Modeling the implicit brand: capturing the hidden drivers

DOIhttps://doi.org/10.1108/JPBM-10-2016-1327
Pages600-615
Date18 September 2017
Published date18 September 2017
AuthorMarco Vriens,Alessandro Martins Alves
Subject MatterMarketing,Product management,Brand management/equity
Modeling the implicit brand: capturing the
hidden drivers
Marco Vriens
Department of Marketing, University of Wisconsin La Crosse, La Crosse, Wisconsin, USA, and
Alessandro Martins Alves
Department of Marketing, Ipsos, Rio de Janeiro, Brazil
Abstract
Purpose – This paper aims to investigate modeling implicit attitudes as potential drivers of overall brand attitudes and stated behavior and
investigate how the results are expected to be different from brand driver models that are based on explicit attitudes.
Design/methodology/approach Data are collected via online surveys in five countries across 15 categories with sample sizes for each
category/country combination in the range of about
N
1,000.
Findings – Implicit attitudes result in a higher number of significant effects than their explicit counterparts when used to explain behavioral
intentions, brand closeness and brand usage in a multivariate situation with potential 12 brand attitude drivers. The authors also find fewer
counter-intuitive effects in the implicit models. The results are consistent across 5 countries and across 15 categories (including CPG products,
services and durable goods). They also show that implicit attitudes are less susceptible to response style effects (e.g. social desirability bias).
Research limitations/implications – The findings have implications for brand building and shopper activation. Further research should look into
the impact of using implicit data on finding different brand segmentation and brand mapping results.
Practical implications – The findings have implications for brand building and shopper activation.
Originality/value – This paper contributes to the fast-growing field of implicit attitudes. The paper confirms and generalizes previous findings. This
is the first paper to the authors’ knowledge that has investigated the impact of implicit attitudes on overall brand attitudes and stated behavior in
a multivariate context.
Keywords Neuroscience, Brand evaluation, Regression analysis, Information processing, Brand performance, Dual processing, Implicit attitudes
Paper type Research paper
1. Introduction
Behavioral economics and neuro science have credibly shown
that human decisions do not always optimize utility, and they
are likely to rely in part or even fully on heuristics (Gigerenzer
and Gaissmaier, 2011). This has led to the description of a
Systems 1 and 2 (Stanovich and West, 2000;Kahneman,
2003;Evans and Stanovich, 2013), where System 1 is more
automatic, autonomous, faster, more intuitive way of making
decisions and choices (Weinberg and Gottwald, 1982;Bargh,
2002). System 1 is supposedly also driven more by emotional
factors (Phelps, 2004;Phelps and LeDoux, 2005;Heath,
2009;Smith and Nosek, 2011). System 2 is more conscious,
controlled, and it is slower and assumed to be more rational.
System 1 decision making is largely unconscious to the
consumer. Emotional processing and response is very fast and
does not seem to require conscious effort (Mast and Zaltman,
2006), may not require attention and can be more important
in creating brand favorability than rational cognitive reasons
(Heath et al., 2006). Bargh (2002) shows that consumers’
behavior can be unconsciously influenced by using priming
effects.
These psychological findings and theories have led to an
innovation in measuring attitudes referred to as implicit
attitude measurement in contrast to the traditional
measurement of attitudes, referred to as explicit attitudes
(Greenwald and Banaji, 1995). Implicit attitudes are referred
to as attitudes that influence our behavior without awareness
(Stanley et al., 2008). Implicit attitude and brand theory
(Krishnan, 1996) states that the brain holds an intricate
network of associations that are the result of experience,
perceptions and repeated exposure to messages (i.e.
advertising) advancing certain perceptions. The richer these
structures are and the more a certain belief is connected to
such experiences and exposures the faster we can respond
when asked if we associate a certain belief with say a specific
brand (Friese et al., 2006;Moses, 2015). The implicit
association test (IAT) is the most used methodological
approach to measure implicit attitudes (Greenwald et al.,
2003;Cunningham et al., 2001). This approach has been used
in most academic research but does not lend itself well for
practical commercial brand research. Hence, in this study, we
deploy a different methodology. Explicit attitudes are those for
which one has had the time to think about before providing the
response (Spence, 2005). Explicit attitudes are usually
The current issue and full text archive of this journal is available on
Emerald Insight at: www.emeraldinsight.com/1061-0421.htm
Journal of Product & Brand Management
26/6 (2017) 600–615
© Emerald Publishing Limited [ISSN 1061-0421]
[DOI 10.1108/JPBM-10-2016-1327]
Received 2 October 2016
Revised 29 March 2017
Accepted 29 March 2017
600
captured in semantic differentials, standard rating scales or
simple yes/no association statements.
There have been studies in marketing that have investigated
the relative impact of explicit and implicit attitudes on
consumer preferences and behavior. However, to our
knowledge in all these previous studies a single variable was
used to measure the explicit attitude toward a brand or a
product, and a single variable was use to represent the implicit
attitude toward a brand or a product. As such these findings
are hard to generalize to practical brand studies. Standard
approaches in brand research typically consider many
potential brand associations which can range from 10 all the
way up to a 100 plus attributes. In this paper, we aim to make
the following contributions. First, we expand on the literature
by investigating how implicit attitudes can be modeled in
typical multivariate situations that are used in brand,
advertising and shopper research (in our case 12 potential
brand associations). Second, we study whether brand driver
models based on implicit responses yield different results as
compared to driver models based on explicit brand association
scores. Specifically, we investigate whether implicit based
driver models differ from explicit based models in terms of fit
as indicated by in-sample and hold-out sample fit, the number
of significant attributes, the number of counter-intuitive
effects in the brand driver models and the average relative
coefficient (across the statistically significant attributes). We
investigate these four questions using three different types of
dependent brand variables: recommendation, brand closeness
and brand usage. Our study covers 5 countries and 15
categories.
2. Literature review and hypotheses
The dual-processing theory (e.g. differentiating between
Systems 1 and 2) states that behavior is driven “by reflective
and impulsive processes” (Friese et al., 2009). Implicit
attitudes have sometimes been found to be better predictors of
actual behavior than explicit attitudes (Greenwald et al.,
2009). Friese et al. (2007) found that an implicit attitude
improved the prediction of future voting behavior over and
above explicit attitudes. Nock et al. (2010) show that the use
of implicit attitudes significantly improved the prediction of
whether patients who were seeking psychological treatment for
depression were going to commit suicide. In the context of
socially sensitive topics, implicit attitudes have indeed proven
to be better predictors of behavior. To generalize that to
marketing situations is not as obvious. Only a handful of
studies have looked at the role and predictive power of implicit
attitudes in the context of consumer behavior. In Karpinski
and Hilton’s (2001) study, implicit attitudes were not able to
predict the choice between an apple and candy. Ayres et al.
(2012) did not find any incremental predictive accuracy of
implicit attitudes over explicit attitudes in terms of predicting
the choice between a healthy snack and an unhealthy snack.
Maison et al. (2004) captured both implicit and explicit
attitudes in three studies pertaining to preferences for yoghurt,
fast food restaurants and soft drinks. Using multiple regression
analysis, they showed that implicit attitudes (measured by the
IAT) can improve the prediction of behavior over and above
the use of explicit attitudes only. In all three studies, the
regression weight for the implicit attitude was smaller than the
regression weight for the explicit variables. We note that both
the implicit and explicit attitudes were captured by a single
variable, so we have no insights into their effects in a
multivariate context. Also, the sample sizes in this study were
very small (50) for two of the three studies, and a N103
for their third study. Perugini (2005) in two studies pertaining
to smoking behavior compared three models:
1 an additive model where both implicit and explicit are
significant drivers;
2 a dissociative model where explicit attitude predicts a
deliberate choice and implicit predicts a spontaneous
choice; and
3 a multiplicative model that contains an interaction term
between implicit and explicit.
He found some support for the multiplicative model (Study 1)
and some support for the dissociative model (Study 2). Friese
et al. (2006) studied preferences for ten CPG products and
tested both branded and generic products. They found that 85
per cent implicitly preferred the branded versions. However,
the explicit responses revealed that only 33 per cent preferred
the branded products. Their results indicated that consumers
choose mostly based on their explicit attitudes except if they
had to make these choices under time pressure. We will
comment on these results further in the discussion section.
Vantomme et al. (2006) tried to predict which consumers had
bought fair-trade products. A logistic regression model was
estimated with as dependent variable whether they bought fair
trade products and with as independent variables one explicit
attitude and one implicit attitude. Both variables yielded
significant regression coefficients, and in this study, the
regression weight for explicit was larger than the regression
weight for implicit. Richetin et al. (2007a) show that explicit
attitudes predicted both incidental and deliberate behavior,
whereas the implicit attitude only predicted the incidental
behavior. Also, the impact of implicit can be moderated by a
person’s decision-making style. Richetin et al. (2007b) tested
whether implicit attitudes would improve a prediction of
whether or not consumers would prefer a fruit over a snack
(binary dependent variable). They estimated a logistic
regression model, and in their model building, they first
included the implicit attitude and then entered the explicit
attitude. Both variables remained in the model as significant
predictors of the food choice. Though the fit of the model was
low (17 per cent explained variance): the explicit variable
received a coefficient of 0.51, and the implicit variable
achieved a coefficient of 0.36. Friese et al. (2012) found that
both implicit and explicit attitudes predicted voting behavior.
A binary model with only one implicit attitude correctly
classified voters’ choices 89.5 per cent of the cases. Even when
an explicit attitude was entered both effects remained
statistically significant. As stand-alone predictors, the explicit
attitude was a stronger predictor.
The empirical evidence suggests that both implicit and
explicit attitudes toward a brand will have an effect on brand
preference and usage metrics and we expect the explicit based
models to have a somewhat higher (predictive) fit:
H1. Models based on implicit attitudes will have somewhat
lower fit and predictive accuracy than models based on
explicit attitudes.
Modeling the implicit brand
Marco Vriens and Alessandro Martins Alves
Journal of Product & Brand Management
Volume 26 · Number 6 · 2017 · 600–615
601

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT