Journal of Consumer Research, Inc.
Shining in the Center: Central Gaze Cascade Effect on Product Choice
Author(s): A. Selin Atalay, H. Onur Bodur, and Dina Rasolofoarison
Reviewed work(s):
Source:
Journal of Consumer Research,
(-Not available-), p. 000
Published by: The University of Chicago Press
Stable URL: http://www.jstor.org/stable/10.1086/665984 .
Accessed: 09/05/2012 11:38
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
The University of Chicago Press and Journal of Consumer Research, Inc. are collaborating with JSTOR to
digitize, preserve and extend access to Journal of Consumer Research.
http://www.jstor.org
000
䉷 2012 by JOURNAL OF CONSUMER RESEARCH, Inc. ● Vol. 39 ● December 2012
All rights reserved. 0093-5301/2012/3904-0013$10.00. DOI: 10.1086/665984
Please use DOI when citing. Page numbers are not final.
Shining in the Center: Central Gaze
Cascade Effect on Product Choice
A. SELIN ATALAY
H. ONUR BODUR
DINA RASOLOFOARISON
Consumers’ tendency to choose the option in the center of an array and the process
underlying this effect is explored. Findings from two eye-tracking studies suggest
that brands in the horizontal center receive more visual attention. They are more
likely to be chosen. Investigation of the attention process revealed an initial central
fixation bias, a tendency to look first at the central option, and a central gaze
cascade effect, progressively increasing attention focused on the central option
right prior to decision. Only the central gaze cascade effect was related to choice.
An offline study with tangible products demonstrated that the centrally located item
within a product category is chosen more often, even when it is not placed in the
center of the visual field. Despite widespread use, memory-based attention mea-
sures were not correlated with eye-tracking measures. They did not capture visual
attention and were not related to choice.
C
onsumers are exposed to horizontally presented arrays
of products in various contexts, such as lunch combos
pictorially presented in the menu of a fast food restaurant,
rows of snack bars in a vending machine, a selection of
bottled drinks at a buffet, or a selection of suggested movies
from an online service provider (see fig. 1). One of the
documented effects when choosing from an array of prod-
ucts is horizontal centrality: the option located in the center
is more likely to be chosen. For instance, Christenfeld
(1995) demonstrated that people chose the middle option
in a number of different contexts, including choice of a
bathroom stall in a public restroom, choice of a toilet paper
dispenser in the restroom, and choice among a row of
A. Selin Atalay (atalay@hec.fr) is assistant professor, department of mar-
keting, HEC Paris, 1 Rue de la Liberation, 78351 Jouy-en-Josas (France).
H. Onur Bodur (bodur@jmsb.concordia.ca) is associate professor, marketing
department, Concordia University, 1455 De Maisonneuve Blv. W., Montreal,
QC, H3G 1M8 (Canada). Dina Rasolofoarison (d.rasolofoarison@aston.ac.uk)
is assistant professor, marketing group, Aston Business School, Aston Triangle,
Birmingham, B4 7ET (UK). The authors would like to thank Hans Baumgartner,
Bianca Grohmann, Tim Heath, Mirko Kremer, Julien Schmitt, and Marc Van-
huele for their feedback on this work. The authors are grateful to Rajdeep
Grewal, Margaret G. Meloy, and Rik Pieters for their guidance. The authors
thank the CMBBR labs for resources and support and Kimberly Duval, Yi Li,
and Shiva Taghavi for their assistance in data collection.
Laura Perrachio served as editor and James Bettman served as associate editor
for this article.
Electronically published May 3, 2012
arbitrary symbols. In the same study, when choosing
among rows of a product displayed on a supermarket shelf,
the middle option was preferred 71% of the time. Similarly,
Shaw et al. (2000) showed that in making a choice among
an array of three alternatives, individuals consistently pre-
ferred the middle option. This effect was replicated with
highlighters, surveys, and chairs.
Location-driven choice patterns have implications for
retail shelf management and point-of-purchase decisions,
among other consumer choice contexts. In efforts to un-
derstand point-of-purchase decisions, researchers have fo-
cused on the drivers (Chandon et al. 2009; Janiszewski 1998;
van der Lans, Pieters, and Wedel 2008a, 2008b), strategies
(Pieters and Warlop 1999), and stages (Clement 2007; Russo
and Leclerc 1994; Russo and Rosen 1975) of the visual
search process and have concluded that the ability of a brand
to capture and hold consumer attention can be a source of
competitive advantage. In fact, using eye-tracking meth-
odology, van der Lans et al. (2008a) demonstrated that com-
petitive advantage on the shelf is generated mostly by in-
store factors rather than out-of-store factors (e.g., search
goals) with a ratio of 2 : 1. Similarly, Chandon et al. (2009)
demonstrated that, at the point of purchase, the brand’s hor-
izontal centrality on the shelf has a strong positive effect
on brand choice. These findings call for a closer investi-
gation of the effects of location on the visual search and
decision-making processes and its relevance to choice.
The current article, in two computer-based eye-tracking stud-
ies and an offline lab experiment, progressively investigates
000 JOURNAL OF CONSUMER RESEARCH
Please use DOI when citing. Page numbers are not final.
FIGURE 1
ARRAY OF MOVIE SUGGESTIONS (FROM NETFLIX.COM)
how horizontal location affects choice. Visual attention is linked
to choice. Interestingly, other research corroborates the impact
of horizontal centrality on choice (Valenzuela and Raghubir
2009) but argues against an attention-based explanation. It is
suggested that inferences made about the option in the hor-
izontal center determine choice. Therefore, the current pro-
ject investigates whether horizontal centrality affects (1)
how much attention a brand gets, (2) the inferences made
about the brand, and (3) whether these factors are related
to the impact of horizontal shelf location on choice. Results
suggest that the indirect impact of horizontal centrality on
choice is through visual attention. The indirect impact of
horizontal centrality on choice through brand inferences was
not significant. A gaze likelihood analysis (Shimojo et al.
2003) was conducted to understand how visual attention is
related to choice. The relationship between visual attention
and choice is not a result of an initial tendency to fixate
on the center but is related to accelerated fixations in the
final seconds of the gaze duration. The brand in the center
benefits from a disproportionately more prominent gaze
cascade, which is coined the central gaze cascade effect.
Given the widespread use of memory-based self-reported
measures of attention, the effects of horizontal centrality
on self-reported measures of attention were also investi-
gated. The results suggested that memory-based attention
measures and eye-tracking measures may be related to dif-
ferent constructs. More specifically, in the current studies,
memory-based attention measures (unaided recall, aided
recall, and multi-item attention measures) did not capture
visual attention and were not related to choice. These find-
ings were replicated (study 1B) with shelf displays shifted
off the center of the visual field (i.e., the computer screen),
confirming that the higher attention paid to the centrally
located option is not due to a tendency to look more at
the center of the computer screen. Extending the findings
to an offline choice context (study 2), it was confirmed
that the centrally located item within a product category
is chosen more often even when it is not placed in the
center of the shelf or the visual field. Taken together, find-
ings emphasize the relationship between horizontal loca-
tion, visual attention, and choice, as well as the critical
role of eye-tracking measures in understanding the visual
search and choice processes. Other possible explanations,
including initial fixation bias, leader-driven primacy,
leader-driven search focus, and marketplace metacogni-
tions are discussed.
EXPLANATIONS FOR THE HORIZONTAL
CENTRALITY EFFECT ON CHOICE
Research in visual perception (Locher and Nodine 1973,
1989) demonstrated that individuals have a propensity to
look longer at the axis of symmetry when exposed to a
symmetric picture. To investigate whether the preference
for symmetry leads to the horizontal centrality effect on
choice, Shaw et al. (2000) conducted an experiment in
which participants were asked to choose a chair among a
row of chairs that were all empty (balanced) or occupied
on one end (unbalanced). The preference for the chair in the
middle was prevalent in both conditions, ruling out the pref-
erence for symmetry as an explanation. Shaw et al. (2000) also
proposed increased attention as an alternative explanation be-
cause the middle option is at the center of the perceptual field.
Findings for the attention explanation, using a recall task to
measure attention, were not conclusive; a call was made for
additional research (Shaw et al. 2000). In recent eye-tracking
CENTRAL GAZE CASCADE EFFECT 000
Please use DOI when citing. Page numbers are not final.
studies, Chandon et al. (2007, 2009) found that brands located
in the (vertical and horizontal) center of a shelf display are
noted more and chosen more often, supporting an attention-
based explanation. Valenzuela and Raghubir (2009) proposed
the center-stage effect as an alternative explanation. According
to Valenzuela and Raghubir (2009), consumers hold the lay
belief that in retail contexts the products placed in central po-
sitions are more popular, reflecting the overall quality of the
product, which leads consumers to systematically prefer items
in the center.
Note that Chandon et al. (2007, 2009) focused on visual
attention as the explanation for horizontal centrality, but they
did not measure inferences made about the chosen brand.
Valenzuela and Raghubir (2009, 2010) reported that the im-
pact of horizontal location on choice is mediated by brand
inferences but not by attention. They obtained these results
with memory-based attention measures (e.g., recall). Extant
research in eye tracking demonstrated that memory-based
self-reported measures do not reflect attention accurately in
the context of brand choice (Aribarg, Pieters, and Wedel
2010; Chandon et al. 2007, 2009), and they concluded that
attention and recall are distinct constructs. In the context of
print advertisement, Aribarg et al. (2010) documented that
consumers falsely report having attended to a given ad when
they have not. In short, the choice or lack of measures and
the distinction between the constructs can be the source of
inconsistent explanations provided for the horizontal cen-
trality effect on choice. The current research investigates
the role of each factor in explaining why horizontal cen-
trality impacts choice, using multiple measures that tap on
different conceptualizations of attention (visual attention and
memory based) and inferences.
While the connection between inferences and choice is
intuitive, could one expect horizontal centrality to lead to
choice through visual attention, independent of conscious
and articulated brand inferences? Although this question has
not been addressed specifically, the connection between ar-
ticulated, conscious inferences and choice has been ques-
tioned. First, individuals are not always able to identify and
report the processes that lead to their choices (Johansson et
al. 2005; Nisbett and Wilson 1977). When asked to explain
the reasons why they have made a specific choice, individ-
uals often provide reports on their cognitive processes that
are simply attributions made post choice (Nisbett and Wilson
1977). These attributions tend to be biased such that having
made a choice changes a person’s preferences (Sharot, Ve-
lasquez, and Dolan 2010) to avoid or reduce dissonance
effects (Festinger 1957). Second, individuals may make
choices without “declarative knowledge” or conscious in-
ferences (Bechara et al. 1997; Shiv et al. 2005). Johansson
et al. (2005) presented participants with their choice and
asked them to explain the reasons for their choice. In reality,
participants were presented with the alternative they had not
chosen and led to believe that it was their choice. Partici-
pants failed to realize this change, pointing to a choice recall
failure. They also provided confabulatory reports when
asked to describe the reasons behind their choices. Congru-
ent with the somatic marker hypothesis, individuals observe
and interpret their own choices to generate emotions that
serve as markers that affect their subsequent judgments (Be-
chara and Damasio 2005; Bechara et al. 1997; Simion and
Shimojo 2006). These findings point to choice without ar-
ticulated inferences.
More relevant to the link between visual attention and
choice, research in visual perception identified a crucial role
of visual attention in the final moments of the choice task
that shapes individual’s preference for the eventually chosen
alternative, independent of the effects of memory or prior
preferences (Shimojo et al. 2003; Simion and Shimojo 2006).
Shimojo et al. (2003) provided gaze pattern analyses that help
in understanding the visual attention process behind choice.
They found that the role of attention on preference for the
eventually chosen option is rapid and concentrated in the final
few seconds of the choice task. This reduces the utility of
memory-based self-reported attention measures to understand
the process (Glaholt and Reingold 2011). Given the low sam-
pling, speed, and accuracy of memory-based self-reports of
attention and the problems in introspective reports of choice
processes, more objective and accurate process measures are
necessary to understand the choice process (Russo 2011).
Next, visual attention and its use as a process measure are
discussed.
VISUAL ATTENTION, CENTRALITY,
AND CHOICE
Process tracing is critical in understanding the elements of
decision making. Methods examining the process of infor-
mation acquisition and decision making (see Kuhlberger,
Schulte-Mecklenbeck, and Ranyard [2011] for an over-
view), ranging from think-aloud procedures and verbal pro-
tocols to skin conductance tests, have been used. Eye-track-
ing procedures and Mouselab have been used to identify the
stages, motivators, and determinants of the decision-making
process (Coupey, Bodur, and Brinberg 1998; Payne et al.
1992; Pieters and Warlop 1999; Russo 2011; Wedel and
Pieters 2000; Willemson, Bo¨ckenholt, and Johnson 2011).
In retail contexts, Chandon et al. (2007, 2009) suggest the
use of eye movements as an indicator of attention as they
are sensitive and objective.
In general, physiological response measures (Krugman
1965) such as eye-tracking procedures are a reliable and
objective measurement of visual attention (Christianson et
al. 1991; Deubel and Schneider 1996; Krugman et al. 1994;
Rosbergen, Pieters, and Wedel 1997; Tsal and Lavie 1993;
van der Heijden 1992; Wedel and Pieters 2000). Research
in neurology supports the link between attention and eye
movements (Kustov and Robinson 1996; Mohler and Wurtz
1976). Eye movements on scenes are composed of fixations
and saccades (Rayner 1998). Fixations are the brief moments
in which the eye is stable and information is extracted from
the scene. Saccades are movements of the eyes that last
about 20–40 milliseconds (Matin 1974; Rayner 1998; Pie-
ters and Wedel 2008). During a saccade, vision is suppressed
000 JOURNAL OF CONSUMER RESEARCH
Please use DOI when citing. Page numbers are not final.
(Matin 1974; Rayner 1998), and the eye is redirected to a
new fixation (van der Lans et al. 2008b). A fixation lasts
about 200–400 milliseconds on average and in general
ranges between 50 milliseconds to a second (Pieters and
Warlop 1999; Rayner 1998). Eye fixations are instrumental
in acquiring information from the perceptual field (Anstis
1974; Sanders and Donk 1996). For example, in reading,
typically, relevant information is acquired during the first
50–70 milliseconds of a fixation (Rayner 1998), while in
viewing web pages visual attention can be assessed within
50 milliseconds (Lindgaard et al. 2006). Recent research in
ad perception documented that an ad can be perceived in a
single fixation of less than 100 milliseconds. Consumers can
identify what they are looking at (an ad vs. editorial material)
even when the information presented is deliberately blurred
(Pieters and Wedel 2012). These findings are consistent with
research in visual perception. In brief, with 90% accuracy,
briefly presented images of natural scenes (e.g., beach, for-
est) can be identified (Castelhano and Henderson 2008;
Rousselet, Joubert, and Fabre-Thorpe 2005). Grill-Spector
and Kanwisher (2005) found that 65 milliseconds were
needed to identify an object with accuracy, while there were
no differences in detection and categorization at 17, 33, 50,
68, or 167 milliseconds. Individuals were accurate in de-
tecting and categorizing typical objects in each time frame.
Both reached a ceiling of 90% accuracy after 80 millisec-
onds of single exposure. The sequence of fixations (gaze
pattern) also provides information about the visual search
process (van der Lans et al. 2008b). In this respect, visual
attention is a systematic process that is accessed via the
study of eye movements: eye fixations, saccades, and gaze
patterns and durations (Christianson et al. 1991; van der
Heijden 1992; van der Lans et al. 2008b).
Visual Attention and Choice
Attention is involved in all marketing efforts (Milosavljevic
and Cerf 2008). Visual attention, captured through eye move-
ments, is a significant predictor of brand choice (Maughan,
Gutnikov, and Stevens 2007; Pieters and Warlop 1999). Recent
evidence suggests that looking at an item for longer can lead
to higher choice likelihood (Armel, Beaumel, and Rangel 2008;
Krajbich, Armel, and Rangel 2010). In a binary choice task
between familiar options, individuals were more likely to
choose the snack food option that they looked at more, after
controlling for preexisting preferences for each option (Krajbich
et al. 2010).
In visual research, Shimojo et al. (2003) demonstrated
that just prior to the choice there is a progressively increasing
bias in the likelihood that the observer’s gaze will be di-
rected toward the chosen stimulus; they termed this the gaze
cascade effect. Particularly, individuals display an “ava-
lanche of fixations on the to-be-chosen object” in the final
seconds of the gaze duration (Changizi and Shimojo 2008,
1512). In explaining this effect, Simion and Shimojo (2006)
argue that the gaze pattern is intrinsically involved in pref-
erence formation by way of a dual process with a feedback
loop: the more the individuals look at a stimulus, the more
they like it, and the more they like it the more they look at
it. This dual process shapes the preference for the eventually
chosen option. Although the gaze cascade effect addresses
the link between choice and final gaze patterns, no argument
has been made about the role of the location of the choice
options, namely, horizontal centrality on the gaze cascade
effect.
Centrality Bias in Visual Attention
In general, eye fixations are directed at the informative
elements of a scene (Antes 1974; Buswell 1935; Mack-
worth and Morandi 1967; Parkhurst, Law, and Niebur
2002; Russo 2011; Yarbus 1967). When a scene appears,
the natural initial response is to look at the center of it;
this has been called the central fixation bias (Tatler 2007;
Tatler, Baddeley, and Gilchrist 2005). The central fixation
bias occurs regardless of how the informative features of
the scene are distributed (Tatler 2007). Two main reasons
motivate this bias. First, the center of the scene is uncon-
sciously considered the optimal location to extract infor-
mation as individuals expect to find informative elements
at this location even when there is none. This is referred
to as the photographer bias and is observed when indi-
viduals have had no previous exposure to the scene (Tseng
et al. 2009). Second, there is a predisposition, called the
orbital reserve, that is an innate preference for eye move-
ments that place the pupils in the central position (looking
straight ahead) rather than elsewhere (Pare´ and Munoz
2001). This causes a re-centering bias to emerge as soon
as the pupils leave the central position. Independent of the
cause, individuals are predisposed to look at the center of
a given visual field (Ho-Phuoc, Guyader, and Gue´rin-Du-
gue´ 2010). Therefore, the tendency to look more at the
center could explain the horizontal centrality effect on
choice. The current research investigates the links between
central fixation bias, gaze cascade effect, and horizontal
centrality in a consumer choice context. Note that the hor-
izontal centrality effect is relevant when the number of
alternatives is greater than two, yet earlier research dem-
onstrating the gaze cascade effect focused on binary choice
sets.
METHODOLOGY
Overview
Using an eye-tracking methodology, studies 1A and 1B
explored the effect of horizontal centrality on choice like-
lihood and how this effect is linked to increased visual at-
tention and/or inferences. These studies also examined the
process of visual attention and how it is involved in the
choice process. Study 2 extended the generalizability of the
findings to an offline context.
