The effect of colors of e-commerce websites on consumer mood, memorization and buying intention

HAL Id: hal-04027333

https://hal.science/hal-04027333

Submitted on 13 Mar 2023

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of sci-

entic research documents, whether they are pub-

lished or not. The documents may come from

teaching and research institutions in France or

abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est

destinée au dépôt et à la diusion de documents

scientiques de niveau recherche, publiés ou non,

émanant des établissements d’enseignement et de

recherche français ou étrangers, des laboratoires

publics ou privés.

Distributed under a Creative Commons Attribution - NonCommercial - NoDerivatives 4.0

International License

The eect of colors of e-commerce websites on consumer

mood, memorization and buying intention

Jean-Éric Pelet, Panagiota Papadopoulou

To cite this version:

Jean-Éric Pelet, Panagiota Papadopoulou. The eect of colors of e-commerce websites on consumer

mood, memorization and buying intention. European Journal of Information Systems, 2012, 21 (4),

pp.438-467. �10.1057/ejis.2012.17�. �hal-04027333�

THE EFFECT OF COLORS OF E-COMMERCE WEBSITES ON

CONSUMER MOOD, MEMORIZATION AND BUYING

INTENTION

Jean-Éric Pelet

& Panagiota Papadopoulou

Nantes University, Nantes, France,

National and Kapodistrian University of Athens, Athens, Greece

Abstract

This paper aims at studying the effect of the colors of e-commerce websites on consumer mood,

memorization and buying intention. Based on a literature review a conceptual model is proposed,

showing the effects of the color of e-commerce websites and specifically of its components, hue and

brightness on the behavioral responses of the consumer, memorization and buying intention. These

responses are conveyed by mood. Data collection was carried out during a laboratory experiment in

order to control for the measurement of the colored appearance of e-commerce websites. Participants

visited one of the 8 versions of a website designed for the research, selling music CDs. Data analysis

using ANOVA, regressions and general linear models (GLM), show a significant effect of color on

memorization, conveyed by mood. The interaction of hue and brightness, using chromatic colors for the

background and foreground supports memorization and buying intention, when contrast is based on low

brightness. A negative mood infers better memorization but a decreasing buying intention. Implications

for theory and practice are discussed.

Keywords: color, consumer behavior, e-commerce, web design, mood, memorization.

0BINTRODUCTION

E-commerce website interfaces seek to entice consumers in a buying intention and manifest a buying

behavior, by activating their senses, specifically their sight or hearing. The perception of a website

atmosphere lies almost exclusively in its visual aspect since 80% of the information processed by Internet

user’s brain comes from sight (Mattelart, 1996). Color constitutes an important sight stimulus for online

consumers, since it is a key website characteristic, associated with the information displayed as well as

with the overall website aesthetics. As such, color is deemed as a significant website factor, positively

influencing the frequency of a consumer visiting a website (Lemoine, 2008) and affecting online shopper

responses (Eroglu et al., 2001, 2003).

Although color is a widely researched topic in various fields (Divard and Urien, 2001), to this day there

is a lack of studies focusing on color in the online context. Color in websites has been studied within

information systems, especially in human-computer interaction, usability and e-commerce, recognized as

a fundamental aspect in web interface design (Lee and Koubek, 2010; Wu et al., 2008; Coursaris et al.,

2008; Kang and Corbitt, 2001). Research has found color to be an important factor in e-commerce,

influencing website aesthetics (Agarwal and Hedge, 2008; Coursaris et al., 2008; Schmidt and Liu, 2005)

e-retailer perceptions (Agarwal and Hedge, 2008), user preference for e-commerce websites (Lee and

Koubek, 2010). Studies are largely associated with the impact of colors on website readability, offering

recommendations about how to choose the most harmonious colors (Hill and Scharff, 1997; Hall and

Hanna, 2003; Nielsen (2000). Yet, color is omnipresent on e-commerce websites. Aware of the

significant and widely known impact of the atmosphere inside stores on consumer activities and behavior

in a traditional buying situation (Kotler, 1973; Donovan and Rossiter, 1982; Filser, 1994, 2003a, 2003b;

Lemoine, 2003), there is a need to investigate the effects of colors as a component of e-commerce

interfaces, on online consumer behavior.

Color has always been used by human beings as an aid to recognize important information among other.

In addition, it can aid an individual’s memory in retaining and recalling information in many activities,

including education or purchases. Similarly, in the online context, the color of an e-commerce website

can possibly improve consumer memorization of information presented in the website.

With the large amount of information presented on e-commerce websites, memorization becomes an

important factor for buying online since consumers are often facilitated in their purchases when they can

retain information from one page to another. This implies that memorization of information in an e-

commerce website may have an impact on consumer buying intention and can potentially be facilitated

by the website colors. However, the relationship between memorization and purchase intention online

has not been investigated. In addition, there is a lack of research regarding color and its effect on

memorization and buying intention in e-commerce websites.

To address this gap, the aim of this paper is to examine how the colors of an e-commerce website can

help consumers memorize information so as to end up buying on the website. The paper presents an

empirical study of the effects of e-commerce website color on the memorization of product information

and buying intention. Our research method includes both a qualitative and a quantitative part. Unlike

most empirical studies dealing with color by comparing warm and cold colors, we examine color by

focusing on its hue, brightness and saturation, following the recommendations of Gorn et al. (2004), so as

to demonstrate that its influence varies according to the intensity of each of these three components. Our

findings show that the colors used on an Internet website have a positive effect on memorization of

product information and buying intention, which is also mediated by mood. They also show that mood

acts as a mediating variable for the effect of colors on memorization.

The structure of the paper is as follows. In the next section a literature review on color and web interface

aspects is provided. The following section presents the research model and hypotheses. The empirical

testing of the model, including the exploratory qualitative study followed by a quantitative study is

described next. The section that follows presents our results which are subsequently discussed in the next

section. The paper ends with the conclusions, implications for theory and practice, limitations and future

research.

1BCOLOR AFFECTS MOOD AND ONLINE PURCHASE DECISIONS

Color contains three principal components (Trouvé, 1999):

• The hue (or chromatic tonality), which is the attribute of the visual sense defined according to the

colors denominations such as blue, green, red;

• The saturation, which provides the proportion of chromatically pure color contained into the total

sense;

• The brightness, which corresponds to the component according to which a surface illuminated by a

source seems to emit more or less light.

To this day, the effects of the three color components on the Internet have been but seldom documented.

In the offline environment, Bellizzi and Hite (1992), Dunn (1992), Drugeon-Lichtlé (1996) and Pantin-

Sohier (2004) chose hue as the main variable in their experiments and showed that brightness and

saturation should be taken into consideration when conducting experiments about color. As Valdez and

Mehrabian (1994), Drugeon-Lichtlé (2009), Camgöz et al. (2002) and Gorn et al. (2004) show about the

brightness component of color, an experiment involving color should compare hue and brightness rather

than warm and cold colors in order to understand what consumers recall and what influences their buying

intention.

On a website, the interface represents a graphic chart, which refers to a collection of website elements. A

graphic chart includes two colors, the foreground color and the background color, both of which

constitute the color scheme. These colors reveal the contrast, which corresponds to a strong opposition

between the foreground and the background colors, as defined by W3C (W3C, 2008). Its main function

relies on facilitating the readability of the displayed information, and a fortiori the memorization process.

A summary of the main studies on color in a computerized context is presented in Table 1.

Table 1 : Studies on computer interface color and its effect on other variables

Critique

Color as part of

aesthetic

aspect, not

specific

variable

Cold vs. warm

colors were used

rather than the

components of

colors

The sites were

identical in

content varying

only in color,

balance, or

combination of

the two.

Experiment

based on

aesthetics, not

on memorization

or buying

intention

process.

Results

(1) pre-use usability and task

completion time were correlated;

(2) the relationship between pre

use

usability and preference was greater

an that of task completion time

and preference; (3) design attribute

assessments after actual use were

highly intercorrelated; and (4)

organizational structure and layout

had a greater effect on user

preference than aesthetic aspects,

such as color and typ

ography

More favorable perceptions

regarding a website design’ s

aesthetics when cool color

combinations (blue-

light blue), as

opposed to warm color

combinations (red

orange), are

used.

This study focused on the role

aesthetics play in website usability.

rceived usability was measured

as the design principles of color and

balance were manipulated. No

statistical differences in user

satisfaction between the four sites.

13 among 23 factors showed

differences based on subjects

related to navigation, language

used

and visual content

Experiment

Experiments on nine online

bookstore websites with ten

participants

A 2 x 2 between

subject

research design manipulates

the temperature of a

Website’ s primary and

secondary colors.

Viewing screenshots of four

homepages

nalysis of 23 factors intended

to compare the results

according to the gender

Scales

Objective performance measures

(task completion time), evaluation

of the design factors, and decisions

of preference on the websites

point Likert scales (anchored

“Stro

ngly Disagree/Agree” )

measured responses to the question

“My perception of this Website is

that it is…”

for each of the

following items: clean, clear,

symmetric, aesthetic, pleasant for

classical aesthetics, original,

creative, fascinating, sophisticated,

and uses special effects for

expressive aesthetics

A usability checklist was developed

by selecting 36 guidelines for web

design and usability (Koyani

et al.

2006)

Interviews

Independent

variable

Color (Aesthetic

aspects)

Color temperature

and gender

lor and balance

Internet sites Color,

Gender

Dependent

variable

User preference

Perceptions of

Website

aesthetics

Perceived

usability

Aesthetics

Reference

Lee and Koubek

(2010)

Coursaris

et al.

(2008)

Brady and

Phillips, (2003)

Moss

et al.

(2006)

ritique

The experimentation

does not take the

memorization process

into account

Color combinations

rather than color

components were used

Interesting experience

to prepare the

laboratory conditions

of our experimentation.

No variable related to

the psychological

aspects of the

respondent was

measured.

Work on time, nothing

on memorization of

information and on

buying intention

Results

Moderate or even high color

contrast does not guarantee quick

visual perception. With black and

white information, the spee

d of

visual perception decreases with

decreasing contrast. Visual

search times, number of eye

fixations, and mean fixation

durations increased strongly with

decreasing luminance contrast

despite the presence of color

contrast.

The speed of reading text in

different color combinations

cannot be described as one

dimensional problem.

Mathematical metrics were

mostly in contradiction with the

judgment of the observers

Red background infers a

perceived duration longer than a

blue one

Experiment

Eye movements d

uring the visual

search experiment were recorded

simultaneously with threshold

search time measurement by

using an SMI (Sensomotoric

Instruments Inc.)

Testing 56 color combinations

and identifying 21 uppercase

alphabetic characters, selected

and presented

in conformance

with the Snellen chart

Observ

tions

;

quantification of treatments of

the digital image color (e.g.

contrast, brightness)

;

55 observers making

comparative tests and tests of

absolute measure around various

configurations of 8 colors

Time perception

: Students

placed in front of a program and

to whom questions were asked

concerning the program and the

execution of certain tasks

Scales

Content analysis

ANOVA, t

test to study

the differences in legibility

among pairs of color

combinations

None

3 items of speed

Independent

variable

Luminance contrast

Color combination of

a text color and

background color

Colors and colored

appearances of

compared images

following various

compression levels

Foreground and

background color

Dependent

variab

Speed of visual

search and reading

Speed of reading

Number of correctly

identified characters

as a measure of

legibility

performance

Appearance of

compressed images

Comparative tests

(organized, forced

choice) and absolute

measure tests were

used

Per

ceived download

speed

Reference

Ojanpää and Näsänen,

(2003)

Humar

et al.

, (2008)

Fernandez

Maloigne,

(2004)

Gorn et al.

, (2004)

Critique

Cold vs. warm

colors were

used rather

than the

components of

colors

Nothing about

the buying

intention

variable

ld vs. warm

colors were

used rather

than the

components of

colors

Only a few

colors were

used for this

experiment, it

is not easily

generalizable.

Results

Music and color factors had a significant effect on

participants’ emotional response Participants fe

more aroused and experienced greater pleasure

when they were exposed to a warm color website.

Absence of significant effect of peripheral color and

its constituents

It seems to be preferable to use more than four

colors or monochromatic and analogical

It seems to be preferable to use low contrast levels

between colors

Cold colors are perceived as more suitable than

warm ones

For monochromatic color scheme, it is preferable

to have high contrast in brightness

Regarding saturation, it was found that the b

est

combination is high saturated background color

with low saturated secondary color.

1) Time passes more slowly with white, blue,

green backgrounds

2) No effect on color on time pass

A white background allows easier reading but a

slower search and more d

ifficult co

prehension

Experiment

Laboratory experiment with a 2

(music: fast/slow) x 2 (color:

warm/cool) between-

subjects

factorial design

Self

-estimation of the duration

of their own test;

Study of various personality

traits in interaction with com

plex

mental processes such as

attention capture, reading,

semantic memorization

A machine learning algorithm

generates a network that relates

the color model of a website

with the emotional values that

are attributed to it by its users

1) Effect of 5 backg

round

colors (white, red, green, blue,

yellow) on 50 subjects who had

to search for information in a

directory

2) Search and

understanding

Scales

item semantic

differential scale

(Mehrabian and

Russell’ s, 1974)

modified for online

shopping to measure

participants’

emotional response to

the surroundings in

terms with a 7

point

Likert scale, Arousal

dimension: happy

–

unhappy, pleased–

annoyed, satisfied

–

unsatisfied,

contented

–

melancholic, hopeful

–

despairing, and relax

–

bored.

SSS version V of

Zuckerman

(1994)

(opt

mal level of

stimulation)

Locus of control

(LOC) (Rotter, 1966)

Androgynous of Bem

(BSRI) (Bem, 1974)

Confidence level (9

point Likert scale)

Bayesian Belief

Network (BBN)

Time error, Response

time, Devotion,

Interest, Irritation,

Fatigu

e, Achievement,

Difficulty

Independent

variable

Website music and

color

Website colors

Color

characteristics

Website colors

Dependent

variable

Emotional responses

and subsequent

shopping behavior

Sense

motivation

Mouse manipulation

Cognitive

memorization

Explanatory

: sex,

initial mood, opt

mal

level of stimulation

Affective state (12

most emotional

descriptors: Pleasant,

Formal, Fresh,

Modern, Friendly,

Aggressive,

Professional,

Attractive, Calming,

Dynamic, Reliable and

Sophisticated)

Tim

e spent on a

website

Time spent for

searching information

Understanding

found information

Reference

et al.

(2008)

Roullet, (2004)

Papachristos,

et al.

(2005)

Kiritani and

rai, (2003)

Critique

No measure of

memorization

Testability of a colo

vision screening test in

a population with

mental retardation, no

measure of

memorization

Discriminating abi

ity

measures for predicting

rea

ability of text on

textured backgrounds,

no measure of

memorization

Testing the rea

ability

of web page colors, n

measure of

memorization

Results

Both experiments showed

significant main effects for

all variables, and that all

interactions were

significant

The overall rate of

testability was 93.2% for

the 1078 athletes screened.

The frequency of males

identified a

s color

deficient was similar to

that expected in the

general population; only

two females (in Spain)

failed the color vision

screening.

Search times indicate that

these background

variations only affect

readability when the text

contrast is low, and that

spatial frequency content

of the background affects

readability.

Readability increases as

the difference of clarity

between the text color and

the background color

increases.

Regardless of the direction

of this difference, clarity is

an important factor f

determining the readability

of a text on a colored

background

Experiment

Experiment I: Measuring the effects of

transparent text:

It employed a 2 (text

transparency type) x 2 (text contrast) x 3

(background) within participants design

Experiment II: Measuring the effects of

very low contrast one color combination

(gray on gray) and one

background luminance level

Readability / Mentally handicapped

persons

Readability / Screen background with

texture

149 subjects and 42 samples

Images having different te

xt colors on

different background colors. It was

necessary to note the readability of

images.

Images in GIF format (modification of

text impossi

ble) using the 216 safe colors.

Scales

Text Contrast, Adjusted

Text, Contrast,

Background, RMS

Contrast, Adjus

ted

RMS, Contrast, Global

Masking, Index,

Adjusted Index

Testability of the

"Color Vision Testing

Made Easy" color

vision test and with a

test using simple

geometric figures that

are easily identified

Several

discriminability

measures were

examined (plain,

periodic texture, and

four spatial

frequency

filtered textures

created from the

periodic texture).

Scale ranging from

“impossible to read”

to “ readable without

effort

» presents on

screen

Independent

variable

Text color with different

modes

Transpa

rent

Opaque

Weak contrast (with

screen background)

Colors

Text color and screen

background color

Text color and image

background color

Dependent

variable

Text readability

Effects of color

readability on

subjects with

mental

deficiencies

Text readab

ility

Readability of

images containing

text

Reference

Scharff and Ahumada,

(2002)

Erickson and Block,

(1999)

Scharff and Hill,

(2000)

Ridpath

et al.

, (2000)

Critique

Readability of websites

with various

for

ground/back-

ground color

combin

tions, no

memo

rization process

measured.

Background color only

Comparison based on

chromatic vs. a

chromatic colors only

Color and pattern were

measured rather than

colors only

Comparison of graphic

charts with no

modification of the

levels of brightness

and saturation.

Results

In general these results

suggest that there is no

one

foreground/background

combination, font, or

word style which leads

to the fastest RT (i.e.

best readability), but

rather a designer must

consider how each

variable affects the

other(s).

Subtle

design

manipulations had

significant effects on

consumer evaluations of

web page aesthetics and

perceptions of the e

retailer

Women tend to use

color in their histogram

choice more than men

and color seems less

important for men

Subjects mention in

advanc

e the attractive

products, the picture of

which kept their

attention more than the

rest

Elimination of different

quite incoherent types of

memorization (retention

due to aesthetics and not

always to contrast)

Experiment

The effects of 6

foreground/backgro

und color

combinations (color), 3 font types

(Arial, Courier New,

& Times

New Roman), and 2 word styles

(Italicized & Plain) on readability

of websites were investigated.

Participants (N=42) scanned

simulated websites for a target

word; readability was inf

erred

from reaction time (RT).

Conjoint analysis and optimal

design methodologies sixteen web

page prototypes assessed through

an online survey

93 subjects have to measure the

complexity of color and black &

white histogram information

385 subjects have to allocate 100

points according to their

preferences; A

ministration of

online questionnaire

Readability test of 136 partic

pants

of 2 web pages with chromatic

manip

lation

multiple choice questionnaire

Scales

Readability / Text

color and screen

background color

Background

color, white

space, thumbnail

image location

and thumbnail

image size were

varied

Likert scale

indicating their

confidence in

their choice

point Likert

scale on the

importance of

security and price

point Likert

scales

Independe

variable

Foreground and background

colors

Background color

Graphics color

(histograms)

Appearance / choice

according to the color of the

screen background of

attractive products (or not?)

Car

: 1) orange red & 2)

green with dollars.

Sofa

: 1) blue w

ith clouds &

2) green with cents

Black text /White

background

White text /Black

background

Light blue text /Dark Black

background

quoise text/Black

background

Dependent

variable

Reading speed on screen

Aesthetic evaluation and

perception of e

retailer

consumer trust product

preference and purchase

intention

Decision performance and

accuracy

Gender, decision

making,

task analysis

Purchase intention

Reading speed

Information memorization

after reading

Reference

Hill and Scharff,

(1997)

Agar

wal and Hedge,

(2008)

So and Smith, (2002)

Mandel and Johnson,

(2002)

Hall and Hanna,

(2003)

Gorn et al., (2004) focusing on the impact of the three color components on downloading time perception

demonstrate that a lengthy waiting time influences the user’s appraisal of the Internet site and can lessen

his/her desire to recommend it to others. Kiritani and Shirai (2003) show that the effects of screen

background colors on time perception vary according to the tasks performed by Internet users. When

reading a text written on a white, blue or green screen background, users have the feeling that time passes

more slowly. When users merely conduct a simple search and only need to understand the meaning of a

sentence, then the screen background color does not have any impact on how they perceive time duration.

Hill and Scharff (1997) have demonstrated the importance of contrast (foreground color vs. background

color) when searching for information within a page. They obtained better readability scores when

resorting to chromatic colors (green foreground color on yellow background color). During an

experiment where colored labels had been placed on screen backgrounds, Camgöz et al. (2002) observed

that brightness, saturation and hue had a specific impact on each colored screen background.

Biers and Richards (2002) have studied the impact of background color on the perception of promoted

products and found that backgrounds with cold hues, such as blue, increased product value and reduced

the risk of purchase postponement, especially with regards to regular Internet users. Hall and Hanna

(2003) studied the impact of background and foreground colors on how readability was perceived and

aesthetic aspect experienced, as well as on the retention of information and on intentions. According to

them, sites promoting knowledge transfer must display black texts on white backgrounds, achromatic

colors with maximum contrast. In addition, they indicate that e-commerce websites should merely use

chromatic colors due to the higher aesthetic appreciation score which is correlated to higher purchase

intention. Blue is the favorite hue for purchase intention. These results underline that when studying

color on a website, it is important to take into consideration color components (hue, brightness and

saturation), as well as the contrasts of the foreground and background colors.

A number of studies have shown that color has a positive effect on consumer mood (Wu et al., 2008) and

buying intention (Wu et al., 2008 ; Roullet, 2004). However, color has mainly been addressed in terms of

warm and cold hues (Coursaris et al., 2008; Papachristos et al., 2005) and has not examined with respect

to its components, hue, brightness and saturation. Most of the studies linking color and e-commerce take

into account balance or brightness as variables of colors (Brady and Phillips, 2003) or combinations of

colors (Humar et al., 2008), which do not allow for comparing the effects of the components of the

colors. Memorization doesn’t seem to be recognized as an important variable in e-commerce considering

the poor amount of research on this topic. For example, Hamilton and Luo (1999) have shown that the

degree of complexity of an e-commerce website varied according to the colors of the interface which

helped the consumer to be concentrated. However, many studies on the effects of color on readability

exist (Hill and Scharff, 1997, 1999; Hall and Hanna, 2003). In addition, research has not studied how

memorization affects buying intention.

2BRESEARCH MODEL

The proposed model follows the general pattern of consumer behavior. It follows the one from Engel et

al. (1978) and introduces three innovative contributions in the consumer behavior analysis (Filser, 1994).

It analyzes the variables that influence the consumer's decision process by distinguishing three

categories:

• Characteristics of the individual;

• Characteristics of its social environment;

• Situational factors.

This model offers precision about the different stages of the perception of stimuli process, which

comprise exposure, attention, understanding, acceptance and retention. It suggests measures of these

different levels of perception. It finally breaks down the decision process into a sequence of five steps.

The latter are commonly used in research on consumer behavior. The model explains how the colors of

an e-commerce website and specifically their components - hue, brightness and saturation - can have an

impact on the buyer’s affective state of mood and cognitive states of memorization and buying intention

(Figure 1).

Figure 1: Research model

9BMemorization

Memorization is a very important factor for the large number of information-based websites that

currently exist. It is important for e-learning applications, since the user goal is usually to retain the

information beyond the time the page is being read. This also applies to information included in e-

commerce websites, since consumer tasks are often facilitated by memorizing information while

navigating. Drawing on offline settings, we argue that memorization can be influenced by the colors of

an e-commerce website.

In order to understand the effects of color on consumer memorization we have to take into account the

quality and quantity of information a consumer has memorized while visiting an e-commerce website.

We posit that memorization varies according to the colors of the website, and especially according to the

contrast between background and text colors, in agreement with the work of Hall and Hanna (2003). As

we stated earlier, the aim of this research is to investigate the effects of the components of colors rather

than colors themselves.

In general, information is stored according to an encoding process enabling one to sort out information

thanks to criteria which will then allow one to retrieve this information. The role of these criteria is to

connect a piece of information to other similar information already stored (Ladwein, 1999). In order to

examine the information memorized by each participant, we resort to recognition and recall, two

procedures belonging to a method of information retrieval based on overall stimulus in long-term

memory. Be it free or cued, recall enables individuals to mimic mentally a stimulus to which they are not

exposed during the evocation, for instance, their past reaction to a promotional action (Filser, 1994).

Thus, we can hypothesize:

H1: The components (hue, brightness, saturation) of colors of an e-commerce website will have a

positive effect on memorization

10BBuying intention

Intention is activated by a desire or a need (Darpy, 1997) and desire is viewed as an active process

(O'Shaughnessy, 1992). Although buying intention is more than a mere desire, it is not a promise to buy

(O'Shaughnessy, 1992), it is the outcome of a cognitively handled desire. According to Darpy (1997),

building on the studies of O'Shaughnessy (1992), Howard (1994) and Belk (1985), “Intention is the result

of a desire or a need handled on the cognitive level and leading to purchase planning”.

Among the environmental factors recognized to produce important emotional and behavioral reactions on

the consumer, color seems to play an important role. It serves to retain consumers longer on the e-

commerce website according to certain criteria related to their perception of the interface. In particular,

pleasure is increased with use of colors whereas boredom can result from a weak use of them (Lemoine,

2008). This duration can help maintaining user interest in a site (Bucklin and Sismeiro, 2003, Hanson,

2000) and give users more time to consider and complete purchase transactions (Bucklin and Sismeiro,

2003). By enhancing consumer interest, it helps to generate repeat visits, which lead to greater long-term

sales (Moe and Fader, 2004). From a business investment point of view, Demers and Lev (2001) show

that sites with longer visit duration also have higher monthly stock returns. Therefore, it can be assumed

that e-commerce website colors are likely to have an impact on buying intention, as they can prolong the

visit duration. According to Agarwal and Hedge (2008), the background color of e-commerce web pages

is an important factor affecting purchase intention. Wu et al., (2008) have found that warm colors of e-

commerce websites have a positive effect on purchase intention. As already mentioned, we are interested

in the effect of color components, hue, brightness and saturation. Therefore, we propose:

H2: The components (hue, brightness, saturation) of color of an e-commerce website will have a positive

effect on consumer buying intention

There are many entries which are available in the memory and in the external environment. They can

potentially be considered in the decision task, but only a few will be used to make a choice in a given

situation. Tactical choices effectively originate from decision making regarding the products we buy,

including:

• Considerations linked to the price (cheaper, use less of it, cost a cheaper price);

• Considerations linked to the performance (the product functions in these conditions, it owns

these qualities);

• Considerations linked to the affect (I like the product, I love the product);

• Normative considerations (my father advised me to buy it, my mother always uses this product);

It is important to understand the procedures which determine which part from memorized information

can be used for making a choice. For these reason, we propose:

H3: Memorization will have a positive effect on consumer buying intention

11BMood, a mediating variable

We wish to bring to the fore the effects of colors on affect, which includes mood experienced when

visiting the e-commerce website. Mood refers to affective states of mind less likely to reach our

conscience. Moreover they last longer than emotions but are less intense (Forgeas, 1999).

Mood is linked to a color perceived as positive or negative according to the personal experience of an

individual with a color (Boyatzis and Varghese, 1993). According to Odom and Sholtz (2004), different

colors tend to incur different moods. Studies have demonstrated the association of colors and mood by

using diverse methods such as the objective impressions (printings), the clinical observations, the

introspection and the experimental investigations (Wexner, 1954). Chebat and Morrin (2006) measured

the effects of cold vs. warm colors of a mall decoration on consumer perceptions. They showed that these

were mostly driven by affective mechanisms such as mood, or by other cognitive states, such as the

evaluation of the mall environment quality. Similar mechanisms can exist in an online context. Wu et al.

(2008) have found that e-commerce website color has an important effect on consumer mood. However,

their study focuses on warm and cold hues of website color and their influence on mood and does not

examine the hue, brightness and saturation color components. In this study, we investigate the direct and

interaction effects of the components of colors. Hence, we suggest the following hypotheses:

H4: The components (hue, brightness, saturation) of color of an e-commerce website will have a positive

effect on consumer mood

Mood is generally considered as a mild affective state that may influence cognitive processes such as

evaluation, memory and decision strategies (Gardner, 1985). However, the observed effects of negative

moods have been less consistent than those of positive moods. For example, Cialdini et al.’s (1973)

negative state relief model of helping asserts that people in a negative mood will behave more charitably

than others if the opportunity has potential for direct social or egoistic approval, suggesting that helping

behavior may be quite a complex phenomenon not fully addressed by simpler explanations such as mood

states (Swinyard, 1993). Gardner (1985) observed that the effects of mood may have special impact in

retail or service encounters because of their interpersonal or dyadic nature, a view also supported by

others (Isen et al., 1978; Westbrook, 1980).

H5: Consumer mood will have a positive effect on memorization

When people are in a good mood, they tend to have more favorable expectations for the future (Eysenck,

1976; Masters and Wyndol, 1976). Drawing on research in offline stores, good mood may contribute to

consumers intention to return to a website at which they have made previous purchases (Isen et al.,

1978). We expect mood to be important in the context of online purchase since many consumers make

purchases after having a pleasant time visiting a website. This is in line with the study of Wu et al. (2008)

suggesting mood as a significant predictor of purchase intention. Thus we propose:

H6: Consumer mood will have a positive effect on buying intention

3BRESEARCH METHOD

Our research method includes both a qualitative and a quantitative study. An exploratory qualitative

study was conducted first to allow for verifying the importance of the research variables and the necessity

of including them in our model to be tested. The proposed research hypotheses were then empirically

tested through a quantitative study conducted in a laboratory setting.

Qualitative study

The main objective of the exploratory phase was to investigate the empirical knowledge gained by

regular consumers and expert users when browsing e-commerce websites. It mainly sought to confirm

that colors have an impact on users’ perceptions, so as to prepare our quantitative study for data

collection. In this direction, the study explored long-term memory and the website factors that influence

it, drawing from Eroglu et al., (2003) suggesting that atmospheric cues affect consumer behaviour. The

study was based on semi-structured interviews conducted with regular consumers and web designers,

where we asked interviewees to speak about past visits to websites of their choice. The interview guide

and the specific research objectives of the questions are shown in Table 2. From these interviews, topics

referring to the affective states lived by the consumer in an online shopping situation emerged. These

topics relate to the emotions and moods and show the importance attached by the consumers to the ease-

of–use of a website. They also reinforce the proposed effects of variables such as color, as well as the

quality of the images perceived by the consumers.

Table 2: Interview guide questions used for data collection and relationship to research objectives

Question

Research objective

1. Phase of introduction: use of Internet in general:

a. Could you speak to me about your use of Internet?

b. Do you often use the Internet for shopping?

Learning about the consumption and

familiarity’s experience of the consumer.

2. Phase of centering the subject:

Investigation into the long term memory in

a. Now I would like you to remember your last visits to e-

commerce websites to carry out activities for shopping on

the Internet. By e-commerce websites, I understand the sites

which offer products like books, CDs of music, DVDs, cars,

travels, hotels, flights and train booking tickets, banking

services, etc. Can you remember your visit to a particular

website to seek information on a product or to buy this

product?

b. What do you think of this site in general ?

order to learn what is remembered by

consumers after their visit to an e-commerce

website.

The consumer’s thinking is also interesting to

assess since we want to understand what is

related with what is memorized.

3. Phase of deepening

Topic 1: atmospheric elements of the commercial sites

a. What do you think of the design of this website?

b. What represents a commercial website which you have

found rich, captivating and pleasant to visit?

Topic 2: Emotions and feelings felt following the

consultation of a commercial site

a. Could you describe the feelings that you associate with an e-

commerce website?

b. user called his friend:

“It sometimes happens, that while surfing on an e-commerce

website, the site holds my attention because it is beautiful,

and easy to surf and to consult. It enables me to go quickly

because I easily find the links which are of interest to me and

I never feel lost when I use it. Links are easy to locate and

help me when I am lost. The facility of reading the site gives

me desire to spend even more time on it: the links are quite

visible, the images are very clean, and I can perfectly

distinguish such and such part of the page of a site, because

the colors are used in an intelligent way, to delimit the

parts”.

c. (To be read again several times if necessary)

Your opinion?

a. Do you think you have already experienced these states?

b. If so, could you speak to me about an experience of this kind

by describing an e-commerce website for example?

c. According to you, why did you live these states?

Topic 3: Antecedents of the behavioral approach

a. What encourages you to buy on a particular website and not

another?

b. According to you what makes you spend more time on one

website than on another?

c. In your opinion what makes you switch to an e-commerce

website?

d. What are the factors which encourage you or discourage you

from revisiting a particular website?

The aspects in terms of design counts in the

consumer’s appreciation. But at which level ?

Qualitative adjectives (rich, captivating…) can

enhance the consumer’s description of a

commercial website, as a way to help him.

The vocabulary dedicated to feelings

contributes to feed our understanding of

aspects linked to the perception.

The projective technique is helpful since

consumers do not have to invent anything.

They can try to remember a past navigation,

which seemed close to a perfect one in order to

compare it with a normal navigation on an e-

commerce website.

The interview is then based on the

interviewee’s perception after he has visited

such a website. He is asked to describe this

period in order to relate particular lived states.

The reason why he experienced these states is

then questioned with the objective to elicit

sentences dedicated to emotional states.

The question we asked in order to understand

behavior when facing an e-commerce website

targets the consumer’s loyalty. Questions were

thus oriented towards the ease of use and ease

of surfing the website as well as questions

linked to ergonomic and interest to the website

aspects.

Phase of conclusion

What makes you say, after having visited an e-commerce

website: “I have visited a really good website which inspires

me to buy from it and to explore it more”?

Can you think of any ideal e-commerce website?

During the conclusion phase, we tried to get

ideas related to the best conditions participants

would like to have experience when shopping

on the Internet.

An emphasis on an ideal shopping experience

on the Internet was then explored.

Participants

Participants were online shoppers with different levels of expertise in terms of Internet use. Sample

selection was primarily based on qualitative criteria. Since the qualitative study aimed at exploring online

consumer perceptions from e-commerce websites, our sample had to have previous experience with

online shopping. In order to qualify for participation in the study, a respondent would have to reply

positively to the question “Have you ever bought a product or service from an e-commerce website?”.

Twenty one subjects, 10 females and 11 males were interviewed. Participants were grouped according to

their expertise with e-commerce websites into expert users and regular users. A participant was selected

as an expert user or not based on the answer to the question “Are you a professional web designer?”.

Subjects characterised as expert users represented 29% of the sample, whereas most of them were regular

Internet users. Finally, the sample was selected pursuing balance in terms of age and socio-professional

background. The sample characteristics are presented in Table 3.

Table 3: Characteristics of the sample of respondents

Demographic

characteristics

Category

Number of

respondents

Gender

Female

Male

160

136

Age

Under 25 years old

25 - 34 years old

35 - 44 years old

45 - 54 years old

55 - 64 years old

242

Internet

familiarity

Expert

Intermediate

Beginner

Education level

Certificate

High school

2nd-year university diploma

License / Master I /Master II

Postgraduate / doctorate / post-graduate / Master degree

no diploma

166

Profession

Executives and academics

Intermediate occupations

Manuel Workers

Students

Storekeepers and business managers

Pensioners

Other (to clarify)

244

Income per

month

From 0 to 500 euros

Under 1000 euros

From 1000 to 1399 euros

From 1400 to 1799 euros

From 1800 to 2199 euros

From 2200 to 2599 euros

247

The use of students as subjects has often been questioned in terms of their appropriateness as a sample.

However, in our study, students are deemed suitable as a sample, as they share many characteristics with

the profile of Internet users population, such as age. As shown by several studies, Internet users tend to

be young adults, while the Internet usage penetration within the age groups of 18-29 raises up to 95%

(Zickuhr, 2010; Pew Research Center, 2010). Hence, although our sample presents a bias towards

younger subjects, it can arguably be acceptable as representative of Internet users. In addition, our study

benefits from the use of students since they are considered as an important group of online consumers

(Delafrooz et al., 2010) and are useful as a sample for empirical studies in e-commerce, in line with

previous research (e.g. Kim et al., 2008).

Our sample would also arguably be stronger in terms of representativeness if it was cross cultural.

However, although the population of respondents was primarily native Continental French, it included

foreign students, with different cultures, as in any school or university settings. In this sense, our

sampling frame does not imply a serious threat to the validity and generalizability of our results.

However, a replication of the experiment with a more culturally diverse sample would allow for richer

data collection and findings.

13BExperiment design

Carrying out the experiment under laboratory conditions allows us to draw valid conclusions about the

groups surveyed (Jolibert and Jourdan, 2006). Internet enables one to conduct non-intrusive studies,

meaning that Internet users are not even aware that their behavior is being analyzed (Dreze and Zufryden,

1997). However, when conducting a study focusing on color, one has to control and neutralize three

major elements: screens, ambient light, and, above all, the participants’ color perception (Fernandez-

Maloigne, 2004). Since, these elements cannot be controlled in a distance study carried out over the

Internet, a controlled laboratory setting had to be used for our study. Screens were calibrated using a

probe to ensure that the colors were displayed as defined. The color of walls was neutral grey. The grey

color of the wall prevents a bad reflection of ambient light on the screens, which avoids color distortion

on the screen. By neutralizing the colors of both the ambient light and the walls we were confident that

we avoided any distortion. The brightness of the room was then set at 1000 lux to guarantee that the

colored appearance of the websites used for the experiment would not be biased by a too dim or a too

bright room light. Finally participants had to pass the Ishihara test for color blindness. Further, detailed

information how each of the three elements was controlled can be found in Appendix 1.

The experiment was a factorial design which included 8 treatments (4 x 2) related to 8 color schemes

developed for the experiment website (Table 6). In order to measure the differences in color perception,

we have created 8 different color schemes by varying hue and brightness and keeping the saturation rate

constant among the schemes. In accordance with Gorn et al. (2004), we set the saturation levels at 100%,

because at that level, the hues are most distinct. The color stimuli were modified in accordance with

Munsell’s (1969) system, considered as the most accurate one (Aumont, 1994), which enables defining

precisely several levels of brightness and saturation for each hue.

Several studies on the impact of culture on the behavior of Internet users and consumers draw attention to

the varying importance of design elements of a website across cultures (Maguire, 2011). Among these,

colors are a significant variable as their perception and impact is strongly influenced by culture. Colors

can convey different semantics depending on the cultural contexts (Filser 1994). Thus, culture is a

parameter that should be controlled in a study on the effects of colors of an e-commerce website. In order

to neutralize this aspect we based our experiment on a color selection driven by research on readability, a

characteristic which is not culture dependent (Hill and Scharff, 1997) .The design of the website was also

not too busy, leaving sufficient room for reading easily the commercial information as well as for

satisfying principles of website ergonomics such as the navigation bar and the search engine visibility.

An easy-to-read text on a screen, thanks to a sufficiently strong contrast between the background and

foreground color, is easier to memorize (Hall and Hanna, 2003; Scharff and Ahumada, 2002, Scharff and

Hill, 2000; Hill and Scharff, 1997). In order to validate that a combination of background and foreground

color was the best for memorization when the contrast was the strongest, we needed to rate each

combination of colors in terms of contrast ratio. We thus measured the latter to provide a gradient scale

between each combination of background and foreground color, which corresponds to a rank with “-“

and “+” symbols in Tables 5 and 6. We first differentiated each color scheme resulting in Table 5, which

was then used to produce Table 6 summarizing all the information required to explain the contrasts

ratios.

The contrast was measured by the Color Contrast Analyser (CCA) 1.1 for Web Pages of the Web

Accessibility Tools Consortium (WAT-C, 2005). Contrast ranks ranged from “+++” to “+”, for strong

ones and from “-“ to “--“ for low ones, according to the CCA. For example, a contrast ratio of 21 was

ranked as “+++” whereas a contrast of 1,82 was ranked as “-- “. CCA was created with the aim to design

easier to read web interfaces, in accordance with the W3C consortium. By taking into account the W3C

and Web Accessibility Initiative (WAI) guidelines, the use of color becomes more professional and the

choices web designers make are more informed in terms of usability, as well as in terms of human

computer interaction in general.

Table 6: Contrast ratio order according to color schemes

Background

Foreground

Contrast ratio

Rank

+++

6,95

6,71

3,26

2,65

1,82

To set our first experiment treatment we used the color scheme used by Hill and Scharff (1997) which

supported the best readability rate in relation to contrast and we chose as chromatic colors a yellow

background (Magnolia yellow) and a green foreground (Newsvine green). Starting from this scheme, we

reduced the brightness level of the two colors so as to obtain the second experiment treatment. For

experiment treatments 3 and 4 we kept the same colors but switched foreground and background colors.

Experiment treatments 5, 6, 7 and 8 are based on black and white (achromatic colors), the most

frequently used colors on e-commerce websites. Brightness and saturation levels were identical with

those used for chromatic colors in experiment treatments 1, 2, 3 and 4 respectively (Table 7).

Table 7: Factorial Design of the Experiment

Plan

Background

Foreground

Contrast

Plans explanations

Name

Magnolia

Yellow

100

Newsvine

Green

120

100

6,95*

Passed at

Level 2

(++)

(Hill and Scharff, 1997) showed that the sharp

contrasts of this scheme offered users the fastest

reading speed possible among chromatic colors.

Hex.

#FFFFD0

#006700

Magnolia

Yellow

100

Granny Apple

Green

100

1,82*

Fail

(--)

Same color scheme as in the Plan 1 with an

increased foreground color brightness (from 40 to

80).

Hex.

#FFFFD0

#6BD500

Newsvine

Green

120

100

Magnolia

Yellow

100

6,95*

Passed at

Level 2

(++)

Same colors as in Plan 1. Foreground and

background colors were switched.

Hex.

#006700

#FFFFD0

Newsvine

Green

120

100

Sunflower

Yellow

100

6,71*

Passed at

Level 2

(++)

Same color’s scheme as in Plan 3 with an increase

in foreground color saturation (from 20 to 60). The

color which should have been used for the text of

the experimental plan 4, in order to preserve rates of

luminosity and saturation in relation to the

background color, could not be preserved. Indeed,

this scheme could not be used given the lack of

contrast between the two colors

(foreground/background) which made the reading

impossible on a more or less old or difficult screen,

for individual presenting deficiencies with color's

vision we refer to the directives of the w3c. We thus

varied its degree of saturation.

Hex.

#006700

#FFFF6B

White

100

Black

21,00*

Passed at

Level 3

(+++)

This scheme is the most widely used one on e-

commerce websites.

Hex.

#FFFFFF

#000000

White

100

Grey

2,65*

Fail

(--)

Same color scheme as Plan 5 with increased

foreground color brightness (from 0 to 60).

Hex.

#FFFFFF

#9C9E9E

Black

White

100

21,00*

Passed at

Level 3

(+++)

Same colors as in Plan 5. Foreground and

background colors have been switched.

Hex.

#000000

#FFFFFF

Black

Grey

3,26*

Fail

(+)

Same scheme as in Plan 7 with a decrease in

foreground color brightness (from 100 to 60).

Hex.

#000000

#9C9E9E

* Text or diagrams and their background must have a luminosity contrast ratio of at least 5:1 for level 2 conformance to

guideline 1.4

Figure 2 shows a screenshot of our experiment website. In this particular example, the background color

is Magnolia Yellow and the foreground color is Newsvine Green representing experiment plan 1. As it

can be seen, the experiment website didn’t have any other colors than the foreground and background

ones. The design was simple, with very limited use of graphics such as the store logotype and the

shopping cart logotype, in order to reveal only the colors selected for the color scheme of each factorial

plan (background, foreground). Screenshots of the experiment website for each of the 8 plans are

provided in Appendix 2.

Figure 2: Experiment website

14BMeasures

26BMemorization

Memorization was measured by measuring recognition, cued recall and free recall.

To measure recognition, participants were asked to recognize two CD covers, each among two other

covers of different albums by the same artist. Recognition scores ranged from 0 to 2, one for each CD

cover they could recognize. Measuring recognition was not deemed useful since the participants

answered to the questionnaires a few minutes after visiting the e-commerce website and 100% of them

recognized both CD covers. Thus, we decided not to include recognition further in our analysis.

Cued recall was measured by asking the respondents to answer to a question with 3 alternative values

(correct, wrong and “I don’t know”) for each of the seven information items related to a CD cover.

Scores could thus be graded from 0 to 7 for each item visited. Since participants were required to check

out two CD covers, scores for cued recall ranged from 0 to 14.

In order to measure free recall, participants were asked to answer to an open-ended question related to an

image about the CD cover they had just seen. The question was “What do you remember from the

information associated with this CD cover?”. Free recall was measured by counting the number of the

items that participants could recall from those used in the CD description. Since participants could see

two CD covers, each having a 20-element description, free recall value ranged from 0 to 40.

The score of commercial information memorization was the sum of the recognition score, cued recall

score and free recall score, ranging from 0 to 56.

27BMood

To measure mood we used Mayer and Gaschke’s (1988) Brief Mood Introspection Scale (BMIS).

Participants were asked to reply to the question “Do you feel _______?” for a list of 16 items rated on a

5-point Likert scale ranging from ‘Definitely do not feel’ (1) to ‘Definitely feel’ (5), and for each of the

16 items select the answer that best expressed their mood. We selected to use BMIS because it provides a

quite exhaustive range of moods and is easy to supervise. The scale and its validity and reliability are

presented in Appendix 3.

Each participant passed a pre-test administrated before the start of the experiment. This pre-test measured

mood, using BMIS, in order to enable linking the affective states of the participants with the website

characteristics. The results we obtained after the visit of the website then refer to the mood experienced

on the experiment e-commerce website.

As an affective state, mood is characterised by its duration and intensity as mentioned earlier. In addition,

it is not possible to link the affective states of participants with the websites characteristics without

measuring the mood before the experiment. We thus used the induction method before the experiment.

Its objective is to induce an artificial and controlled affective state during a short period of time, in order

to make participants attain a state that should be similar to a natural affective state. For deontological

reasons and in order to avoid cognitive failures in the activity of the participants, the intensity of the

induction cannot be too strong. The induction phase is framed by two BMIS questionnaires in order to

assess its effectiveness. Participants were divided into groups and were asked to complete the first BMIS.

The researcher waited until all participants in the group have completed the questionnaire, then, to his

signal, the induction procedure started. The assignment of this phase was as follows: "report a lifetime

experience that reflects a happy event (or sad). This is anonymous; we encourage you to describe

concrete and precise life experiences. You have 10 minutes and three sheets to make your story. Once

finished, fill in the questionnaire and wait until the researcher tells you to continue”. After ten minutes,

the experimenter indicates the end of writing the autobiography and invites participants to start the

questionnaire. Once the participant finishes to complete the BMIS, the experimental phase, visiting the

experiment e-commerce website, starts. After visiting the website, and participants complete the second

BMIS questions among other questions which are part of the questionnaire used for data collection.

Analytical statistics then take into account results of the two BMIS.

28BBuying Intention

We used a four items scale developed by Yoo and Donthu (2001). The items were measured on a 5-point

Likert scale ranging from strongly disagree (1) to strongly agree (5). Already used in a similar context,

the validity and reliability of the scale was good, presented in Appendix 4.

4BDATA ANALYSIS AND RESULTS

Data were analysed using General Linear Model (GLM) and ANOVA. We used GLM to test the effect of

the color components (hue and brightness) on mood, memorization and buying intention, and variance

analyses (ANOVA) to test the significance of the links between variables and the difference among

treatments. The effect of saturation on the dependent variables of our model was not possible to be tested

because, as explained earlier in the experiment design, saturation was set constant in order to facilitate

the measurement of hue and brightness. We also examined interaction effects between hue and brightness

with a series of regressions on each of the dependent variables. Regression analysis was used to test the

effect of mood on memorization and the effect of memorization on buying intention. The results are

analyzed in the following paragraphs and are summarized in Figure 3:

15B

Figure 3: Summary of GLM, ANOVA and regression results

Direct effects of the colors of the color scheme on memorization

The GLM analysis did not show a significant effect of color on cued recall. The results in Table 8

indicate that neither hue (p=0.750) nor brightness (p=0.381) have a significant direct effect on cued

recall. Furthermore, we did not find an interaction effect between hue and brightness on cued recall.

Hue and brightness were not found to have a significant direct effect on free recall either. However, a

significant interaction effect of hue and brightness on free recall (F=2.484; p ≤ 0.061*) was found. R² is

3% which is low but sufficient. We can thus accept hypothesis H1, indicating that website color

components, hue and brightness, have an interaction effect on the memorization of the consumer.

Table 8: Effects of Color scheme colors on cued and free recalls

Effects of color components (hue, brightness) on cued recall

p-value

Hue

0.404

0.750

Brightness

0.771

0.381

Hue x Brightness

0.616

Effects color components (hue, brightness) on free recall

p-value

Hue

0.288

0.834

Brightness

0.049

0.835

Hue x Brightness

2.484

0.061*

*p<0.1

Based on the ANOVAs carried out, we noted that the interaction effect of hue and brightness on free

recall is most significant when hue 2 (background color = Newsvine Green, foreground color = Magnolia

Yellow or Sunflower Yellow) was employed (F = 4,048; p ≤ 0.048). With a low level of brightness

(brightness 1) participants remember the content of the website better than with a high level of brightness

(brightness 2) (Figure 4).

Figure 4: Effects of brightness on free recall

From this result, we understand that a lower contrast between background color and foreground color

enhances the memorization of the commercial information given on a website.

16BDirect effects of the colors of the color scheme on buying intention

The results of the GLM analysis demonstrate that the color scheme of an e-commerce website is very

influential on buying intention (Table 9). Brightness has a significant positive effect on buying intention

(F = 15.201, p ≤ 0.000) (Table 9). In line with our results for memorization, we note that when the

background and foreground colors’ brightness is not too strong, buying intentions are the highest (Table

8).

Table 9: Effects of color scheme colors on buying intention

p-value

Hue

0.349

0.790

Brightness

15.201

0.000***

Hue x Brightness

3.732

0.012*

*p<0.1; ***p<0.001

The GLM analysis also shows that hue and brightness have a positive interaction effect on buying

intention (F = 3.732; p ≤ 0.012).

The results of the ANOVA show that the effect of brightness on buying intention is only significant for

hues n°1 (background color = Magnolia Yellow, foreground color = Newsvine Green or Granny Apple

Green) and n°2 (background color = Newsvine Green and foreground color = Magnolia Yellow or

Sunflower Yellow), i.e. with chromatic hues. Brightness but has no significant effect on buying intention

in the case of black and white hues. This result differs from Hall and Hanna (2003) finding that color

combinations do not have a significant effect on consumer purchase intention. When contrast is higher

and brightness increases, buying intention increases (Figure 5). Therefore, H2 is accepted.

Figure 5: Effects of brightness on buying intention

17BEffect of memorization on buying intention

A simple regression enables us to observe that free recall has a positive effect on buying intentions

(Table 10). The more information an individual memorizes about a product, the stronger her or his

buying intention will be. Thus, hypothesis H3 is accepted.

Table 10: Regression between memorization and buying intention

Buying intentions

Memorization

0.044*

Constant

2.096**

F = 3.824 ; R² = 0.013

* p < 0.1 ** p < 0.01

18BDirect effects of the colors of the color scheme on mood

GLM analyses show that hue and brightness have a significant interaction effect on negative mood (F =

3.042; p ≤ 0.029) (Table 11). In this sense, we can accept hypothesis H4.

Table 11: Effects of color scheme colors on mood

Effects of color scheme

colors on positive mood

Effects of color scheme

colors on negative mood

p-value

Hue

0.374

0.772

1.159

0.326

Brightness

0.041

0.840

0.334

0.564

Hue x Brightness

0.916

0.434

3.042

0.029*

*p<0.1

ANOVAs show that color schemes based on hues n°1 (foreground = Newsvine Green or Granny Apple

Green and background = Magnolia Yellow) and n°4 (background = Black and foreground = White or

Grey) allow for an interaction effect between hue and brightness. When hue n°1 (Newsvine

Green/Magnolia Yellow and Granny Apple Green/Magnolia Yellow) is used, an increase of the brightness

level entails a significant increase of negative mood (F = 3.066; p ≤ 0.084), while with hue n°4

(White/Black - Grey/Black), an increase of the brightness level decreases negative mood (F = 3.815; p ≤

0.055).

Effect of mood on memorization and buying intention

Since positive mood was not found to be affected by color, the effect of mood on memorization and

buying intention was tested by examining the effect of negative mood on these variables. Two simple

regressions give evidence that negative mood has a significant and negative impact on buying intention

(b = -0.129; p ≤ 0.01), but does not have a significant effect on memorization (free recall) (Table 12).

Therefore hypothesis H5 is not accepted, while hypothesis H6 can be accepted.

Table 12: Regression between negative mood and buying intention

Buying intentions

Negative mood

-0.129**

Constant

- 8.215E-17

F = 4.901 ; R² = 0.017

* p < 0.1 ** p < 0.01

5BDISCUSSION

Experiments on the behavior of Internet users are still relatively rare. At the same time there is a paucity

of research distinguishing among the three components of color, hue, brightness and saturation in e-

commerce websites. Building on Drugeon-Lichtlé (1996, 2007, 2009) and Gorn et al. (2004)'s works and

understanding the importance to distinguish the three components of color, this study makes a

contribution in this research area by examining these three color components in e-commerce. Indeed, it

helps to discover new effects on consumer behaviour and responses that we did not know before. Our

research enabled us to bring to the fore the effects of the colors used on e-commerce websites on

consumer memorization and buying intention. It also shows that negative mood as a mediating variable

reinforces these effects.

We thereby found that a lower contrast between background color and foreground color enhanced

memorization. In other words, the retention of information is higher when the latter is more difficult to

read. Such a result existed in psychology (Hill and Scharff, 1997; Scharff and Ahumada, 2002; Hall and

Hanna, 2003) but there are not any research studies explaining this in e-commerce. In addition, although

hue and brightness were not found to have a significant effect on memorization, they have an interaction

effect on memorization and, in specific, on free recal. This interaction effect is most significant with

chromatic colors. Green and yellow hues, which are chromatic ones, are more likely to enhance the

memorization of the displayed information than black and white (achromatic colors). Chromatic colors

induce better memorization scores, thus the effect of interaction of chromatic hues at particular levels of

brightness, produce better memorization scores. The contribution of hue is thus not neutral; it can lead to

better scores of memorization when it is properly combined with brightness.

Hue and brightness were not found to have a significant direct effect on cued or free recall. This seems

reasonable since the hue or brightness of the text of the website (foreground color) and the hue and

brightness of its background do not necessarily have to facilitate the retention of information.

Participants provided equivalent answers to closed questions about the content of the website (cued

recall), regardless of the colors of the color scheme. These questions actually helped participants to

memorize information in that they accurately added up the information that could be easily memorized.

Instead, colors were more important in free recall, when no help was provided and participants had to

remember what they saw on the website.

Brightness has a positive effect on buying intention, which is only significant for chromatic colors. In

addition, hue and brightness have a significant interaction effect on buying intention. Low brightness

favourites a stronger buying intention. However, a too strong brightness can also discomfort the

consumer and limit its desire to buy. When contrast is higher and brightness increases, buying intention

increases. Similarly, it seems that a high brightness of the color scheme, and therefore a sharp contrast

between background color and foreground (text) do not lead to a stronger buying intention.

Furthermore, we found an interaction effect between hue and brightness on mood. This is even more

significant for chromatic hues. We can argue this by explaining that the decrease in contrast due to the

increased lighting of the foreground color on the background one, discomforts respondents and put them

in a negative mood when colors are chromatic. The visual comfort caused by the use of chromatic colors,

which is not very common as explained earlier, should not be damaged by a too bright appearance..

Indeed, when someone has to guess what is written in front of him, affective and cognitive states are

modified, resulting in a decreasing mood. Our results show that the effect of color was significant only

for negative mood.

Purchase intention was found to be influenced by negative mood. Indeed some color schemes show that

when brightness is high, negative mood increases. On the other hand, when negative mood is high,

buying intention are weaker. It thus seems normal to find that higher brightness leads to weaker buying

intention. The intention to buy diminishes when the colors of the e-commerce website seem unpleasant

(negative mood) in the eyes of the consumer.

However, the effect of mood on memorization was not found to be significant. Customers memorize the

commercial information provided on an e-commerce website better when colors result in a negative

mood. This happens when the background and foreground colors create an unpleasant contrast to the

respondents. Indeed, for some color schemes, the brighter colors are, the stronger the negative mood

becomes. At the same time, a strong negative mood leads to strong memorization. It thus seems normal

to find that a higher brightness of the color scheme leads to weaker memorization.

Memorization and in particular, free recall, was found to have a positive effect on purchase intention.

The more a consumer memorizes information about a product, the more he intends to buy it. This result

should be clarified because it would be interesting to know what information is effectively stored. Further

analysis of the results of this experiment would allow us to answer to this question. Repeating the

experimentation with the same respondents after a certain period of time could effectively help us to

answer more accurately.

These results on color had not been revealed in previous work studying the three components of color in

a similar experiment. Rather than comparing warm and cold hues, we focused on hue and brightness. We

found that hue is important online, especially when measured with brightness, in line with previous

findings about the brightness component of color (Drugeon-Lichtlé, 2009; Valdez and Mehrabian,1994;

Gorn et al.,1997, 2004 and Camgöz et al., 2002). Our results on the effect of color on memorization are

in line with Hall and Hanna (2003), whereas the effect of color on mood as well as the effect of mood

and memorization on buying intention have not been studied in previous research. Although the effect of

color on emotions has been the research topic of several studies such as Valdez and Mehrabian (1994)

and Lichtlé (2007), the effect of color on mood had not been previously examined in literature. Finally,

our results regarding the effect of mood on memorization differ from those of Chebat et al. (1995)

finding that positive mood increases memorization. In comparison to their work referring to offline

shopping environments, our work suggests that it is negative mood that enhances memorization. The

main difference between these two studies stems from the different nature of the shopping environment

(online vs. offline). In particular, the psychological stimulation between a shopper in the offline

environment and an online shopper appears to be different. Online shopping entails a cognitive effort of

consumers to read information on screen, an effort that is missing from traditional shopping. Readability

in the offline environment is not as important as online, as the input consumers receive about products

comes from the physical setting and not from the computer screen, which is the primary means of

receiving information for the online consumer. Arguably, readability can influence affective states, and

mood in particular. Therefore, negative mood can occur more easily online, influenced by readability, by

commercial information which is too difficult to read. However, due to the additional effort required for

reading, information that is difficult to read is also difficult to forget and is retained in consumer

memory. This could explain the paradox of negative mood influencing positively memorization. In

offline shopping, negative mood could arise from factors other than bad readability, lacking, thus, a

positive effect on memorization.

6BCONCLUSIONS, IMPLICATIONS AND FUTURE RESEARCH

According to Gefen et al. (2003), an e-commerce website is an information technology and it is the

primary interface with an online vendor. Therefore the design of an e-commerce website is of paramount

importance for e-commerce success, influencing customer interaction and behavior. This is recognized

by a large stream of research in information systems which have studied web design and have focused on

perceived website quality and website characteristics as factors explaining online consumer behavior

(Torkzadeh and Dhillon, 2002; Singh et al., 2005; Hampton-Sosa and Koufaris, 2005). Color constitutes

an important variable for the design of e-commerce websites, as reported in information system studies,

especially in usability, human-computer interaction and e-commerce literature (Lee and Koubek, 2010;

Coursaris et al., 2008; Agarwal and Hedge 2008; McKracken et al., 2003). In this study, we examined

how color, a key attribute of the technological aspect of a website, influences consumer memorization,

mood and purchase intention. In addition, in doing so, we address color in terms of its components, hue

and brightness.

This study examines color in the context of e-commerce and website design with respect to its

components, brightness and hue. In this approach, it extends previous research which focuses only on the

hue component of color, and which compares warm and cold hues or combinations of hues.

Our study examines memorization in the context of e-commerce. We introduce memorization as a factor

influencing buying intention, contributing to the existing body of literature on antecedents of buying

intention in e-commerce. We also examined two factors influencing memorization, color and mood. We

found that website color components, hue and brightness, have an interaction effect on consumer

memorization. In addition, our results indicate that negative mood also leads to better memorization.

Furthermore, our findings show that the color of e-commerce websites affects consumer mood. While

this is in accordance with Wu et al. (2008), our results also show that color hue and brightness jointly

influence mood. In addition, our study shows that the color of e-commerce websites also influences

purchase intention. Extending previous work showing that warm colors positively influence purchase

intention (Wu et al., 2008) our study contributes as it shows that brightness has a positive effect on

purchase intention for chromatic hues and that hue and brightness have an interaction effect on purchase

intention. Finally, our study shows that purchase intention is influenced by mood, in line with Wu et al.

(2008). Overall, our study also provides easy to replicate standards for laboratory conditions, when

conducting experiments on colors on the Internet. It seems that a study on color should compare hue and

brightness rather than warm and cold colors when trying to examine what consumers recall and what

leads them in purchasing. Indeed, in everyday life there is no support helping consumers to recall the

content of an e-commerce website they visited or to compare it with another offer.

Our study provides findings on color that can inform and guide the design of websites in e-commerce so

that they are effective for attracting customer purchases. By looking into mood and memorization, as

variables which are antecedents of buying intention and are influenced by color, our study further

contributes in information systems. The research implications are most important for memorization, as it

is proposed as a new factor predicting buying intention of online consumers. Our results on the effect of

color hue and brightness on mood and memorization are not only relevant to e-commerce web design but

to web interface design in general and are also of value for research in web aesthetics, usability and

human-computer interaction. Hence, our findings are relevant and important to information systems

researchers, particularly to those active in the areas of usability, human-computer interaction and e-

commerce or the intersection of these areas.

These results must be related to the research conducted by Silverstein (1987) who noticed that

monochrome screens entailed more eyestrain and overall tiredness. Therefore, e-vendors should be aware

of this and choose carefully the hues of the foreground and background colors that they will use on their

site so as to adjust them to their target. They should also take into account the aesthetic and functional

impact of those colors: their contrast facilitates finding information on a webpage. Moreover, low

brightness fosters better memorization scores and stronger buying intention. Consumers recall

information more easily when they had difficulty reading on an e-commerce website, however, they did

not necessarily feel like buying a product from this website. This holds especially for chromatic colors.

As with any research, this study comes with limitations. First of all, data were collected from participants

interacting with a website designed for the experiment and not a real online store. In addition, the

experiment website was an online store selling only CDs and not other products that could possibly

produce different results. Furthermore, saturation as well as recognition variables were not possible to be

measured. Finally, generalization of findings to all colors should be made with caution.

The difference of perception of colors among cultures should also be taken into account. Our sampling

frame was a student population, primarily French, and we did not control for culture as an intervening

variable. Future research with a color scheme varying according to the culture of the respondents could

bring further and more detailed results. This entails the need for a cross-cultural study with samples with

distinct color preferences as well as diverse semantic associations of color. Research across cultures is

also needed to test how online consumer mood, memorization and buying intention and the links among

them may be influenced and vary by cultural characteristics. As implied by a number of studies (Singh

and Baack, 2004; Cyr et al., 2010) e-commerce websites are not culturally neutral and their design

characteristics can have different effects depending on culture. Hence, further work would be necessary

to enhance our understanding towards a culturally informed design of effective e-commerce websites.

Future research related to the measurement of memorization or buying intention of online consumers,

should undoubtedly take into consideration other brightness and saturation rates. As explained earlier,

studying the long memory would probably bring more useful results regarding the effects of colors. This

would require a replication of the experiment some months or years after the first one, with the same

respondents, the same laboratory conditions and the same colors for the website. Coupled with the use of

sound on e-commerce websites, further studies would enable us to reach a better understanding of the

effects of the atmosphere pervading e-commerce websites on consumers, adopting a holistic rather than

an individual approach to the phenomenon. Finally, besides hue, saturation and brightness, further

research should examine the three other color components, texture, brilliance and matt effect, used in

virtual worlds such as Second Life and in new e-commerce websites using the glossy effect for example.

This research illustrated the combinations of the most successful colors for web stores by taking into

consideration the color hue, brightness and saturation. It also underlines the importance of chromatic

colors and reinforces the need to include chromatic colors in web design, especially since most e-

commerce websites use mainly achromatic colors. These results can contribute to help the designers of e-

commerce websites particularly with regards to the combinations of colors of content zones (supporting

the commercial information) and navigation zones (supporting the utilitarian information). The use of a

combination of adequate colors (a) in the contents and (b) in the navigation, by respecting the territories

of each of them by the use of borders or adequate colors when they are close one to the other, will allow

designers to make the color scheme more pertinent in transforming the intention into act of purchase.

The ability of people to distinguish colors, or the inability of some to differentiate among them

constitutes an issue to be considered during the design of an interface in order to increase its usability.

With 8% of color blind people in the male population (Lanthony, 2005), brands have to use proper colors

that are easily distinguishable by them. The W3C and WAI provide solutions for web designers on this

topic by suggesting the use of “web safe colors”. Among the 256 colors that can be displayed on

traditional computer screens, only 10 are viewed by color blind people on the Internet. Colors should be

manipulated, so that the consumers lose no information because of a contrast or of a harmony which they

do not perceive. This should be taken into account early in the design phase of any e-commerce website.

It seems evident that this applies to people affected by color blindness or by another disease affecting the

vision of colors, for whom access to information outweighs the aesthetics of an e-commerce website.

Besides, the growing percentage of seniors using e-commerce, requires taking more seriously into

consideration the management of colors within the wider context of enabling access to the information.

This can be improved with the choice of better contrasts between the foreground color (text) and the

background color. As sight decreases with age, e-sellers should provide an even easier way for seniors to

read information, as assistance in their shopping experience. Thus, playing with brightness and saturation

levels of safe colors carefully selected for the text and background, making the website readable for old

and color blind people, can be a critical success factor in terms of ROI.

7BREFERENCES

AGARWAL A and HEDGE A (2008) The impact of web page usability guideline implementation on aesthetics and

perceptions of the e-retailer. Proceedings of the Human Factors and Ergonomics Society, September 2008, vol.

52, n°6,, 528-532.

AUMONT J (1994) Introduction à la couleur : des discours aux images, Paris, Armand Colin.

BELK R W (1985) Issues in the Intention-Behavior Discrepancy, In Sheth Jagdish N., eds. Research in Consumer

Behavior - Volume 1. CN:Greenwich: JAI Press: 1-34.

BELLIZZI JA and HITE RE (1992) Environmental Color, Consumer Feelings, and Purchase Likelihood,

Psychology and Marketing, 9, September-October, 347-64.

BEM SL (1974) The Measurement of Psychological Androgyny, Journal of Consulting and Clinical Psychology,

1974 Apr, 42(2),, 155-162.

BIERS K and RICHARDS L (2002) Web Page Background Color Evaluative Effect On Selected Product

Attributes, working paper, Utah State University.

BRADY L, and PHILLIPS C (2003) Aesthetics and usability: a look at colour and balance. Usability News, 5(2),

http://www.surl.org/usabilitynews/112/pdf/Usability%20News%20112%20-%20Phillips.pdf

BREMOND É. (2002) L'intelligence de la couleur, Paris, éds. Albin Michel.

BOYATZIS CJ and VARGHESE R (1993) Children's emotional associations with colors, The Journal of Genetic

Psychology, 155(1), 77-85.

BUCKLIN R E and SISMEIRO C (2003) A model of web site browsing behavior estimated on clickstream data,

Journal of Marketing Research, 40, 3, 249-267.

CAMGÖZ N, YENER N and GÜVENÇ D (2002) Effects of hue, saturation, and brightness on preference, Color

Research & Application, 27, 3, 199-207.

CHEBAT, J.-C., GELINAS-CHEBAT, C., VANINSKI, A., AND FILIATRAULT, P. (1995) The Impact of Mood

on Time Perception, Memorization, and Acceptance of Waiting. Genetic, Social, and General Psychology

Monographs 121(4), 411-4.

CHEBAT J-C and MORRIN M (2006) Colors and cultures: exploring the effects of mall decor on consumer

perceptions. Journal of Business Research, 60(3), 189-196.

CIALDINI RB, DARBY BL and VINCENT JE (1973) Transgression and Altruism: A Case for Hedonism. Journal

of Experimental Social Psychology, 9(6), 502-516.

COURSARIS C, SWIERENGA S and WATRALL E (2008) An Empirical Investigation of Color Temperature and

Gender Effects on Web Aesthetics, Journal of Usability Studies, (3)3, 103-117.

CYR D, HEAD M and LARIOS H (2010) Colour appeal in website design within and across cultures: A multi-

method evaluation, International Journal of Human-Computer Studies, 68(1-2), 1-21.

DARPY D (1997) Une variable médiatrice du report d'achat : La procrastination, Communication au 13ème

Congrès International de l'AFM, Toulouse.

DELAFROOZ N, PAIM LH and KHATIBI A (2010) Students’ Online Shopping Behavior: An Empirical Study,

Journal of American Science, 6(1), 137-147.

DEEB SS (2004) Molecular genetics of color-vision deficiencies, Visual Neuroscience, 21(3), 191-6.

DEEB SS and MOTULSKY AG (2011) Red-Green Color Vision Defects, In: Pagon RA, Bird TD, Dolan CR,

Stephens K, (eds). GeneReviews, Seattle (WA): University of Washington, Seattle; 1993-2005 [updated 2011

Sep 29].

DEMERS E and LEV B (2001) A rude awakening: internet shakeout in 2000, Review of Accounting Studies, 6(2-3),

331–359.

DIVARD R and URIEN B (2001) Le consumer vit dans un monde en couleurs, Recherche et Applications en

Marketing, 3-24.

DONOVAN RJ and ROSSITER JR (1982) Store atmosphere: an environmental psychology approach, Journal of

Retailing, 58(1), 34-57.

DREZE X and ZUFRYDEN F (1997) Testing Web Site Design and Promotional Content, Journal of Advertising

Research, 37 (2), 77-91.

DRUGEON-LICHTLÉ M-C (1996) Les effets des couleurs d'une annonce magazine sur les émotions du consumer :

conceptualisation et résultats d'une étude exploratoire, Actes de l'Association Française de Marketing, 12,

Poitiers, éd. Michel Kalika, 445-458.

DRUGEON-LICHTLÉ M-C (2002) Couleur d'une annonce publicitaire, goûts des individus et perception des

marques, Décisions Marketing, April/June, 26, abi/inform global, 29.

DUNN B (1992) Choice of Color for Product Can Be Critical Factor, The Gazette, August 10, 6.

ENGEL JF, KOLLAT DT and BLACKWELL RD (1978) Consumer Behavior, 3rd edition, Hinsdale, Dryden Press,

II.

ERICKSON G B and Block S S (1999) Testability of a color vision screening test in a population with mental

retardation, Journal of the American Optometric Association, 70(12), 758-763.

EROGLU S A, MACHLEIT K A and DAVIS L M (2001) Atmospheric qualities of online retailing: a conceptual

model and implications, Journal of Business Research, special issue on Retail Strategy and Consumer Decision

Research, 54, November, 177-184.

EROGLU S A, MACHLEIT K A and DAVIS L M (2003) Empirical Testing of a Model of Online Store

Atmospherics and Shopper Responses, Psychology & Marketing, 20, 2, 139-50.

EYSENCK M W (1976) Arousal, learning, and memory, Psychological Bulletin, 83, 389-404.

FERNANDEZ-MALOIGNE C (2004) Quelle métrique pour l'évaluation de la qualité couleur d'une image

numérique ? Application à la compression JPEG2000, CAM Conférence, Paris, octobre.

FILSER M (1994) Le comportement du consommateur, Précis Dalloz, Paris.

FILSER M (2003a) Le marketing sensoriel: la quête de l'intégration théorique et managériale, Revue Française du

Marketing, 194, 4/5, Septembre, 5-11.

FILSER M (2003b) Vingt ans de recherches en comportement du consumer, in : Rémy, I. Garabuau-Moussaoui, D.

Desjeux and M. Filser, (eds), Sociétés, Consommation et Consommateurs, L'Harmattan, 15-20.

FORGEAS J P (1999) Network theories and beyond, T. Dalgleish and M. J. Power (eds.), Handbook of Cognition

and Emotion, 591-612, Chichester, Wiley.

GARDNER MP (1985) Mood states and consumer behavior: a critical review, Journal of Consumer Research,

12(3), 281-300.

GEFEN D, KARAHANNA E and STRAUB DW (2003) Trust and TAM in Online Shopping: An Integrated Model.

MIS Quarterly, (27)1, 51-90.

GORN G, CHATTOPADHYAY A, SENGUPTA J and TRIPATHI S (2004) Waiting for the web : how screen

color affects time perception, Journal of Marketing Research, XLI, May, 215-225.

GORN G, CHATTOPADHYAY A T YI and DAHL D W (1997) Effects of color as an executional cue in

advertising: they're in the shade, Management Science, 43, 10, 1387-1400.

HALL RH and HANNA P (2003) The Impact of Web Page Text-Background Color Combinations on Readability,

Retention, Aesthetics, and Behavioral Intention, Behavior and Information Technology, 23(3), 183-195.

HAMILTON M and LUO R (1999) Impact of Animation and Color Complexity on Web Site Effectiveness, Working

Paper, Department of Communication Sciences, University of Connecticut, 1-10.

HAMPTON-SOSA W and KOUFARIS M (2005) The Effect of Web Site Perceptions on Initial Trust in the Owner

Company, International Journal of Electronic Commerce, 10(1), 55-81.

HANSON, WA (2000) Principles of Internet Marketing, South-Western College Publishers, Cincinnati, Ohio.

HILL A and SCHARFF L V (1999) Readability of computer displays as a function of color, saturation, and

background texture, éd. D. Harris, Engineering Psychology and Cognitive Ergonomics, 4 , 123-130.

HILL A and SCHARFF L V (1997) Readability of websites with various foreground/background color

combinations, font types and word styles, Proceedings of 11th National Conference in Undergraduate Research,

2, 742-746.

HOWARD J A (1994) Buyer Behavior in Marketing Strategy, Englewood Cliffs, NJ: Prentice Hall.

HUMAR I, GRADISAR M and TURK T (2008) The Impact of Colors on the Speed of Reading Web Pages on a

CRT monitor, International Journal of Industrial Ergonomics, 38 (2008) 885-899.

ISEN A M, SHALKER T E, CLARK M and KARP L (1978) Affect, accessibility of material in memory, and

behavior : A cognitive loop, Journal of Personality and Social Psychology, 36, 1-12.

ITTEN J (1970) The elements of Color, New York, Van Nostrand Reinhold Company.

JOLIBERT A and JOURDAN P (2006) Marketing Research, Méthodes de recherche et d'études en marketing, éd.

Dunod, Paris.

KANG K-S and CORBITT B (2001) Effectiveness of graphical components in web site e-commerce application – a

cultural perspective, Electronic Journal on Information Systems in Developing Countries, 7(2), 1-6.

KIRITANI Y and SHIRAI S (2003) Effects of background colors on user's experience in reading website, Journal

of the Asian Design International Conference, Academic Journal, 1, 64.

KIM D, FERRIN DL and RAO HR (2008) A trust-based consumer decision-making model in electronic commerce:

The role of trust, perceived risk, and their antecedents. Decision Support Systems, 44(2), 544-564.

KOTLER P (1973) Atmosphere as a marketing tool, Journal of Retailing, 49, Winter, 48-64.

KOYANI S, BAILEY R and NALL P. (2006) Research-Based Web Design and Usability Guidelines.

www.usability.gov.

LADWEIN R (1999) Le comportement du consommateur et de l'acheteur, Paris, Economica.

LANTHONY P (2005) La perception des couleurs sur écran, Intervention dans le cadre d'un séminaire sur la

couleur, 3C S.A., Abbaye de Royaumont - France, Juin 2005.

LEE S and KOUBEK RJ (2010) The effects of usability and web design attributes on user preference for e-

commerce web sites, Computers in Industry, 61, 329-341.

LEMOINE J-F (2003) Vers une approche globale de l'atmosphère du point de vente, Revue Française du

Marketing, 194, Septembre, 83-101.

LEMOINE J-F (2008) "L'influence de l'atmosphère des sites web marchands sur les réponses des internautes",

24ème congrès international de l'association française du marketing, Paris, 15 et 16 mai 2008, CDRom.

LICHTLÉ M-C (2009) Pour favoriser la mémorisation d'une marque, un annonceur a-t-il intérêt à utiliser des

couleurs étonnantes ?, Gestion 2000, Vol. 26 n° 5, sept.-oct. 2009.

LICHTLÉ M-C (2007) The effect of an advertisement’s colour on emotions evoked by an ad and attitude towards

the ad The moderating role of the optimal stimulation level, International Journal of Advertising, 26(1), 37–62.

MAGUIRE M (2011) Guidelines on Website Design and Colour Selection for International Acceptance, 6769

LNCS (PART 1), 162-171

MANDEL N and JOHNSON E J (2002) When web pages influence choice: effects of visual primes on experts and

novices, Journal of Consumer Research, 29(2), 235-245.

MASTERS J and WYNDOL F (1976) Effects of Affective States on Noncontingent Outcome Expectancies and

Beliefs in Internal or External Control, Developmental Psychology, 12, 481-82.

MATTELART A (1996) The Invention of Communication, London/Minneapolis, University Minnesota Press.

MAYER JD and GASCHKE YN (1988) The experience and meta-experience of mood, Journal of Personality and

Social Psychology, 55, 102-111.

MCCRACKEN, DD ,WOLFE RJ and SPOOL JM (2003) User-centered Website Development: A Human–

Computer Interaction Approach, Pearson Prentice Hall Inc., Upper Saddle River, NJ.

MEHRABIAN A and RUSSELL JA (1974) An Approach to Environmental Psychology, MIT Press, Cambridge,

MA.

MOE WW and FADER PS (2004) "Dynamic Conversion Behavior at E-Commerce Sites", Management Science,

50 (3), 326-35.

MOSS G, GUNN R and HELLER J (2006) Some men like it black, some women like it pink: consumer

involvements of differences in male and female website design, Journal of Consumer Behavior, London, 5, 4,

328-342.

MUCCHIELLI A (1991) Les méthodes qualitatives, Que sais-je?, Presses Universitaires de France, Paris.

MUNSELL A (1969) The Munsell Colour Atlas, Munsell Color Corp.

NIELSEN J (2000) Designing Web Usability, Indianapolis, Indiana, New Riders Publishers.

ODOM AS and SHOLTZ SS (2004) The reds, whites, and blues of emotion: examining color hue effects on mood

OJANPÄÄ H and NÄSÄNEN R (2003). Effects of luminance and colour contrast on the search of information on

display devices, Displays, 24, 167-178.

O'SHAUGHNESSY J (1992) Explaining Buyer Behavior: Central concepts and Philosophy of Science issues, NY:

Oxford University Press.

PANTIN-SOHIER G (2004) Le rôle de la couleur dans la perception des traits de personnalité de la marque : une

étude comparative Etats-Unis / Suède, 1ère journée de recherche AFM-AUDENCIA sur le Marketing et le

Design - Janvier.

PAPACHRISTOS E, TSELIOS N and AVOURIS N (2005) Inferring relations between color and emotional

dimensions of a Web site using Bayesian Networks, Proceedings of Interact 2005.

PEW RESEARCH CENTER (2010) Millennials. Pew Research Center, February 24. Available at:

http://pewresearch.org/millennials/

RIDPATH C, TREVIRANUS J and WEISS PLT (2000) Testing The Readability Of Web Page Colors, University

of Toronto's Adaptive Technology Resource Centre, WWW9 conference in Amsterdam

Hhttp://www.aprompt.ca/WebPageColors.htmlH.

ROTTER JB (1966) Generalized Expectancies For Internal Versus External Control Of Reinforcement,

Psychological Monographs, 80(1), Whole N° 609.

ROULLET B (2004) L'influence de la couleur en marketing: vers une neuropsychologie du consommateur, Thèse

de doctorat en sciences économiques et de gestion, Université de Rennes 1, Rennes, décembre.

SCHARFF LV and AHUMADA AJ JR (2002) Predicting the readability of transparent text, Journal of Vision 2(9),

653-666.

SCHARFF L V and HILL AL (2000) Discriminability measures for predicting readability of text on textured

SCHMIDT KE and LIU Y (2005) Design of consumer product webpages: Experimental investigations of aesthetic

and performance factors. Proceedings of the Human Factors and Ergonomics Society, 1743-1746.

SILVERSTEIN LD (1987) Human Factors for Color Display System: Concepts, Methods, and Research, Color and

the Computer, éd. J. Durret, San Diego, CA, Academic Press, 27-61.

SINGH N and BAACK DW (2004) Web Site Adaptation: A cross-cultural comparison between U.S. and Mexican

Web sites, Journal of Computer Mediated Communication, 9(4).

SINGH N, DALAL N and SPEARS N (2005) Understanding Web home page perception, European Journal of

Information Systems, 14, 288-302.

SO S and SMITH M (2002) Colour graphics and task complexity in multivariate decision making, Accounting,

Auditing & Accountability Journal, 15(4) 565-593.

SWINYARD WR (1993) The Effects of Mood, Involvement and Quality of Store Experience on Shopping

Intentions. Journal of Consumer Research, 20, 271-280.

TORKZADEH G and DHILLON G (2002) Measuring Factors that Influence Success of Internet Commerce,

Information Systems Research, 13(2), 187-204.

TROUVÉ A (1999) La mesure de la couleur, Paris, CETIM, éd. Association française de normalisation (AFNOR).

VALDEZ P and MEHRABIAN J (1994) Effects of color on emotions, Journal of Experimental Psychology,

General, 123, 4, 394-409.

WAT-C. (2005) Tools - Colour Contrast Analyser 1.1. Retrieved from the Web Accessibility Tools Consortium

Web Site: http:/www.wat-c.org/tools/CCA/1.1

W3C (2008) Publication Version 1.1 of the AccessiWeb repository created by the W3C, of the June 9, 2008,

available at: Hhttp://www.w3c.org/H.

WESTBROOK R (1980) Intrapersonal Affective Influences on Consumer Satisfaction with Products. Journal of

Consumer Research, 7, 49-54.

WEXNER L B (1954) The degree to which colors (hues) are associated with mood-tones, Journal of Applied

Psychology, 6, 432-436.

WU C-S, CHENG F-F and YEN DC (2008) The atmospheric factors of online storefront environment design: An

empirical experiment in Taiwan. Information & Management, 45, 493-498.

YOO B and DONTHU N (2001) Developing and validating multidimensional consumer-based brand equity scale,

Journal of Business Research, 52(1), 1-14.

ZICKUHR K (2010) Generations Online in 2010, Pew Internet & American Life Project, Pew Research Center.

ZUCKERMAN M (1994) Behavioral Expressions and Biosocial Bases of Sensation Seeking, Cambridge:

Cambridge University Press.

8BAPPENDICES

19BAppendix 1: Devices and installation required to conduct the experiment properly

Experiment

room

(Fernandez-

Maloigne,

2004)

Measurements were taken at different intervals with a

luxmeter:

- Keep a distance of about one meter between the back

of the room and the screen,

- Room lighting (ambient illumination),

- High-quality assessment monitor, size 50-60 cm (22"

- 26").

The luxmeter

Participants

(Lanthony,

2005)

- An Ishihara’s test for color blindness was conducted

to guarantee that participants were not color-blind and

thus in a position to provide valid answers.

The Ishihara’s test

Screens

All the screens used during the experiment were

calibrated

- The screens must warm up for an hour before

calibration;

- A CRT display must be used rather than a plasma

screen;

- Ambient light compensation must be disabled;

- The probe allows to generate the ICC profile;

- Save the ICC profile which will be set automatically

afterwards.

The probe

20BAppendix 2: Experiment website screenshots for the 8 color schemes

Experiment Plan 1

Experiment Plan 2

Experiment Plan 3

Experiment Plan 4

Experiment Plan 5

Experiment Plan 6

Experiment Plan 7

Experiment Plan 8

21BAppendix 3: Brief Mood Introspection Scale (BMIS) – (Mayer and Gaschke, 1988)

22BThe scale is based on the question “Do you feel _______?” for each of the 16 items shown below. Each

item was measured on a 5-point Likert scale ranging from 1 – ‘Definitely do not feel’ to 5 – ‘Definitely

feel’

N°item

Item

Corrected

Correlation

With

out

item

Quality of

representation

Contribution

to the factor 1

Contribution

to factor 2

Happy (Hum1)

.377

.779

.505

Fed up (Hum2)

.412

.846

.540

.658

Caring (Hum4)

.512

.838

.526

.725

Nervous (Hum5)

.537

.834

.712

.751

Satisfied (Hum6)

.411

.776

.631

Grouchy (Hum7)

.486

.759

.628

Sad (Hum9)

.588

.826

.653

.807

Jittery (Hum10)

.571

.772

.755

Loving (Hum11)

.407

.776

.514

Drowsy (Hum12)

.394

.854

.753

.583

.643

Lively (Hum13)

.544

.853

.600

.605

Gloomy (Hum14)

.634

.819

.746

.850

Tired (Hum15)

.459

.842

.661

.683

Active (Hum16)

.475

.778

.631

Cronbach Alpha

0.814

Eliminated Items

Peaceful – Hum3, Affectionate – Hum8

Bartlett Sphericity Test

Approximate Chi² = 985.340 – ddl = 28 – Sig. = .000

KMO value

0.854

Recommended value

3.304

1.120

% of information

64.882

23BItems were measured on a 5-point Likert scale ranging from 1 – Definitely do not feel to 5 – Definitely

feel

24BAppendix 4: Buying intention scale (Yoo and Donthu, 2001)

N°item

Item

Corrected

Correlation

Without

item

Quality of

representation

Contribution to

the factor

I will certainly buy products

coming from this website in

the near future.

.791

.864

.825

.908

I intend to buy on this

website in the near future.

.834

.837

.863

.929

It is likely that I buy on this

website in the near future.

.791

.870

.824

.908

I plan to buy on this website

in the near future.

.832

.865

.842

.927

Cronbach Alpha

0.899

Eliminated Items

Bartlett Sphericity Test

Approximate Chi²= 563.367 – ddl 3 = – Sig. = .000

KMO value

.749

Recommended value

2.512

% of information

83.739

25BItems were measured on a 5-point Likert scale ranging from 1 – ‘Strongly agree’ to 5 – ‘Strongly

disagree’