Positive, Energetic Multisensory Stimuli: When Ads Can Hurt Your

ASSOCIATION FOR CONSUMER RESEARCH
Labovitz School of Business & Economics, University of Minnesota Duluth, 11 E. Superior Street, Suite 210, Duluth, MN 55802
Positive, Energetic Multisensory Stimuli: When Ads Can Hurt Your Brand
Dhruv Grewal, Babson College
Keith Wilcox, Columbia University, USA
Nancy Puccinelli, University of Oxford, UK
The effects of multi-sensory cues are often assumed to be static. This research demonstrates that cues that typically lead consumers to
feel positive and upbeat, can create cognitive conflict among sad consumers that renders the stimuli difficult to process and leads to
more negative brand attitudes.
[to cite]:
Dhruv Grewal, Keith Wilcox, and Nancy Puccinelli (2015) ,"Positive, Energetic Multisensory Stimuli: When Ads Can Hurt Your
Brand", in AP - Asia-Pacific Advances in Consumer Research Volume 11, eds. Echo Wen Wan and Meng Zhang, Duluth, MN :
Association for Consumer Research, Pages: 9-12.
[url]:
http://www.acrwebsite.org/volumes/1018748/volumes/ap11/AP-11
[copyright notice]:
This work is copyrighted by The Association for Consumer Research. For permission to copy or use this work in whole or in
part, please contact the Copyright Clearance Center at http://www.copyright.com/.
Consumer Responses to Multisensory Inputs
Rhonda Hadi, University of Oxford, UK
Paper #1: Virtual Touch: How Computer Interfaces Impact
Consumer Choice
Hao Shen, Chinese University of Hong Kong, China
Meng Zhang, Chinese University of Hong Kong, China
Aradhna Krishna, University of Michigan, USA
et al.), affective state (Puccinelli, et al.), and dispositional sensationseeking (Hadi, et al.).
We are honored to have Rashmi Adaval, an eminent scholar
specializing in affect, sensory information, and consumer judgments,
as a discussant. Consistent with the conference’s “Vibrant Integration” theme, Dr. Adaval’s expertise and the progressive approaches
from this session are sure to induce a lively discussion, and are likely to appeal not only to sensory researchers but also to researchers
grounded in more traditional approaches to cognitive psychology
and consumer behavior.
Paper #2: Positive, Energetic Multisensory Stimuli: When Ads
Can Hurt Your Brand
Nancy M. Puccinelli, University of Oxford, UK
Keith Wilcox, Columbia University, USA
Dhruv Grewal, Babson College, USA
Virtual Touch: How Computer Interfaces Impact
Consumer Choice
Paper #3: Boost the Brightness, But Turn Down the Volume:
Cross-modal Compensation for Meta-Sensory Homeostasis
Rhonda Hadi, University of Oxford, UK
Lauren Block, Baruch College, USA
Suresh Ramanathan, Texas A&M University, USA
EXTENDED ABSTRACT
In the digital world, purchase decisions and product choices are
increasingly made on a variety of computer interfaces. For example,
people can order take-out meals on their smartphone, ipad, laptop,
desktop computer, etc. Some of these interfaces are touch-based
(e.g., iPhone, iPad) whereas others are not (e.g., laptop, desktop
computers). When using a touch-interface, consumers often make
a choice by touching the image of the option they want; on a desktop, they typically indicate their choice by using a mouse to click.
Will the act of choosing on different interfaces lead to differences in
choices? This paper examines such a possibility.
We focus our investigation on a binary context similar to the
ones widely examined by prior researchers (Dhar & Wertenbroch
2000; Shiv & Fedorikhin 1999), that is, one alternative is affectladen (i.e., it elicits a strong positive affect, but has fewer positive
cognitive associations; e.g., cheese cake) and the other is cognitively
superior (i.e., it has better cognitive benefits, but does not induce
strong positive affect; e.g., fruit salad). The pair of products is presented pictorially on an iPad or a desktop. We predict that choosing
on an iPad (vs. a desktop) enhances the choice share of the affectladen alternative over the cognitively-superior one.
People may mentally stimulate the expected interactions with
an object upon seeing it (Elder & Krishna, 2011; Tucker & Ellis,
1998; Shen & Sengupta, 2012). For example, the sight of a soccer
ball makes people mentally simulate the action of kicking it with
their foot. Decades of dual-system models (Esptein, 1993; Loewenstein, 1996; Metcalfe & Mischel, 1999; Pham, 1998) have shown a
key difference between the reason-based, more utilitarian alternative
and the feeling-based, more hedonic one: Relative to the former, the
latter induces a spontaneous action tendency within consumers to
“grab it directly”. This is, when consumers reach out to touch an option by their hand on an iPad, this action is very similar to the action
that they would spontaneously simulate in their mind upon seeing
the image of a hedonic product (i.e., an impulse to grab it – Shiv &
Fedorikhin 1999). Reaching out to touch the hedonic product on a
touch-interface is thus more congruent with the impulse than using a
mouse to click on a non-touch interface, leading to a greater choice
share.
We conducted five experiments to test our prediction. In Study
1 (N = 85), participants used either an iPad or a desktop to make a
binary choice that was visually presented on the screen. The results
were as predicted: Participants were more likely to choose cheesecake over salad when they made their choice on an iPad (95%) than
on a desktop (73%; p <.01).
Paper #4: Discussant
Rashmi Adaval, Hong Kong University of Science &
Technology, China
SESSION OVERVIEW
Consumers are continuously exposed to sensory cues, both in
their environments (e.g., retail atmospheres) and while interacting
with products (e.g., touching products). While it is well documented
that sensory cues impact meaningful consumer behaviors, researchers have identified the need to more critically examine the mechanisms through which sensory cues exert their effects (Krishna, 2012).
The proposed session addresses this call and advances sensory
research by combining three papers that move research on sensory
cues away from mere effect demonstrations and on to a more refined understanding of when and why sensory cues are relevant to
consumers’ evaluations and behavior. Specifically, each of the three
papers draws upon different supporting theories to explore the impact of various sensations (haptic, visual and auditory) on important
marketing outcomes: product choice, product evaluations, and brand
attitudes.
The first paper examines the impact of a single-modality cue:
haptics. Shen and colleagues find that touch interfaces (e.g., digital
tablets) lead to more affect-laden (as opposed to cognitively-superior) choices when compared to non-touch interfaces (e.g., computer with a mouse). They find that this effect is driven by increased
mental simulation. Puccinelli and colleagues demonstrate positive,
energetic multisensory stimuli can elicit a negative reaction when
they conflict with the emotional state of the consumer. This research
demonstrates the dynamic nature of multisensory cues. Finally, Hadi
and colleagues bring the session to a close by demonstrating that
sensory exposure in one modality may trigger general sensory goals.
These goals might be satisfied along other modalities, in order to aid
in meta-sensorial homeostasis. All of these papers are in advanced
stages of completion, with multiple studies run.
Together, these papers highlight that there are many different
nuanced processes through which sensory cues impact consumer
responses. Particularly, they draw upon three different theoretical
frameworks to explain the sensory effects: mental simulation (Shen,
et al.), cognitive conflict (Puccinelli, et al.), and homeostasis (Hadi,
et al.). Additionally, these papers explore distinct moderators to help
define boundary conditions of these effects: depiction mode (Shen,
9
Asia-Pacific Advances in Consumer Research
Volume 11, © 2015
10 / Consumer Responses to Multisensory Inputs
Study 2 (N = 192) further demonstrated the effect of interface
on reaction time. In that study, participants were told that after they
saw any pair of pictures, they should choose the smaller picture as
quickly as possible. After some practices, they went to the target pair
(cheesecake vs. salad). We found that when participants made the
choice on an iPad, they were quicker to make a choice when the cake
(vs. the salad) was the correct response (p < .05), whereas this difference disappeared when they made choice on desktop (F <1). The
interaction effect was also significant (p < .05).
Study 3 (N = 132) further replicated our findings using similar
stimuli (ice-cream vs. USB) and went a step forward to show that
such an effect was not simply due to any other differences between
an iPad or a desktop. Specifically, participants were assigned to
three conditions: one on desktop and two on iPad. In the iPad conditions, half the participants made their choices by directly touching
the pictures using their fingers (compatible with “grabbing the product directly”), whereas the other half made their choices by touching the pictures via an iPad touch pen (incompatible). As the latter
iPad condition bears little resemblance to the mentally stimulated
responses upon seeing a hedonic product, the choice of the hedonic
product in this condition (58%) was similar to the desktop condition
(60%) - both being significantly lower than the other iPad condition
(80%; ps<.05).
Study 4 then tested the predicted effect using a moderator.
Since people typically use their dominant hand to grab the object
they want, using the dominant hand to choose on an iPad might
match their spontaneous impulse of grabbing the hedonic product,
whereas using the non-dominant hand to do so, might not. We recruited only participants those who were right-handed (N = 226).
Participants were either instructed to use their right or left hand to
choose between pictures of two types of food on the screen. As predicted, when they used their right hand to choose, they were more
likely to choose a cupcake over blueberries on an iPad (69%) than on
a desktop (52%, p = .07). However, when they used their left hand
to make their choice, such a pattern was not observed (35% vs. 48%,
p = .16). The interaction effect was significant (p = .02). In addition, we also demonstrated that our effect in right hand condition was
mediated by the increased desire to reach toward the hedonic food.
Study 5 (N = 224) examined whether our predicted effect held
only when the products were presented pictorially. This is because
that prior evidence only shows a strong link between “vision” and
“action” such that the visual depiction of an object can activate the
mental simulation of interacting with the object, but not so strong between the name of the objects and the mentally simulated responses
(Tucker & Ellis, 1998). As expected, when the two options were
visually presented on the screen, participants were more likely to
choose an ice-cream cup over a corn cup on an iPad than a desktop
(71% vs. 51%, p = .03). However, when the two foods were verbally
presented (i.e., their names were shown), such an effect disappeared
(60% vs. 69%, respectively, n.s.). The interaction effect was significant (p = .02).
Taken together, the results of five experiments provide converging evidence for our prediction. These findings advance the current
literature of mental simulation, affect-cognition choice conflict, and
have rich implications in the new area concerning the influence of
different digital interfaces on consumer behavior.
Positive, Energetic Multisensory Stimuli: When Ads Can
Hurt Your Brand
EXTENDED ABSTRACT
Sensory experience is often assumed to be static. This research
demonstrates the dynamic nature of sensory experience, in particular, the moderating role of consumer emotional state. While positive, energetic stimuli generally make consumers feel positive and
upbeat, there are conditions under which the reverse occurs. From
a managerial perspective, there is a growing paradox within media:
while ads remain uniformly positive and upbeat, the media content
they sponsor grows increasingly negative. Eleven out of the top 25
programs are negative (Schneider 2013). By comparison, only about
10% of the top programs from 1990-1999 where negative (TV.com
2013).
The negative emotions engendered by media content are likely
to affect consumers since emotions have been shown to influence
judgment and decision-making (Adaval 2001; Cohen & Andrade
2004; Raghunathan & Pham 1999). However, it is less clear how
consumers experiencing a negative emotion, such as sadness, will
respond to positive marketing communications. One perspective argues that people want to feel good so they will seek ways to uplift
themselves when they are feeling down (Zillmann 1988), suggesting
that consumers experiencing negative emotions should respond favorably to positive marketing communications. Nevertheless, there
is evidence that people experiencing negative emotions often avoid
activities that might improve their affective state (Erber, Wegner &
Therriault 1996). For instance, people frequently choose to listen to
sad songs instead of happy songs when they are sad (Cohen & Andrade 2004). Similarly, consumers often prefer negative salespeople
over positive salespeople when they are experiencing sadness (Puccinelli 2006). Thus, a better understanding of how and why people
experiencing negative emotions respond to positive sensory cues can
help managers improve on the execution of their marketing strategies.
We argue that because the experience of sadness is characterized by a preference for inactivity (Rucker & Petty 2004), sad
consumers experience conflict when exposed to positive, energetic
content. Consequently, they find the experience of watching such
content more difficult, making them more likely to avoid watching
it compared to positive content that is less energetic. Additionally,
we propose that the same conflict from watching highly energetic
commercials will not be observed among consumers experiencing a
neutral affective state or another negative emotion not characterized
by deactivation (e.g., anger) because such states are not characterized by a preference for inactivity. While highly energetic positive
commercials may not be inherently difficult to watch, we show that
sad consumers find such commercials difficult because the characteristics of the commercial conflict with their emotional state.
Four studies were conducted to test this theory. In Study 1 (N
= 94), we induced sadness or a neutral affective state using a video
clip before having participants watch a positive commercial that was
either high or low energy. We show that people experiencing sadness spend less time watching the positive, high energy commercial
compared to a positive low energy commercial (p < .05). The tendency to avoid the high energy commercial was not observed for
people in a neutral affective state. Study 2 (N = 80) demonstrates
that people experiencing sadness have less favorable attitudes (p <
.01) and marginally less favorable behavioral intentions (p < .10)
toward an advertiser after watching a positive high energy commercial by the advertiser compared to a positive low energy commercial
by the advertiser. Studies 3a (N = 154) and 3b (N = 178) provide
Asia-Pacific Advances in Consumer Research (Volume 11) / 11
evidence that the difficulty viewing high energy content by those experiencing sadness is due to a preference for inactivity. Specifically,
we show that the effect of sadness on the difficulty of exposure to
positive, high energy cues is attenuated when people are primed with
an action goal, that reduces people’s preference for inactivity (Study
3a interaction p = .001; Study 3b interaction p < .05). Additionally,
we demonstrate that the same aversion to high energy cues is not
observed when people are in a neutral affective state (Study 3a) or
experience another negative emotion not characterized by deactivation (i.e., anger; Study 3b). Studies 3a and 3b suggest that the phenomenon generalizes to the broader media context. Finally, Study
4 (N = 197) was a field study on Hulu (hulu.com) to demonstrate
that sad consumers perceive positive, high energy commercials to
be more difficult to watch than low energy commercials (p < .05).
This research makes several contributions. This research demonstrates that under certain conditions, consumers can find positive
stimuli difficult to process, choose to reduce their exposure to the
stimulus, and exhibit more negative attitudes towards the stimulus.
Specifically, when consumers are sad they respond more negatively
to positive, upbeat stimuli which may cause positive, upbeat ads to
backfire.
Boost the Brightness, But Turn Down the Volume:
Cross-modal Compensation for Meta-Sensory Homeostasis
EXTENDED ABSTRACT
While much of marketing practice seems to stem from the basic
tenet that more sensory stimulation is better (Lutz & Lutz 1978),
research has documented the potential risks of sensory overload
(Baker 1984; Malhotra 1984; Morrin & Chebat 2005; Soars 2009),
and the importance for individuals to achieve and maintain relatively
homeostatic levels of sensory stimulation (Berlyne 1960; Steenkamp
& Baumgartner 1992). Everyday observation suggests that we do
indeed seek balance within each sensory mode. When music is too
loud (quiet), we lower (increase) the volume. We dislike food that is
too bland or too spicy. However, no research has examined whether over-stimulation in one sensory mode (e.g., audition) may lead
individuals to compensate by seeking reduced stimulation in other
sensory modes (e.g., vision or olfaction). In this research we demonstrate that this may be the case: when individuals are exposed to
sensory stimuli sufficiently above their optimal homeostatic level in
one sense, they may seek to compensate across other sensory dimensions in an effort to achieve meta-sensory homeostasis.
Individuals often face fluctuations in their external environment
that may originate from natural sources (e.g., changing weather) or
marketing-driven sources (e.g., music in a retail store). In response
to such variations, we have a tendency to regulate our internal environment to ensure stability- a dynamic, iterative process known
as homeostasis (Jänig 2008; Marieb & Hoehn 2007). While this is
often done via automatic physiological processes, research suggests
that decision-making can also function as a homeostatic process
(Bechara, Damasio & Damasio 2000; Paulus 2007). Specifically,
our brains encourage behavioral homeostasis via state-dependent
alterations of hedonic perception, by signaling pleasure or aversion
towards stimuli depending on our internal state and corresponding
homeostatic needs, and we are accordingly led to make decisions
that promote stability (Cabanac 1971; Rolls 2005, Paulus 2007). For
example, if an individual is exposed to a sub-optimally cold environment, she will be motivated to seek warmth and will make decisions to help achieve that warmth goal. This homeostasis-driven
decision-making account is consistent with optimal stimulation level
(OSL) theory, which posits that if environmental stimulation is too
low (high), individuals will attempt to increase (decrease) stimulation (Berlyne 1960, 1974; Steenkamp & Baumgartner 1992).
Within the sensory domain, research has demonstrated the existence of optimal levels of visual (Berlyne 1971), auditory (Knöferle
et al. 2012) and olfactory (Doty et al., 1984) stimulation. Thus the
majority of research on sensory OSL has examined the construct
within independent sensory modes. However, perception is fundamentally multisensory (Krishna 2012), and one area of the brain
(the insular cortex) is responsible for integrating multimodal sensory
inputs (Wheeler 2003). Accordingly, cross-modal correspondences
exist between the sensory modalities (e.g., haptic and auditory sensations influence taste perceptions, Krishna & Morrin 2008; Spence et
al. 2013). While research suggests that cross-modal congruency is
generally favorable (Holmes & Spence 2005), the synaesthetic quality of sensory inputs should also imply that if input in one modality
exceeds an individual’s optimal level and she is unable to quell input
in that modality, she may seek to reduce inputs in alternative modalities. This notion of homeostasis-driven, cross-modal compensation
is precisely the phenomenon we propose in this research.
In Study 1 (N = 95), we randomly assigned participants to a
condition based on a 2 (music valence: unfavorable vs. favorable)
x 2 (music volume: low vs. high) between-subjects design. Music
valence was manipulated by playing music that pretested as either
unfavorable or favorable, and music volume was manipulated by
pre-adjusting the volume of the stereo system in the lab. While being exposed to the music, participants evaluated a series of (nonauditory) products that differed in sensory intensity. We found that
when the music played was unfavorable and high in volume, participants showed a greater preference for products low in sensory-intensity (e.g., mild food, unscented products, visually-simple objects)
as compared to products high in sensory-intensity (e.g., spicy food,
scented products, visually-complex objects). This suggests that the
unfavorably high auditory stimulation led to a reduced desire for
stimulation in other modalities. A follow-up study followed the same
procedure, but included a condition in which individuals were able
to adjust the music volume to their liking before evaluating products.
This study demonstrated that when individuals are able to regulate
within the disturbed modality, they no longer show the tendency to
compensate across other modalities.
Our first two studies suggest that if we are exposed to aboveoptimal environmental sensations in one modality, we may seek
decreased sensory input across other modalities. Interestingly however, sensory inputs also contain semantic meaning, due to their
common associations with experiences (e.g., cinnamon is a “warm”
scent and mint is a “cool” scent, Krishna, Elder, & Caldara 2010).
In fact, cross-modal congruence can be assessed by the consistency
of the semantic associations among sensory characteristics and has
been shown to impact behavior (Holland, Hendriks, & Aarts 2005).
If this is the case, then according to our earlier theorizing, we should
expect the cross-modal compensation phenomenon to additionally
apply across conceptual dimensions. Thus in Study 3 (N = 130), we
wished to examine if above-optimal levels of an environmental sensation would lead individuals to compensate along the same dimension but in other sensory modes. We manipulated the temperature
of the laboratory (between 64-91°F) to create a variety of conditions
that would lead to variations in thermal comfort. Participants indicated the extent to which they wished the room was colder, and
then indicated their current desire for products that varied in terms
of conceptual temperature (e.g. conceptually cold products included
mint-flavored gum and Piano Jazz music, while perceptually warm
products included cinnamon-flavored gum and Latin Salsa music).
We found that individuals who wished the room was colder were
12 / Consumer Responses to Multisensory Inputs
more likely to show preference for the conceptually cool (vs. warm)
products than those individuals who did not want the room colder.
In this research, we uncover a cross-modal compensation phenomenon, by demonstrating that when individuals are exposed to
sensory stimuli sufficiently above their optimal homeostatic level in
one sense, they may seek to compensate other sensory modes, either
in terms of intensity or across a conceptual dimension. This research
has important implications for marketers targeting consumers within
contextual and dynamic environments. Future studies will measure
arousal to examine its role in the underlying process and explore the
moderating role of dispositional sensation-seeking.
Discussant
Drawing upon the findings across these three papers, we will
discuss comparative as well as overarching issues pertaining to the
antecedents of well-being. Comments and suggestions from the audience will be sought so as to enhance the session’s interactivity and
the discussion’s overall quality.