Professional Documents
Culture Documents
Package Color Saturation and Food Healthfulness Perceptions
Package Color Saturation and Food Healthfulness Perceptions
A R T I C L E I N F O A B S T R A C T
Keywords: Vivid, highly saturated colors are often perceived as exciting and arousing, making them popular in branding
Food and package design. However, are foods packaged in vivid colors also perceived as unhealthful? Across four
Color saturation experiments, we demonstrate that consumers appear to perceive foods in vivid, highly color-saturated food
Conceptual fluency packaging as less healthful than foods in muted, less color-saturated packaging. Further, we demonstrate that
Healthfulness
conceptual fluency mediates the effect, subjective nutrition knowledge weakens the effect, and restrained eating
Restrained eaters
behavior strengthens the effect. We contribute to the color literature that explores the distinct effects of different
Subjective knowledge
color elements on consumer perceptions. We also advance the food well-being literature by identifying a new
heuristic that affects food well-being, and in doing so, join other researchers who have connected learned color
associations to substantive consumer outcomes. Finally, we offer food marketers new insights into consumers'
evaluations of their products.
1. Introduction understand, any factor that may undermine the nutritional assessment
of food, we attempt to answer these questions.
Vivid, highly saturated colors are often perceived as exciting (Aaker, In this research, we investigate and identify a previously unknown
1997; Labrecque & Milne, 2012) and arousing (Gorn, Chattopadhyay, bias that affects consumers' perceptions of food healthfulness, the vivid
Yi, & Dahl, 1997), making them popular in branding and package de- packaging equals unhealthful heuristic. While previous research has
sign. However, in the context of packaged food, could package color identified how color can bias consumers' perceptions of food attributes
saturation also serve as an incidental cue that also affects consumers' (Huang & Lu, 2016; Levy, Riis, Sonnenberg, Barraclough, & Thorndike,
evaluation of the food item's healthfulness? For example, when grocery 2012; Mai, Symmank, & Seeberg-Elverfeldt, 2016; Temple et al., 2011),
shopping, consumers are repeatedly exposed to food packages with this research investigates how one important parameter1 of color, sa-
varying degrees of color saturation, from vivid and vibrant (high sa- turation, or the vividness or purity of a color hue, biases food health-
turation) to muted and washed-out (low saturation). However, vivid or fulness perceptions. Across four experiments, we demonstrate that
muted colors do not, at least anecdotally, appear to be evenly dis- package color saturation can influence consumers' perceptions of food
tributed among all types of packaged food products. Rather, un- healthfulness across a variety of color hues. Further, we reveal that the
healthful foods are often displayed in vivid, highly saturated packages, heuristic affects healthfulness perceptions of both healthful and un-
while healthful foods are often showcased in muted, less color-saturated healthful products, and we provide evidence that conceptual fluency
packages. Could repeated exposures to chips, candies, and sodas dis- accounts for the relationship between package color saturation and
played in vivid, highly saturated packages cause consumers to con- healthfulness perceptions. We also examine how the concepts of
ceptually link vivid, highly saturated colors with less healthful foods? package color vividness and unhealthfulness may have become asso-
Could such a conceptual link unknowingly bias consumers' evaluation ciated, and we identify subjective nutrition knowledge and restrained
of packaged food items? Would all consumers be equally affected by eating behavior as moderating factors that respectively weaken or en-
such a bias? Following the perspective of Bublitz et al. (2013) that it is hance the effects of the vivid packaging equals unhealthful heuristic on
critical for researchers to identify, and for consumers and marketers to consumers' food perceptions.
☆
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
⁎
Corresponding author.
E-mail addresses: jmead@uwf.edu (J.A. Mead), wrricherson@salisbury.edu (R. Richerson).
1
As noted in Labrecque and Milne (2012) and Labrecque, Patrick, and Milne (2013), saturation (vividness, sometimes also referred to as intensity or purity) is one of several color
appearance parameters. A color's saturation refers to the amount of neutral gray “mixed” into a hue. A color's hue refers to its tone, or place on the color wheel. A color's lightness
measures the relative amount of white or black that has been “mixed” into a given hue. While at first glance these terms may appear somewhat interchangeable, they refer to specific
elements of color appearance.
http://dx.doi.org/10.1016/j.jbusres.2017.08.015
Received 31 January 2017; Received in revised form 16 August 2017; Accepted 17 August 2017
Available online 01 September 2017
0148-2963/ Published by Elsevier Inc.
J.A. Mead, R. Richerson Journal of Business Research 82 (2018) 10–18
Our work contributes to the literature in several ways. First, we have repeatedly been exposed to unhealthful foods presented in vivid
identify the unknown and distinct effect of one specific element of packages, and those exposures may have created a previously uni-
color, color saturation, on consumer outcomes, broadening our under- dentified association between the two concepts (Elliot et al., 2007).
standing of how specific color elements affect consumers' perceptions. Further, perhaps to evoke a contrast with unhealthful foods packaged in
Second, we theoretically investigate and empirically demonstrate that vivid colors (Babin & Babin, 2001), many food marketers appear to
conceptual fluency is the mechanism underlying the effect. In doing so, favor muted, less color-saturated packages for more healthful packaged
we advance the color literature by demonstrating how a color asso- foods. Given that food item assessment, choice, and consumption are
ciation (vivid equals unhealthful) can feel accurate and expected within tied to consumers' well-being and health (Bublitz et al., 2013;
a context (indulgent food), thereby influencing consumers' perceptions. Koenigstorfer, Groeppel-Klein, & Kamm, 2014), and that well-being and
Third, we support this theorizing by demonstrating two boundary health are areas of significant theoretical and substantive focus, we
conditions of the vividness equals unhealthful heuristic. Fourth, we attempt to address how package color saturation can bias consumer
advance the food well-being literature by identifying an unknown perceptions and potentially undermine the pursuit of health goals.
heuristic that affects food well-being. In doing so, we contribute to the Specifically, we propose that vivid, highly color-saturated packaging
work of other researchers who have connected learned color associa- may prime the concept of unhealthful foods in consumers' memory,
tions to substantively important outcomes (e.g., Elliot, Maier, Moller, which may bias their perceptions of a food item.
Friedman, & Meinhardt, 2007). It is our hope that with an increased
awareness of this bias, consumers may be better positioned to correct 2.2. Memory and consumer perceptions
for its effect. Finally, for food marketers, we provide new insights into
factors that influence consumer evaluations of their products, allowing Marketers have long recognized that memory plays a key role in
these marketers to make more informed use of color saturation in consumer choice and decision-making (Bettman, 1979). For example,
package designs (Bublitz & Peracchio, 2015). concepts stored in memory can be activated or primed to affect product
The remainder of this manuscript proceeds as follows. Section 2 choice and other consumer responses (Sela & Shiv, 2009). Further, re-
provides a brief conceptual review in the areas of package design and search has shown that memory activations spread through networks of
color as they relate to food item perceptions. Further, we draw from associations, linking seemingly unrelated concepts, like vivid colors and
research in memory, conceptual fluency, restrained eating, and sub- unhealthfulness, together in chains of connectedness (Balota & Lorch,
jective knowledge to develop our hypotheses. Section 3 details our 1986). Even seemingly unrelated concepts can become connected
experimental work and findings. Section 4 offers conclusions, limita- through repeated and frequent co-occurrences (Berger & Fitzsimons,
tions, and potential areas of further research. 2008; Elliot et al., 2007). Indeed, researchers have uncovered memory
connections among color hues and concepts. For example, consumers
2. Conceptual background often perceive the color green as positive and the color red as negative
(De Bock et al., 2013). Consumers' association between green labels and
2.1. The influence of package cues and color in biased food judgments healthful foods (Temple et al., 2011) and red labels with unhealthful
foods (Levy et al., 2012) may be driven in part by consumers' associa-
Many consumers lack the knowledge, time, or motivation to process tion of green with “safety” or “go” and red with “risk” and “stop”
nutritional information deliberately and accurately (Cornish & Moraes, (Koenigstorfer et al., 2014; Schuldt, 2013). Similarly, we propose that
2015; Raghunathan, Naylor, & Hoyer, 2006). As such, consumers often consumers have encoded a link between package color saturation and
employ evaluative shortcuts or heuristics, triggered by product or unhealthfulness in memory. Thus, when consumers see foods packaged
packaging cues, to inform food purchase decisions (Bublitz et al., 2013). in vivid colors, the concept of indulgent, unhealthful foods may also be
However, such heuristics can systematically bias consumers' judgments, retrieved and passed into working memory, which may influence con-
often with negative outcomes for consumer well-being. For example, sumers' perceptions of the food. As such, we contend that:
heuristics can lead consumers to infer that organic foods are healthful
H1. Consumers will perceive food items as less (more) healthful if they
(Sundar & Kardes, 2015), assume unhealthful foods are filling (Suher,
are displayed in packaging with higher (lower) color saturation.
Raghunathan, & Hoyer, 2016), equate nutrition information displayed
in green with healthfulness (De Bock, Pandelaere, & Van Kenhove,
2013), or misjudge the healthfulness of products in portion control 2.3. Conceptual fluency
packages (i.e., small 100 calorie packs; Coelho do Vale,
Pieters, & Zeelenberg, 2008; Scott, Nowlis, Mandel, & Morales, 2008). Further, when memory associations are frequently encountered and
Prior research has demonstrated that certain package colors or color reinforced, the association between two stimuli may begin to evoke a
properties of food packaging can lead to biased judgments about a sense of conceptual fluency (Berger & Fitzsimons, 2008; Lee & Labroo,
food's attributes and can affect related behaviors. For example, con- 2004). When associations are conceptually fluent to consumers, the
sumers perceive food in red-colored packaging as sweeter than food in associations fit with what consumers already know about the world
blue or green packaging (Huang & Lu, 2015) and as less healthful than around them (Kidwell, Farmer, & Hardesty, 2013). Fluent concepts and
food presented in blue-colored packaging (Huang & Lu, 2016). Simi- associations feel familiar and accurate (Whittlesea, 1993), and fluent
larly, consumers have been found to associate green labels with associations are easily processed, categorized, and integrated into
healthful foods (Temple et al., 2011) and red labels with unhealthful consumers' established networks of associations (Winkielman, Schwarz,
foods (Levy et al., 2012). Research has also demonstrated that con- Fazendeiro, & Reber, 2003).
sumers perceive darker colors as heavy, which can lead consumers to Previous research has demonstrated that a sense of conceptual flu-
assess food items as calorically heavy, affecting perceptions of food ency can result from recent or frequent exposures to stimuli
healthfulness and taste (Mai et al., 2016). In total, it seems clear that (Berger & Fitzsimons, 2008; Lee & Labroo, 2004). However, recent and
consumers often infer certain food item attributes based on package frequent exposure can also result in a priming effect, when two stimuli
color cues, yet, the specific influence of package color saturation re- are presented together, which can indicate the presence of such an
mains unknown. association. Yet, the influence of conceptual fluency goes beyond that of
Vivid, highly saturated colors can generate arousal (Gorn et al., a priming effect alone (Alter & Oppenheimer, 2009; Schwarz et al.,
1997) and perceptions of excitement (Aaker, 1997; Labrecque & Milne, 1991). Primed associations that are also conceptually fluent are more
2012). Therefore, it is not surprising that marketers frequently use such likely to be perceived as true and accurate (Alter & Oppenheimer,
colors on the packages of unhealthful foods. As a result, consumers may 2009). As such, the sense of truth and accuracy that accompanies
11
J.A. Mead, R. Richerson Journal of Business Research 82 (2018) 10–18
conceptual fluency may dissuade consumers from further elaborating 2.5. The moderating effect of restrained eating
on the validity of a primed association. Therefore, a sense of conceptual
fluency is not only indicative of the retrieval of information from Conversely, we expect that consumers who exhibit restrained eating
memory (likely due to a priming effect), but also a conceptual fit be- behavior are more susceptible to the vivid packaging equals unhealthful
tween the retrieved information and the particular context at hand. food heuristic. Restrained eaters represent a group of consumers who
Taken together, when consumers encounter foods in vivid, highly make deliberate efforts to combat their internal feelings of hunger to
color-saturated packaging, we expect two responses occur. First, the control their intake of calories, often in pursuit of weight loss goals
associated concept of unhealthfulness is activated from memory. (Federoff, Polivy, & Herman, 1997). To that end, restrained eaters at-
Second, feelings of conceptual fluency and accuracy are experienced. tempt to disregard their desires for specific, often indulgent, foods. To
That is, when seemingly unhealthful foods are presented in vivid control their caloric intake, restrained eaters tend to devote consider-
packaging, the pairing of an unhealthful food with a vivid package feels able attention to external food cues, like packaging (Schachter,
right to consumers as the pairing aligns with previous experiences and Goldman, & Gordon, 1968; Tom & Rucker, 1975). Therefore, restrained
expectations (Whittlesea, Jacoby, & Girard, 1990). The two responses eaters should be more likely to attend to and evaluate package elements
work together to affect perceptions, as the feeling of accuracy dis- when assessing the healthfulness of foods. Thus, they may have stronger
courages consumers from more carefully processing package informa- associations between packaging color saturation and food healthful-
tion to determine if the unhealthfulness assumption is correct. There- ness, which should strengthen the vivid packaging equals unhealthful
fore, we predict: heuristic.
Further, restrained eaters' constant attention to food choices may
H2. Conceptual fluency mediates the relationship between consumer's
leave them in a state of general cognitive depletion (Scott et al., 2008).
perception of food healthfulness and package color saturation.
Depletion is associated with an increased use of heuristics or judgment
Effects like the vivid equals unhealthful heuristic can undermine shortcuts in evaluation (Shah & Oppenheimer, 2008), and due to cog-
consumer intentions to eat healthful foods. As such, to advance con- nitive depletion, restrained eaters may lack the cognitive resources to
sumer food well-being, it is important to understand factors that may deliberate about the accuracy of the vivid packaging equals unhealthful
enhance or weaken the vivid equals unhealthful heuristic in addition to food association at times of evaluation and food choice. In total, re-
the underlying process by which it operates. In the sections that follow, strained eaters are likely more aware of subtle food package cues which
we explore two such moderating factors, subjective nutrition knowl- may indicate product healthfulness, such as color saturation, while also
edge and restrained eating behavior. being more likely to rely on heuristic judgments when evaluating
packaged food, due to a sense of conceptual fluency
(Whittlesea & Leboe, 2000). As such, we predict that:
2.4. The moderating effect of subjective knowledge
H4. The main effect of package color saturation on food healthfulness
perceptions will be increased among restrained eaters.
Subjective knowledge broadly refers to one's self-assessed knowl-
edge and evaluative skills in a domain (Alba & Hutchinson, 2000;
Brucks, 1985; Park, Mothersbaugh, & Feick, 1994). We contend that a 3. Experimental materials and methods.
high level of subjective nutrition knowledge weakens the vivid equals
unhealthful heuristic for two reasons. First, consumers with higher le- We provide support for these hypotheses using online and con-
vels of subjective knowledge rely less on marketer-supplied evaluations trolled laboratory experiments with both student and Mturk samples. In
of products than do other consumers (Brucks, 1985). Instead, these experiment 1, results indicate that consumers who viewed a food item
consumers employ their own judgment criteria when assessing products displayed in a vivid, highly color-saturated package perceived the food
(Brucks, 1985; Moorman, Diehl, Brinberg, & Kidwell, 2004). As sub- item as significantly less healthful than consumers who viewed the food
jective knowledge is associated with perceptions of ability (Brucks, item in a muted, less color-saturated package. Experiment 2 identifies
1985), consumers with higher levels of subjective nutrition knowledge conceptual fluency as the meditating psychological mechanism re-
should ignore less diagnostic but easily accessible incidental cues, like sponsible for the vivid packaging equals unhealthful heuristic and ex-
package color saturation (Mackenzie & Spreng, 1992; Petty & Cacioppo, tends the generalizability of the effect in the context of a different
1986). Further, as incidental cues are often less associated with specific packaged food product (potato chips). Experiment 3 replicates our
product attributes (Mackenzie & Spreng, 1992), consumers with higher mediation result and demonstrates a moderated mediation effect, such
levels of subjective nutrition knowledge should form weaker associa- that consumers with higher levels of subjective nutrition knowledge are
tions between package color saturation and food healthfulness, which less likely to rely on the vivid equals unhealthful heuristic when as-
should diminish the conceptual fluency they experience when they sessing packaged food products. Finally, experiment 4 demonstrates
encounter less healthful food in vivid packaging (Winkielman et al., that restrained eaters are more affected by the vivid packaging equals
2003). unhealthful heuristic than unrestrained eaters. A description of the
Second, consumers with higher levels of subjective knowledge may stimuli used throughout this research immediately follows.
not be as exposed to unhealthful foods in vivid packaging as other
consumers. While exposure is an incidental byproduct of shopping, 3.1. Experimental stimuli creation and pretests
research suggests that due to a need for consistency between their
knowledge and behavior, consumers with higher levels of subjective To create our experimental stimuli, we searched for packaged food
nutrition knowledge tend to concentrate their shopping in areas of brands in four steps. First, to address potential bias from previous
supermarkets that are more likely to contain healthful foods (Moorman product experience, several food brands were identified that were likely
et al., 2004). Thus, the vivid equals unhealthful mental association may unfamiliar to our sample population. Next, from among the considered
not be as frequently reinforced among consumers with higher levels of brands, we selected brands with moderate levels of color saturation in
subjective nutrition knowledge. Therefore, we predict that: their package designs. With moderate saturation levels, a package's
color saturation level could be digitally increased or decreased while
H3. Subjective nutrition knowledge attenuates the relationship
remaining within the saturation norms for the product category. Then,
between vivid package colors and food healthfulness by reducing
we narrowed down the brands to only include foods whose packages
perceptions of conceptual fluency.
minimally displayed the food itself, as food color has been shown to
influence a variety of consumer perceptions (Labrecque et al., 2013).
12
J.A. Mead, R. Richerson Journal of Business Research 82 (2018) 10–18
Finally, among remaining brands, we selected one ostensibly less they viewed an image of “Frank*” brand protein bars. Participants in
healthful packaged food category, potato chips, and an ostensibly more the muted condition saw the low saturation version of the “Frank*”
healthful packaged food category, nutrition bars, to increase the gen- image, while participants in the vivid condition viewed the high sa-
eralizability of the experimental results across different food categories. turation version of the “Frank*” image. Apart from the package color
We then conducted a pretest to confirm that the selected brands manipulation, the images were identical. After viewing the image,
(“Maui Style” brand potato chips and “Frank*” brand nutrition bars) participants were asked to evaluate the food on a 5-item, 7-point se-
were unfamiliar to the research participant pool and that the product mantic differential scale of perceived healthfulness adapted from pre-
categories were perceived to be significantly different in healthfulness. vious research (Raghunathan et al., 2006; Scott et al., 2008): un-
Ninety-five undergraduate business students from a large Southeastern healthful/healthy, tastes bad/tastes good (reverse coded), not
University completed a lab study to satisfy a research requirement. nutritious/nutritious, indulgent food/diet food, high in calories/low in
First, participants viewed an image of each brand's package and rated calories (α = 0.75).
each food item on perceived familiarity via a 3-item, 7-point Likert
scale (“I am very familiar with this exact product,” “I eat this exact 3.2.2. Results
product often,” “I see this exact product frequently when shopping”; We performed a regression analysis where the muted condition was
αchips = 0.92, αbars = 0.88). As intended, one-sample t-tests indicated coded as 0 and the vivid condition was coded as 1. The results indicate
that perceived familiarity for each food item was significantly below the that participants in the vivid condition perceived the protein bars as
scale midpoint of 4 (t(94) = − 14.41, p < 0.001; Mchips famil- significantly less healthful than participants in the muted condition
iarity = 1.94, SD = 1.39) and (t(94) = − 17.81, p < 0.001; Mbars famil- (β = − 0.276, t(57) = −2.17, p < 0.05; Mmuted = 6.02, SD = 1.38
iarity = 1.81, SD = 1.18). Next, participants were asked to rate the vs. Mvivid = 5.26, SD = 1.27), in support of H1.
healthfulness of the two product categories, chips and nutrition bars, on
a 5-item, 7-point semantic differential scale adapted from previous re- 3.2.3. Discussion
search (Raghunathan et al., 2006; Scott et al., 2008): unhealthful/ Results demonstrate that consumers perceived a nutrition bar pre-
healthy, tastes bad/tastes good (reverse coded), not nutritious/nu- sented in vivid, more highly color-saturated packaging as a less
tritious, indulgent food/diet food, high in calories/low in calories healthful food than an otherwise identical nutrition bar presented in
(αchips = 0.75, αbars = 0.82). A paired samples t-test indicated that muted, less color-saturated packaging, supporting H1. Thus, consumers,
participants viewed the chips product category as significantly less who aspire to make more healthful packaged food choices may inad-
healthful than the nutrition bars product category (t(93) = −11.25, vertently find their perceptions guided by package color saturation,
p < 0.001; Mchips = 3.43, SD = 1.18 vs. Mbars = 5.63, SD = 1.60). which may not be a true reflection of food healthfulness. While these
Following the pretest, Adobe Photoshop was used to adjust the color results are consistent with the larger assertion that through repeated
saturation of the packaging of each food brand down (referred to exposure, consumers have learned to associate vivid packaging with
hereafter as muted) or up (referred to hereafter as the vivid) 50 points unhealthful foods, this experiment is unable to speak directly to this
from its original value using the color saturation slider tool. The color association. As such, in experiment 2, conceptual fluency is investigated
hue and color lightness sliders were left unaltered. Please see Fig. 1 for as the mechanism underlying the vivid packaging equals unhealthful
details. effect. Given that conceptual fluency often manifests through repeated
We also conducted a second pretest among 128 Mturk participants association and exposure (Berger & Fitzsimons, 2008), and that per-
located in the U.S. to confirm that consumers perceived differences in ceptions of conceptual fluency are also indicative of the retrieval of
color saturation distinctly from differences in color lightness (Mai et al., information from memory (Alter & Oppenheimer, 2009), if the vivid
2016). After sampling an orange hue from our stimuli, we used Adobe packaging equals unhealthful heuristic is the result of a learned asso-
Photoshop to independently adjust the lightness property ± 50 points ciation, conceptual fluency should mediate the relationship between
and the saturation property by ± 50 points, creating four new color package color saturation and perceptions of food healthfulness.
stimuli – light, dark, vivid, and muted. Study participants viewed and
rated each color (order randomized) on a variety of perceptual di- 3.3. Experiment 2
mensions which, based on established literature, we expected would
vary between light/dark and vivid/muted colors (Labrecque & Milne, In experiment 2 we investigate conceptual fluency as the meditating
2012; Mai et al., 2016). As expected, each color was viewed as unique psychological mechanism underlying the vivid packaging equals un-
and was perceived distinctly. For example, the dark color was perceived healthful heuristic. We propose that repeated exposure to unhealthful
as heavier and more masculine, the light color was lighter and more foods displayed in vivid, highly color-saturated packaging has created
feminine, and the vivid color was more exciting and arousing. The an association between less healthful foods and vivid, highly color-sa-
muted color was the least strong and vigorous color, while also some- turated packages. As such, when less healthful foods (in the case of
what light and feminine. Overall, these results align with previous re- experiment 2, potato chips) are displayed in highly color-saturated
search (Labrecque & Milne, 2012; Mai et al., 2016). Please see Appendix packaging, consumers should experience a sense of conceptual fluency,
Table A1 for details. which is expected to decrease consumers' perceptions of food health-
fulness. This experiment also attempts to replicate the vivid packaging
3.2. Experiment 1 equals unhealthful main effect identified in experiment 1 with a dif-
ferent food product to increase the generalizability of the results.
Experiment 1 tests the effect of package color saturation on con-
sumers' perceptions of food healthfulness. We expect consumers to 3.3.1. Method
judge food items with vivid, highly color-saturated packages as less One hundred and sixty-two undergraduate students from a large
healthful than food items displayed in muted, less color-saturated Southeastern University completed an online experiment to satisfy a
packages. research requirement. Participants were randomly assigned to one of
two experimental conditions (package colors: muted vs. vivid) where
3.2.1. Method they viewed an image of “Maui Style” brand potato chips. Participants
Fifty-nine undergraduate business students from a large in the muted condition saw the low saturation version of the “Maui
Southeastern University completed an online experiment to satisfy a Style” image, while participants in the vivid condition viewed the high
research requirement. Participants were randomly assigned to one of saturation version of the “Maui Style” image. Apart from the package
two experimental conditions (package colors: muted vs. vivid) where color manipulation, the images were identical. As in experiment 1,
13
J.A. Mead, R. Richerson Journal of Business Research 82 (2018) 10–18
B
-50 Points Color Saturation (Muted) +50 Points Color Saturation (Vivid)
perceived product healthfulness was measured with a 5-item, 7-point and perceptions of the healthfulness of the “Maui Style” brand potato
semantic differential scale (α = 0.70). After participants had provided chips as the 95% confidence interval of both the indirect effect (− 0.23
their healthfulness perceptions, conceptual fluency was measured with to −0.004) and direct effect (−0.75 to − 0.002) of the experimental
a 12-item, 7-point semantic differential scale adapted from previous condition excluded zero, in support of H2. That is, participants, per-
research (Kidwell et al., 2013; Lee & Aaker, 2004; White, ceived the vivid, highly color-saturated image as more conceptually
MacDonnell, & Dahl, 2011): unclear/clear, not compelling/compelling, fluent (Mmuted = 3.84, SD = 1.36 vs. Mvivid = 4.24, SD = 1.33), and
not credible/credible, did not flow/flowed well, difficult to follow/easy higher levels of fluency were associated with lower levels of perceived
to follow, not plausible/plausible, not easy to relate to/easy to relate to, healthfulness when viewing the “Maui Style” brand potato chips. Please
difficult to process/easy to process, difficult to understand/easy to see Fig. 2 for details.
understand, difficult to comprehend/easy to comprehend, does not
make sense/makes sense, seemed unfamiliar/seemed familiar
(α = 0.93). 3.3.3. Discussion
Experiment 2 provides additional evidence that consumers use the
color saturation of food packaging to inform their food healthfulness
3.3.2. Results perceptions. Again, participants who viewed the vivid, highly color-
Again, a regression analysis indicates that participants in the vivid saturated package of potato chips perceived the chips to be less
condition (coded as 1) perceived the chips as significantly less healthful healthful than participants who viewed a muted, less color-saturated
than participants in the muted condition (β = − 0.18, t(160) = − 2.38, bag of chips, suggesting an association between the concepts. Further,
p < 0.05; Mmuted = 3.17, SD = 1.27 vs. Mvivid = 2.72, SD = 1.15), the partial mediation result supports conceptual fluency as the psy-
replicating the results from experiment 1. chological mechanism behind the vivid packaging equals unhealthful
We used the Process macro (Hayes, 2013) to run a mediation ana- food heuristic. That is, it appears that consumers not only associate
lysis (model 4), with 5000 bootstrapped samples, where perceived food vivid package colors with unhealthfulness, but also experience a sense
healthfulness served as the dependent variable, participants' experi- of conceptual fluency when an ostensibly indulgent food (potato chips)
mental condition was entered as the independent variable (muted is presented in a color saturated package. Therefore, it appears that
coded as 0, vivid coded as 1), and conceptual fluency was entered as the when consumers encounter a seemingly indulgent food in a vivid food
mediator. Results revealed that participants' conceptual fluency par- package, not only is the concept of unhealthfulness activated in
tially mediated the relationship between the experimental condition memory, but also the pairing of a vivid package with an indulgent food
14
J.A. Mead, R. Richerson Journal of Business Research 82 (2018) 10–18
conceptually fits with expectations. knowledgeable than others about nutrition information,” 7-point Likert
In experiment 3, we explore how subjective nutrition knowledge scale, adapted from Moorman et al., 2004).
moderates the mediation result demonstrated in experiment 2. This
additional variable may account for some of the variance in our model,
3.4.2. Results
given the partial mediation results reported. We expect that consumers
A regression analysis indicates that participants in the vivid condi-
with higher levels of subjective knowledge will have a weaker asso-
tion (coded as 1) perceived the chips as marginally less healthful than
ciation between vivid package colors and unhealthful food, thereby
participants in the muted condition (β = 0.15, t(133) = 1.76,
diminishing the sense of conceptual fluency they experience when such
p = 0.08; Mmuted = 6.25, SD = 1.64 vs. Mvivid = 6.72, SD = 1.45).
a pairing is encountered. Further, if, as we expect, subjective nutrition
We used the Process macro (Hayes, 2013) to run a moderated
knowledge moderates the link between vivid package colors and con-
mediation analysis (model 7), with 5000 bootstrapped samples, where
ceptual fluency, the results will provide additional support for our ac-
perceived food healthfulness served as the dependent variable, parti-
count of conceptual fluency underlying our effect.
cipants' experimental condition as the independent variable (muted
coded as 0, vivid coded as 1), conceptual fluency (α = 0.90) as the
3.4. Experiment 3 mediator, and subjective nutrition knowledge as the moderator.
Results revealed that participants' conceptual fluency fully mediated
Experiment 3 tests the moderating role of subjective nutrition the relationship between the experimental condition and perceptions of
knowledge on consumers' tendency to view food in vivid, highly color- the healthfulness of the “Maui Style” brand potato chips, but only
saturated packages as less healthful than food in muted, less color-sa- among participants low in subjective nutrition knowledge, that is,
turated packages. We expect that higher levels of subjective nutrition participants with scores below the Johnson-Neyman point of 3.14
knowledge attenuate the relationship between package color saturation (Spiller, Fitzsimons, Lynch, & McClelland, 2013). The 95% confidence
and conceptual fluency, which will weaken the main effect of package interval of the index of moderated mediation (− 0.23 to − 0.0005)
vividness on food healthfulness perceptions. Additionally, to extend the excluded zero, while the 95% confidence interval of the direct effect
generalizability of the vivid equal unhealthful heuristic, we assess food (− 0.13 to 0.89) of participants' experimental condition on to food
healthfulness based on perceptions of specific nutrition information healthfulness perceptions included zero, supporting H3 and H4. That is,
(e.g., fat, sodium) and employ an Mturk sample. the link between vivid, highly color-saturated food packaging and
conceptual fluency (Mmuted = 4.78, SD = 1.45 vs. Mvivid = 5.08,
SD = 1.30) was attenuated among participants with moderate to high
3.4.1. Method
levels subjective of subjective nutrition knowledge. Please see Fig. 3 for
One hundred thirty-five Mturk participants (mean age = 39.26,
details.
52.6% female) from the U.S. participated in an online survey in ex-
change for a small honorarium. Experiment 3 replicates the design of
experiment 2 with two additions, a 5-item attribute-based measure of 3.4.3. Discussion
food healthfulness (α = 0.92; 9-point systematic differential scale: very Experiment 3 provides additional evidence to support our theo-
low in calories/very high in calories, very low in fat/very high in fat, rizing that conceptual fluency mediates the relationship between vivid
very low in sodium/very high in sodium, very low in carbohydrates/ packaging and unhealthful food perceptions. Further, moderated med-
very high in carbohydrates, very low in cholesterol/very high in cho- iation analysis indicates that consumers with higher levels of subjective
lesterol) and a measure of subjective nutrition knowledge (“I am more nutrition knowledge are less likely to judge a food item based on the
15
J.A. Mead, R. Richerson Journal of Business Research 82 (2018) 10–18
16
J.A. Mead, R. Richerson Journal of Business Research 82 (2018) 10–18
encountered, consumers' semantic association with unhealthful food is needed to test these claims. Additionally, since this research asserts that
activated, the association feels familiar and fluent, and the food is the vivid packaging equals unhealthful heuristic is the result of a
perceived as unhealthful (Stafford & Grimes, 2012). learned semantic association between unhealthful foods and highly
Additionally, results demonstrate that higher levels of subjective color-saturated packaging, these results may not generalize to con-
nutrition knowledge can weaken the vivid packaging equals un- sumers in other countries or from different cultures (Brunsø & Grunert,
healthful heuristic by attenuating the relationship between package 1998). For example, if unhealthful foods are less prevalent or less likely
color vividness and conceptual fluency. We contend that consumers to be displayed in highly color-saturated packaging relative to healthful
with higher levels of subjective nutrition knowledge are less likely to foods in a given part of the world, the effects demonstrated in this re-
accept incidental package cues as diagnostic of food healthfulness and search would likely weaken or even vanish.
are less likely to categorize foods in vivid packages as either healthful or
unhealthful based on vividness alone. Further, these consumers may 4.3. Further research
have had fewer exposures to unhealthful foods in vivid packaging due
to deliberate efforts to maintain consistency between their perceived The link between color saturation and packaged food healthfulness
knowledge and behavior (Moorman et al., 2004). Conversely, results may represent one of many learned associations consumers have
indicate that restrained eaters are more likely than unrestrained eaters formed with food. For example, if marketers often favor primary colors
to be affected by the vivid packaging equals unhealthful heuristic. First, in their packaging of indulgent foods, or disproportionally display or-
higher motivation to attend to and monitor food choices (Schachter ganic foods in packages colored in certain hues (e.g., green or brown),
et al., 1968; Tom & Rucker, 1975) means that restrained eaters have consumers may have similarly learned to associate primary colors or
likely spent more time examining food and food packages than un- certain color hues with healthfulness, taste, or heaviness. Researchers
restrained eaters. This extra focus should result in a stronger semantic may also wish to investigate if the vivid packaging equals unhealthful
association between healthfulness and package color saturation heuristic extends beyond perceptions to also affect consumers' food
(Berger & Fitzsimons, 2008). Additionally, with higher levels of psy- consumption.
chological depletion, restrained eaters should also be more likely to rely Finally, this research explored how the overall color saturation of
on heuristic cues to inform their food healthfulness perceptions (Scott food packaging influenced consumer perceptions of a food item.
et al., 2008). Depletion should increase their sensitivity to the semantic However, it is possible that individual elements of the package design
association between unhealthful foods and highly color-saturated could be separately manipulated to produce differential effects on
packaging, as fluency effects are strongest when consumers rely on consumers' food perceptions. For example, research has demonstrated
heuristic processing methods (Fukawa & Niedrich, 2015). that when health information is displayed in green, rather than red,
consumers judge the food item as more healthful (Schuldt, 2013). Re-
4.2. Limitations latedly, research has indicated that vivid, highly saturated colors are
often reflective of high nutritional value in the context of fruits and
While this research provides repeated experimental evidence for the vegetables (Goldman, 2003). Therefore, consumers may perceive a
presence of a vivid packaging equals unhealthful heuristic when eval- packaged food as more healthful if the package displays the food itself
uating food items, this work is not without limitations. First, our results in vivid, highly saturated colors, while the rest of the package employs
do not address the influence of packaging color saturation on con- muted, less saturated colors. As such, further research may wish to
sumers' frequency of purchase, frequency of consumption, or quantity explore the more nuanced effects which may emerge if the color sa-
of consumption. However, as perceptions tend to influence behaviors, it turation of the package and the color saturation of the food item pic-
seems reasonable to expect such effects, though further research is tured within are manipulated separately.
Appendix A
Table A1
Color perception pretest.⁎
A B C D
17
J.A. Mead, R. Richerson Journal of Business Research 82 (2018) 10–18
18