See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/233924863

Brand Emotional Connection and Loyalty

Article  in  Journal of Brand Management · February 2012

DOI: 10.1057/bm.2012.3

CITATIONS READS

220 19,341

3 authors:

Sandra Maria Correia Loureiro Hans Ruediger Kaufmann

ISCTE-Instituto Universitário de Lisboa University of Nicosia
367 PUBLICATIONS   7,134 CITATIONS    134 PUBLICATIONS   2,324 CITATIONS   

SEE PROFILE SEE PROFILE

Demetris Vrontis
University of Nicosia
472 PUBLICATIONS   9,987 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Economics & Business Ph D. View project

Family businesses View project

All content following this page was uploaded by Sandra Maria Correia Loureiro on 23 May 2014.

The user has requested enhancement of the downloaded file.

 Original Article

Brand emotional connection and

loyalty
Received (in revised form): 4th January 2012

Sandra Maria Correia Loureiro

is Professor of Marketing at the University of Aveiro and a researcher in the GOVCOPP and SOCIUS research center. Her
current research interests include consumer-brand relationship, online environment and purchase, tourism experience,
and corporate social responsibility issues. Her papers were published in a variety of peer reviewed journals and at well
reputed international conferences. She has acted as a reviewer for several international journals and conferences. She has
participated in several research projects funded by the EU and FCT (Foundation for Science and Technology).

Kaufmann Hans Ruediger

worked in various functions for Manchester Metropolitan University and at several European academic institutions
in Budapest and Liechtenstein, after extensive experience in German Bank Management. Since October 2006 he is
Associate Professor at the University of Nicosia, was a launching member and President (2007–2009) of CIRCLE and
Vice-President of EMBRI, two research networks on Consumer Behavior and Management, respectively. He is Associate
Editor of the World Review of Entrepreneurship, Management and Sustainable Development.

Vrontis Demetris
is Professor of Marketing and Dean of the School of Business at the University of Nicosia in Cyprus. He is the Editor of the
EuroMed Journal of Business (EMJB) and the President of the EuroMed Research Business Institute (EMRBI). His prime research
interests are on international marketing, marketing planning, branding and marketing communications, areas that he has
widely published in over 70 refereed journal articles, contributed chapters and cases in books and presented papers to global
conferences. Dr Demetris is also the author of 11 books in the areas of international marketing and marketing planning.

ABSTRACT This study, for the first time, aims to integrate brand attachment as an
antecedent of brand love and both, affective commitment and brand trust as
mediators between brand love and loyalty. It is also the first time that differences
between male and female consumers’ perceptions on this topic have been compared
in literature. On the basis of previous research, a model was developed and validated
using consumers who bought one of three car brands: Toyota, Ford and Renault.
Living in three regions of Portugal, the sample consisted of consumers who bought
a new car 2 years before the research and had declared being satisfied with the
purchase. The model is tested in the context of a non-hedonic product, differing
from the common perspective that focuses on hedonic products. Hypotheses were
tested by employing multi-group structural equation modeling. Findings suggest that
brand attachment is positively related to brand love. Brand love reinforces the trust,
interest in continuing a relationship and faith in the future of the brand. Women
exhibited trust and placed more importance on dyadic relationships than did men.
Correspondence: Men, however, desired to identify socially with the brand.
Vrontis Demetris
School of Business, University of Journal of Brand Management advance online publication, 24 February 2012;
Nicosia, 46 Makedonitissas Ave.,
PO Box 24005, doi:10.1057/bm.2012.3
1700 Nicosia, Cyprus
E-mails: vrontis.d@unic.ac.cy;
sandra.loureiro@ua.pt Keywords: brand love; brand attachment; loyalty intentions

© 2012 Macmillan Publishers Ltd. 1350-23IX Journal of Brand Management 1–15

www.palgrave-journals.com/bm/
 Loureiro et al

INTRODUCTION high perception of brand love reinforces

In recent years, branding theory literature the faith, the relationship, and the trust
has focused on the customer–brand rela- regarding the brand and increases the con-
tionship, which is based on a metaphor that sumer’s retention.
suggests consumers form relationships with The findings will be particularly inter-
brands in the same way they form relations esting to researchers and managers in
in a social context (Fournier, 1998; Parvatiyar organizations that are seeking to understand
and Sheth, 2001; McAlexander et al, 2002). the relationship between their brands
The concept is multifaceted in nature and (corporate or product, hedonic or utili-
reflects affective and socio-emotional attach- tarian) and consumers and the importance
ments, behavior ties and supportive cogni- of properly communicating the brand
tive beliefs. Related to this stream of literature, values and characteristics to create prox-
several relevant concepts have been devel- imity with consumers (self-connection and
oped by relationship marketing research, attraction), which leads to a strong passion
such as trust or commitment (for example, and a continuous commitment that retain
Morgan and Hunt, 1994; Garbarino and the consumer and lead him/her to forgive
Johnson, 1999; Johnson et al, 2001; Hansen possible brand (organization) mistakes.
et al, 2003; Fullerton, 2005). Emotional ties, The article begins with the presentation
connection and attachment characteristics of and discussion of theories and models
the customer–brand relationship together that embrace the concept of love in the
with providing a positive experience to consumer–brand relationship. Second, we
customers and creating the possibility for propose and empirically test a model that
consumers to have an intense emotional extends the Carrol’s and Ahuvia’s model
attachment to loved objects led Carroll and (2006) to non-hedonic products and
Ahuvia (2006) to introduce the concept of includes some antecedents (brand attach-
brand love. ment) and outcomes (commitment and
The purpose of the present study is to trust) of brand love. Finally, we discuss the
integrate the so far scattered brand attach- results, point out limitations, and provide
ment, commitment and trust concepts into implications for brand managers and further
Carroll and Ahuvia’s (2006) model of brand research opportunities.
love. It is the first time that the model is
tested in a context of non-hedonic products. THEORETICAL BACKGROUND AND
The second intention is to compare the dif- HYPOTHESES DEVELOPMENT
ferences between male and female consumer In the field of the dyadic relationship
perceptions; to find if significant differences between brand and consumer, extant studies
exist between the sexes in the context of use a relationship metaphor from interper-
Portugal, a Mediterranean country. sonal relationship theory. According to
The Partial Least Squares (PLS) technique Fournier’s (1998) brand relationship quality
was used to test the model in the context model, there are six dimensions including
of a product category that can be classified attachment (love/passion and self-connec-
as high involvement but also utilitarian tion), behavioral ties (interdependence and
(auto industry). The new model implies commitment) and supportive cognitive
that emotional identification and attach- beliefs (intimacy and brand partner quality).
ment improves the relationship between These dimensions are based on theories
the consumer and the brand, strengthens such as interdependence theory and theo-
the positive feeling toward a brand, and ries of attraction (Kelley and Thibaut, 1978;
leads to passion and love. In this wave, a Hinde, 1979; Kelley, 1979; Rusbult and

2 © 2012 Macmillan Publishers Ltd. 1350-23IX Journal of Brand Management 1–15

 Brand emotional connection and loyalty

Van Lange, 2003). However, love and brand and declaration of love for the brand.
passion are the essential concepts in the More recently, Albert et al (2008, 2009)
attachment theory (Bowlby, 1979). propose two main components of brand
Attachment theory is based on studies of love: six first-order dimensions (idealiza-
parent–child relationships (Bowlby, 1979) tion, intimacy, pleasure, dream, memories
and continues through adulthood to romantic and uniqueness) and two second-order
relationships (Hazan and Shaver, 1994), kin- dimensions (passion and affection).
ships and friendships (Weiss, 1988; Trinke Regarding models that integrate the
and Bartholomew, 1997). Several studies brand love concept, we can find three: the
suggest that emotional attachment is formed holistic approach proposed by Fournier
between human beings and animals, places, (1998), the causal approach of Carroll and
destinations, special objects, brands (for example Ahuvia (2006), and the systemic approach
Richins, 1994; Schouten and McAlexander, of Batra et al (2008). As mentioned above,
1995; Price et al, 2000; Yuksel et al, 2010), Fournier (1998) suggests six strong relation-
and even human brands or celebrities ships and is interested in the different rela-
(Thomson, 2006). In consumer research, tionships a consumer may have with a
Thomson et al (2005) identified a higher- brand. Batra et al (2008) integrate various
order emotional attachment construct con- constructs like brand loyalty, satisfaction
sisting of three factors: affection, passion and and attitude strength into a model of brand
connection. Thus, attachment reflects an love relationship; however, they do not dif-
emotional bond similar to love. ferentiate between antecedents and out-
The interpersonal triangular theory of comes of the consumer’s love for a brand
love (Sternberg, 1986) adapted to the con- and do not explicitly use the construct of
sumption context is closely related to the love. Reversely, Carrol and Ahuvia (2006)
emotional attachment construct; it con- model the relationships between brand love
siders that brand love is made up of dimen- and both its antecedents and consequences.
sions such as: passion, intimacy and They propose that brand love is influenced
commitment (Kamat and Parulekar, 2007; by a hedonic product and self-expressive
Keh et al, 2007). Ahuvia (1993) suggests brand and has a positive effect on brand
that consumers can have real feelings of loyalty and positive word-of-mouth.
love toward an object and conceptualizes Carroll and Ahuvia (2006) highlight that
the love as having two dimensions: real and brand love includes a willingness to declare
desired integration. Batra et al (2008) pro- love (as if the brand were a person) and
pose that love is made up of seven dimen- involves integration of the brand into a
sions: perceived functional quality, consumer’s identity. A consumer’s love
self-related cognitions, positive affect, neg- toward a brand is greater for brands that
ative affect, satisfaction, attitude strength play a significant role in shaping the con-
and loyalty. sumer’s identity. Therefore, a self-expres-
Given the complexity of the construct, sive brand is the consumer’s perception of
interpersonal love theory alone does not the degree to which the specific brand
allow for the discovery of specific dimen- enhances one’s social self and/or reflects
sions of love. On the basis of the brand– one’s inner self (Carroll and Ahuvia, 2006,
consumer relationship paradigm, Carroll p. 82); it includes two dimensions: inner
and Ahuvia (2006) suggest that brand love self and social self. Consumers satisfied and
is composed of five dimensions: passion, in love with a brand are more willing to
attachment, positive evaluation of the repurchase and to recommend it to others.
brand, positive emotions in response to the On the basis of the above considerations,

© 2012 Macmillan Publishers Ltd. 1350-23IX Journal of Brand Management 1–15 3

 Loureiro et al

Connection

Self- Commitment
H3 H4
expressive H2
brand
H7
H5
Brand love f. Loyalty I.

H1 H8
Brand H6 Brand
Attachment trust

Figure 1: Proposed conceptual model.

this research proposes a model shown in the brand is irreplaceable and miss it when
Figure 1 that links brand attachment, com- they do not have the brand.
mitment and trust to Carrol and Ahuvia’s As we stated previously, Carroll and
model (2006). Ahuvia (2006) suggest that consumer love
The proposed model presents brand love is greater for brands that reflect their inner
feelings as a central construct. In this study self and/or their social self. Thus, self-
we conceptualize the brand love feelings as expressive brand concerns the fit between
the passion, positive evaluation of the the brand and the person (Power et al,
brand, positive emotions in response to the 2008; Vlachos et al, 2010). Given this con-
brand and declaration of love for the brand, ceptualization, the following hypotheses are
as proposed by Carroll and Ahuvia (2006) proposed:
for love brand; however, we do not include
the attachment component. In the Chang Hypothesis 1: Brand attachment has a
and Chieng (2006) study based on con- positive impact on brand love feeling.
sumer–brand relationship quality of Fournier
(1998), love and attachment are regarded Hypothesis 2: Self-expressive brand has a
as separate dimensions. In this perspective, positive impact on brand love feeling.
our study points out brand love feelings and
brand attachment as two variables. Regarding the consequence of brand
Serving as antecedents are self-expressive love, Assael (1987) defines brand loyalty as
brand and brand attachment; both compose commitment to a certain brand with com-
the connection to the brand representing mitment being defined as ‘an enduring
the factor known as Consumers’ Emotional desire to maintain a valued relationship’
Brand Attachments proposed by Thomson (Moorman et al, 1992, p. 316). In later
et al (2005). Hazan and Shaver (1990) the- research, brand commitment was consid-
orize that love is fundamentally an attach- ered a relational dimension of loyalty
ment process through which attachment and as a variable at the core of the con-
bonds are formed. The attachment process sumer–brand relationships (Fournier, 1998;
and the fear to lose the partner should Chaudhuri and Holbrook, 2001).
increase the passionate love (Hatfield and Commitment is different from brand
Rapson, 1993).To love a brand, consumers love; Fehr (1988) demonstrates that they are
need to be attached to it, that is feel that two distinct constructs. In consumer–brand

4 © 2012 Macmillan Publishers Ltd. 1350-23IX Journal of Brand Management 1–15

 Brand emotional connection and loyalty

relationship, quality love/passion and Moreover, trust has been considered an

commitment are two different dimensions important determinant of commitment
(Fournier, 1998) and Chang and Chieng (Achrol, 1991; Morgan and Hunt, 1994;
(2006) also formulate love and commit- Garbarino and Johnson, 1999; Chaudhuri
ment as two relationship dimensions. A and Holbrook, 2001; Kingshott and
loyal consumer can be regarded as a com- Pecotich, 2007). Theories of trust and com-
mitted consumer (Oliver, 1999), suggesting mitment also propose that both constructs
that a consumer with faith in his/her future are influential in future loyalty intentions of
with a brand wants to continue the rela- an exchange partner (Morgan and Hunt,
tionship with a brand (being committed), 1994; Garbarino and Johnson, 1999; Dagger
and is more willing to buy the same brand and O’Brien, 2010). Satisfied consumers
in the future and to recommend it to others who also express brand love are more con-
(being loyal). fident in their relationship with the brand;
Brand love is an antecedent of loyalty and they consider that the brand is reliable.
word-of-mouth (Thomson et al, 2005; Therefore, the following hypotheses are
Carroll and Ahuvia, 2006). Similarly, Johnson proposed.
et al (2006) demonstrated that commitment
(the affective component, as we mention in Hypothesis 6: Brand love feeling has a
this study) has a direct and positive effect on positive impact on brand trust.
loyalty intentions (intention to re-purchase
and word-of-mouth). Consequently, the Hypothesis 7: Brand trust has a positive
following hypotheses are proposed: impact on commitment.

Hypothesis 3: Brand love feeling has a Hypothesis 8: Brand trust has a positive
positive impact on commitment. impact on loyalty intentions.

Hypothesis 4: Commitment has a posi- Several theories offer evidence to explain

tive impact on loyalty intentions. different perceptions and behaviors of men
and women. Cross and Madson (1997) sug-
Hypothesis 5: Brand love feeling has a gested that women are more interdependent
positive impact on loyalty intentions. whereas men are more independent. How-
ever, self-construal theory (Baumeister and
Trust is another construct regarded in Sommer, 1997) suggests that a fundamental
this study as a consequent of love, passion, difference between men and women lies in
and positive evaluation and emotions of the duality relational versus collective interde-
brand. Moorman et al (1992, p. 82) define pendence. Seeley et al (2003) examined
trust as ‘a willingness to rely on an exchange whether gender differences in interdepend-
partner in whom one has confidence’. ence were due to differences in attachment
More specifically, Morgan and Hunt (1994, to groups rather than differences in the type
p. 23) define trust as the perception of ‘con- of group (relational or collective). Women
fidence in the exchange partner’s reliability are more likely to establish and maintain
and integrity’. These definitions highlight relationship with friends, family and social
the importance of confidence and reliability contexts, and focus on dyadic bonds
in the conception of trust. Trust exists more than men do (Gardner et al, 2002;
between the two love partners (Regan et al, Markovits et al, 2006). Men focus more on
1998) implying that trust is a consequence a group that includes identification with a
of having feelings of love for a brand. social category such as college student,

© 2012 Macmillan Publishers Ltd. 1350-23IX Journal of Brand Management 1–15 5

 Loureiro et al

hometown or gender group (Sedikides and lation of consumption experience. Three

Brewer, 2001). hundred and twenty-nine car users were
Hence, women may place greater value in willing to fill in the questionnaire and did
attachment to individual members of a group so between May 2009 and September of
(relational attachment) whereas men may 2009 (response rate of 88 per cent). The
place greater value in attachment to the group brands selected for this study were chosen
identity (collective attachment). These find- because they are car brands that sold the
ings have implications on how people per- most during 2007 and 2008 (ACP, 2009)
ceive information. Schema theory ‘construes and because they originate from three dif-
perception as a constructive process wherein ferent countries. The reason for that lies in
what is perceived is a product of the interac- findings from Zeugner-Routh et al (2008)
tion between the incoming information and arguing that mere product association with
the perceiver’s pre-existing schema’ (Bem, a particular country significantly impacts
1981, p. 355). Consequently, people tend to consumers’ perceived quality, loyalty and
process information based on the sex-linked awareness/associations of such products.
associations that constitute gender schema, that Therefore, in this research we tried to con-
is gender roles. Men are guided by self, agentic trol such an effect.
goals while women pursue communal con- Over 50 per cent of the sample fell into
cerns (Meyers-Levy, 1988). According to He the age group 31–50 years. About 20 per
et al (2008), men are sensitive to achievement cent of respondents were under 31 years
due to agentic orientation whereas women and approximately the same percent of
are sensitive to the prevention of losses. respondents were over 50 (Table 1). The
car owners of the three brands were split:
METHODS 35.0 per cent Renault, 33.4 per cent Ford
and 31.6 per cent Toyota.
Sample and data collection
The questionnaire that captured both latent Measurement
and demographic variables was pre-tested The constructs under study were measured by
using six car owners personally interviewed means of multi-item scales adapted from the
in three car dealers (car stands that sell cars existing literature. The items in the question-
and perform car maintenance). Then, 374 naire were first written in English, translated
car owners who lived in three regions of into Portuguese, and then back translated to
Portugal (Oporto, Baixo Vouga and English. Back translation was used to ensure
Tâmega) were contacted through the car that the items in Portuguese communicated
brand dealers. The car owners purchased a similar information as those in English (Brislin,
new car 2 years prior and – in surveys car- 1970; Sekaran, 1983) meaning that conceptual
ried out by the car manufacturers – declared equivalence was assured.
being satisfied with their purchase. The
2-year delay between the purchase and the Table 1: Demographic characteristic of car users and
study was necessary to ensure that the sub- number of respondents by region

jects gained sufficient experience with the Gender (%) Age (%) Region (%)
car and the services subsequently provided.
Male: 59.6 18–20: 2.4 OPorto: 30.4
As Evard and Aurier (1996) point out, Female: 40.4 21–30: 31.0 Baixo Vouga: 48.6
truthful and strong relationships between a 31–40: 27.4 Tâmega: 21.0
consumer and a company or between the 41–50: 24.0
brand and the consumer do not occur in a 51–60: 11.2
> 60: 4.0
single encounter; they result from accumu-

6 © 2012 Macmillan Publishers Ltd. 1350-23IX Journal of Brand Management 1–15

 Brand emotional connection and loyalty

As suggested by Carroll and Ahuvia (2006), to explain the variance of the constructs in
the pre-test is also used to adapt the constructs the model (Chin, 1998). In terms of analysis
to the European context and non-hedonic advantages, PLS simultaneously estimates
products. Especially, in the case of brand path coefficients and individual item load-
love feeling, our intention was to capture ings in the context of a specified model. As
the essence of feelings of love with a brand, a result, it enables researchers to avoid biased
and so, items with ‘neutral feelings’ or not and inconsistent parameter estimates. On
directly related to love feelings were not the basis of recent developments (Chin et al,
used. As to what concerns to commitment, 2003), PLS has been found to be an effective
the study intends to analyze the more analytical tool to test interactions by reducing
affective components and not the calculative Type II errors. By creating a latent construct
components of commitment, as do Chang that represents an interaction term, a PLS
and Chieng (2006) and Thomson et al approach significantly reduces this problem
(2005). by accounting for error related to the meas-
In this vein, self-expressive brand was ures (Echambadi et al, 2006).
measured using six items (see items in Tenenhaus et al (2005) propose the geo-
Table 2), and brand love feeling with five metric mean of the average communality
items (Carroll and Ahuvia, 2006; Vlachos (outer mode) and the average R2 (inner
et al, 2010). Attachment to brand was meas- model) that is varied between the values of
ured with three items based on Fournier 0 and 1 to be used in calculation of the overall
(1998), Chang and Chieng (2006), and goodness-of-fit (GoF) measure for PLS (cross
Thomson et al (2005). Commitment and validated PLS GoF). The relationship between
brand trust were measured with two items these measures is given as:
each (Morgan and Hunt, 1994; Fournier,
1998; Garbarino and Johnson, 1999; Johnson GoF = communality .R 2 (1)
et al, 2006). Finally, loyalty intentions were
assessed with five items adapted from scales Finally, the differences between male and
previously developed and used to measure female are compared by splitting the full
loyalty intentions (Zeithaml et al, 1996; sample by sex and using a t-test (where
Johnson et al, 2006). Respondents rated all m = male subsample size and n = female sub-
measures on a 5-point Likert scale ranging sample size), according to the equation 2.
from 1 (Completely Disagree) to 5 (Com- This test uses the path coefficients and the
pletely Agree). standard errors of the structural paths cal-
culated by PLS with the two subsamples.
( b Male 1 − b Female 2 )
t=
⎛ 1 1⎞
Sp ⎜ + ⎟
⎝ m n⎠

⎡ (m − 1)2 (n − 1)2 ⎤
Sp = ⎢ × SE Male
2
+ × SE Female
2
2⎥ ( 2)
⎣ (m + n − 2) (m + n − 2)
1
⎦

Data analysis
A structural equation model approach using Owing the small sample size of each
PLS (Ringle et al, 2005) was employed to subsample and the exploratory nature of
test the hypotheses of this study. PLS is based the research, the PLS approach becomes
on an iterative combination of principal the most adequate technique (Chin et al,
components analysis and regression; it aims 2003).

© 2012 Macmillan Publishers Ltd. 1350-23IX Journal of Brand Management 1–15 7

 Loureiro et al

Table 2: Measurement results

Construct Mean LV index Item Composite AVEa

(Standard values loading reliability
deviation)

Brand loveb — 2.6 — 0.929 0.723

This is a wonderful brand. 2.8 — 0.863 — —
(1.13) — — — —
This brand makes me feel good. 2.8 — 0.834 — —
(1.11) — — — —
This brand makes me very happy. 2.6 — 0.863 — —
(1.09) — — — —
This brand is a delight. 2.6 — 0.856 — —
(1.10) — — — —
I am passionate about this brand. 2.4 — 0.846 — —
(1.17) — — — —

Self-expressive brandb — 2.2 — 0.957 0.787

This brand symbolizes the kind of person I really. 2.1 — 0.924 — —
am inside (1.12) — — — —
This brand reflects my personality. 2.1 — 0.940 — —
(1.16) — — — —
This brand is an extension of my inner self. 2.1 — 0.932 — —
(1.11) — — — —
This brand mirrors the real me. 2.0 — 0.926 — —
(1.14) — — — —
This brand contributes to my image. 2.4 — 0.816 — —
(1.14) — — — —
This brand has a positive impact on what others 2.4 — 0.770 — —
think of me. (1.16) — — — —

Brand attachmentb — 2.4 — 0.854 0.670

No other brand can take the place of this brand. 2.2 — — — —
(1.19) — 0.722 — —
I feel missing if I haven’t this car brand. 2.6 — — — —
(1.20) — 0.833 — —
I’m very attached to this brand. 2.5 — — — —
(1.25) — 0.860 — —

Commitmentb — 3.0 — 0.919 0.850

I have a lot of faith in my future with this brand. 2.9 — — — —
(1.20) — 0.918 — —
I want to continue my relationship with this brand. 3.1 — — — —
(1.14) — 0.926 — —

Brand trustb — 3.6 — 0.848 0.736

This brand is reliable and dependable. 3.4 — — — —
(1.10) — 0.857 — —
This brand is trustworthy. 3.7 — — — —
(0.93) — 0.859 — —
The price of my car is more than fair for the. 3.4 — — — —
performance I receive (0.95) — 0.798 — —

Loyalty I.b — 3.1 — 0.901 0.647

Next time I will definitely buy the same car. 3.0 — — — —
(or its successor) again (1.19) — 0.820 — —
If I have a problem with my car (eg, stolen) . 2.7 — — — —
I will definitely buy it again (1.20) — 0.809 — —

8 © 2012 Macmillan Publishers Ltd. 1350-23IX Journal of Brand Management 1–15

 Brand emotional connection and loyalty

Table 2 continued

Construct Mean LV index Item Composite AVEa

(Standard values loading reliability
deviation)

If I got any car for free. I would choose my car. 3.0 — — — —

(1.36) — 0.795 — —
I recommend my car to other people. 3.4 — — — —
(1.09) — 0.861 — —
I talk to other people about my car. 3.4 — — — —
(1.05) — 0.741 — —

a
AVE Average Variance Extracted.
b
The items of these constructs were evaluated with 5 point Likert scale (1=strongly disagree. 5=strongly agree).

RESULTS of manifest variables extracted by constructs

A PLS model should be analyzed and inter- (AVE) was at least 0.5; this indicates that
preted in two stages. First, the adequacy of more variance was explained – rather than
the measures (see Tables 2 and 3) is assessed unexplained – in the variables associated
by evaluating the reliability of the indi- with a given construct. Table 2 shows that
vidual measures and the discriminant the lowest value of AVE is 0.647, higher
validity of the constructs (Hulland, 1999). than the minimum value of 0.5.
Then, the structural model is evaluated. Finally, to assess discriminant validity,
Item reliability is assessed by examining the the square root of AVE should be greater
loadings of the measures on their corre- than the correlation between the construct
sponding construct. Each item loading of and other constructs in the model (Fornell
scales measuring respective constructs and Larcker, 1981). Table 3 shows that this
should approach or exceed 0.707, which criterion has been met.
indicates that over 50 per cent of the var- The full structural results are presented
iance in the observed variable is explained in Figure 2. In this study, a non-parametric
by the construct (Carmines and Zeller, approach called Bootstrapping was used to
1979). Table 2 shows that the lowest assess the precision of the PLS estimates
value of item loading is 0.722. Thus, all and support the hypotheses. Five-hundred
individual measures are reliable as all item samples sets were created to obtain 500
loadings exceed the value of 0.707. estimates for each parameter in the PLS
Composite reliability was used to analyze model. Each new sample was obtained
the internal consistency of the constructs as by a resample process and replacement of
this method is considered a more exact the original data set (Fornell and Larcker,
measure than Cronbach’s  (Fornell and 1981; Chin, 1998). Path coefficients,
Larcker, 1981). Table 2 shows that all con- interpreted similar to standardized betas,
structs are reliable as the composite relia- indicate the strength of the direct rela-
bility values exceed the threshold of 0.7 tionship between constructs. All path
and exceed even the stricter threshold of coefficients were found to be significant
0.8 (Nunnally, 1978). In fact, the lowest at the 0.001 (Hypotheses 1, 2, 3, 4 and 6)
value of composite reliability is 0.848 (see or 0.05 (Hypotheses 5 and 7) levels
Table 2). except for the relationship between brand
The measures also demonstrate that the trust and loyalty intentions. Consequently,
convergent validity as the average variance all hypotheses were supported except

© 2012 Macmillan Publishers Ltd. 1350-23IX Journal of Brand Management 1–15 9

 Loureiro et al

Table 3: Discriminant validity analysis: Square root of AVE and correlations of constructs

Construct Commitment Brand love F. Loyalty I. Brand Self-expressive Brand trust

attachment brand

AVE ½ 0.92 0.85 0.80 0.82 0.89 0.86

Commitment 1.00 0.54 0.63 0.64 0.57 0.41
Brand love F. 0.54 1.00 0.57 0.76 0.77 0.41
Loyalty I. 0.63 0.57 1.00 0.58 0.52 0.46
Brand attachment 0.64 0.76 0.58 1.00 0.80 0.37
Self-expressive brand 0.57 0.77 0.52 0.80 1.00 0.25
Brand trust 0.41 0.41 0.46 0.37 0.25 1.00

Connection
Commitment
0.450*** R2= 33.4%
Self- 0.401***
0.399*** 24.3% Q2= 0.1
expressive 25.1%
30.2%
brand
0.224* 0.278*
Brand love f. 9.1% Loyalty I.
15.8%
R2= 64.7% R2= 49.3%
Q2= 0.4 Q2= 0.1

0.449*** 0.183 NS
Brand Brand trust
34.5% 0.406***
Attachment R2= 16.5% 8.4%
16.5% Q2= 0.1
GoF = 0.55

Figure 2: Structural results.

Source: *** P<0.001; *P<0.05; NS not significant.

Hypothesis 8. In addition, all the signs were 34.5 per cent of variability of brand love
in the expected direction. feeling is explained by brand attachment
The Q2 statistic is used to evaluate the and 25.1 per cent of variability of loyalty
predictive relevance of the model. As all intentions is explained by commitment.
Q2 are positive (Fornell and Cha, 1994) in Correspondingly, the strength of the direct
the model (see Figure 2), the relations in relationship between commitment and loy-
the model have predictive relevance. The alty intentions were found to be higher
model also demonstrates predictive power than that of the relationship between brand
(R2) as the modeled constructs explained trust and brand love feeling and loyalty
64.7 per cent of the variance in brand love intentions. Considering the determinants of
feeling and 49.3 per cent of the variance in brand love feeling, self-expressive brand
loyalty intentions. As the objective of PLS and brand attachment have nearly the same
is prediction in a regression sense, there is impact.
no overall GoF indices. However, the Globally, the results are consistent with
overall GoF (Gof = 0.55 for this study) pro- the literature review but reflect the impor-
posed by Tenenhaus et al (2005) reveals tance of commitment as to intentions to
good fit (Figure 2). buy again or to recommend to others.
The multiplications between Pearson’s Thus, commitment reflects a greater degree
correlation value and path coefficient value of faith and desire to continue with the
of each of the two constructs reveal that brand than trust, and when trust reinforces

10 © 2012 Macmillan Publishers Ltd. 1350-23IX Journal of Brand Management 1–15

 Brand emotional connection and loyalty

the strength of commitment, the effect on between brand trust and loyalty intentions.
loyalty intention is stronger than the direct Therefore, all hypotheses were supported
effect of trust on loyalty intentions. Besides, except Hypotheses 7 and 8.
most of the empirical studies analyzed do In this way, the results reflect what we
not employ the two constructs together expected from literature review. For
whereas the findings of Dagger and O’Brien women, brand attachment has a stronger
(2010) are in line with the results of this effect on love for the brand than the self-
article. expressive brand. Trust in the brand and
Finally, the differences between male considering it reliable and trustworthy is
and female are shown in Table 4. The t-test particularly important for women, as trust
results suggest that there are significant strongly impacts on commitment. Women,
differences (critical t-value = |1.960|) in more than man, need to trust the brand to
the causal order relationships brand be committed.
love feeling;commitment and brand These results are in accordance with the
trust;commitment. theory of self-construal, as men are more
For the female subsample, all path coe- guided by self (inner and social) and focus
fficients were found to be significant at the on group identification with a social cate-
0.001 (Hypotheses 1, 4, 6, 7 and 8) or 0.05 gory. Their identification refers to their
(Hypotheses 2, 3 and 5) levels. Conse- social group such as a sport club members,
quently, all hypotheses were supported. college student, hometown or gender
However, for the male subsample, all path group. This identification is very important
coefficients were found to be significant at to drive the passion for the car brand and
the 0.001 (Hypotheses 1, 2, 3, 4 and 6) leads to commitment. For females, trust is
or 0.05 (Hypothesis 5) levels except for very important to be committed to the car
the relationship between brand trust and brand because they need to ensure that the
commitment and for the relationship car is safe for themselves and their family.

Table 4: Multi-group analysis

Structural paths Male Female Standard Standard Sp1 t-test

error male error female

Commitment - > Loyalty I. 0.420*** 0.371*** 0.112 0.105 1.432 0.306 NS

Brand attachment - > Brand 0.330*** 0.504*** 0.104 0.125 1.454 − 1.066 NS
love F.
Brand love F.- > Commitment 0.609*** 0.219* 0.088 0.100 1.202 2.893*
Brand love F. - > Loyalty. I. 0.286* 0.255* 0.124 0.106 1.551 0.180 NS
Brand love F.- > Brand trust 0.410*** 0.398*** 0.095 0.091 1.228 0.086 NS
Self-expressive brand- > Brand 0.560*** 0.283* 0.097 0.140 1.470 1.681NS
love F.
Brand trust- > Commitment 0.096 NS 0.398*** 0.099 0.095 1.277 − 2.105*
Brand trust- > Loyalty I. 0.149 NS 0.252* 0.092 0.101 1.241 0.739 NS
R2 B. Love F. 0.723 0.553 — — — —
R2 Commitment 0.429 0.726 — — — —
R2 B. Trust 0.168 0.158 — — — —
R2 Loyalty I. 0.516 0.480 — — — —
GoF 0.58 0.52 — — — —

1
Unbiased estimator of average error standard variance ***P < 0.001; *P < 0.05;
NS=Not significant.

© 2012 Macmillan Publishers Ltd. 1350-23IX Journal of Brand Management 1–15 11

 Loureiro et al

DISCUSSION AND IMPLICATIONS relationship with customers and create an

Following the model proposed by Carroll identity that leads them to self-identifica-
and Ahuvia (2006), this research examines tion and social-identification with the
the construct attachment to brand as an brands. The results also highlight the signi-
antecedent to brand love feeling and inte- ficant and positive causal order between
grates commitment and brand trust as trust and commitment (Morgan and Hunt,
mediators between brand love feeling and 1994; Garbarino and Johnson, 1999).
loyalty intentions. The model was tested With regard to gender differences, the
contextually with a group of satisfied car results suggest that female brand attachment
owners of three different brands with each has a stronger effect on love for the brand
brand originating from different countries. than the self-expressive brand. Trust in the
The product category can be classified as brand and considering it reliable and trust-
high involvement as a big economic or worthy is particularly important for women.
psychological loss is at stake. It should be Thus, the commitment of women to the
noted that even for utilitarian car, car taxes brand is heavily influenced by trust. For
are exorbitant in Portugal. The car owners men, it is especially important that inner
who bought cars from these three brands and social identification with the brand is
feel satisfied with them and all the cars are established to increase passion and love for
utilitarian, meaning they are used daily to the brand, ultimately leading to commit-
go to work, to go to the supermarket, to ment. These results are in agreement with
take children to school and other daily the theory of self-construal.
activities.
The major findings of this study are as Managerial implications
follows. Brand attachment has a significant The considerations discussed above lead to
positive effect on brand love feeling. The some important managerial recommenda-
feeling of love toward a brand can reinforce tions. Corporate and local managers should
trust, the interest in continuing the rela- always be aware of opportunities to build
tionship and faith in the future regarding and to consistently re-build brand identity
the brand. Brand love feeling is influenced and communicate a message to consumers
nearly equally by self-expressive brand that leads present and future customers to
and brand attachment. Finally, loyalty feel identification with the brand. Cus-
intentions are less influenced by brand tomers need to feel that the brand symbol-
trust than by commitment or brand love izes what they are; this contributes to the
feeling. customer’s image and helps to develop a
These results corroborate the findings sense of belonging. Identification with the
of Carroll and Ahuvia (2006) because self- brand coupled with good experiences with
expressive brand is an important antecedent the brand work to strengthen the attraction,
to brand love feeling with loyalty inten- passion and love for the brand. Proximity
tions an outcome. Carroll and Ahuvia between the brand and the customers is a
(2006, p. 87) suggest that ‘more hedonic continuous process. Passion leads customers
products and/or self-expressive brands tend to be more open to a stronger relationship,
to be more loved; managers may find that which leads to more confidence in the
enhancing these aspects of their offerings brand. This way, customers will be more
increases this intense emotional response in likely to re-buy products of the same brand,
consumers’. Nevertheless, this study sug- to recommend it to others, to choose the
gests that even brands associated with util- brand over another and even to depreciate
itarian cars with high involvement build a a problem if it occurs.

12 © 2012 Macmillan Publishers Ltd. 1350-23IX Journal of Brand Management 1–15

 Brand emotional connection and loyalty

Manager should also be aware that Ahuvia, A.C. (1993) I love it! Towards an unifying theory
of love across divers love objects. Ph. Dissertation,
women tend to value the dyadic relation- Northwestern University, Evanston/Chicago, IL.
ship between the individual and the brand Albert, N., Merunka, D. and Valette-Florence, P.
to build confidence and commitment. (2008) When consumers love their brands:
Exploring the concept and its dimensions. Journal
Managers should plan personalized com- of Business Research 61(10): 1062–1075.
munication and maintain close contact with Albert, N., Merunka, D. and Valette-Florence, P.
these consumers. Men need to feel that (2009) The feeling of love toward a brand: Concept
they are connected to the brand and that and measurement. In: A.L. McGill and S. Shavitt,
(eds.), Proceedings of the Advances in Consumer
they have made the right choice; they truly Research. October, Duluth, MN: Association for
want to identify to the brand. The brand, Consumer Research, Vol. 36, pp. 300–307.
both product and company, must be con- Assael, H. (1987) Consumer Behavior and Marketing
Action, 3rd edn. Boston, MA: PWS-Kent.
sistent with this sense of belonging. Thus, Batra, R., Ahuvia, A.C. and Bagozzi, R. (2008) Brand
communication should reflect a lifestyle, a Love: Its Nature and Consequences. Michigan Dear-
way to be in accordance with the inner and born University: Ann Arbor, MI. Working paper.
Baumeister, R.F. and Sommer, K.L. (1997) What do
social self of the men who buy the brand. men want? Gender differences and two spheres of
belongingness: Comment on Cross and Madson.
Psychological Bulletin 122( 1): 38–44.
Limitations and future research Bem, S.L. (1981) Gender schema theory: A cognitive
This research did not consider variables account of sex typing. Psychological Review 88(4):
such as brand hate or others that express 354–364.
Bowlby, J. (1979) The Making and Breaking of Affectional
neutral or negative affect about the brand Bonds. London: Tavistock.
to define brand love. The model presented Brislin, R.W. (1970) Back-translation for cross-cultural
in this study was applied in an innovative research. Journal of Cross-cultural Psychology 1(3):
185–216.
manner, in the context of non-hedonic Carmines, E.G. and Zeller, R.A. (eds.) (1979) Reliability
products, yet the results are associated with and Validity Assessment. London: Sage Publications.
a product category and cannot be appro- Carroll, B.A. and Ahuvia, A.C. (2006) Some anteced-
priately generalized to other. ents and outcomes of brand love. Marketing Letters
17(2): 79–89.
In future research, some other moder- Chang, P.-L. and Chieng, M.-H. (2006) Building con-
ating variables should be considered such sumer-brand relationship: A cross-cultural experiential
as socio-demographic (for example, view. Psychology and Marketing 23(11): 927–959.
Chaudhuri, A. and Holbrook, M.B. (2001) The chain
incomes). Other antecedents and conse- effects of brand trust and brand affect to brand
quents of brand love should be tested such performance: The role of brand loyalty. Journal of
as brand personality, credibility and con- Marketing 65(2): 81–93.
Chin, W.W. (1998) The partial least squares approach
sumer’s willingness to sacrifice. In addition, to structural equation modeling. In: G.A. Marcoulides
different product categories, brands and dif- (ed.) Modern Methods for Business Research. Mahwah,
ferent levels of involvement should be a NJ: Lawrence Erlbaum Associates Publisher,
part of the research context. When enough pp. 295–336.
Chin, W.W., Marcolin, B.L. and Newsted, P.R. (2003)
information is gathered on the subject of A partial least squares latent variable modelling
brand love, a typology of the different types approach for measuring interaction effects: Results
of brand love may be possible. from a Monte Carlo simulation study and an
electronic mail emotion/ adoption study. Informa-
tion Systems Research 14(2): 189–217.
Cross, S.E. and Madson, L. (1997) Models of the self:
REFERENCES Self-construals and gender. Psychological Bulletin
Achrol, R. (1991) Evolution of the marketing organi- 122(1): 5–37.
zation: New forms for turbulent environments. Dagger, T.S. and O’Brien, T.K. (2010) Does experi-
Journal of Marketing 55(4): 77–93. ence matter? Differences in relationship benefits,
ACP – Automóvel Clube de Portugal. (2009) satisfaction, trust, commitment and loyalty for
http://www.acp.pt/index.php?template_id=6165, novice and experienced service users. European
accessed 2 March 2009. Journal of Marketing 44(9/10): 1528–1552.

© 2012 Macmillan Publishers Ltd. 1350-23IX Journal of Brand Management 1–15 13

 Loureiro et al

Echambadi, R., Campbell, B. and Agarwal, R. (2006) future of national customer satisfaction index models.
Encouraging best practice in quantitative manage- Journal of Economic Psychology 22(April): 217–245.
ment research: An incomplete list of opportunities. Johnson, M.D., Herrmann, A. and Huber, F. (2006)
Journal of Management Studies 43(8): 1801–1820. The evolution of loyalty intentions. Journal of
Evard, Y. and Aurier, P. (1996) Identification and Marketing 70(2): 122–132.
validation of the components of the person– Kamat, V. and Parulekar, A. (2007) Brand love – The
object relationship. Journal of Business Research 37(2): precursor to loyalty. In: J.R. Priester, D.J. MacInnis
127–134. and C.W. Park (eds.) New Frontiers in Branding;
Fehr, B. (1988) Prototype analysis of the concepts of Attitudes, Attachments, Relationships. Advertising and
love and commitment. Journal of Personality and Consumer 2007 Psychology Proceedings. 7–9 June,
Social Psychology 55(4): 557–579. Santa Monica, CA, pp. 94–98.
Fornell, C. and Cha, J. (1994) Partial least squares. In: Keh, H.T., Pang, J. and Peng, S. (2007) Understanding
R.P. Bagozzi (ed.) Advanced Methods of Marketing and measuring brand love. In: J.R. Priester, D.J.
Research. United Kingdom: Blackwell, Cambridge, MacInnis and C.W. Park (eds.). New Frontiers in
pp. 52–78. Branding; Attitudes, Attachments, Relationships. Adver-
Fornell, C. and Larcker, D.F. (1981) Evaluating struc- tising and Consumer 2007 Psychology Proceedings, 7–9
tural models with unobservables variables and June, Santa Monica, CA, pp. 84–88.
measurement error. Journal of Marketing Research Kelley, H.H.(1979) Personal Relationships: Their Struc-
18(1): 39–50. tures and Processes. Erlbaum, NJ: Hillsdale.
Fournier, S.M. (1998) Consumers and their brands: Kelley, H.H. and Thibaut, J.W. (1978) Interpersonal Rela-
Developing relationship theory in consumer tions: A Theory of Interdependence. New York: Wiley.
research. Journal of Consumer Research 24(4): Kingshott, R.P.J. and Pecotich, A. (2007) The impact
343–373. of psychological contracts on trust and commitment
Fullerton, G. (2005) The impact of brand loyalty in supplier-distributor relationships. European Journal
commitment on loyalty to retail service brands. of Marketing 41(9/10): 1053–1072.
Canadian Journal of Administrative Sciences 22(2): Markovits, H., Benenson, J. and White, S. (2006) Gender
97–110. and priming differences in speed of processing of
Garbarino, E. and Johnson, M.S. (1999) The differen- information relating to social structure. Journal of
tial roles of satisfaction, trust, and commitment in Experimental Social Psychology 42(5): 662–667.
customer relationships. Journal of Marketing 63(2): McAlexander, J.H., Schouten, J.W. and Koenig, H.F.
70–87. (2002) Building brand community. Journal of
Gardner, W.L., Gabriel, S. and Hochschild, L. Marketing 66( 1): 38–54.
(2002) When you and I are ‘We,’ You are not Meyers-Levy, J. (1988) Influence of sex roles on judg-
threatening: The role of self-expansion in social- ment. Journal of Consumer Research 14(4): 522–530.
comparison. Journal of Personality and Social Psy- Morgan, R.M. and Hunt, S.D. (1994) The commit-
chology 82(2): 239–251. ment–trust theory of relationship marketing. Journal
Hansen, H., Sandvik, K. and Selnes, F. (2003) Direct of Marketing 58( 3): 20–38.
and indirect effects of commitment to a service Moorman, C., Gerald, Z. and Rohit, D. (1992)
employee on the intention to stay. Journal of Service Relationships between providers and users of
Research 5(4): 356–368. market research: The dynamic of trust within and
Hatfield, E. and Rapson, R.L. (1993) Love and attach- between organizations. Journal of Marketing Research
ment processes. In: M. Lewis and J.M. Haviland 29(3): 314–328.
(eds.) Handbook of Emotions. New York: Guilford Nunnally, J.C. (1978) Psychometric Theory, 2nd edn.
Press, pp. 595–604. New York: McGraw-Hill.
Hazan, C. and Shaver, P.R. (1990) Love and work: Oliver, R.L. (1999) Whence consumer loyalty? Journal
An attachment theoretical perspective. Journal of of Marketing 63: 33–44.
Personality and Social Psychology 59(2): 270–280. Parvatiyar, A. and Sheth, J.S. (2001) Customer relation-
Hazan, C. and Shaver, P.R. (1994) Attachment as ship management: Emerging practice, process, and
an organizational framework for research on close discipline. Journal of Economic and Social Research 3(2):
relationships. Psychological Inquiry 5(1): 1–22. 1–34.
He, X., Inman, J.J. and Mittal, V. (2008) Gender jeop- Power, J., Whelan, S. and Davies, G. (2008) The attrac-
ardy in financial risk taking. Journal of Marketing tiveness and connectedness of ruthless brands: The
Research 45(4): 414–424. role of trust. European Journal of Marketing 42(5/6):
Hinde, R.A. (1979) Towards Understanding Relationships. 586–602.
London: Academic Press. Price, L.L., Arnould, E.J. and Curasi, C.F. (2000) Older
Hulland, J. (1999) Use of partial least squares (PLS) in consumers’ disposition of special possessions. Journal
strategic management research: A review of four of Consumer Research 27(2): 179–201.
recent studies. Strategic Management Journal 20(2): Regan, P.C., Kocan, E.R. and Whitlock, T. (1998)
195–204. Ain’ love grand! A prototype analysis of the concept
Johnson, M.D., Gustafsson, A., Andreassen, T.W., of romantic love. Journal of Social and Personal Rela-
Lervik, L. and Cha, J. (2001) The evolution and tionships 15(3): 411–420.

14 © 2012 Macmillan Publishers Ltd. 1350-23IX Journal of Brand Management 1–15

 Brand emotional connection and loyalty

Richins, M.L. (1994) Special possessions and the Trinke, S.J. and Bartholomew, K. (1997) Hierarchies
expression of material values. Journal of Consumer of attachment relationships in young adulthood.
Research 21(3): 522–533. Journal of Social and Personal Relationships 15(5):
Ringle, C.M., Wende, S. and Will, A. (2005) SmartPLS 603–625.
2.0 (beta), www.smartpls.de, Hamburg, accessed Thomson, M. (2006) Human brands: Investigating
2 August 2010. antecedents to consumers’ strong attachments to
Rusbult, C.E. and Van Lange, A.M.(2003) Interde- celebrities. Journal of Marketing 70(3): 104–119.
pendence, interatction and relationships. Annual Thomson, M., MacInnis, D.J. and Park, C.W.
Review of Psychology 54(February): 351–375. (2005) The ties that bind: Measuring the strength
Schouten, J.W. and McAlexander, J.H. (1995) Subcul- of consumers’ emotional attachment to brands.
tures of consumption: An ethnography of the new Journal of Consumer Psychology 15(1): 77–91.
bikers. Journal of Consumer Research 22(1): 43–61. Vlachos, P.A., Theotokis, A., Pramatari, K. and
Sedikides, C. and Brewer, M.B. (2001) Individual self, Vrechopoulos, A. (2010) Consumer-retailer
relational self, and collective self. In: C. Sedikides emotional attachment: Some antecedents and the
and M.B. Brewer (eds.) Individual Self, Relational moderating role of attachment anxiety. European
Self, Collective Self. Philadelphia, PA: Psychology Journal of Marketing 44(9/10): 1478–1499.
Press, pp. 1–4. Weiss, R.S. (1988) Loss and recovery. Journal of Social
Seeley, E.A., Gardner, W.L., Pennington, G. and Issues 44(3): 37–52.
Gabriel, S. (2003) Circle of friends or members of Yuksel, A., Yuksel, F. and Bilim, Y. (2010) Destination
a group? Sex differences in relational and collective attachment: Effects on customer satisfaction and
attachment to groups. Group Processes and Intergroup cognitive, affective and conative loyalty. Tourism
Relations 6(3): 251–263. Management 31(2): 274–284.
Sekaran, U. (1983) Methodological and theoretical Zeithaml, V.A., Berry, L.L. and Parasuraman, A. (1996)
issues and advancements in cross-cultural research. The behavioral consequences of service loyalty.
Journal of International Business Studies 14(2): 61–73. Journal of Marketing 60(2): 31–46.
Sternberg, R.J. (1986) A triangular theory of love. Zeugner-Routh, K.P., Diamantopoulos, A. and Mon-
Psychologist Review 93(2): 119–135. tesinos, M.A. (2008) Home country image, country
Tenenhaus, M., Vinzi, V.E., Chatelin, Y.-M. and brands equity, and consumers’ product preferences:
Lauro, C. (2005) PLS path modeling. Computational An empirical study. Management International Review
Statistics & Data Analysis 48(1): 159–205. 48(5): 577–602.

© 2012 Macmillan Publishers Ltd. 1350-23IX Journal of Brand Management 1–15 15

View publication stats