Content uploaded by Kosuke Motoki
Author content
All content in this area was uploaded by Kosuke Motoki on Nov 04, 2018
Content may be subject to copyright.
1
Light colors and comfortable warmth: Crossmodal correspondences between thermal
sensations and color lightness influence consumer behavior
Forthcoming in Food Quality and Preference
Names and affiliations:
Kosuke Motoki1,2, Toshiki Saito1, Rui Nouchi1,3,4, Ryuta Kawashima1,3, and Motoaki
Sugiura1,4
1Institute of Development, Aging and Cancer, Tohoku University, Sendai, Japan; 2Japan
Society for the Promotion of Science, Tokyo, Japan; 3Smart-Aging Research Center, Sendai,
Japan; 4International Research Institute of Disaster Science, Tohoku University, Sendai, Japan
Correspondence should be addressed to Kosuke Motoki, Institute of Development, Aging and
Cancer, Tohoku University, Seiryo-machi 4-1, Aoba-ku, Sendai 980-8575, Japan.
E-mail: kosuke.motoki.p2@dc.tohoku.ac.jp
2
ABSTRACT
Consumers are often surrounded by somatosensory (e.g., warmth) and visual (e.g., color)
information. For example, consumers often see light-colored goods under warm conditions.
Previous studies have shown that sensory interactions, such as those involving auditory and
visual stimuli, influence consumer behaviors. However, it remains unknown whether
somatosensory–visual information (e.g., warmth and color lightness) interactively guide
consumer behaviors. Additionally, the conditions under which sensory interactions increase
consumer preferences are also unclear. This study focused on how the effects of the novel
correspondences between somatosensory and visual (warmth and color lightness) perceptions
extend from the capture of visual attention to the formation of preferences, as well as on how
attitudes toward sensory experiences (i.e., positive reactions to sensory experiences) play
critical roles in preference formation. The results showed the existence of crossmodal
correspondences between feeling warm and light colors (Study 1), and such crossmodal
correspondences influenced consumers’ visual attention. Physical warmth increased the visual
attention directed toward light-colored goods (Study 2). Although this correspondence did not
directly influence consumer preferences (Study 3), it did increase consumer preferences for
light-colored goods under conditions of comfortable (but not uncomfortable) warmth (Study
4). These results reveal novel crossmodal correspondences between thermal sensations and
levels of color lightness and demonstrate how such correspondences have consumer-relevant
consequences.
Keywords: Crossmodal correspondences; Color lightness; Thermal sensation; Thermal
comfort; Sensory marketing
3
INTRODUCTION
Consumers are surrounded by both somatosensory (e.g., warmth) and visual (e.g., color)
information. For example, when consumers enter a warm shop (somatosensory) selling
products with various levels of lightness (visual), to which stimulus do they attend, and which
do they prefer? Consumers may differ in their degree of comfort in a warm shopping
environment; one person may feel comfortable, whereas another may experience discomfort.
To what extent do consumers’ attitudes toward warmth affect their buying decisions regarding
colored products?
The thermal environment and color of products are important considerations when designing
marketing strategies for many business settings. Stores can usually control ambient
temperatures using air-conditioners, and light- and/or dark-colored products are displayed in
such venues. Recognition of the impact of sensory experiences in natural shopping
environments has led to increased attention to the effects of such experiences on consumer
behavior (Krishna, 2012; Krishna, Cian, & Sokolova, 2016; Krishna & Schwarz, 2014). For
example, thermal sensations influence gift-giving (Williams & Bargh, 2008). After use of a hot
therapeutic pad, participants were more likely to give goods to others than to themselves. Note
that these findings could not be replicated; they must thus be considered with caution (Lynott,
Corker, Wortman, Connell, Donnellan, Lucas, & O’Brien, 2014). Color lightness also
influences consumer behavior, and research has shown that advertisements including lighter
colors are preferable (Gorn, Chattopadhyay, Yi, & Dahl, 1997). Despite such findings,
interactions between the senses (i.e., thermal and color perceptions) have been studied only in
terms of thermal and color hue perception (Ho, Iwai, Yoshikawa, Watanabe, & Nishida, 2014;
Ho, Van Doorn, Kawabe, Watanabe, & Spence, 2014). Neither the interactions of thermal and
color lightness nor the effects thereof on consumer behavior have been investigated to date.”
4
We investigated how thermal sensation and color lightness interacted and explored how
such interactions influenced lower perceptual (information processing) behavior and
higher-level, cognitive consumer behavior (preference formation).”
Many previous studies have tested only the “match” between senses (Deroy, Crisinel, &
Spence, 2013; Velasco, Woods, Petit, Cheok, & Spence, 2016), but few have reported on how
sensory interactions affect attention and preferences (Hagtvedt & Brasel, 2016). Most
importantly, to our knowledge, no study has investigated how sensory interactions guide
consumer-relevant outcomes in a way that is dependent on sensory attitudes (e.g., thermal
comfort). Indeed, it is possible that sensory interactions affect consumer-relevant outcomes
(e.g., preferences for products) only when consumers have a positive attitude toward the
sensory experiences involved. Thus, this research examined how sensory matching
(temperature and color lightness) affects consumer-relevant consequences in terms of the
relationship between information processing and preference formation as well as how sensory
attitudes play a critical role in preference formation.
THEORETICAL BACKGROUND
Visual (e.g., color) and somatosensory (e.g., warmth) information are important in a variety
of marketing domains. Sensory marketing has recently evolved, and research has shown that
visual and somatosensory information influences consumer attention and preferences (Krishna,
2012; Krishna, Cian, & Sokolova, 2016; Krishna & Schwarz, 2014). However, previous
studies investigated visual or somatosensory influences on consumer behaviors separately
(Gorn, Chattopadhyay, Yi, & Dahl, 1997; Motoki, Saito, Nouchi, Kawashima, & Sugiura,
2018; Singh, 2006; Williams & Bargh, 2008; Zwebner, Lee, & Goldenberg, 2014). Thus, how
multisensory interactions between visual and somatosensory perceptions influence consumer
behaviors remains unstudied. Given that consumers often perceive visual (e.g., colored
products) and somatosensory (e.g., ambient warmth) stimuli at the same time, more systematic
5
research on visual–somatosensory interactions is needed to expand our understanding of
sensory influences on consumer behaviors.
The literature on crossmodal correspondences, defined as apparently arbitrary associations
among sensory features of different modalities (Spence 2012; Parise 2016), suggests that
thermal sensation and color are interactively matched. Consider, for example, thermal–hue
correspondence, according to which people are likely to match temperature-related words
(warm/cool) with different hues (red/blue) (Ho, Van Doorn, Kawabe, Watanabe, & Spence,
2014).
These crossmodal correspondences may be attributable to statistical learning (Spence, 2012),
as individuals may learn that certain thermal (warm/cool) and color (red/blue) properties
co-occur and interact. Many crossmodal correspondences are likely to result from
internalization of statistical regularities in the environment (i.e., frequent pairing between
different senses) (Spence & Deroy, 2013). Such information can be used to identify sensory
signals that normally occur simultaneously or separately (Spence & Deroy, 2013). Also, the
brain may use predictive models to minimize errors; the brain may adapt new sensory signals to
internal predictions (Piqueras-Fiszman & Spence, 2015). In support of this theory, it has been
shown that sensory integration of two arbitrary sensory signals (e.g., visual and somatosensory
signals) can be learned if they statistically co-occur (Ernst 2007).
Indeed, this interplay of thermal sensation and hue are common features in the fields of
industrial and interior design in many parts of the world (e.g., the color schema of hot and cold
water taps) (Fenko, Schifferstein, & Hekkert, 2010). The other explanation for these
interactions involves structural similarity. That is, these correspondences may occur due to
similarities in the dimensions of the stimuli (intensity, arousal, positivity, length, etc.) or in the
neural connections they elicit (Spence, 2012) .
Although thermal–hue correspondences have been demonstrated, questions regarding
whether thermal sensations and other types of color dimensions (i.e., color lightness) are
matched remain unanswered. Color can be divided into three dimensions: value (lightness),
hue, and saturation (Labrecque, Patrick, & Milne, 2013). The current research focuses only on
6
value because this dimension has been shown to influence visual attention (Gorn et al., 1997)
and consumer preferences (Hagtvedt & Brasel, 2016), which are our main variables of interest.
Given that the term “value” has other connotations in marketing (e.g., in the context of pricing),
following a previous study (Hagtvedt & Brasel, 2016), we use “lightness” to refer to this
variable.
Study 1: Crossmodal Correspondences between Color lightness and Thermal sensation
Thermal sensation and color lightness are likely to be matched. In a Munsell color system,
value (color lightness) can be defined as the degree of darkness or lightness in a given color
(Labrecque et al., 2013). Black is on the low-value end of this continuum, whereas white is on
the high-value end. In terms of associations between thermal sensation and color lightness,
statistical rules may be in play that associate warmth/coldness with light/dark colors. Thus,
people often experience co-occurring thermal sensations and color lightness perceptions, and
may internalize the statistical regularities of these pairings to, in turn, develop crossmodal
correspondences. In our daily lives, thermal sensation (warm/cool) is often referenced in terms
of color lightness (light/dark), with light colors typically associated with warmth and dark
colors with coolness. Lighter colors seem to be associated with sunshine and daytime, whereas
darker colors seem to be associated with night, and evening temperatures are usually cooler
than daytime temperatures. In the context of natural science, color lightness is an effect of
heating (MacIsaac, Kanner, & Anderson, 1999), and some animals (treefrogs) turn a light color
in warm temperatures (King, Hauff, & Phillips, 1994). Given that color lightness can be
defined as the degree of lightness in a given color, it would not be surprising if there were
associations between thermal sensation and color lightness. Thus, people often experience the
co-occurrence of thermal sensations and perceptions of color lightness, and may internalize the
statistical regularities of these pairings, in turn leading to the development of crossmodal
correspondences. However, to our knowledge, no study has examined such possible
correspondences.
7
Prior to examining the effects of such interactions on consumer-relevant outcomes, we
investigated whether novel correspondences between thermal sensation and color lightness
could be identified. Thus, Study 1 investigated whether people reliably matched thermal
sensations (warmth/coolness) with color lightness (light/dark colors)
Study 2: Thermal sensation–color lightness correspondences influence consumers’ visual
attention
Crossmodal correspondences between thermal sensation and color lightness are likely to
increase the visual attention devoted by consumers to the stimuli to which these properties
belong. Consumers’ visual attention (e.g., the target and duration of their gaze) is a precursor of
memory, preference, and buying behaviors (Wedel & Pieters, 2008). Thus, the importance of
understanding consumers’ visual attention has been recognized for a very long time (Wedel &
Pieters, 2008). In general, crossmodal correspondences have been shown to enhance
performance, enhancement that, possibly, requires attention (e.g., perceptual discrimination,
accelerated classification) (Spence 2011). This may be because people respond to incoming
sensory signals more rapidly and more correctly when the signals are congruent with
predictions (Spence 2011). Recently, it has been shown that sensory-matching captures
consumer visual attention (Hagtvedt & Brasel 2016). For example, vision–audition
correspondence enhanced visual attention (Hagtvedt & Brasel, 2016) according to a study in
which participants simultaneously listened to high/low tones and viewed light/dark stimuli
(Hagtvedt & Brasel, 2016). When the sounds matched the colors (high tone–light color, low
tone–dark color), participants viewed the stimuli longer (Hagtvedt & Brasel, 2016). However,
the previous study focused on the relationship between vision and sound, and whether
correspondence between thermal sensation and vision (color lightness) interactively capture
visual attention remains unknown. Given evidence that thermal sensation influences
subsequent behavior (Williams & Bargh, 2008), we hypothesized that thermal sensation would
increase the attention devoted to goods of the corresponding color.
8
To elucidate the implications of these correspondences for consumer-relevant outcomes, we
investigated whether they affect the lower cognitive functions involved in consumer behaviors
(visual attention). Thus, Study 2 investigated whether thermal sensation and color–lightness
correspondence enhance consumers’ visual attention. Similar to the experimental design used
by Williams and Bargh (2008), participants in this study experienced a thermal sensation by
briefly wearing either a warm or a cold pad. Soon after removing the pad, participants viewed
light- versus dark-colored goods while their eyes were tracked. We expected the primed
thermal sensation to affect the visual attention devoted to goods of the corresponding color.
Study 3: Thermal sensation–color lightness correspondences influence consumer preferences
In addition to visual attention, the crossmodal correspondences between thermal sensation
and color lightness also likely influence consumer preferences. Consumers’ attention to
products enhances the likelihood of preference and choice, although this is not always the case,
because the duration of fixation influences the value of an item depending on its valence. If the
item is neutral or positive, greater gaze time should result in a greater probability of choosing it,
whereas if the target is aversive, greater gaze time should result in less probability of choosing
it (Armel, Beaumel, & Rangel, 2008). For example, increases in relative visual attention
increase the probability of choosing neutral or appetitive items, such as faces (Shimojo, Simion,
Shimojo, & Scheier, 2003) or junk food (Armel, Beaumel, & Rangel, 2008; Krajbich, Armel,
& Rangel, 2010), and the opposite is true for aversive items (e.g., baby food (Armel, Beaumel,
& Rangel, 2008). It has also been shown that well-matched experiences such as the judgment
mode and regulatory orientations (e.g., eager strategy/promotion; vigilant strategy/prevention)
increase consumer preference (Higgins, Idson, Freitas, Spiegel, & Molden, 2003). Given that
crossmodal correspondences are derived from good matches between senses (Spence, 2011),
the feeling associated with this experience, “well-matching,” would be expected to increase
consumer preferences (e.g., thermal environment and product color). In fact, vision–audition
correspondence increases visual attention as well as preference (Hagtvedt & Brasel, 2016). The
longer people visually attend to light (vs. dark) products in warm environments (vs. cool), the
9
more likely they are to choose the product. However, whether the crossmodal correspondences
between thermal sensations and level of color lightness increase consumer preferences remains
unknown.
In addition to the application of these correspondences to the lower cognitive functions
involved in consumer-relevant outcomes (visual attention), we investigated whether these
correspondences affected the higher cognitive functions involved in consumer behaviors
(preference formation). Thus, using the same procedure as in Study 2, Study 3 investigated
whether the correspondences between thermal sensation and color lightness increased
consumer preferences. Participants evaluated their preference for light- and dark-colored
goods after experiencing physical warmth or coolness.
Study 4: Thermal comfort mediates the influence of thermal sensation–color lightness
correspondences on consumer preferences
Although several studies have found that crossmodal correspondences increase the value
consumers place on goods (Hagtvedt & Brasel, 2016; Sunaga, Park, & Spence, 2016), another
study produced the opposite results (Lowe & Haws, 2017). These inconsistent findings may be
attributable to the fact that previous studies considered only sensory perceptions and
overlooked the affective components of sensory experiences (e.g., thermal comfort in the
context of thermal experience). We assume that sensory attitudes play critical roles in the
formation of subsequent product preferences that are congruent with sensory perceptions.
Thermal experiences involve perception (e.g., thermal sensation: degree of warmth) as well
as attitudes (e.g., thermal comfort: degree of positive response to degree of warmth) (Arens,
Zhang, & Huizenga, 2006). There are individual differences in the degree to which people are
comfortable with given thermal sensations (Nikolopoulou & Lykoudis, 2006; Gagge, Stolwijk,
& Hardy, 1967; Nikolopoulou & Lykoudis, 2006). It seems that humans tend to seek
comfortable thermal sensations and to avoid uncomfortable ones. Indeed, in the same warm
store, some consumers feel comfortable, whereas others do not.
10
According to feelings-as-information theory, people rely on their current affective states to
provide information about product preferences (Schwarz, 2013). Specifically, consumers in a
pleasant (unpleasant) mood tend to prefer (dislike) products that are congruent with prior
information to maintain (mitigate) their emotions. For example, consumers who were in a
pleasant mood (excited or relaxed) showed greater preference for congruent products (exciting
or relaxing, respectively) (Kim, Park, & Schwarz, 2009). In contrast, consumers who were in
an unpleasant mood (guilty or ashamed) showed decreased preferences for congruent
advertising (guilt- or shame-related, respectively) (Agrawal, & Duhachek, 2010). According to
the feelings-as-information framework, consumers who feel thermal comfort (i.e., pleasant
thermal sensations) show increased preferences for the products with congruent colors. Thus,
we predicted that consumers would prefer light-colored (vs. dark-colored) goods only when
they were comfortable (vs. uncomfortable) in a warm (vs. cool) environment.
To examine the potential mediating influence of thermal–color lightness correspondences on
the higher functions involved in consumer behaviors (preference formation), we investigated
how positive responses to thermal sensation (i.e., thermal comfort) affect the formation of
preferences based on such correspondences. The feelings-as-information framework suggests
that positive responses to a thermal sensation (i.e., thermal comfort) may trigger a greater
preference for correspondingly colored products. Thus, Study 4 investigated whether thermal
comfort mediates the effects of the interactions between thermal environment and color
lightness on consumer preferences, hypothesizing that consumers who feel thermal comfort
will prefer correspondingly colored goods. For example, those who feel comfortable when
warm will exhibit a greater preference for light-colored products, whereas those who feel
comfortable when cool will exhibit a greater preference for dark-colored products.
Overview of empirical studies
In daily life, temperature is often discussed in terms of color lightness, with light colors
typically associated with warm/hot and dark colors with cool/cold conditions. For example, it
seems natural that lighter colors should be associated with sunshine and daytime, when
11
temperatures are usually warm, whereas darker colors are associated with night and evening,
when temperatures are usually relatively cool. However, this has not been empirically
investigated. Here, we report on a series of four experiments designed to assess whether
thermal sensation and color lightness were interactively matched and, if so, to examine the
effects of such matching on consumer behavior in terms of the roles played by lower and higher
cognitive functions (Table 1). Study 1 tested the crossmodal correspondences between
temperature (warm/cool conditions) and color lightness (light/dark). Given that sensory
correspondences influence information-processing and preference formation (Hagtvedt &
Brasel, 2016), we predicted that thermal-color lightness correspondences would increase
attention and preference. Study 2 investigated whether this correspondence influenced one of
the lower cognitive functions involved in consumer behavior (visual attention), using physical
temperature. Study 3 examined whether this correspondence influenced one of the higher
functions involved in consumer behaviors (preference formation), via manipulation of ambient
temperature. Study 4 investigated the role of sensory attitudes (thermal comfort) in the
formation of preferences based on the abovementioned correspondences.
STUDY 1
In Study 1, we tested the existence of crossmodal correspondences between thermal
sensation (warmth/coolness) and color lightness (light/dark).
Participants
In total, 21 healthy participants (13 females, Mage = 20.048±1.687 years) were recruited
using a bulletin board and a mailing list of students. It has been suggested that
between-subjects deigns require at least 20 observations per condition to have the power to
detect effects (Simmons et al., 2011). Our study used a within-subject design, which afforded a
substantial increase in the statistical power in comparison to a between-subjects design
12
(Greenwald, 1976). We included 21 participants in our study to achieve enough power to detect
effects with a reasonable degree of confidence. The number of participants included in studies
2–4 (N ≤ 20) was determined in the same way. This study (including all four separate Studies)
was approved by the Ethics Committee and was conducted in accordance with the Declaration
of Helsinki.
Design
The study used a within-subject design, with color lightness (from dark to light) as the
dependent variable and temperature (warm/cool) as the independent variable. Participants were
asked to match the thermal words with the degree of color lightness.
Procedure
Participants were asked to match thermal words (warm words: “hot” and “warm,” and cool
words: “cool” and “cold”) with color-lightness scales anchored by dark and light colors (Figure
1). In total, there were four trials in which participants matched each word with a value on a
visual analog scale (VAS) ranging from 0 (dark color) to 100 (light color). The order of the
trials was randomized across participants. All tasks were presented using PsychoPy (Peirce,
2007). Word stimuli were presented in Arial font of 0.2 norm (the normalized unit of
PsychoPy).
Statistical analyses
We calculated the degree to which warm (an average of “hot” and “warm”) and cool (an
average of “cool” and “cold”) words were matched with the corresponding colors and then
compared these results using paired t tests. Additionally, one-sample t-tests were performed to
13
test whether the ratings for each thermal word differed from the mid-point of the scale (50).
This procedure was similar to that used in previous research on crossmodal correspondences
(e.g., see Spence & Ngo, 2012, for a review). Data were analyzed using R statistical software
(R Core Team, 2013).
Results and discussion
Warm words were matched with lighter colors significantly more frequently than were cool
words (Mwarm words =73.31±17.82 vs. Mcool words =42.47±30.83; t20 = 3.174, p = 0.005, dz =
0.693) (Figure 2), and the same results were obtained from a separate analysis of each word. A
warm word (“hot”) (Mhot=75.14±22.01) was matched with lighter colors significantly more
frequently than were cold (Mcold =40.57±32.20; t20 = 3.163, p = 0.005, dz = 0.690) or cool
words (Mcool =44.38±33.92; t20 = 2.877, p = 0.009, dz = 0.628) words. The other warm words
(“warm”) (Mwarm=71.47±26.14) were matched with lighter colors significantly more
frequently than were cold (Mcold =40.57±0.322; t20 = 3.127, p = 0.005, dz = 0.682) or cool
words (Mcool =44.38±0.33; t20 = 2.435, p = 0.024, dz = 0.531)
Additionally, one-sample t-tests were performed to test whether the ratings for each
temperature-related word differed from the mid-point of the scale (50). This analysis revealed
that participants rated warm words significantly closer to the light-color end of the scales (t20 =
5.992, p < 0.001, d = 1.308), whereas the degree to which cool words were rated as being closer
to the dark-color end of the scales did not reach significance (t20 = −1.118, p = 0.277, d =
–0.244). The same results were obtained when each word was analyzed separately. Participants
rated hot and warm significantly closer to the light-color end of the scales (hot: t20 = 5.234, p <
0.001; warm: t20 = 3.765, p = 0.001). In contrast, the extents to which “cold” and “cool” were
rated as closer to the dark-color end of the scale did not reach significance (cold: t20 = −1.342, p
= 0.195; cool: t20 = −0.759, p = 0.457).
14
These results suggest the existence of a crossmodal correspondence between thermal
sensation and color lightness, with warm words more strongly associated with light colors than
cool words. Thus, Study 1 showed that people associate warmth with light color, suggesting
that consumers visually attend to information in a crossmodally matched manner even in the
context of purely linguistic stimuli. Indeed, a previous study found that the correspondence
between pitch and color lightness influenced visual attention (Hagtvedt & Brasel, 2016). To
extend the findings of Study 1, Study 2 used eye tracking to test whether people visually attend
to information that is crossmodally matched.
STUDY 2
Study 2 used eye tracking to investigate whether the crossmodal correspondences between
thermal sensation (warm/cool) and color lightness (light/dark) influence visual attention.
Participants
In total, 39 healthy participants (23 females, Mage = 21.102±2.441 years) were recruited
using a bulletin board and a mailing list of students.
Design
The study used a 2 (thermal sensation: warm/cool) × 2 (color lightness: light/dark color)
mixed-subjects design, with temperature treated as a between-subjects factor and color
lightness treated as a within-subject factor.
Procedure
15
The procedure followed that of a previous study featuring physical temperature
manipulation (Williams and Bargh, 2008). In the cited work, participants briefly held either a
hot or cold therapeutic pad under the guise of product evaluation and, after removing the pad,
the authors investigated how temperature influenced product choice.” Participants were
randomly allocated to the warm (n = 21; female = 12, male = 9) or the cool (n = 18; female = 9,
male = 9) condition and wore a hot (warm condition) or a cold (cool condition) pad (Figure 3)
around their neck. The temperatures of the hot and cold pads were about 50°C and –5°C
respectively. Although it seems that the pad surface temperatures were somewhat extreme, we
used the pad following the manufacturer’s instructions. The participants also did not rate the
temperature as extreme (please see the results of Studies 2 and 3). The warm and cool pads
were of the same design. While wearing the pad, participants rated their reactions to it (“How
much do you like the pad?”), its temperature (“To what extent do you feel warm?” under the
warm condition and “To what extent do you feel cool?” under the cool condition), their
attitudes toward it (“How positive is your attitude toward the pad?”), and its usefulness (“How
useful is the pad?”). Attitude and usefulness were included as filler questions; the answers were
not relevant to the hypothesis under test. This procedure is similar to that of Williams and
Bargh 2008. The rating procedure employed a seven-point Likert scale ranging from 1 (not at
all) to 7 (very much) for each question. Each question was answered within 10 seconds, and
participants wore the pad for about 40 seconds.
Soon after the warm/cool manipulation (i.e., the pad was removed), participants engaged in
the goods-viewing task, which relied on eye-tracking technology. Participants were instructed
to view two consumer products presented on a screen. The two products were the same but
differed in color lightness (one was light and the other was dark). Colors of goods were
manipulated with the aid of gimp2 software. We applied the Colors->Colorify filter, and then
chose default colors: RBG color Red for “red goods”, RBG color Green for “green goods”, and
CMY color Cyan for “cyan goods. Then, the lightness from light (100%) to dark (50%) was
manipulated. Saturation was not manipulated. Thus, goods were colored light/dark red,
light/dark cyan, or light/dark green. The variety of color was the same under warm and cold
16
conditions. In total, eight consumer goods (Figures 4; mug cup, cap, T-shirt, curtain, umbrella,
towel, tote bag, and cushion) were taken from the Internet for use in this study. First, a fixation
cross appeared for 2 s, and consumer goods (the same goods but light or dark in color, each
with a 3.34° × 3.34° visual angle) were presented on the left and right of a white square against
a black background for 8 seconds (Figure 5). Between trials, a fixation cross appeared for 2 s.
The images were presented on the screen using Tobii Studio software (ver. 3.3.2; Tobii
Technology, Stockholm, Sweden). The positioning of light-/dark-colored goods was balanced
across participants. The total fixation times on the light- and dark-colored goods were
measured using an eye-tracking device, and all data from both measures were log-transformed
to modify skewed distributions. Visual attention was defined as the total viewing time for each
item.
Eye tracking was performed using an LCD monitor with the Tobii pro X2–60 (60 Hz) set at
a resolution of 1920 × 1080; the distance between the participant and the display was
approximately 60 cm. Prior to the task, a nine-point manual calibration was performed in the
Tobii studio to ensure that the eye tracker accurately recorded eye position. After the
calibration, the participants were instructed to try their best not to move their heads during the
task.
Statistical analyses
We first checked on whether the manipulation of thermal sensations was successful. A
one-sample t test assessed whether the warm and cool ratings for words differed from the origin
of the scale (0: participants did not feel warm/cool at all). A paired t test was conducted to
compare subjective ratings between warm and cool conditions.
To assess the effects of thermal sensation–color lightness correspondences on consumer
visual attention, we used a generalized linear mixed model (GLMM), which allows the nesting
of a level-1 variable (participant) within a level-2 variable (colored goods) and has the
17
advantage of efficiently utilizing all the available data in a set. The dependent variable in the
GLMM was the difference between the times (ms) spent viewing light- and dark-colored goods,
which was treated as an indicator of the extent to which participants visually attended to
light-colored goods. Sensory congruency was set as a fixed factor (match: warm-light and
cool-dark = 1, mismatch: warm-dark and cool-light = 0), and participant was set as a random
factor.
To assess the effects of thermal sensation–color lightness correspondences on visual
attention in greater detail, another GLMM was also applied to the data. The dependent variable
in this GLMM was the same as that in the previous one (difference between times spent
viewing (ms) light- and dark-colored goods). Temperature (warm =1, cool = 0), color lightness
(light = 1, dark = 0), and their interaction were set as fixed factors, and participant was set as a
random factor. When the interaction term was significant, we conducted a post hoc analysis to
analyze the data from warm and cool conditions separately.
All statistical analyses were conducted using R software. For the GLMMs, we used the
lme4 package in R (Bates, Maechler, Bolder, & Walker, 2015).
Results and discussion
Participants felt warm (Mwarm = 5.381±1.071, t20 = 4.810, p < 0.001) rather than neutral
(warm rating = 0: participants did not feel warm at all) under the warm condition, whereas
participants felt cool (Mcool = 4.631±1.257, t17 = 4.320, p < 0.001) rather than neutral (cool
rating = 0: participants did not feel cold at all) under the cool condition. Neither the extent to
which participants liked the product (Mwarm = 4.900±1.586 vs. Mcool = 4.250±1.390, t34 = 1.289,
p = 0.206) nor the degree to which they reported positive reactions to it (Mwarm = 5.476±1.030
vs. Mcool = 5.316±1.057, t38 = 0.486, p = 0.630) differed under the warm and cool conditions.
Perceived usefulness also did not differ under warm and cool conditions (Mwarm = 5.000±1.183
vs. Mcool = 5.632±1.012, t38 = −1.804, p = 0.079).
18
The first analysis showed that sensory congruency (warm–light and cool–dark) attracted
more visual attention than those that were mismatched (warm–dark and cool–light).
Participants in the warmth (coolness) condition viewed the corresponding goods longer (light
color for warmth and dark color for coolness) (B = 0.238, SE = 0.083, z = 2.857, p = 0.004).
The second analysis revealed significant effects of the interactions between temperature and
color lightness. Next, we conducted post hoc comparisons to test which correspondence
(warm–light color/cool–dark color) had the strongest influence on visual attention (B = 0.404,
SE = 0.169, z = 2.393, p = 0.017) (Figure 6). According to the results, the only sensory
congruency that influenced visual attention was warm–light color. Under the warm condition,
participants spent more time viewing light-colored (vs. dark-colored) goods) (B = 0.574, SE =
0.114, z = 5.042, p < 0.001), whereas participants did not spend more time viewing the
dark-colored (vs. light-colored) goods under the cool condition (B = 0.168, SE = 0.125, z =
1.338, p = 0.182).
These results suggested that the crossmodal correspondence between thermal sensation and
color lightness influenced consumer visual attention. Specifically, warmth was associated with
increased visual attention to light-colored goods, which may trigger preference formation. The
results were consistent with those of Study 1, which showed that warmth was significantly
matched with light colors, but cold was not matched with dark colors. The allocation of visual
attention influences subsequent preference formation (Shimojo, Simion, Shimojo, & Scheier,
2003). Also, vision–audition correspondence increases both visual attention and preference
(Hagtvedt & Brasel, 2016). Study 3 tested the possibility that people might prefer
crossmodally-matched products.
STUDY 3
Study 3 investigated whether the crossmodal correspondences between thermal
sensation (warm/cool) and color lightness (light/dark) influence consumer choice.
19
Participants
In total, 47 healthy participants (21 females, Mage = 21.149±2.469 years) were recruited
using a bulletin board and a mailing list of students.
Design
The study used a 2 (thermal sensation: warm/cool) × 2 (color lightness: light/dark)
mixed-subjects design, with temperature treated as a between-subjects factor and color
lightness treated as a within-subject factor.
Procedure
The warm/cool manipulation was the same as used in Study 2. Participants were randomly
allocated to the warm (n = 24; female = 11, male = 13) or cool (n = 23; female = 12, male = 11)
condition, in which they wore a hot (warm condition) or cool (cool condition) pad of the same
design around their neck. While wearing the pad for about 40 s (the same duration as in Study
2), they rated the pad in terms of liking, warmth, positivity, and usefulness.
Soon after warm/cool manipulation, participants were asked to choose between light- and
dark-colored consumer goods. The goods used in this study were the same as those used in
Study 2. First, a fixation cross appeared for 2 s, and consumer goods (the same goods but light
or dark in color, 3.34° × 3.34° visual angle for each) were presented on the left and right of the
computer screen, and participants were asked to choose the one they wanted. Between trials, a
fixation cross appeared for 2 s. The positioning of light-/dark-colored goods was balanced
across participants. The images were presented on the screen using Tobii Studio software (ver.
3.3.2; Tobii Technology, Stockholm, Sweden).
20
Statistical analyses
To assess the effects of temperature–color lightness correspondences on consumer choice, a
hierarchical logistic regression analysis was performed. The dependent variable in the analysis
was the choice of goods (light color = 1, dark color = 0), temperature was set as a fixed factor
(warmth = 1, coolness = 0), and participant was set as a random factor.
All statistical analyses were conducted using R software.
Results and discussion
Participants felt warmer under the warm condition than under the cool condition (Mwarm =
6.150±0.671 vs. Mcool = 2.261±0.964; t41 = 15.129, p < 0.001). However, the preferences
(Mwarm = 4.900±1.410 vs. Mcool = 4.870±1.290; t41 = 0.074, p = 0.942), positivity (Mwarm =
4.833±1.308 vs. Mcool = 5.043±1.492; t45 = −0.514, p = 0.610), and perceived usefulness (Mwarm
= 5.364±1.049 vs. Mcool = 5.261±1.888; t43 = 0.224, p = 0.824) of goods did not differ under
warm and cool conditions.
Participants did not differ in their choices for light-colored (vs. dark-colored) goods (B =
−0.312, SE = 0.338, z = −0.924, p = 0.356) according to the temperature of their condition
(Figure 7).
These results indicate that sensory congruency did not affect consumer preference, as
participants did not choose more light-colored (vs. dark-colored) goods under the warm than
under the cool condition. As it is possible that a mediator affected the influence of thermal
sensation on color lightness, Study 4 examined whether thermal comfort acted as a mediator on
the effects of thermal sensation on preferences for color lightness.
21
STUDY 4
In Study 4, we investigated the role of sensory attitude (thermal comfort) in preference
formation based on thermal sensation–color lightness correspondences. The hypothesis was
derived through consideration of the affect-as-information framework (Schwarz, 2011).
According to this framework, consumers exhibit positive attitudes to products when they
themselves feel positive. Thermal sensations are unique in that they include affective
information (i.e., thermal comfort; a positive reaction to a thermal sensation). However, any
role for affect in terms of sensory marketing has previously been ignored. Using the
affect-as-information framework, we predicted that when consumers felt positive (i.e.,
thermally comfortable), they were more likely to exhibit a greater preference for
correspondingly colored products.
We also changed the evaluation task and thermal sensation imparted in Study 4. In the
forced-choice task (Study 3), participants were more likely to choose light-colored than
dark-colored products (63, 37%, respectively). Participants compared options in the
forced-choice task. Even if one option (a light-colored product) was greatly preferred to
another (a dark-colored product), the influence of thermal sensation on light-colored product
choice might be negligible. To reduce the intrinsically greater preference for light-colored
products, we changed the index of consumer preference from joint evaluation (the
forced-choice task of Study 3) to separate evaluation (the rating task of Study 4). Given that
stores can control thermal sensations using air-conditioning, we manipulated thermal sensation
using ambient temperature, which is regarded as a more ecologically valid manipulation in
consumer settings.
Participants
In total, 51 healthy participants (25 females, Mage = 21.320±1.823 years) were recruited
using a bulletin board and a mailing list of students.
22
Design
The study used a 2 (ambient temperature: warm/cool) × 2 (color lightness: light/dark color)
mixed-subjects design, with temperature treated as a between-subjects variable and color
lightness treated as a within-subject variable.
Warm/cool manipulation
Participants were randomly allocated to a warm (n = 27; female = 13, male = 14) or cool (n
= 24; female = 13, male = 11) room, with between two and six participants allocated to rooms
accommodating a maximum of about 10 people. The ambient temperature was set at 27–30°C
(MWarm = 28.900±1.393) under the warm condition and at 20–23°C (MCool = 21.812±0.747)
under the cold condition. To establish ecological validity, we set both temperature conditions
to be slightly warm/cool and within the comfortable range (i.e., stores seem to adjust their
temperature only minimally). The temperature disparity (the difference between the cool and
warm conditions) was comparable to that used in previous studies (Bell & Baron, 1977; Huang,
Zhang, Hui, & Wyer, 2014; Ijzerman & Semin, 2009). The baseline-temperature room, where
the ethical guidelines were explained and the basic data for each participant were collected,
was set at 24–27°C.
Buying intentions
The participants rated their buying intentions (“To what extent do you want to buy the
goods?”) for eight light-colored and eight dark-colored goods (Figure 8). The goods were the
same as those used in Studies 2 and 3. The rating procedure employed a seven-point Likert
scale ranging from 1 (not at all) to 7 (very much) for each self-paced question.
23
Statistical analyses
To assess the effects of temperature–color lightness correspondences on buying intentions,
a GLMM was applied to the data. The dependent variable in the GLMM was buying intention,
sensory congruency was set as fixed factors (match: warm–light and cool–dark = 1, mismatch:
warm–dark and cool–light = 0), and participant was set as a random factor.
Results and discussion
We used paired t tests to examine differences in the actual room temperature (°C) under warm
and cool room conditions. The temperature under the warm condition was significantly higher
than that under the cool condition (MWarm = 28.900°C vs. MCool = 21.812°C, t50 = 22.603, p <
0.001). Next, we investigated differences in the perceived warmth under the warm and cool
conditions using a paired t test. The perceived warmth (“To what extent do you feel warm?”
rated from 1 = cool to 7 = warm) was greater under the warm than under the cool condition
(MWarm = 5.222±1.121 vs. MCool = 2.250±0.676, t49 = 11.288, p < 0.001). Although
respondents’ thermal comfort (“How do you feel about the current temperature?” rated from 1
= negative to 7 = positive) differed between conditions (MWarm = 4.259±1.403 vs. MCool =
3.250±1.260, t49 = 2.689, p = 0.001), their general emotional states (“How would you rate your
emotional state?” rated from 1 = negative to 7 = positive) did not (MWarm = 4.074±1.141 vs.
MCool = 3.792±1.103, t49 = 0.896, p = 0.375). These data suggest that the temperature
manipulation was successful.
Sensory congruency did not lead to greater buying intentions (B = −0.052, SE = 0.103, z =
−0.500, p = 0.618). When the interaction term for sensory congruency and thermal comfort was
entered into the model, it significantly increased buying intentions (B = 0.188, SE = 0.048, z =
3.913, p < 0.001) (Figure 9).
Next, we performed post hoc comparisons to identify which sensory congruency
24
(warm–light color/cool–dark color) influenced buying intentions via interaction with thermal
comfort. Consumers who felt comfortable under the warm condition expressed greater buying
intentions for light-colored (versus dark-colored) consumer goods (B = 0.167, SE = 0.066, z =
2.513, p = 0.012). However, consumers who reported feeling comfortable under the cool
condition did not express greater buying intentions for dark-colored (versus light-colored)
consumer goods (B = 0.017, SE = 0.079, z = 0.216, p = 0.829).
These results suggest that crossmodal correspondences did not directly influence consumer
choices but that they did influence buying intentions via interaction with thermal comfort.
When positive attitude toward a thermal sensation (i.e., thermal comfort) was held constant,
those in a warm (versus a cool) room were more likely to express greater buying intentions for
light-colored (versus dark-colored) consumer goods.
GENERAL DISCUSSION
The thermal environment and color of products are commonly manipulated design variables
in many business settings, and consumers react to the interactions between their sensory
experiences of these variables. However, the ways in which the senses are interactively
matched and the consumer-relevant consequences (visual attention, preference formation) of
such interactions remain unknown. Based on a crossmodal correspondence framework, the
present research investigated how thermal sensation and color lightness are matched, how such
sensory matching (correspondences) has consumer-relevant consequences in terms of lower
(visual attention) and higher (preference formation) cognitive functions, and how attitudes
toward sensory experiences affect preference formation. A warm sensation was reliably
matched with a light color even in the context of purely linguistic stimuli (Study 1). Following
this experience of a physical warm sensation delivered by a neck pad, consumers devoted more
visual attention to light-colored (versus dark-colored) goods (Study 2). Although the
25
correspondence between physical temperature imparted by a neck pad and light-colored goods
did not directly influence consumer choices (Study 3), it did influence buying intentions via its
interaction with thermal comfort (Study 4). Specifically, when a positive attitude toward
thermal comfort was held constant, those in a warm (versus a cool) ambient temperature were
more likely to express greater buying intentions for light-colored (versus dark-colored)
consumer goods. These results reveal a novel crossmodal correspondence between thermal
sensation and color lightness and demonstrate how such correspondences affect consumers’
visual attention and preferences.
Theoretical contributions
Our findings contribute to the literature on crossmodal correspondence and its influence on
consumer behaviors. Although it has been recently shown that different senses are matched
(Deroy et al., 2013; Spence, 2011; Velasco et al., 2016), crossmodal correspondences between
thermal sensations and levels of color lightness remain largely uninvestigated. Additionally,
how such correspondences operate on lower- (visual attention) and higher-level (preference
formation) consumer-relevant behaviors remains unknown. The present research showed that
warmth and light color were crossmodally matched and that this association was transferrable
to consumers’ visual attention and preference formation. Importantly, the correspondence was
transferable to preference formation when the sensory attitude was positive (i.e., feeling
comfortable under the warm condition). These results confirm that thermal–lightness
correspondences occur and demonstrate that such correspondences are transferrable to
consumer attention and preferences.
The effect of thermal-color lightness correspondences on attention may be derived from
learned predictiveness. Attention can, in fact, be so derived (how much information do stimuli
provide about other outcomes?), and/or via learned values (how rewarding are stimuli?).
Although people who felt warmer were more likely to view light-colored products, this did not
extend to preferential choice. People may learn from experience that warm cues are reliably
associated with light colors, and such learned predictiveness triggers attention. Thus, the
26
effects of thermal-color lightness correspondences on attention may be derived from learned
predictiveness, not learned values.
The affect-as-information framework may explain the influence of the warmth–color
lightness correspondence on buying intentions via interaction with thermal comfort. According
to this framework, affect provides information, and consumers attend to their feelings as
sources of information while making judgments (Schwarz, 2011). Incidental affects have
valence-congruent influences on evaluations of products when consumers use their emotions
as sources of information for judgments. Positive emotional states are generally interpreted as
indicative of liking, whereas negative emotional states are generally interpreted as indicative of
disliking (Schwarz, 2011). People who feel positive emotions increase their valuations of
goods (Sherman, Mathur, & Smith, 1997), whereas people who feel negative ones decrease
them (Axelrod, 1963). Thus, people who have positive attitudes toward warmth might increase
their buying intentions for light-colored products, whereas those with negative attitudes toward
warmth might not. By integrating the affect-as-information framework with data on
crossmodal influences on preferences, the present findings advance our understanding of the
role of such correspondences in preference formation.
We observed only warm-light color correspondences and, thus, not cool-dark color
correspondences. Usually, correspondences are found between polar dimensions (Schietecat,
Lakens, IJsselsteijn, & de Kort, 2018). Given the theoretical foundations of the work, coolness
and dark colors should correspond. We assume that the intensities of coolness/darkness affect
the results. Coolness was set about 3°C lower than the indoor temperature, and darkness at
half-maximum lightness. If the temperature had been set to colder and the color darker,
cool-dark color correspondences might emerge. In addition, the indices of consumer
preference differed between Studies 3 and 4. Study 3 featured a forced-choice task and Study 4
a rating task. Differences in the evaluation mode (joint evaluation in Study 3; separate
evaluation in Study 4) may have influenced the results. A future study should explore how
sensory intensity influences cool-dark color correspondences.
27
The results imply that the contradictory findings of warmth and prosocial behaviors may be
resolved by considering thermal comfort. Although Williams and Bargh (2008) found that
warm temperatures led to more prosocial behaviors, the independent replications did not
support the findings (Lynott, Corker, Wortman, Connell, Donnellan, Lucas, & O’Brien, 2014).
Based on the feelings-as-information theory (Schwarz, 2011), people may rely on their
comfortability of warmth to provide information about subsequent prosocial behaviors. One of
motives for prosocial behaviors is to receive pleasant feelings from the prosocial act itself (i.e.,
warm-glow motive) (Andreoni, 1990). Only feeling comfortable warmth may trigger a
warm-glow motive, while only feeling an uncomfortable warmth may not. This is because
experiencing (un)comfortable warmth may offer information that subsequent prosocial
behaviors will be (un)pleasant. Similar to the relationship between warmth and preference for
light-colored products, the results suggest that thermal comfort may be a key moderator of the
influence of warmth on prosocial behavior.
Limitations
The study has several limitations that should be noted. First, based on the affect-as appraisal
framework, we used a mediator as a proxy for thermal comfort. However, there are other
potential mediators, such as fluency. Although the mechanism underlying warmth–color
lightness correspondences are unknown, the link may derive from “statistical learning”
whereby consumers may learn the association between warmth and color lightness through
repeated pairings of these two senses. This may lead to perceptual fluency (Michael, Galich,
Relland, & Prud’hon, 2010) and manifest in a predictable correspondence. Indeed, a previous
study suggested that fluency may constitute a sensory influence on consumer preferences
(Sunaga, 2018). Further research should investigate whether fluency mediates the influence of
the warmth–color lightness correspondence on consumer preferences.
Although we suggest that transfer of warm-light color correspondences from attention to
preference constitutes a positive response to warmth (i.e., thermal comfort), the
methodological differences between physical warming (Studies 2–3) and ambient temperature
28
(Study 4) may have influenced the results. For example, the extent of physical touching and the
temporal sequence of events differed between the studies. Warming physical manipulation
(Study 2) was followed by attention to light-colored goods in the absence of the neck pad. In
contrast, in Study 4, participants rated colored goods in a warm room. Further research should
explore how methodological differences contribute to the effects of warm-light color
correspondences on attention and preference.
The statistical power of our present findings requires attention. We performed post-hoc
power analysis using G*Power (Faul, Erdfelder, Lang, & Buchner, 2007) on Study 1 data and
similar analyses of Study 2–4 data using SIMR (Green and MacLeod, 2016). Although the
sample size of Study 1 was relatively small (n = 21), the statistical power (1 – β) was over 0.80,
suggesting that a repeat study with the same number of participants would be associated with
an 80% probability of a significant result in terms of thermal-color lightness correspondence.
In Studies 2–4, we used GLMMs that nested a level-1 variable (the participants) within a
level-2 variable (colored goods). Thus, the sample size was relatively large (the number of
participants multiplied by the number of stimuli), although the number of participants was
limited. As G*Power cannot subject GLMM data to power analysis, we applied the post-hoc
power analysis method of the R package SIMR. We focused on the principal choice in each
study: participants viewed more light- than dark-colored goods when warm (Study 2);
participants chose more light- than dark-colored goods when warm (Study 3); participants
exhibited a higher-level buying intention for light-colored than dark-colored goods when warm
(with consideration of individual differences in thermal comfort) (Study 4). Thus, Study 2
afforded a power of over 90%, Study 3 a power of about 20%, and Study 4 a power of about
70%. The data suggest that the sample size was appropriate to detect significant results in
Studies 1–2 (the existence of warmth-color lightness correspondence and the influence thereof
on visual attention) and Study 4 (correspondences increase preferences when participants feel
comfortable). However, the findings require replication using a larger sample.
Practical contributions
29
This work has implications for how thermal environments and colored products should be
designed. The current findings indicate that the creation of a sensory association between light
colors and warmth should contribute to effective store design. For example, if a marketing
manager wants customers to attend to products, light-colored goods should be displayed in a
warm environment. Additionally, managers who want to increase the sales of light-colored
products should also ensure that the ambient temperature is in a comfortably warm range, as it
is only when the warmth is experienced as pleasant that buying intentions for light-colored
goods increase.
Conclusions
The present study showed that warmth and color lightness interactively influence consumer
behaviors in a hierarchical manner. Based on this warmth–light color correspondence, this
research also found that this matching had consumer-relevant consequences in terms of both
lower (visual attention) and higher (preference formation) cognitive functions. Consistent with
the affect-as-information framework, positive attitudes toward warmth play a critical role in
preference formation. These findings elucidate consumers’ multisensory experiences of
warmth and color lightness and provide theoretical and practical insights into the effects of
crossmodal correspondence and the importance of effective store design.
References
Andreoni, J. (1990). Impure altruism and donations to public goods: A theory of warm-glow
giving. The Economic Journal, 100(401), 464–477.
Agrawal, N. & Duhachek, A. (2010). Emotional compatibility and the effectiveness of
antidrinking messages: A defensive processing perspective on shame and guilt. Journal
of Marketing Research, 47(2), 263–273.
Arens, E., Zhang, H., & Huizenga, C. (2006). Partial-and whole-body thermal sensation and
comfort—Part I: Uniform environmental conditions. Journal of Thermal Biology,
31(1), 53–59.
Armel, K. C., Beaumel, A., & Rangel, A. (2008). Biasing simple choices by manipulating
relative visual attention. Judgment and Decision making, 3(5), 396-403.
Axelrod, J. N. (1963). Induced Moods and Attitudes toward Products. Journal of Advertising
Research, 3(2), 19–24.
30
Bell, P. A. & Baron, R. A. (1977). Aggression and ambient temperature: The facilitating
and inhibiting effects of hot and cold environments. Bulletin of the Psychonomic
Society, 9(6), 443–445.
Deroy, O., Crisinel, A.-S., & Spence, C. (2013). Crossmodal correspondences between odors
and contingent features: Odors, musical notes, and geometrical shapes. Psychonomic
Bulletin & Review, 20(5), 878–896.
Ernst, M. O. (2007). Learning to integrate arbitrary signals from vision and touch. Journal of
Vision, 7(5), 7:1–14.
Faul, F., Erdfelder, E., Lang, A. G., & Buchner, A. (2007). G* Power 3: A flexible statistical
power analysis program for the social, behavioral, and biomedical sciences. Behavior
Research Methods, 39(2), 175–191.
Fenko, A., Schifferstein, H. N., & Hekkert, P. (2010). Looking hot or feeling hot: What
determines the product experience of warmth? Materials & Design, 31(3), 1325–1331.
Gagge, A. P., Stolwijk, J., & Hardy, J. (1967). Comfort and thermal sensations and associated
physiological responses at various ambient temperatures. Environmental Research,
1(1), 1–20.
Gorn, G. J., Chattopadhyay, A., Yi, T., & Dahl, D. W. (1997). Effects of color as an
executional cue in advertising: They’re in the shade. Management Science, 43(10),
1387–1400.
Green, P., & MacLeod, C. J. (2016). SIMR: An R package for power analysis of generalized
linear mixed models by simulation. Methods in Ecology and Evolution, 7(4), 493–498.
Greenwald, A. G. (1976). Within-subjects designs: To use or not to use? Psychological
Bulletin. 83, 314–320.
Hagtvedt, H., & Brasel, S. A. (2016). Cross-Modal Communication: Sound Frequency
Influences Consumer Responses to Color Lightness. Journal of Marketing Research,
53(4), 551–562.
Hensen, J. (1990). Literature review on thermal comfort in transient conditions. Building and
Environment, 25(4), 309–316.
Higgins, E. T., Idson, L. C., Freitas, A. L., Spiegel, S., & Molden, D. C. (2003). Transfer of
value from fit. Journal of Personality and Social Psychology, 84(6), 1140.
Ho, H. N., Iwai, D., Yoshikawa, Y., Watanabe, J., & Nishida, S. Y. (2014). Combining colour
and temperature: A blue object is more likely to be judged as warm than a red
object. Scientific Reports, 4, 5527.
Ho, H.-N., Van Doorn, G. H., Kawabe, T., Watanabe, J., & Spence, C. (2014).
Colour-temperature correspondences: When reactions to thermal stimuli are influenced
by colour. PloS One, 9(3), e91854.
Huang, X. I., Zhang, M., Hui, M. K., & Wyer, R. S. (2014). Warmth and conformity: The
effects of ambient temperature on product preferences and financial decisions.
Journal of Consumer Psychology, 24(2), 241–250.
IJzerman, H. & Semin, G. R. (2009). The thermometer of social relations: Mapping social
proximity on temperature. Psychological Science, 20(10), 1214–1220.
31
Kim, H., Park, K., & Schwarz, N. (2009). Will this trip really be exciting? The role of
incidental emotions in product evaluation. Journal of Consumer Research, 36(6),
983–991.
King, R. B., Hauff, S., & Phillips, J. B. (1994). Physiological color change in the green
treefrog: Responses to background brightness and temperature. Copeia, 422–432.
Krajbich, I., Armel, C., & Rangel, A. (2010). Visual fixations and the computation and
comparison of value in simple choice. Nature Neuroscience, 13(10), 1292.
Krishna, A. (2012). An integrative review of sensory marketing: Engaging the senses to affect
perception, judgment and behavior. Journal of Consumer Psychology, 22(3), 332–351.
Krishna, A., Cian, L., & Sokolova, T. (2016). The power of sensory marketing in advertising.
Current Opinion in Psychology, 10, 142–147.
Krishna, A., & Schwarz, N. (2014). Sensory marketing, embodiment, and grounded cognition:
A review and introduction. Journal of Consumer Psychology, 24(2), 159–168.
Labrecque, L. I., Patrick, V. M., & Milne, G. R. (2013). The marketers’ prismatic palette: A
review of color research and future directions. Psychology & Marketing, 30(2),
187–202.
Lowe, M. L., & Haws, K. L. (2017). Sounds Big: The Effects of Acoustic Pitch on Product
Perceptions. Journal of Marketing Research, 54(2), 331–346.
Lynott, D., Corker, K. S., Wortman, J., Connell, L., Donnellan, M. B., Lucas, R. E., & O’Brien,
K. (2014). Replication of “Experiencing physical warmth promotes interpersonal
warmth” by Williams and Bargh (2008). Social Psychology, 45, 216–222.
MacIsaac, D., Kanner, G., & Anderson, G. (1999). Basic physics of the incandescent lamp
(lightbulb). The Physics Teacher, 37(9), 520–525.
Michael, G. A., Galich, H., Relland, S., & Prud’hon, S. (2010). Hot colors: The nature and
specificity of color-induced nasal thermal sensations. Behavioural Brain Research,
207(2), 418–428.
Motoki, K., Saito, T., Nouchi, R., Kawashima, R., & Sugiura, M. (2018). The paradox of
warmth: Ambient warm temperature decreases preference for savory foods. Food
Quality and Preference. 22(3), 1–9
Nikolopoulou, M., & Lykoudis, S. (2006). Thermal comfort in outdoor urban spaces: Analysis
across different European countries. Building and Environment, 41(11), 1455–1470.
Parise, C. V. (2016). Crossmodal correspondences: Standing issues and experimental
guidelines. Multisensory Research, 29(1–3), 7–28.
Peirce, J. W. (2007). PsychoPy—psychophysics software in Python. Journal of Neuroscience
Methods, 162(1–2), 8–13.
Piqueras-Fiszman, B., & Spence, C. (2015). Sensory expectations based on product-extrinsic
food cues: An interdisciplinary review of the empirical evidence and theoretical
accounts. Food Quality and Preference, 40, 165–179.
Schwarz, N. (2011). Feelings-as-information theory. Handbook of Theories of Social
Psychology, 1, 289–308.
Schwarz, N. (2013). Feelings as information: Implications for affective influences on
information processing. Theories of mood and cognition: A user’s guidebook, 159.
32
Sherman, E., Mathur, A., & Smith, R. B. (1997). Store environment and consumer purchase
behavior: Mediating role of consumer emotions. Psychology and Marketing, 14(4),
361–378.
Schietecat, A. C., Lakens, D., IJsselsteijn, W. A., & de Kort, Y. A. (2018). Predicting
context-dependent cross-modal associations with dimension-specific polarity
attributions Part 1–brightness and aggression. Collabra: Psychology, 4(1), 14.
Shimojo, S., Simion, C., Shimojo, E., & Scheier, C. (2003). Gaze bias both reflects and
influences preference. Nature Neuroscience, 6(12), 1317.
Simmons, J. P., Nelson, L. D., and Simonsohn, U. (2011). False-positive psychology:
Undisclosed flexibility in data collection and analysis allows presenting anything as
significant. Psychological. Science. 22, 1359–1366.
Singh, S. (2006). Impact of color on marketing. Management Decision, 44(6), 783–789.
Spence, C. (2011). Crossmodal correspondences: A tutorial review. Attention, Perception, &
Psychophysics, 73(4), 971–995.
Spence, C. (2012). Managing sensory expectations concerning products and brands:
Capitalizing on the potential of sound and shape symbolism. Journal of Consumer
Psychology, 22(1), 37–54.
Spence, C., & Deroy, O. (2013). How automatic are crossmodal
correspondences? Consciousness and Cognition, 22(1), 245–260.
Sunaga, T. (2018). How the sound frequency of background music influences consumers’
perceptions and decision making. Psychology & Marketing, 35(4), 253–267.
Sunaga, T., Park, J., & Spence, C. (2016). Effects of Lightness-Location Congruency on
Consumers’ Purchase Decision-Making. Psychology & Marketing, 33(11), 934–950.
Velasco, C., Woods, A. T., Petit, O., Cheok, A. D., & Spence, C. (2016). Crossmodal
correspondences between taste and shape, and their implications for product
packaging: A review. Food Quality and Preference, 52, 17–26.
Wedel, M., & Pieters, R. (2008). A review of eye-tracking research in marketing Review of
marketing research (pp. 123–147): Emerald Group Publishing Limited.
Williams, L. E., & Bargh, J. A. (2008). Experiencing physical warmth promotes interpersonal
warmth. Science, 322(5901), 606–607.
Zwebner, Y., Lee, L., & Goldenberg, J. (2013). The temperature premium: Warm temperatures
increase product valuation. Journal of Consumer Psychology, 24(2), 251–259.
Table 1. Overview of studies
Main outcome measure
Hypothesis
Study
1
Existence of crossmodal
matching
Match between warmth (coolness) and light (dark)
colors
Study
Crossmodal visual
Visual attention directed toward light (dark) colors
33
2
attention to consumer
goods
following physical warmth (coolness)
Study
3
Crossmodal preferences
Preference for light-colored (dark-colored) goods
following physical warmth (coolness)
Study
4
Thermal comfort
mediates crossmodal
preferences
Preference for light-colored (dark-colored) goods
in warm (cool) ambient temperature when
comfortable
Figure 1. Task 1
Figure 2. The correspondence between thermal sensation and color lightness.
The error bars represent standard deviations.
34
Figure 3. Hot/cool pad.
Figure 4. List of colored goods.
35
Figure 5. Sample goods-viewing/choice task.
Figure 6. The influence of sensory congruence on visual attention.
36
Figure 7. The influence of warmth (vs. coolness) on a preference for light-colored (vs.
dark-colored) goods.
Figure 8. Sample buying-intention task.
37
Figure 9. The influence of thermal-color lightness congruency on buying intention modulated
by thermal comfort.
The English in this document has been checked by at least two professional editors, both native
speakers of English. For a certificate, please see:
http://www.textcheck.com/certificate/g2S7u3
