of routine

( Greenwald and Banaji, 1995 ; Pelzmann

et al , 2005 ). People have a tendency to

use simple heuristics, theorised as Boun-

ded Rationality ( Selten, 2002 ; Todd and

Gigerenzer, 2003 ). In the Bounded Ration-

ality theory, individuals learn social rules,

which are obedient to general standards

of their culture (such as those astutely

discerned by Goffman (1959) ), and each

indivi dual constructs a repertoire of social

beha viours, which is adapted to each situ-

ation ( Gigerenzer, 2001 ). It is then expected

that humans act socially mostly by implicit

rather than by explicit strategies, that is,

without full cognitive awareness ( Critchley

et al , 2000 ). Indeed, Critchley and his co-

workers believe that common experience

reveals that individuals form impressions of

their peers implicitly, and implicitly use this

information when they interact ( Critchley

et al , 2000 ). In spite of this, most of the

studies in neuroscience use explicit para-

digms. For this reason, we designed an fMRI

experiment to study whether implicit and

explicit brands ’ appraisal is in fact different.

It is worth emphasising the methodo-

logy of our approach. The purpose of the

present study is not to ﬁ nd proofs on the

dimensions of brands from an exact disci-

pline. This is rather the ﬁ rst study of a series

that aims at an experimental use of neuro-

scientiﬁ c techniques and knowledge to

investigate brand perception. Three princi-

ples from Grounded Theory are recognis-

able in our strategy ( Strauss and Corbin,

1990 ; Corbin and Strauss, 2007 ). First, we

do not have previously constructed models

from neuroscience, nor psychology, nor

sociology, nor, of course, from marketing.

Previous models tend to introduce a bias

in the studies and, if we aim to capture a

different perspective on brands, in this case

a neuroscientiﬁ c perspective, such biases

could introduce inﬂ uences from established

knowledge pertaining to other disciplines.

That is why this ﬁ rst study is broad, loosely

bounded and very simple; it contrasts assor ted

brands ’ logos with non-emotional words,

and uses simple, yet very robust, fMRI

block design. However, we introduced

some level of complexity in the fMRI data

analysis by employing a model-free tool

that is not commonly used. The results

produ ced by this tool are stressed exactly

because it is model-free and diminishes

eventual inﬂ uences originated by the para-

digm structure. Second, all the ﬁ ndings

must be data grounded. The activated brain

structures will be used to infer concepts that

support brands ’ dimensions in the same

way when texts are coded to generate

higher level concepts, categories and theo-

ries. The difference is that the researcher ’ s

subjec tive perspective is considerably

reduced as the ‘ coding ’ (activated and deac-

tivated brain structures) will be done by

computer programs. Nevertheless, compa-

risons with simi lar neuroscientiﬁ c studies

will be present throughout to guarantee

nomological validity, and this speciﬁ c work

will be carried by human researchers, who

will decide on the more pertinent according

to their perceptions. Third, we aim to

produce a theory on consumers ’ brands

AUTHOR COPY

Santos et al

4

be acquired. In each slide, a brand logo was

placed over a black background. It is worth

noting that none of the participants in the

screening procedure took part in the actual

experiment.

As a baseline, we used words without

emotional content that could not evoke

objects or actions. These words were deter-

miners, conjunctions, prepositions or

adverbs, and they were written in white

(lower case, font Arial, 150, bold) over

a black background. Because this was the

ﬁ rst study of a series, an option for an

elaborated baseline (for example, brands ’

logos of products and services from other

markets) could disguise important activa-

tions, and eliminate possible relevant trends

prematurely ( Matthews et al , 2003 ). The

natural option would be a baseline that

could achieve high contrasts with the

logos, as a ﬁ xation cross, despite knowing

that looking at a ﬁ xation cross is not resting

at all ( D ’ Argembeau et al , 2005 ). Other

studies on passive viewing and on the

default mode have reported cortical activa-

tions in structures related to self-referential

reﬂ ective activity ( Iacoboni et al , 2004 ;

Schilbach et al , 2008 ). Thus, the use of

a baseline that could induce self-referential

reﬂ ective acti vity, such as a ﬁ xation cross

or chequered patterns, would cancel such

an important characteristic. Hence, for the

baseline, we ultimately chose words that

could not evoke emotions, in the hope of

retaining an eventual emotional content

associated only with the brands, and that

simultaneously could provide some inno-

cuous activity that would swerve self-

referential thoughts from the participant ’ s

mind.

perception. The ﬁ ndings of the present

study will be used to design future inves-

tigations, which will challenge previous

concepts and constructed concepts ’ links.

Initial studies will examine general brands ’

perception, but subsequent ones will intro-

duce the necessary reﬁ nement to produce

a grounded theory.

METHOD

Experimental design

We designed an fMRI experiment made

up of two identical runs where commercial

brands ’ logos were the stimuli visually pre-

sented to the study subjects. In the ﬁ rst

run, stimuli were presented without pre-

vious instructions, which aimed to capture

implicit behaviours. Before the second

run, participants received explicit instruc-

tions, which aimed to capture overt behav-

iours. The scheme of the study is depicted

in Figure 1 .

The experiment was designed in blocks,

where the slide set used was the same for

both runs, employing as stimuli brands ’

logos, with their characteristic shapes, col-

ours and wording in everyday life. Before

the scanning sessions, 237 commercial

brands ’ logos were screened by a question-

naire delivered to 147 volunteers. The pur-

pose of this preliminary screening was to

decide on the brands that were most well

known to the population from which the

study sample was to be taken, thus mini-

mising the risk of including unknown

brands in the slide set. It was our assump-

tion that unknown brands would elicit dif-

ferent brain processes, and thus would

introduce ‘ cognitive noise ’ in the data to

Figure 1 : Complete sequence of the experiment. Duration is approximate.

AUTHOR COPY

Neuroscience in branding

5

In each run, the slide set was composed

of 16 baseline periods alternated with 16

stimuli periods, starting with a baseline

period. The stimuli period had the same

duration as the baseline period, 30 seconds

long. Within each 30-second period, ﬁ ve

slides were visually displayed, each for

6 seconds. Thus, the slide set contained

80 stimulus slides and 80 baseline slides, and

lasted for exactly 16 min. Figure 2 depicts

the structure of the paradigm showing some

examples of the stimulus and baseline.

In the ﬁ rst run, the implicit one, volun-

teers only had to look at the screen during

the scanning procedure. Nothing was men-

tioned with regard to what they were about

to see. In the interval between the ﬁ rst

and the second runs, the subjects completed

a questionnaire on the brands ’ logos they

saw in the ﬁ rst run, and which they would

see during the second run. The volunteers

were asked to evaluate each brand hedoni-

cally, rating them as unknown, or negative,

or indifferent, or positive. In this way, they

were trained in the brand assessment they

were asked to do in the second run.

The order of the runs was crucial; as the

implicit task was ﬁ rst, we hoped to avoid

any expectations and strategies from the

participants, in capturing covert evalua-

tions, to then compare with the explicit

assessment of the same brands. If the explicit

task had been ﬁ rst, the next task could

never have been implicit, as the participants

would have guessed the intention owing

to the biasing effect of previous exposure

to instructions. Although it is good practice

to randomise or alternate runs, this would

have spoiled the intended effect in the

present study.

Human subjects

The participants were six healthy male and

eight healthy female volunteers, right handed,

with neither a history of neurological nor

psychiatric disturbances (mean age: 28.4

years, 5.4 SD ; mean education: 16.2 years,

1.5 SD ). None of the participants was taking

psychoactive medication. Informed consent

was obtained in all cases. A safety form for

MRI was completed by the participants.

This research project complied with the

Declaration of Helsinki and was approved

by the local ethics committee.

Two female participants were excluded

from the analysis, one due to excessive

head movement and the other due to

claustrophobia.

Data acquisition

Functional images were obtained using a

T2 * -weighted EPI sequence in a Siemens

®

Magnetom Trio 3 Tesla MRI scanner

(Siemens AG, Germany) (TR = 3000 ms,

TE = 30 ms, 64 × 64 matrix, FOV = 192 mm,

36 axial slices with 3.0 mm thickness). A

whole brain structural scan was also acqui red

for each volunteer, using a T1-weighted

MPRAGE protocol (256 × 256 matrix,

FOV = 192 mm, 36 axial slices with 3.0 mm

thickness), for co-registration purposes. Both

acquisitions were interleaved. Gradient ﬁ eld

mapping was additionally obtained. In each

run (implicit and explicit), 340 functional vol-

umes were acquired. The ﬁ rst 20 volumes

were discarded because of pulse stabilisation.

Figure 2 : Structure of the paradigm with some examples of the baseline (non-emotional words).

Note : Slides with real brand logos were used in place of LOGO.

AUTHOR COPY

Santos et al

6

outlier detection ( Woolrich, 2008 ). In this

level, group means were calculated from

the ﬁ rst-level contrasts.

z (Gaussianised T / F) statistic images were

thresholded using clusters determined by

z > 2.3 and a (corrected) cluster signiﬁ cance

threshold of p = 1.00 ( Worsley, 2001 ). Only

clusters with more than 50 voxels survived

the threshold.

Conjunction analysis was performed acc-

or ding to Nichols and colleagues ’ method

( Nichols et al , 2005 ), that is, the voxels con-

sidered active in the conjunction were those

that cumulatively were statistically signiﬁ -

cant in the implicit and explicit analysis.

In the MELODIC analysis, 24 data sets

(12 implicit and 12 explicit) were com-

puted, aiming to extract independent spatial

components common to both runs. The

following data pre-processing was applied:

masking of non-brain voxels, voxel-wise

de-meaning of the data and normalisation

of the voxel-wise variance. Pre-processed

data were whitened and projected into a

113-dimensional subspace using proba bi-

listic Principal Component Analysis where

the number of dimensions was estimated

using the Laplace approximation to the

Bayesian evidence of the model order

( Minka, 2000 ; Beckmann and Smith, 2004 ).

The whitened observations were decom-

posed into sets of vectors, which describe

signal variation across the temporal domain

(time-courses), the session / subject domain

and across the spatial domain (maps) by

optimising for non-Gaussian spatial source

distributions using a ﬁ xed-point iteration

technique ( Hyv ä rinen, 1999 ). Estimated com-

ponent maps were divided by the standard

deviation of the residual noise and thresh-

olded by ﬁ tting a mixture model to the

histogram of intensity values ( Beckmann

and Smith, 2004 ).

The identiﬁ cation of the main ana-

tomical structures in the clusters was made

with masks based on the statistical para-

metric maps produced by both analysis

Image analysis

FMRI data processing was carried out

using FEAT (FMRI Expert Analysis Tool)

Version 5.98, a model-based GLM (Gene ral

Linear Model) analysis tool, and also

using Tensorial Independent Component

Analysis ( Beckmann and Smith, 2005 ) as

implemented in MELODIC (Multivariate

Exploratory Linear Decomposition into

Independent Components) Version 3.09, a

model-free analysis tool, both part of FSL –

FMRIB ’ s Software Library, www.fmrib

.ox.ac.uk/fsl ( Smith et al , 2004 ).

In the FEAT analysis, the following pre-

statistics processing was applied: motion

correction using (Motion Correction using

FMRIB ’ s Linear Image Registration Tool)

( Jenkinson et al , 2002 ); slice-timing correc-

tion using Fourier-space time-series phase-

shifting; non-brain removal using BET

(Brain Extraction Tool) ( Smith, 2002 ); spa-

tial smoothing using a Gaussian kernel of

full width half maximum 5 mm; grand-mean

intensity normalisation of the entire 4D data

set by a single multiplicative factor; highpass

temporal ﬁ ltering (Gaussian-weighted least-

squares straight line ﬁ tting, with sigma =

30.0 s). Time-series statistical analysis was

performed using FILM (FMRIB ’ s Improved

Linear Modelling) with local autocorrela-

tion correction ( Woolrich et al , 2001 ). Reg-

istration to high resolution structural and / or

standard space images was done with FLIRT

(FMRIB ’ s Linear Image Registration Tool)

( Jenkinson and Smith, 2001 ; Jenkinson et al ,

2002 ).

At the ﬁ rst-level analysis and separately

for each run, stimuli and baseline were sub-

tracted, resulting in the contrasts implicit >

baseline and explicit > baseline, and also

stimuli were subtracted between them,

resulting in the contrasts implicit > explicit

and explicit > implicit.

Higher-level analysis was performed

using FLAME (FMRIB ’ s Local Analysis

of Mixed Effects) stage 1 ( Beckmann et al ,

2003 ; Woolrich et al , 2004 ) with automatic

AUTHOR COPY

Neuroscience in branding

7

tools (GLM and model-free). The masks

were designed according to the probabil-

istic atlases Harvard-Oxford Cortical Structural

Atlas and Harvard-Oxford Subcortical Struc-

tural Atlas provided by the Harvard Centre

for Morphometric Analysis ( www.cma

.mgh.harvard.edu ), which are part of FSL

View v3.0.2, part of FSL 4.1.2. Each voxel

of each cluster was assigned to a single brain

structure. In cases where several structures

could be probabilistically attributed to a

voxel, the structure that had the highest

probability was chosen.

RESULTS

Behavioural results

The subjects reported as ‘ negative ’ 14 per

cent of the brands, as ‘ indifferent ’ 32 per cent

of the brands and as ‘ positive ’ 53 per cent of

the brands. The ‘ unknown ’ answers were

negligible (1 per cent).

Activations produced in the brain

common to the implicit and explicit

paradigms

The main activations produced in the brain

common to both the implicit and explicit

paradigms are depicted in Figure 3 . The

activations found in the paracingulate gyrus,

medial frontal pole, left frontal orbital cortex,

hippocampus and fusiform gyrus (occipital

fusiform gyrus and temporal occipital fusi-

form cortex) are of special interest.

Figure 4 illustrates the hemodynamic

response of the medial frontal pole and the

paracingulate gyrus. The response in the

frontal pole is similar in both runs (implicit

and explicit), only pointing the decay in

the implicit response along the stimulus

block. In the paracingulate gyrus we draw

attention to a decay along the stimulus

block in both runs, with a very similar pat-

tern. In the graph, the results show that the

signal change is stronger in the explicit run

Figure 3 : Activations obtained with the conjunction analysis (statistical parametric maps produced by FEAT).

Note : In each pane, the left column refers to the thresholded map ( z > 2.3), and the right column refers to the thresholded activations

with the brain structures highlighted with false colours (R: right; P: posterior; FFG: fusiform gyrus; FOC: frontal orbital cortex; Hip:

hippocampus; mFP: medial frontal pole; pCG: paracingulate gyrus; MNI152 coordinates ).

AUTHOR COPY

Santos et al

8

synchronous activity in the following areas:

amygdala, fusiform gyrus, frontal medial

cortex, frontal orbital cortex, frontal oper-

culum cortex, insular cortex, medial frontal

pole and paracingulate gyrus The activation

of these structures had a unique period of

60 seconds (1.67 Hz / 100), which was exactly

the same of the stimulus onset.

than in the implicit run, although both are

activating.

In the model-free analysis with MELODIC,

the independent component # 32 was selected

due to the high correlation of its timecourse

with the block-design sequence of both runs

(implicit and explicit) of the experiment

( P < 0.00001). This component included

Figure 4 : Selected peristimulus hemodynamic response in two voxels. ( a ) ( − 08, 62, 30) with 67 per cent probability in the frontal

pole, and ( b ) ( − 06, 16, 44) with 61 per cent probability in the paracingulate gyrus (MNI152 coordinates).

AUTHOR COPY

Neuroscience in branding

9

The statistical parametric maps and cor-

responding graphs are shown in Figure 5 .

Activations produced in the brain that

characterise the implicit paradigm

(contrast implicit > explicit)

The amygdala, the parahippocampal gyrus

and a ventral medial region comprising the

ventral medial frontal pole, the frontal

medial cortex and subcallosal cortex were

brain structures signiﬁ cantly activated when

the implicit run was contrasted with the

explicit run (see Figure 6 ).

Activations produced in the brain that

characterise the explicit paradigm

(contrast explicit > implicit)

The activations produced in the brain

when the explicit paradigm was contrasted

with the implicit are shown in Figure 7 .

Figure 5 : Independent component # 32 selected from the model-free analysis with MELODIC. This component explains 1.04 per

cent of the total variance. ( a ) The top row depicts thresholded activations and deactivations. The bottom row refers only to the

thresholded activations with brain structures highlighted in false colours (R: right; P: posterior; Amy: amygdala; FFG: fusiform gyrus;

FMC: frontal medial cortex; FOC: frontal orbital cortex; FOp: frontal operculum cortex; Ins: insular cortex; mFP: medial frontal pole;

pCG: paracingulate gyrus; MNI152 coordinates). ( b ) Timecourse of the independent component # 32 and full model ﬁ t; F -test on the

full model ﬁ t: P = 686.01 (dof1 = 2; dof2 = 317) P < 0.00001; Contrast of parameter: z = 22.96; P < 0.00001. ( c ) Powerspectrum of the

timecourse. The frequency of the peak is 1.67 Hz / 100, which corresponds to a period of 60 seconds.

AUTHOR COPY

Santos et al

10

Figure 6 : Activations that characterise the implicit task when contrasted with the explicit (statistical parametric maps produced by FEAT).

Note : The left column refers to the thresholded map ( z > 2.3), and the right column refers to the thresholded activations with the

brain structures highlighted in false colours (R: right; P: posterior; Amy: amygdala; FMC: frontal medial cortex; pHG: parahippocampal

gyrus; MNI152 coordinates).

Figure 7 : Activations that characterise the explicit task when contrasted with the implicit (statistical parametric maps produced by FEAT).

Note : The left column refers to the thresholded map ( z > 2.3), and the right column refers to the thresholded activations with the brain

structures highlighted in false colours (R: right; P: posterior; FOp: frontal operculum cortex; IFG: inferior frontal gyrus (comprising the

pars opercularis and the pars triangularis); Ins: insular cortex; lFP: lateral frontal pole; Pal: pallidum; Put: putamen; MNI152 coordinates).

AUTHOR COPY

Neuroscience in branding

11

The brain structures identiﬁ ed include the

inferior frontal gyrus (comprising the pars

opercularis and pars triangularis), insular

cortex, frontal operculum and nucleus

lentiformis (comprising the pallidum and

putamen).

DISCUSSION

As the slide set that contained the brands ’

logos that served as stimulus was the same

for all subjects, there was the possibility that

unknown brands / logos could introduce

cognitive processes that would interfere with

brand appraisals. However, as the stimuli

were pre-screened, the unknown logos were

negligible (1 per cent of total stimuli). There-

fore, we can conclude that subjects per-

formed only brand assessments, at least during

the explicit scanning session.

When contrasting the two tasks, implicit

and explicit, our goal was to capture speciﬁ c

processes in the hope of a better under-

standing of how individuals deal implicitly

with brands ’ logos, and of how they assess

the brands, in an explicit and purposeful

manner. There is, however, a methodo-

logical difﬁ culty. In order to allow subjects

to make free implicit assessments of brands,

they could not be instructed beforehand,

and it was not also possible to control

the execution of the task. Nevertheless, our

ﬁ rm hypothesis is that, on a daily basis,

we evaluate brands implicitly more often

than explicitly. As such, we circumvented

this problem by doing a conjunction analysis

of the two runs, and therefore uncovered

a general mechanism of brands ’ assess-

ment, common to both implicit and exp-

licit situations.

Common processes: The self-related

dimension

We registered a common activation to the

implicit and explicit tasks in the medial

frontal pole. The location of this activation

is consistent with what Amodio and Frith

named as the anterior rostral medial frontal

cortex, arMFC ( Amodio and Frith, 2006 ).

From their meta-analysis results, this region

was found to be important in the neural

processing of different categories of tasks:

self-knowledge, person knowledge and

mentalising. All these categories are crucial

for social interactions; the ability to read

how others evaluate our self-image is an

example. Self-knowledge is pivotal for an

individual to be able to differentiate himself /

herself from others ( Ruby and Decety,

2004 ), and promotes the capacity to self-

attribute preferences and dispositions ( Kelley

et al , 2002 ). In addition, activations in this

region were found to occur when trying

to differentiate people from objects ( Mitchell

et al , 2005 ), which raises the possibility

that brands are not considered mere objects,

but are judged to be closer to people, as

distinctive components. Self-knowledge is

a reference to self-concept, so that the moti-

vations self-esteem and self-consistence can

act purposefully ( Sirgy, 1982 ; Banister and

Hogg, 2004 ), mainly in the social envi-

ronment ( Grubb and Grathwohl, 1967 ;

Johar and Sirgy, 1991 ). This self-referential

processing in the social domain has been

shown to have neural correlates, again in

the medial frontal pole ( Northoff et al ,

2006 ). Schaefer and colleagues have dem-

onstrated that culturally self-relevant familiar

cars ’ brands, displayed implicitly, also acti-

vate the medial frontal pole in a way

similar to our study ’ s ﬁ ndings ( Schaefer

et al , 2006 ). All of these ﬁ ndings suggest

that commercial brands, together with

their symbolic content, are landmarks that

each individual recognises as useful for

self-characterisation, and are used to con-

struct his / her identity within the social

milieu ( Elliott and Wattanasuwan, 1998 ).

Other phenomenological studies have been

reporting the role that brands and com-

modities have in self-construal ( Belk, 1988 ;

Fournier 1998 ; Ahuvia, 2005 ; Escalas and

Bettman 2005 ; Allen et al , 2008 ). There is

therefore convergence between this body

AUTHOR COPY

Santos et al

12

action, the self-reﬂ exive nature of the

individual and the negotiation of each indi-

vidual ’ s self-concept in the social context.

Consequently, the Theory of Mind plays

a crucial role, as every individual, during

a social transactional process, must infer the

mental state of his / her peers (namely beliefs,

aims, intentions and strategies), and brands ’

socially relevant meanings may have a con-

tribution in such inferences. Considering

our ﬁ ndings and the supportive literature,

we hypothesise that brands are meaningful

utensils that each individual gathers and

uses to diffuse his / her own identity and to

perceive and interpret the messages ema-

nated by his / her peers. Our interpretation

is that brands are a culturally accepted social

currency, for an individual to make infer-

ences reliably of others: brands are indeed

social tools.

Common processes: A probable

general valuation mechanism also

used for brands

Damasio established the connection bet-

ween damage of the orbitofrontal cortex,

emotions and decision-making ( Dam á sio,

1994 ). Other neurological cases of lesions

in the same cortical area have been reported

to have similar consequences, for example

the inability to perform advantageously in

the Iowa Gambling Task ( Bechara, 2004 ),

and inappropriate social behaviour, in spite

of the conservation of the awareness of

social norms ( Beer et al , 2006 ). The modu-

lation of the orbitofrontal cortex extends

to non-conscious brain areas, with indivi-

duals being able to anticipate rewards whilst

performing economic decisions ( Bechara

and Dam á sio, 2005 ). It also participates in

emotion modulation and behaviour condi-

tioning, through a top-down control over

structures such as the insula or the amygdala

( Adolphs, 2001, 2003 ). The orbitofrontal

cortex is subdivided into three regions: one

medial, which comprises the ventral medial

frontal pole, the frontal medial cortex and

of knowledge and the ﬁ ndings of our study

obtained with fMRI. Interestingly, these

processes seem to happen both consciously

and also beyond conscious awareness, that

is, explicitly and implicitly, which supports

our initial assumption.

Common processes: The social

dimension

Another important activation common to

both runs was found in the paracingulate

gyrus. In accordance with the discussion in

the previous paragraph, functional investi-

gations that study social interactions usually

report activations in this area. The Theory

of Mind, mentalising, meta-representations

and second-order meta-representations have

been linked to the paracingulate gyrus

( Gallagher and Frith, 2003 ; Rilling et al ,

2004 ; Amodio and Frith, 2006 ; Brunet-

Gouet and Decety, 2006 ; Saxe, 2006 ; Frith,

2007 ). The same brain area is thought to

be involved when subjects make judge-

ments about similar and dissimilar individ-

uals ( Mitchell et al , 2006 ), and again when

forming impressions of people as opposed

to objects ( Mitchell et al , 2004, 2005 ). The

Theory of Mind is important in making

predictions about others ’ behaviour on the

basis of their mental states ( Baron-Cohen

et al , 1985 ). Stone defends this as one of

the underpinnings of the complexity of our

social groups ( Stone, 2006 ), attributing to

humans a social cognition ( Adolphs, 2001,

2003 ). The reﬂ exive meta-representations,

or second-order representations, where an

individual predicts what other individuals

think about himself / herself, are essential for

communicative intentions between indi-

viduals ( Ermer et al , 2006 ; Frith, 2007 ), and

brings up the triadic social interaction: Ide

ntity ↔ Communication ↔ Image. On the

other hand, according to the theory of

the Symbolic Interactionism, the value of

a brand is asserted within the social group

( Ligas and Cotte, 1999 ). Symbolic inter-

actionism is a complex play among social

AUTHOR COPY

Neuroscience in branding

13

the subcallosal cortex as delineated in the

Harvard-Oxford Cortical Structural Atlas , and

two lateral relative to the medial, corre-

sponding to the lateral frontal pole and the

frontal orbital cortex. Stimuli valence rep-

resentations are usually assigned to the late ral

regions ( O ’ Doherty et al , 2001 ; Rolls,

2004 ; Ursu and Carter, 2005 ), although

this is still controversial ( Elliott et al , 2000 ).

The medial region decodes rewards and

implements a reinforced learning mecha-

nism that monitors and sustains the relevant

reinforcers ( Rolls, 2004 ; Windmann et al ,

2006 ). In our study, the lateral regions parti-

cipated extensively in both runs, although

speciﬁ c sub-regions activated more signi-

ﬁ cantly in the explicit run. A small area of

activation in the medial region was regis-

tered only in the model-free analysis, and

was more extensive in the contrast implicit >

explicit. More studies are needed to explore

these ﬁ ndings further. Perhaps the orbitof-

rontal cortex, represented by these medial

frontal areas, participates in the perception

and valuation of brands.

The common activation between the

implicit and explicit paradigm that we have

found in the hippocampus was expected,

because of its function in declarative and

mnemonic memories ( Critchley et al , 2000 ).

This structure participates in the process

of recall based on recognition ( Paller and

Wagner, 2002 ; Yonelinas, 2002 ; Bailey and

Kandel, 2004 ; Fortin et al , 2004 ), and in

a study performed on culturally familiar

sodas, both the right and left hippocampi

responded preferentially to brand-cued versus

light-cued soda delivery ( McClure et al ,

2004 ).

Interpretation of the processes

observed in the implicit run

Looking at the results of our study, it is

apparent that the implicit and explicit para-

digms recruited a network of brain regions

that does not completely overlap. In other

words, our data strongly suggests that the

neural substrates of forming impressions

on commercial brands are different accor-

ding to whether or not the participants are

given instructions. The same may happen

in other types of experiments. For example,

games are often used in neuroeconomic

research, where the participants are previ-

ously instruc ted on their rules (for a review,

see Montague et al , 2006 ). In some cases,

the participants are even trained before the

study session, as we did in our second –

explicit – fMRI run. Considering our results,

we would suggest caution in the interpre-

tation of such studies, where conditioning

the subjects ’ performance may modulate

the resulting neural activation. Economic

behaviour is not always conscious and

rational. Emotions drive most of the deci-

sions ( Dam á sio 1994 ; Bechara and Dam á sio,

2005 ), and emotions tend to induce beha-

viours implicitly ( Critchley et al , 2000 ;

Pelzmann et al , 2005 ).

We consider the activation of the amy-

gdala, just observed in the implicit run and

in the model-free analysis, a key result, due

to its role in primary emotional processing

( Adolphs et al , 1998 ; Bechara et al , 1999 ;

Adolphs, 2003 ; Zald, 2003 ; Norris et al ,

2004 ; Ashwin et al , 2007 ; Beaucousin et al ,

2007 ), with connections to the frontal medial

cortex and the hippocampus ( Stefanacci

and Amaral, 2002 ). This suggests that the

human emotional network can be involved

in the perception of brands, although the

logos that were chosen for our study were

varied and were not screened purposefully

according to their emotional content. Adolphs

(2006) proposed that the amygdala is neces-

sary for humans to probe the social envi-

ronment systematically, in the search for

clues that allow them to make inferences

about the minds of others and to use ‘ other

people as a collective resource ’ . This also

supports the social role that we hypothesise

that brands have. Interestingly, the amy-

gdala activated in the implicit run, but

not in the explicit one. It may happen that

AUTHOR COPY

Santos et al

14

Sirgy, 1991 ). Thus, tracking him / herself in

the social milieu is crucial for a purposeful

navi gation ( Blumer, 1969 ; Stryker, 1990 );

the individual ’ s emotional system is the best

adaptive behavioural trump to be successful

( Rolls, 2000 ). Brands supply socially rele-

vant meanings to help individuals construct

their self-concepts ( Sirgy, 1982 ; Belk, 1988 ;

Kleine III et al , 1993 ). This study allowed

for the emergence of the social dimension

in the perception of brands as one of the

most relevant. Based on our ﬁ ndings, new

hypotheses can be formulated and further

studies should investigate the involvement

of each of the concepts discussed in more

detail (and respective neuroanatomic cor-

relates). To strengthen our ﬁ ndings, brands ’

logos should be contrasted against diverse

baselines. This is particularly relevant for

the inferior frontal gyrus, which activated

unexpectedly in the explicit run. Was this

due to speech inhibition, as the participants

were instructed to assess the brands covertly,

without speaking? Was it part of the explicit

reasoning process? It should also be further

pursued if the social relevance found is

common to all brands, or if it is speciﬁ c to

some classes.

We used as stimuli assorted brands ’ logos

without any kind of categorisation or

screening (except for their recognition).

Future studies should introduce differences

in brands and should search for anato mical

structures or networks that could be brain

signatures for such categories and which

might surpass traditional limitations of ver-

balising when individuals are faced with

questions in marketing research interviews

the non-natural behaviour that subjects

performed in the explicit run may have

suppressed the activation of the amygdala.

Further studies are required to make clear

the amygdala ’ s role in brands impressions.

Interpretation of the processes

observed in the explicit run

Signiﬁ cantly more in the explicit than in

the implicit paradigm, we have found acti-

vations in the frontal operculum, inferior

frontal gyrus, insular cortex, pallidum and

putamen, structures that are possibly invol ved

in deliberative reasoning. Referring in

more detail to the activation found in the

inferior frontal gyrus, it is well known that

it is part of Broca ’ s area in the left-brain

hemisphere. Intriguingly, the paradigm ’ s

baseline was composed of words that had

neither emotional content, nor suggested

objects or actions, and every stimulus had

only the wording correspondent to its

respective brand. Therefore, in theory, non-

emotional language areas should have not

produced activations in the brain. We ack-

nowledge though that language processing

is complex and far from being completely

understood, and that the participation of

Broca ’ s area obtained in our study, together

with other brain regions that activated sig-

niﬁ cantly more in the explicit run, may have

other explanations and should therefore be

further investigated.

CONCLUSIONS AND FURTHER

RESEARCH

Although without obtaining deﬁ nite answers

in our ﬁ rst approach to brands ’ logos per-

ception using neuroscientiﬁ c knowledge

and methods, we inferred several abductive

concepts ( Peirce, 1931 , CP:5, pp. 188 – 191)

that can be part of such a process: emotions,

self-reference and social relevance (see

Figure 8 ). The interplay among these con-

cepts is logical under a social cognition

umbrella: any individual seeks to attain his /

her ideal social self-concept ( Johar and

Figure 8 : Constructs and model inferred in the experiment.

AUTHOR COPY

Neuroscience in branding

15

or when they are asked to report about

their own emotions ( Chamberlain and

Broderick, 2007 ).

In the present study, the meaning of the

brands was conveyed through their respec-

tive logos. Although it may be expected

that logos with their usual vivid colours,

shapes, fonts and wording express a large

amount of brand experience, in fact only

the visual sense was recruited (in addition

to the noise, the scanner imposes some

limitations to interact with subjects, which

largely compromise the recruitment of the

other senses, although some researchers

have tried to overcome such limitations

with success ( Plassmann et al , 2008 )). How-

ever, it may be expected that brand

know ledge is also suggested through other

means, for example sounds or smells that

were not tested in the present study. It

would be interesting to have a parallel

investigation of the neural bases of these

brand imprints, either by fMRI, or by

other techniques. Such investigations could

reveal if it is possible to deliver the brand ’ s

meaning through several senses and if these

channels converge or act in tandem provi-

ding complementary dimensions. In such a

case, it would be also important to probe

whether the social brain still has a role in

brand perception, in the same way as in

our study.

Strengthening further the model of brand

perception that we have proposed may

yield useful guidelines for practitioners

in the near future. It can be useful in brand

construction, focusing on emotional, self-

related and socially relevant content, which

may be assessed with neuroscientiﬁ c tech-

niques. There is already sufﬁ cient litera-

ture that considers these constructs in

branding, but novelty lies in the possibility

of accessing these dimensions with the

exemption of self-reporting and indirect

techniques. In fact, it may be now possible

to question directly the source of all beha-

vioural responses: the brain.

REFERENCES

Adolphs , R . ( 2001 ) The neurobiology of social cognition .

Current Opinion in Neurobiology 11 (2) : 231 – 239 .

Adolphs , R . ( 2003 ) Cognitive neuroscience of human

social behaviour . Nature Reviews Neuroscience 4 (3) :

165 – 178 .

Adolphs , R . ( 2004 ) Processing of emotional and

social information by the human amygdala . In:

M.S. Gazzaniga (ed.) The Cognitive Neurosciences III .

Cambridge: MIT Press , pp. 1017 – 1030 .

Adolphs , R . ( 2006 ) How do we know the minds

of others? Domain-speciﬁ city, simulation, and

enactive social cognition . Brain Research 1079 (1) :

25 – 35 .

Adolphs , R . , Tranel ,

D . and Dam á sio , A . R . ( 1998 )

The human amygdala in social judgment . Nature

393 (6684) : 470 – 474 .

Ahuvia , A . C . ( 2005 ) Beyond the extended self: Loved

objects and consumers ’ identity narratives . Journal

of Consumer Research 32 (1) : 171 – 184 .

Allen , C . T . , Fournier , S . and Miller , F . ( 2008 ) Brands

and their meaning makers . In: C.P. Haugtvedt,

P.M. Herr and F.R. Kardes (eds.) Handbook of

Consumer Psychology . New York: Taylor & Francis ,

pp. 781 – 822 .

Alvarez , J .

and Emory , E . ( 2006 ) Executive function

and the frontal lobes: A meta-analytic review . Neuro-

psychology Review 16 (1) : 17 – 42 .

Amaro Jr , E . and Barker , G . J . ( 2006 ) Study design in

fMRI: Basic principles . Brain and Cognition 60 (3) :

220 – 232 .

Ambler , T . ( 2008 ) What will neuroscience do for adver-

tisers? Is brain-scanning the answer to marketers ’

prayers? Admap 43 (490) : 24 – 26 .

Amodio , D . M . and Frith , C . D . ( 2006 ) Meeting of

minds: The medial frontal cortex and social cogni-

tion . Nature Reviews Neuroscience 7

(4) : 268 – 277 .

Ashwin , C . , Baron-Cohen , S . , Wheelwright , S . ,

O ’ Riordan , M . and Bullmore , E . T . ( 2007 ) Dif-

ferential activation of the amygdala and the ‘ social

brain ’ during fearful face-processing in Asperger

Syndrome . Neuropsychologia 45 (1) : 2 – 14 .

Bailey , C . H . and Kandel , E . R . ( 2004 ) Synaptic growth

and the persistence of long-term memory: A

molecular perspective . In: M.S. Gazzaniga (ed.)

The Cognitive Neurosciences III . Cambridge: MIT

Press , pp. 647 – 663 .

Banister , E . N

. and Hogg , M . K . ( 2004 ) Negative

symbolic consumption and consumers ’ drive for

self-esteem: The case of the fashion industry . European

Journal of Marketing 38 (7) : 850 – 868 .

Baron-Cohen , S . , Leslie , A . M . and Frith , U . ( 1985 )

Does the autistic child have a ‘ theory of mind ’ ?

Cognition 21 (1) : 37 – 46 .

Beaucousin , V . , Lacheret , A . , Turbelin , M .- R . , Morel , M . ,

Mazoyer , B . and Tzourio-Mazoyer , N . (

2007 )

FMRI study of emotional speech comprehension .

Cerebral Cortex 17 (2) : 339 – 352 .

Bechara , A . ( 2004 ) The role of emotion in decision-

making: Evidence from neurological patients with

AUTHOR COPY

Santos et al

16

orbitofrontal damage . Brain and Cognition 55 (1) :

30 – 40 .

Bechara , A . and Dam á sio , A . R . ( 2005 ) The somatic

marker hypothesis: A neural theory of econo mic

decision . Games and Economic Behavior 52 (2) :

336 – 372 .

Bechara , A . , Dam á sio , H . , Dam á sio , A . R . and Lee ,

G . P . ( 1999 ) Different contributions of the human

amygdala and ventromedial prefrontal cortex to

decision-making . The Journal of Neuroscience 19 (13) :

5473 – 5481 .

Beckmann , C . F .

, Jenkinson , M . and Smith , S . M .

( 2003 ) General multilevel linear modeling for group

analysis in FMRI . Neuroimage 20 (2) : 1052 – 1063 .

Beckmann , C . F . and Smith , S . M . ( 2004 ) Probabilistic

independent component analysis for functional

magnetic resonance imaging . IEEE Transactions on

Medical Imaging 23 (2) : 137 – 152 .

Beckmann , C . F . and Smith , S . M . ( 2005 ) Tensorial

extensions of independent component analysis for

multisubject FMRI analysis . Neuroimage 25 (1) :

294 – 311 .

Beer , J . S . , John

, O . P . , Scabini , D . and Knight , R . T .

( 2006 ) Orbitofrontal cortex and social behavior:

Integrating self-monitoring and emotion-cognition

interactions . Journal of Cognitive Neuroscience 18 (6) :

871 – 879 .

Belk , R . W . ( 1988 ) Possessions and the extended self .

Journal of Consumer Research 15 (2) : 139 – 168 .

Bernard , H . R . , Killworth , P . , Kronenfeld , D . and

Sailer , L . ( 1984 ) The problem of informant accu-

racy: The validity of retrospective data . Annual

Review of Anthropology 13 (1) : 495 – 517 .

Blumer

, H . ( 1969 ) Symbolic Interactionism: Perspective

and Method . Los Angeles: University of California

Press .

Brett , M . ( 1999 ) The MNI brain and the Talairach

atlas , http://imaging.mrc-cbu.cam.ac.uk/imaging/

MniTalairach , accessed 7 May 2010 .

Broca , P . ( 1861 ) Remarques sur le si é ge de la facult é

du langage articul é , suivies d ’ une observation

d ’ aph é mie (perte de la parole ) . Bulletin de la Soci é t é

Anatomique 6 : 330 – 357 .

Brodmann , K . ( 1909 ) Vergleichende Lokalisationslehre der

Grobhirnrinde in ihren Prinzipien dargestellt auf Grund

des Zellenbaues . Leipzig, Germany: Verlag von

Johann Ambrosius Barth .

Brunet-Gouet , E . and Decety , J . ( 2006 ) Social brain

dysfunctions in schizophrenia: A review of neuro-

imaging studies .

Psychiatry Research: Neuroimaging

148 (2 – 3) : 75 – 92 .

Chamberlain , L . and Broderick , A . ( 2007 ) The appli-

cation of physiological observation methods to

emotion research . Qualitative Market Research: An

International Journal 10 (2) : 199 – 216 .

Corbin , J . and Strauss , A . ( 2007 ) Basics of Qualitative

Research: Techniques and Procedures for Developing

Grounded Theory , 3rd edn. Thousand Oaks, CA:

Sage Publications Ltd .

Critchley , H . D . et al ( 2000 ) Explicit and implicit neural

mechanisms for processing of social information

from facial expressions: A functional magnetic

resonance imaging study . Human Brain Mapping

9 (2) : 93 – 105 .

D ’ Argembeau , A . et al (

2005 ) Self-referential reﬂ ective

activity and its relationship with rest: A PET study .

Neuroimage 25 (2) : 616 – 624 .

Dam á sio , A . R . ( 1994 ) Descarte’s Error – Emotion, Reason

and the Human Brain . New York: Penguin Putnam .

Elliott , R . and Wattanasuwan , K . ( 1998 ) Brands as

symbolic resources for the construction of identity .

International Journal of Advertising 17 (2) : 131 – 144 .

Elliott , R . , Dolan , R . J . and Frith , C . D . ( 2000 ) Dis-

sociable functions in the medial and lateral orbitof-

rontal cortex: Evidence from human neuroimaging

studies . Cerebral Cortex 10 (3) : 308 – 317 .

Ermer , E . , Guerin , S . A . , Cosmides , L . , Tooby , J . and

Miller , M . B . ( 2006 ) Theory of mind broad and

narrow: Reasoning about social exchange engages

ToM areas, precautionary reasoning does not . Social

Neuroscience 1 (3 – 4) : 196 – 219 .

Escalas , J . E . and Bettman , J . R . ( 2005 ) Self-construal,

reference groups, and brand meaning . Journal of

Consumer Research 32 (3) : 378 – 389 .

Fortin , N . J . , Wright , S . P . and Eichenbaum , H . ( 2004

)

Recollection-like memory retrieval in rats is

dependent on the hippocampus . Nature 431 (7005) :

188 – 191 .

Fournier , S . ( 1998 ) Consumers and their brands: Devel-

oping relationship theory in consumer research .

Journal of Consumer Research 24 (4) : 343 – 353 .

Frith , C . D . ( 2007 ) The social brain? Philosophical Trans-

actions of the Royal Society B 362 (1480) : 671 – 678 .

Gallagher , H . L . and Frith , C . D . ( 2003 ) Functional

imaging of ‘ theory of mind ’ . Trends in Cognitive

Sciences 7 (2) : 77 – 83 .

Gigerenzer , G .

( 2001 ) Decision making: Nonrational

theories . In: N.J. Smelser and P.B. Baltes (eds.)

International Encyclopedia of the Social and Behavioral

Sciences . Oxford: Elsevier Science , pp. 3304 – 3309 .

Goffman , E . ( 1959 ) The Presentation of Self in Everyday

Life . New York: Anchor Book .

Greenwald , A . G . and Banaji , M . R . ( 1995 ) Implicit

social cognition: Attitudes, self-esteem, and stere-

otypes . Psychological Review 102 (1) : 4 – 27 .

Grubb , E . L . and Grathwohl , H . L . ( 1967 ) Consumer self-

concept, symbolism and market behavior: A theo-

retical approach . Journal of Marketing 31 (4) : 22 – 27 .

Haynes , J .-

D . and Rees , G . ( 2006 ) Decoding mental

states from brain activity in humans . Nature Reviews

Neuroscience 7 (7) : 523 – 534 .

Hubert , M . and Kenning , P . ( 2008 ) A current overview

of consumer neuroscience . Journal of Consumer Beha-

viour 7 (4 – 5) : 272 – 292 .

Huettel , S . A . , Song , A . W . and McCarthy , G . ( 2004 )

Functional Magnetic Resonance Imaging . Sunderland,

MA: Sinauer Associates .

Hyv ä rinen , A . ( 1999 ) Fast and robust ﬁ xed-point algo-

rithms for independent component analysis . IEEE

Transactions on Neural Networks 10

(3) : 626 – 634 .

Iacoboni , M . et al ( 2004 ) Watching social interactions

produces dorsomedial prefrontal and medial parietal

AUTHOR COPY

Neuroscience in branding

17

BOLD fMRI signal increases compared to a resting

baseline . Neuroimage 21 (3) : 1167 – 1173 .

Jenkinson , M . and Smith , S . M . ( 2001 ) A global opti-

misation method for robust afﬁ ne registration of

brain images . Medical Image Analysis 5 (2) : 143 – 156 .

Jenkinson , M . , Bannister , P . R . , Brady , J . M . and

Smith , S . M . ( 2002 ) Improved optimization for

the robust and accurate linear registration and

motion correction of brain images . Neuroimage 17 (2) :

825 – 841 .

Jezzard , P . , Matthews , P . M . and

Smith , S . M . ( 2001 )

Functional MRI: An Introduction to Methods . Oxford:

Oxford University Press .

Johar , J . S . and Sirgy , J . M . ( 1991 ) Value-expressive

versus utilitarian advertising appeals: When and

why to use which appeal . Journal of Advertising 20 (3) :

23 – 33 .

Kelley , W . M . , Macrae , C . N . , Wyland , C . L . , Caglar ,

S . , Inati , S . and Heatherton , T . F . ( 2002 ) Finding

the self? An event-related fMRI study . Journal of

Cognitive Neuroscience 14 (5) : 785 –

794 .

Kleine III , R . , Kleine , S . and Kernan , J . ( 1993 )

Mundane consumption and the self: A social-iden-

tity perspective . Journal of Consumer Psychology :

209 – 235 .

LeDoux , J . ( 2007 ) The amygdala . Current Biology

17 (20) : R868 – R874 .

Lee , N . , Broderick , A . and Chamberlain , L . ( 2007 )

What is ‘ neuromarketing ’ ? A discussion and agenda

for future research . International Journal of Psycho-

physiology 63 (2) : 199 – 204 .

Ligas , M . and Cotte ,

J . ( 1999 ) The process of negoti-

ating brand meaning: A symbolic interactionist

perspective . In: E.J. Arnould and L.M. Scott (eds.)

Advances in Consumer Research . Provo, UT: Asso-

ciation for Consumer Research , pp. 609 – 614 .

Logothetis , N . K . ( 2008 ) What we can do and what we

cannot do with fMRI . Nature 453 (7197) : 869 –

878 .

Matthews , P . M . et al ( 2003 ) Towards understanding

language organisation in the brain using fMRI .

Human Brain Mapping 18 (3) : 239 – 247 .

McClure , S . M . , Li , J . , Tomlin , D . , Cypert , K . S . ,

Montague

, L . M . and Montague , P . R . ( 2004 )

Neural correlates of behavioral preference for cul-

turally familiar drinks . Neuron 44 (2) : 379 – 387 .

Miller , E . K . and Cohen , J . D . ( 2001 ) An integrative

theory of prefrontal cortex function . Annual Reviews

in Neuroscience 24 : 167 – 202 .

Minka , T . P . ( 2000 ) Automatic Choice of Dimen-

sionality for PCA . MIT Media Lab Vision and

Modeling Group, Cambridge: MIT. Technical

Report 514 .

Mitchell , J . P . , Macrae , C . N . and Banaji , M . R . ( 2004 )

Encoding-speciﬁ c effects of social cognition on the

neural correlates of subsequent memory . The Journal

of Neuroscience 24 (21) : 4912 – 4917 .

Mitchell , J . P . , Macrae , C . N . and Banaji , M . R . ( 2005 )

Forming impressions of people versus inanimate

objects: Social-cognitive processing in the medial

prefrontal cortex . Neuroimage 26 (1) : 251 – 257 .

Mitchell , J . P . , Macrae , C . N . and Banaji , M . R . ( 2006 )

Dissociable medial prefrontal contributions to

judgments of similar and dissimilar others . Neuron

50 (4) : 655 – 663 .

Montague , P . R .

, King-Casas , B . and Cohen , J . D .

( 2006 ) Imaging valuation models in human choice .

Annual Reviews in Neuroscience 29 : 417 – 448 .

Mueller , S . , Lockshin , L . and Louviere , J . J . ( 2010 )

What you see may not be what you get: Asking

consumers what matters may not reﬂ ect what they

choose . Marketing Letters 21 (4) : 335 – 350 .

Murphy , K . and Garavan , H . ( 2005 ) Deriving the

optimal number of events for an event-related

fMRI study based on the spatial extent of activation .

Neuroimage 27 (4) : 771 – 777 .

Nichols , T . E . , Brett

, M . , Andersson , J . , Wager , T .

and Poline , J .- B . ( 2005 ) Valid conjunction infe-

rence with the minimum statistic . Neuroimage 25 (3) :

653 – 660 .

Norman , K . A . , Polyn , S . M . , Detre , G . J . and Haxby , J . V .

( 2006 ) Beyond mind-reading: Multi-voxel pattern

analysis of fMRI data . Trends in Cognitive Sciences

10 (9) : 424 – 430 .

Norris , C . J . , Chen , E . E . , Zhu

, D . C . , Small , S . L . and

Cacioppo , J . T . ( 2004 ) The interaction of social and

emotional processes in the brain . Journal of Cognitive

Neuroscience 16 (10) : 1818 – 1829 .

Northoff , G . , Heinzel , A . , de Greck , M . , Bermpohl ,

F . , Dobrowolny , H . and Panksepp , J . ( 2006 )

Self-referential processing in our brain – A meta-

analysis of imaging studies on the self . Neuroimage

31 (1) : 440 – 457 .

O ’ Doherty , J . , Kringelbach , M . L

. , Rolls , E . T . ,

Hornak , J . and Andrews , C . ( 2001 ) Abstract reward

and punishment representations in the human orbit-

ofrontal cortex . Nature Neuroscience 4 (1) : 95 – 102 .

Paller , K . A . and Wagner , A . D . ( 2002 ) Observing the

transformation of experience into memory . Trends

in Cognitive Sciences 6 (2) : 93 – 102 .

Peirce , C . S . ( 1931 ) Collected Papers of Charles Sanders

Peirce [1] – Pragmatism and Pragmaticism , Harvard

University Press [online], http://www.textlog.de/

peirce_pragmatism.html , accessed 4 April 2009 .

Pelzmann , L . , Hudnik , U .

and Miklautz , M . ( 2005 )

Reasoning or reacting to others? How consumers

use the rationality of other consumers . Brain Research

Bulletin 67 (5) : 438 – 442 .

Penn , D . ( 2008 ) Beyond neuroscience – Whatever hap-

pened to neuromarketing? Admap 43 (490) : 27 – 29 .

Plassmann , H . , Ambler , T . , Braeutigam , S . and

Kenning , P . ( 2007 ) What can advertisers learn

from neuroscience? International Journal of Adver-

tising 26 (2) : 151 – 175 .

Plassmann , H . , O ’ Doherty , J . , Shiv , B .

and Rangel , A .

( 2008 ) Marketing actions can modulate neural rep-

resentations of experienced pleasantness . Proceedings

of the National Academy of Sciences of the USA 105 (3) :

1050 – 1054 .

Poldrack , R . A . ( 2006 ) Can cognitive processes be

inferred from neuroimaging data? Trends in Cognitive

Sciences 10 (2) : 59 – 63 .

AUTHOR COPY

Santos et al

18

Poldrack , R . A . ( 2008 ) The role of fMRI in cognitive

neuroscience: Where do we stand? Current Opinion

in Neurobiology 18 (2) : 223 – 227 .

Rilling , J . K . , Sanfey , A . G . , Aronson , J . A . , Nystrom ,

L . E . and Cohen , J . D . ( 2004 ) The neural correlates

of theory of mind within interpersonal interactions .

Neuroimage 22 (4) : 1694 – 1703 .

Rolls , E . T . ( 2000 ) Precis of ‘ the brain and emotion ’ .

Behavioral and Brain Sciences 23 (2) : 177 – 191

, discussion

192 – 233 .

Rolls , E . T . ( 2004 ) The functions of the orbitofrontal

cortex . Brain and Cognition 55 (1) : 11 – 29 .

Ruby , P . and Decety , J . ( 2004 ) How would you feel

versus how do you think she would feel? A neu-

roimaging study of perspective-taking with social

emotions . Journal of Cognitive Neuroscience 16 (6) :

988 – 999 .

Saxe , R . ( 2006 ) Four brain regions for one theory of

mind? In: J.T. Cacioppo, P.S. Visser and C.L.

Pickett (eds.) Social Neuroscience: People Thinking

about Thinking People . Cambridge: MIT Press ,

pp. 83 – 101 .

Schaefer , M . , Berens , H . , Heinze , H .- J .

and Rotte , M .

( 2006 ) Neural correlates of culturally familiar

brands of car manufacturers . Neuroimage 31 (2) :

861 – 865 .

Schilbach , L . , Eickhoff , S . B . , Rotarska-Jagiela , A . ,

Fink , G . R . and Vogeley , K . ( 2008 ) Minds at rest?

Social cognition as the default mode of cognizing

and its putative relationship to the ‘ default system ’

of the brain . Consciousness and Cognition 17 (2) :

457 – 467 .

Selten , R . ( 2002 ) What is bounded rationality? In:

G. Gigerenzer and R. Selten (eds.) Bounded Ration-

ality: The Adaptative Toolbox . Cambridge: MIT

Press , pp. 13 – 36 .

Sirgy , M . J . ( 1982 ) Self-concept in consumer behavior:

A critical review . Journal of Consumer Research 9 (3) :

287 – 300 .

Smith , S . M . ( 2002 ) Fast robust automated brain extrac-

tion . Human Brain Mapping 17 (3) : 143 – 155 .

Smith , S . M . et al ( 2004 ) Advances in functional and

structural MR image analysis and implementation as

FSL . Neuroimage 23 (Supplement 1) : S208 – S219 .

Squire , L . R . ( 2004 ) Memory systems of the brain: A

brief history and current perspective . Neurobiology

Learning and Memory 82 (3) : 171 – 177 .

Stefanacci , L .

and Amaral , D . G . ( 2002 ) Some observa-

tions on cortical inputs to the macaque monkey

amygdala: An anterograde tracing study . Journal of

Comparative Neurology 451 (4) : 301 – 323 .

Stone , V . E . ( 2006 ) Theory of mind and the evolution

of social intelligence . In: J.T. Cacioppo, P.S. Visser

and C.L. Pickett (eds.) Social Neuroscience: People

Thinking about Thinking People . Cambridge: MIT

Press , pp. 103 – 129 .

Strauss , A . and Corbin , J . ( 1990 ) Basics of Qualitative

Resea rch: Techniques and Procedures for Developing

Groun ded Theory . Thousand Oaks, CA: Sage Pub-

lications .

Stryker , S . ( 1990 ) Symbolic interactionism: Themes and

variations . In: M. Rosenberg and R. Turner (eds.)

Social Psychology: Sociological Perspectives . New Brunswick,

NJ: Transaction Publishers , pp. 3 – 29

.

Talairach , J . and Tournoux , P . ( 1988 ) Co-planar Stere-

otaxic Atlas of the Human Brain: 3-dimensional Pro-

portional System: An Approach to Cerebral Imaging .

New York: Thieme Medical Publishers .

Todd , P . M . and Gigerenzer , G . ( 2003 ) Bounded

rationality to the world . Journal of Economic Psy-

chology 24 (2) : 143 – 165 .

Tong , F . ( 2011 ) Aligning brains and minds . Neuron

72 (2) : 199 – 201 .

Ursu , S . and Carter , C . S . ( 2005 ) Outcome representa-

tions, counterfactual comparisons and the human

orbitofrontal cortex: Implications for neuroimaging

studies of decision-making . Cognitive Brain Research

23 (1) : 51 – 60 .

Walvis , T . H . ( 2007 ) Three laws of branding: Neuro-

scientiﬁ c foundations of effective brand building .

Journal of Brand Management 16 (3) : 176 – 194 .

Windmann , S . et al ( 2006 ) On framing effects in decision

making: Linking lateral versus medial orbitofrontal

cortex activation to choice outcome processing .

Journal of Cognitive Neuroscience 18 (7) : 1198 – 1211 .

Woolrich , M . W . ( 2008 ) Robust group analysis using

outlier inference . Neuroimage 41 (2) : 286 – 301 .

Woolrich , M . W . , Behrens , T . E . , Beckmann , C . F . ,

Jenkinson , M . and Smith

, S . M . ( 2004 ) Multilevel

linear modelling for FMRI group analysis using

Bayesian inference . Neuroimage 21 (4) : 1732 – 1747 .

Woolrich , M . W . , Ripley , B . D . , Brady , J . M . and Smith ,

S . M . ( 2001 ) Temporal autocorrelation in univariate

linear modeling of FMRI data . Neuroimage 14 (6) :

1370 – 1386 .

Worsley , K . J . ( 2001 ) Statistical analysis of activation

images . In: P. Jezzard, P.M. Matthews and S.M. Smith

(eds.) Functional MRI: An Introduction to Methods . New

York: Oxford University Press , pp. 251 – 270 .

Yonelinas , A . P

. ( 2002 ) The nature of recollection and

familiarity: A review of 30 years of research . Journal

of Memory and Language 46 (3) : 441 – 517 .

Zald , D . H . ( 2003 ) The human amygdala and the emo-

tional evaluation of sensory stimuli . Brain Research

Reviews 41 (1) : 88 – 123 .

Zaltman , G . ( 2003 ) How Customers Think: Essential

Insights into the Mind of the Market . Boston, MA:

Harvard Business School Press .

AUTHOR COPY

Neuroscience in branding

19

APPENDIX

This section is intended to give a basic

knowledge in order to interpret the results

and conclusions in this article correctly. We

recommend complementary readings that

focus on some important aspects on this

technique so that deeper understanding

may be achieved ( Jezzard et al , 2001 ; Huettel

et al , 2004 ).

What is measured in fMRI?

The last word in fMRI, imaging, means

that this technique delivers mainly images.

In this article, the ﬁ gures of brain images

included are used to report the results. In

fact, these pictures are not like common

photos; they result from the overlaying of

two distinct images. The one that lies below

is the anatomic and is an image of the

physical brain, that is, its biological tissues.

It is acquired using a T1-weighted

MPRAGE protocol (see Methods section)

and is ﬁ nely detailed. The side of a pixel

is usually around 1 mm (in the present study

it is 0.75 mm). The image that is above

corresponds to the ﬁ rst letter of fMRI:

functional. This image represents the blood-

oxygen-level-dependent (BOLD) signal,

is acquired at a different stage (although

usually within the same session) using a

T2 * -weighted EPI sequence (see Methods

section), and it is not as detailed as the ana-

tomical; usually the side of the pixels is

around 3 mm (as in the present study).

The purpose of the anatomical image is

to indicate the place in the brain to which

the functional data belongs. As the brain is

anatomically divided, it is then possible to

assign a certain function to a speciﬁ c brain

region. However, this matter is not con-

sensual and there is enough evidence that

the functional organisation does not coin-

cide with the anatomical division ( Tong,

2011 ). In any case, naming structures in the

brain where there is activity helps in com-

municating the ﬁ ndings among the scien-

tiﬁ c community.

The BOLD signal, the essence of fMRI

technique, is far from being understood

( Logothetis, 2008 ). It is believed that it

indirectly represents the neural activity.

Supposedly, the BOLD signal is generated

when oxyhaemoglobin molecules in red

blood cells release oxygen, transforming

into deoxyhaemoglobin. The former mole-

cule is diamagnetic, whereas the latter is

paramagnetic. This distinct magnetic beha -

viour is detectable and measurable in mag-

netic resonance scanners and constitutes

the core of the T2 * -weighted EPI (func-

tional) acquisitions.

Thus, in fMRI, perturbations in the

magnetic ﬁ eld that occur when oxyhaemo-

globin liberates oxygen are the target of

measurement. This is an indirect assessment

of neuronal activity. Supposedly, the con-

sumption of glucose and oxygen increases

through the delivery of energy when neu-

rons are recruited for a cognitive process,

and promotes the chemical reaction that

transforms oxyhaemoglobin in a paramag-

netic compound (deoxyhaemoglobin). All

this process has a magnetic imprint that is

measurable.

This means that absolute quantiﬁ cations

of neuronal activity in fMRI are neither

practical nor useful. If the purpose of the

study is to realise whether a certain brain

region participates or not in a process, it

has to be put in two (at least) different stages

of oxygen consumption (activation and

resting state). It is the contrast between

these two stages that will allow the com-

putation of the differences between image

signal intensities, supposedly proportional

to active neuronal performance. Activated

brain areas will show signal intensity vari-

ations, whereas non-requested areas will

not have fMRI-detectable hemodynamic

changes. Moreover, when two stages are

compared, if the signal in the stage of

interest is stronger than in the other stage,

activation in that brain region has occurred;

when the opposite is shown, it is called

AUTHOR COPY

Santos et al

20

deactivation. Thus, there is activation of

a certain brain region if the stimulus of

interest participates more than the contrast

situations, and deactivation when the oppo-

site happens. These phenomena are marked

in functional images. Contrarily to the ana-

tomical image, which is usually shown in

grey scale, the functional image is colour-

coded. Two ranges of colours are estab-

lished: red to yellow representing increasing

activations, and blue to light blue repre-

senting increasing deactivations.

Functional data are commonly analysed

with the GLM. This method yields statis-

tical signiﬁ cances, which means that the

more the signal follows the model, the

higher is the positive signiﬁ cance, which is

coded with the range red to yellow; on the

contrary, the higher is the negative signiﬁ -

cance, the more the signal follows the

inverse o

Neuroscience in branding: A functional magnetic resonance imaging study on brands' implicit and explicit impressions

Abstract and Figures