2 ADVERTISEMENT

HOME !SCIENCE ! VOL. 323, NO. 5918 ! BLUE OR RED? EXPLORING THE EFFECT OF CO… * . / 1 4 !

2
2 Current Issue First release papers Archive About
3
Submit manuscript

HOME ! SCIENCE ! VOL. 323, NO. 5918 ! BLUE OR RED? EXPLORING THE EFFECT OF COLOR ON COGNITIVE TASK PERFORMANCES

+ REPORTS
&'() 󡀆 *
Blue or Red? Exploring the Effect of Color on Cognitive Task
Performances
RAVI MEHTA AND RUI (JULIET) ZHU Authors Info & Aoliations

SCIENCE 27 Feb 2009 Vol 323, Issue 5918 pp. 1226-1229 DOI: 10.1126/science.1169144

5681 412
. / 0 1

Abstract
$ Existing research reports inconsistent findings with regard to the effect of color on cogni- %
tive task performances. Some research suggests that blue or green leads to better perfor- &
mances than red; other studies record the opposite. Current work reconciles this discrep-
ancy. We demonstrate that red (versus blue) color induces primarily an avoidance (versus '
approach) motivation (study 1, n = 69) and that red enhances performance on a detail-ori- (
ented task, whereas blue enhances performance on a creative task (studies 2 and 3, n = 208
)
and 118). Further, we replicate these results in the domains of product design (study 4, n =
42) and persuasive message evaluation (study 5, n = 161) and show that these effects occur 5
outside of individuals' consciousness (study 6, n = 68). We also provide process evidence
*
suggesting that the activation of alternative motivations mediates the effect of color on
cognitive task performances.

SIGN UP FOR THE SCIENCEADVISER NEWSLETTER

The latest news, commentary, and research, free to your inbox daily
SIGN UP !

Color is a fundamental aspect of human perception, and its effects on cognition and behavior
have intrigued generations of researchers. Although a large amount of research has been done
in this domain, the psychological processes through which color operates have not been ex-
plored fully. As a result, the field has observed certain conflicting results. One inconsistency,
which is the focus of this report, concerns the effect of color on cognitive task performance.
Most research examining this topic has focused on two of the three primary colors—red versus
blue (or green). Some have proposed that red enhances cognitive task performance as com-
pared with blue or green (1, 2); others have shown exactly the opposite (3, 4).

This report details our effort to understand the theory behind the psychological process
through which color affects cognitive task performances. Based on our theorizing, we are able
to reconcile the above-described inconsistency. We demonstrate that red and blue activate dif-
ferent motivations and consequently enhance performances on different types of cognitive
tasks. In line with most of the extant research, we limit our research to the two primary colors,
red and blue.

Color theorists believe that color influences cognition and behavior through learned associa-
tions (3). When people repeatedly encounter situations where different colors are accompa-
nied by particular experiences and/or concepts, they form specific associations to colors. Red
and blue have been shown to have different associations within the cognitive domain. Red is
often associated with dangers and mistakes [e.g., errors that are circled with a red ink pen,
stop signs, and warnings (3)]. Claims have been made linking the color red to the highest level
of hazard and also the highest level of compliance (5, 6). In contrast, blue is often associated
with openness, peace, and tranquility [e.g., ocean and sky (7)]. A word association test con-
firmed that people indeed generate these different associations to red versus blue color in the
cognitive task domain (8, 9).

We propose that these different associations related to red versus blue color can induce alter-
native motivations. Specifically, red, because of its association with dangers and mistakes, Hypothesis
should activate an avoidance motivation, which has been shown to make people more vigilant
and risk-averse (10–12). Thus, red, compared with blue, should enhance performance on de-
tail-oriented tasks (i.e., tasks that require focused, careful attention). In contrast, because blue
is usually associated with openness, peace, and tranquility, it is likely to activate an approach
motivation, because these associations signal a benign environment that encourages people to
use innovative as opposed to “tried-and-true” problem-solving strategies (13). Indeed, an ap-
proach motivation has been shown to make people behave in a more explorative, risky manner
(10, 11). Thus, blue versus red should enhance performance on creative tasks.

We report six studies (14) that offered systematic support to our hypotheses. Most studies
were conducted on computers, and color was manipulated through the background screen col-
or. Color is usually defined along three dimensions: hue (the pigment of the color, e.g., blue,
red, etc.), chroma (saturation of color), and value (degree of darkness or lightness of the color)
(15). In order to reduce confounds and to be consistent with prior research, we manipulated
only hue (i.e., red versus blue) and kept chroma and value constant. For computer-based stud-
ies, we employed the HSL (hue-saturation-lightness) scheme (red: hue = 0, saturation = 240,
lightness = 120; blue: hue = 160, saturation = 240, lightness = 120) (8) (table S1). To provide a
baseline for comparison, we also included a neutral condition in some studies where computer
background color was set to be white.

Study 1 tested our hypothesis that red color will induce primarily an avoidance motivation,
whereas blue will activate an approach motivation. Sixty-nine participants were randomly as-
description
signed to the red, blue, or neutral background color condition and completed a computer- of the 1st
based study that consisted of two tasks. In the first task, participants solved a series of 12 ana- task; 3
kinds of
grams, with three of them having target words related to avoidance motivation (e.g., prevent), anagrams
another three having target words related to approach motivation (e.g., adventure) (16, 17),
and the remaining six that were unrelated to either motivation (e.g., computer) (8). The re-
sponse times for each type of correctly solved anagrams were averaged to create three reaction
*the rest turns out
time (RT) indices—avoidance, approach, and neutral RT indices. Faster reaction time to ap- to be measured in a
proach-related (or avoidance-related) anagrams would imply a stronger activation of an ap- crazy way so id just
ignore this study
proach (or avoidance) motivation (18). As anticipated (Fig. 1), for approach-related anagrams,
those in the blue condition [10.93 ± 5.51 s (mean ± SD)] responded faster than those in the red
[18.53 ± 12.25 s; t(66) = 2.81, P < 0.01; Cohen's d = 0.81] or neutral condition [17.50 ± 9.17 s;
t(66) = –2.29, P < 0.03; Cohen's d = 0.91]. For avoidance-related anagrams, the reverse pattern
appeared, such that those in the red condition (10.40 ± 3.64 s) responded faster than those in
the blue [20.39 ± 14.73 s; t(66) = –3.21, P < 0.01; Cohen's d = 0.96] or neutral condition [19.14 ±
11.48 s; t(66) = –2.67, P < 0.01; Cohen's d = 1.1]. For neutral anagrams, however, no effect of
color was observed (red, 10.56 ± 5.48 s; blue, 12.64 ± 7.53 s; neutral, 11.58 ± 4.37 s; t < 1).

! OPEN IN VI

Fig. 1 . Participants' response times to approach-related, avoidance-related, and neutral anagrams under red, blue, and
neutral color conditions (study 1). F2,66 = 8.79, P < 0.001. Error bars, ±1.00 SD.

In the second task, participants read descriptions of three pairs of brands and reported their
preferences along a scale from 1 (prefer brand A) to 7 (prefer brand B). Within each pair, one
brand highlighted a negative outcome people try to avoid, whereas the other brand highlight-
ed a positive outcome people try to approach. For example, one pair featured two brands of
toothpastes, with brand A being particularly good for cavity prevention (avoidance focused),
and brand B being particularly good for tooth whitening (approach focused) (8). Across three
pairs, we found that those in the blue color condition (4.03 ± 1.55) indicated greater prefer-
ence for brands that were approach-oriented (highlighting positive benefits) than those in the
red [2.79 ± 1.65; t(66) = –2.80, P < 0.01; Cohen's d = 0.79] and the neutral condition [3.05 ±
1.43; t(66) = 2.08, P < 0.05; Cohen's d = 0.67]. Thus, this study demonstrated that within a cog-
nitive task domain, red (versus blue) can activate an avoidance (versus approach) motivation.
A post hoc study ruled out mood as an alternative explanation (8).

The next two studies tested whether red (versus blue), because of its activation of avoidance
(versus approach) motivation, enhances performance on a detail-oriented (versus a creative)
task. Study 2 (n = 208) contained two tasks, a detail-oriented and a creative task. A set of par-
ticipants completed the detailed-oriented task (i.e., a memory exercise) presented on comput-
ers with red, blue, or neutral background color. They studied a list of 36 words for 2 min and
were asked to recall as many words as they could after a 20-min delay. Three measures con-
firmed that red indeed enhanced performance on this memory task. Those in the red condition
(15.89 ± 5.90) recalled more correct items than those in the blue condition [12.31 ± 5.48; t(100)
= 2.50, P < 0.02; Cohen's d = 0.64] (Fig. 2). Furthermore, blue led to more false recalls (0.86 ±
1.29) than red [0.34 ± 0.64; t(100) = –2.42, P < 0.02; Cohen's d = 0.52] or neutral [0.38 ± 0.55;
t(100) = 2.21, P < 0.03; Cohen's d = 0.48] condition. These two measures have been shown to
reflect people's attention to details (19, 20). However, color manipulation did not affect the
total number of items recalled (P > 0.11) (8).

! OPEN IN VI

Fig. 2 . Total number of correct recalls for the memory task (study 2). F2,100 = 3.15, P < 0.05. Error bars, ±1.00 SD.

Another set of participants completed a creative task where they were asked to generate as
many creative uses for a brick as they could think of within 1 min (21). Consistent with prior
research (21), each participant's responses were coded into three categories: (i) total number
of uses generated, (ii) mean creativity score as rated by a panel of judges, and (iii) total number
of creative uses. Participants in the three color conditions produced equal number of uses in
total (F <1; red, 4.83 ± 2.31; blue, 4.67 ± 2.62; neutral, 4.94 ± 1.68). However, the quality of
these uses differed by color conditions. Those in the blue condition (3.97 ± 0.99) demonstrated
a higher mean creativity score than those in the red [3.39 ± 0.97; t(102) = –2.81, P <0.01; Co-
hen's d = 0.6] or neutral color condition [3.50 ± 0.63; t(102) = 2.23, P < 0.03; Cohen's d = 0.57]
(Fig. 3). Similarly, those in the blue (1.64 ± 1.46) condition produced more creative uses than
those in the red [0.86 ± 0.97; t(102) = –2.93, P < 0.01; Cohen's d = 0.64] or neutral condition
[0.91 ± 0.83; t(102) = 2.70, P < 0.01; Cohen's d = 0.62] (8). Findings from this study suggest that,
although color did not affect the amount of processing, as shown in the equal recall level for
the memory task and comparable uses generated for the brick task, it affected the quality of
responses, i.e., red led to superior performances on detail-oriented tasks and blue, on creative
tasks.

! OPEN IN VI

Fig. 3 . Mean creativity scores for the brick task (study 2). F2,102 = 4.43, P < 0.02. Error bars, ±1.00 SD.

Study 3 (n = 118) tested the generalizability of results observed in study 2 by using two differ-
ent tasks. Moreover, it aimed to demonstrate that the activation of avoidance or approach mo-
tivations is the underlying force that drives our results. As in study 2, this study was computer-
based, and color was manipulated using the computer background screen color. The detail-ori-
ented task in this study was a proofreading task (22). Participants examined five sets of items,
with each set containing a pair of names or addresses, which were either identical or slightly
different (8). Participants' task was to judge whether items within each pair were identical or
not. To assess whether color-induced motivations drive our expected effects, we also asked
participants to answer three questions concerning the extent to which they focused on accura-
cy (mistake-avoidance motivation) versus speed (approach motivation). Results revealed that
red color condition (4.33 ± 0.77) led to more correct responses than blue [3.53 ± 0.80; t(51) =
2.49, P <0.02; Cohen's d = 1.05] or neutral [3.68 ± 1.20; t(51) = 2.07, P < 0.05; Cohen's d = 0.66]
condition (F2,51 = 3.56, P < 0.04). Further, mediation analysis (23) revealed that approach ver-
sus avoidance motivations were indeed the driving force for the observed effect.

The Remote Associates Test (RAT), which is widely used as a test of creative thinking (24), was
used as the creative task. Each RAT item consists of three or four stimulus words (e.g., “Shelf,”
“Read,” and “End”) that are in some way related to a fourth or fifth unreported word (e.g.,
“Book”). Participants were presented with five RAT items and were asked to determine what
the target words were. As predicted, those in the blue condition (4.00 ± 0.74) produced more
correct answers (thus exhibiting higher creativity) than those in the red [3.45 ± 0.89; t(61) = –
2.35, P < 0.03; Cohen's d = 0.69] or neutral [3.38 ± 0.67; t(61) = 2.67, P < 0.01; Cohen's d = 0.9]
condition (F2,61 = 4.33, P < 0.02). Mediation analysis again confirmed that the alternative moti-
vations activated by color drive the observed effect (8).

Study 4 aimed to further extend the previous studies by using a single task that could examine
both people's creativity level and their attention to details. For this purpose, participants were
presented with a sheet of paper with drawings of 20 different parts (fig. S1). Participants (n =
42) were required to use any five parts and draw a design of a toy a child (age 5 to 11) could use
to play with (8, 25). Unlike previous studies, the color manipulation in this study was done by
presenting the 20 parts either in red or blue color. Twelve judges evaluated black-and-white
copies of each design on two dimensions, one assessing the originality and novelty (reflecting
creativity level) and the other assessing the practicality and appropriateness (reflecting atten-
tion to details) of the design. The correlation between these two dimensions was 0.29 (P =
0.08) and interjudge reliability was 0.75 for originality and 0.83 for appropriateness. As ex-
pected, toys designed by those in the red color condition were judged to be more practical and
appropriate (3.47 ± 0.79) than those in the blue condition (2.95 ± 0.88; F1,40 = 4.16, P < 0.05;
Cohen's d = 0.64), but were judged less original and novel (2.94 ± 0.55) than those in the blue
condition (3.37 ± 0.76; F1,40 = 4.46, P < 0.05; Cohen's d = 0.67). A set of anagrams similar to
those described in study 1 were included in this study to test whether avoidance or approach
motivations were the driving force for the effect. Results revealed that those in the red versus
blue condition responded faster to the avoidance-related anagrams (which indicated an acti-
vation of an avoidance motivation) and, subsequently, exhibited a higher score on practicality
and appropriateness. In contrast, those in the blue versus red condition responded faster to
the approach-related anagrams (which indicated an activation of an approach motivation)
and, subsequently, exhibited a higher score on originality and novelty.

Study 5 tested our theorizing in yet another domain, namely persuasive message evaluation.
Participants (n = 161) evaluated one of two versions of an advertisement for a camera on a
computer screen with the background color set to be either red or blue. The two versions of
the advertisement were identical except for the visuals (26). In one version, the visuals repre-
sented specific product details of the camera (e.g., lens) and thus fitted a detail-oriented pro-
cessing style (27). We expected that red, which enhances attention to details, would lead to
higher persuasion for this version. The other version included visuals that represented rather
remotely related associations (e.g., a road sign, a dining table in a restaurant, and a map),
which would require creative thinking to connect all these images to a camera-related theme,
e.g., travel (fig. S2). Thus, we expected that blue, which appears to enhance creative cognition,
would lead to more persuasion for this version (27). Participants evaluated one of the ads on
three seven-point items assessing its appeal, favorability, and effectiveness. For the red color
computer background screen, participants formed more favorable evaluations when the ad in-
cluded visuals representing specific product details (4.69 ± 1.26) as opposed to remotely relat-
ed associations (4.11 ± 1.28; F1,150 = 3.41, P < 0.07; Cohen's d = 0.42). In contrast, when the
background color was blue, the reverse occurred, i.e., more favorable evaluations emerged
when the ad contained visuals representing remotely related associations (4.41 ± 1.47) versus
specific product details (3.60 ± 1.59; F1,150 = 6.01, P < 0.02; Cohen's d = 0.56) (8).

In the last study, we tested whether people are aware of the differential effects of red versus
blue color. No color manipulation was done for this study, and all the instructions and the fo-
cal task were presented in black color font with white background screen color on computers.
Participants (n = 68) were told that one of these tasks they would complete requires detailed,
careful, and systematic processing of information, and it could be presented to them with ei-
ther a red or a blue background color. Participants' task was to select one color that they
thought would enhance their performance on that task. A sample of the red and blue colors
was presented (fig. S3). On the next screen, participants were told that another task in this
study would require creative, imaginative, and outside-of-the-box thinking and were asked to
select one of the two colors that they thought might enhance their performance on the cre-
ative task.

The data revealed that significantly more participants chose the blue (66%) versus red (34%)
color when the task was described to be creative [χ2 (1) = 7.12, P < 0.01]. However, interestingly,
the same pattern of results emerged when the task was described to be detail-oriented, i.e.,
more people thought the blue (74%) versus red (26%) background color would enhance their
performance even on the detail-oriented task [χ2 (1) = 15.06, P < 0.001] (8). These results are
consistent with the general belief that people have an overall preference for blue versus red
color, although we found that red can be beneficial when the focal task requires detailed atten-
tion.

From a series of six studies, using various tasks covering a number of different domains, we
demonstrate that red (versus blue) can activate an avoidance (versus approach) motivation
and subsequently can enhance performance on detail-oriented (versus creative) cognitive
tasks. This research thus offers a reconciliation of the conflicting results reported in the ex-
tant literature and advances current research on the effect of color on cognition and behavior
[e.g., (3)]. More important, our findings offer a wide range of implications for daily human life.
What wall color do we pick for an educational facility? What color enhances persuasion in a
consumption context? What color enhances creativity in a new product design process? Re-
sults from this research suggest that, depending on the nature of the task, different colors
might be beneficial. If the task on hand requires people's vigilant attention (e.g., memorizing
important information or understanding the side effects of a new drug), then red (or another
color that activates an avoidance motivation) might be particularly appropriate. However, if
the task calls for creativity and imagination (e.g., designing an art shop, or a new product idea
brainstorming session), then blue (or another color that activates an approach motivation)
would be more beneficial.

Supplementary Material

File (mehta.som.pdf)

DOWNLOAD 338.10 KB

References and Notes

1 N. Kwallek, C. M. Lewis, Appl. Ergon.21, 275 (1990).

" GO TO REFERENCE • CROSSREF • PUBMED • ISI • GOOGLE SCHOLAR

2 N. J. Stone, J. Environ. Psychol.23, 63 (2003).

" GO TO REFERENCE • CROSSREF • ISI • GOOGLE SCHOLAR

3 A. J. Elliot, M. A. Maier, A. C. Moller, R. Friedman, J. Meinhardt, J. Exp. Psychol. Gen.136, 154 (2007).

# SEE ALL REFERENCES • CROSSREF • PUBMED • ISI • GOOGLE SCHOLAR

4 S. Soldat, R. C. Sinclair, M. M. Mark, Soc. Cogn.15, 55 (1997).

" GO TO REFERENCE • CROSSREF • ISI • GOOGLE SCHOLAR

SHOW ALL REFERENCES

eLetters (0)

eLetters is a forum for ongoing peer review. eLetters are not edited, proofread, or indexed, but they are
screened. eLetters should provide substantive and scholarly commentary on the article. Embedded @gures
cannot be submitted, and we discourage the use of @gures within eLetters in general. If a @gure is essen-
tial, please include a link to the @gure within the text of the eLetter. Please read our Terms of Service before
submitting an eLetter.

SUBMIT A RESPONSE TO THIS ARTICLE

No eLetters have been published for this article yet.

Recommended articles from TrendMD

Modafinil Shifts Human Locus Coeruleus to Low-Tonic, Autocompletion of repetitive stroking with image guid-
High-Phasic Activity During Functional MRI ance
Michael J. Minzenberg et al., Science, 2008 Yilan Chen et al., Computational Visual Media, 2023

Comment on "Wandering Minds: The Default Network and Empirical Evidence of Idea Generation in Open Innova-
Stimulus-Independent Thought" tion Community
Sam J. Gilbert et al., Science, 2007 Zhengfa Yang et al., International Journal of Crowd Sci-
ence, 2023
Recruitment of an Area Involved in Eye Movements Dur-
ing Mental Arithmetic Ongoing pain facilitates emotional decision-making be-
André Knops et al., Science, 2009 haviors
Chennan Lin et al., Brain Science Advances, 2022
Neuronal Activity Related to Reward Value and Motivation
in Primate Frontal Cortex Let’s all dance: Enhancing amateur dance motions
Matthew R. Roesch et al., Science, 2004 Qiu Zhou et al., Computational Visual Media, 2023

Dynamic Shifts of Limited Working Memory Resources in Cognitive impairment in two subtypes of a single sub-
Human Vision cortical infarction
Paul M. Bays et al., Science, 2008 Tang Yang et al., Chinese Medical Journal, 2021

Powered by

I consent to the use of Google Analytics and related cookies across the TrendMD network (widget, website, blog).
Learn more

Yes No

ADVERTISEMENT

CURRENT ISSUE
ROS-induced ribosome impairment underlies ZAKα-
mediated metabolic decline in obesity and aging
BY GODA SNIECKUTE, LAURA RYDER, ET AL.

Spatial transcriptomics of B cell and T cell receptors

reveals lymphocyte clonal dynamics
BY CAMILLA ENGBLOM, KIM THRANE, ET AL.

Toward a Cenozoic history of atmospheric CO2

BY
THE CENOZOIC CO2 PROXY INTEGRATION PROJECT (CENCO2PIP)
CONSORTIUM*†

TABLE OF CONTENTS ! Sign up for ScienceAdviser

Subscribe to ScienceAdviser to get the latest news,
commentary, and research, free to your inbox daily.

SUBSCRIBE
!

ADVERTISEMENT

LATEST NEWS
NEWS 12 DEC 2023 NEWS 12 DEC 2023

Pristine asteroid samples offer New class of ‘ozone-safe’

window into the early Solar System refrigerants may have unexpected
downside

SCIENCEINSIDER 12 DEC 2023 SCIENCEINSIDER 12 DEC 2023

New Florida law blocks Chinese Dino extinction researcher

students from academic labs committed research misconduct—
but not fraud, university report ^nds

NEWS 11 DEC 2023 SCIENCEINSIDER 8 DEC 2023

Cheap electricity could recycle Their budget already stretched near

animal waste, recover valuable bursting, U.S. particle physicists
chemicals dream small

ADVERTISEMENT

RELATED JOBS
Clinical Research Senior QA and QC Associate
Baylor College of Medicine
Hybrid within Greater Houston, Texas

Postdoc - The Bhattacharya Lab for Computational Genomics

University of Texas MD Anderson Cancer Center
Houston, Texas

Assistant Professor
University of Kansas Medical Center
Kansas City

MORE JOBS ▶

RECOMMENDED
REPORTS JULY 2004

Cognitive Imitation in Rhesus Macaques

RESEARCH ARTICLE AUGUST 2013

Poverty Impedes Cognitive Function

REPORT JULY 1967

Perceptual De^cit during a Mental Task

REPORTS DECEMBER 1999

Cognitive Modularity and Genetic Disorders

View full text | Download PDF

FOLLOW US

+ " # , $ 󡀆 GET OUR NEWSLETTER

NEWS CAREERS COMMENTARY JOURNALS AUTHORS & REVIEWERS

All News Careers Articles Opinion Science Information for Authors

ScienceInsider Find Jobs Analysis Science Advances Information for Reviewers

News Features Employer Hubs Blogs Science Immunology

Subscribe to News from Science Robotics

Science Science Signaling
News from Science FAQ Science Translational
About News from Science Medicine

Science Partner Journals

LIBRARIANS ADVERTISERS RELATED SITES ABOUT US HELP

Manage Your Institutional Advertising Kits AAAS.org Leadership FAQs

Subscription Custom Publishing Info AAAS Communities Work at AAAS Access and Subscriptions
Library Admin Portal Post a Job EurekAlert! Prizes and Awards Order a Single Issue
Request a Quote Science in the Classroom Reprints and Permissions
Librarian FAQs TOC Alerts and RSS Feeds

Contact Us

© 2023 American Association for the Advancement of Science. All rights reserved. AAAS is a partner of HINARI, AGORA, OARE, CHORUS, CLOCKSS, CrossRef
and COUNTER. Science ISSN 0036-8075.

Terms of Service | Privacy Policy | Accessibility

-