APPLYING KANSEI ENGINEERING TO DETERMINE
EMOTIONAL SIGNATURE OF ONLINE CLOTHING WEBSITES
M. N. Nor Laila, M. L. Anitwati
Faculty of Information Technology and Quantitative Sciences, Universiti Teknologi Mara, Malaysia
Mitsuo Nagamachi
User Science Institute, Kyushu University,Fukuoka 815-8540, Japan
Keywords: Emotional design, Kansei Engineering, E-Commerce.
Abstract: The integration of cognitive, semantic and affective elements is crucial in the conception and development
of designed products. IT artefact design and development has ignored the importance of affective elements
until recent years within the HCI community. Current literature reflects two main foci in the area: emotional
design and its evaluation. Of the two, the later is widely researched and reported. In our paper, we present
our research attempt to establish the design method for organizing the emotional design requirements of E-
Commerce (EC) websites by applying Kansei engineering (KE). We proposed a Kansei EC website design
method and demonstrated the method by conducting the semantic evaluation of pre-selected online clothing
websites using 40 Kansei words as descriptors of emotional sensation which was organized as a 5-point
Semantic Differential (SD) scale to form the Kansei checklist. 120 participants were asked to rate 35 pre-
selected online clothing websites using the Kansei checklist. Cluster analysis and Partial Least Square
method were then performed to identify the Kansei word cluster and from this result we uncover the
relationship between Kansei word cluster and online clothing website design.
1 INTRODUCTION
The discipline of design science emphasizes the
integration of cognitive, semantic and affective
elements in the conception and development of
designed products. Designers of IT artifacts have
begun to address affective or emotional elements in
products such as mobile phones. However, the
literature does not exhibit significant work on
artifacts such as websites. In this paper, we report
the results of our research to establish the design
method for organizing the design requirements
based on the emotional signature of websites. Here,
we demonstrate the use of Kansei Engineering (KE)
to identify the emotional signature of websites and
presents our empirical findings in support of using
Kansei as a means to incorporate the affective or
emotional appeal of websites. The context of web
application chosen for this work is the design of
online clothing E-Commerce (EC) websites where
emotional appeal is assumed to be significant.
2 EMOTIONAL DESIGN OF EC
WEBSITES
Human computer issues related to EC applications
were formerly focused on cognitive aspects of
websites. Since the early work of Nielsen in the
1990s, the emphasis was on the qualities of
usefulness and usability in producing good website
design. Tractinsky, et al, (2000) cited that most
studies dedicated to EC website evaluation are based
on two assumptions. The first assumption is that
target customers spend at least a few minutes on a
website and the second assumption is that good
website features usually elicit positive cognitive
evaluations and shopping experience. They pointed
out that, obviously, these assumptions ignored the
primary affective reaction or emotional responses
towards the website. Echoing this concern, Na Li
and Ping Zhang (2005) stressed that online shopping
behaviour is a complex phenomena and recognized
that affective reaction has been cited to be a factor
that promotes online shopping. This is because EC
142
N. Nor Laila M., L. Anitwati M. and Nagamachi M. (2008).
APPLYING KANSEI ENGINEERING TO DETERMINE EMOTIONAL SIGNATURE OF ONLINE CLOTHING WEBSITES.
In Proceedings of the Tenth International Conference on Enterprise Information Systems - HCI, pages 142-147
DOI: 10.5220/0001680101420147
Copyright
c
 SciTePress

websites have gone beyond the function of
conveying information to the extent of providing
persuasive engagement with website visitors through
the lively process of perception, judgment and
action. Since affect has been found to influence
decision making (Tractinsky, et al 2000; Norman,
2002), we argue that EC websites should induce
desirable consumer experience and emotion that
influences users’ perception of the websites to
extend the outreach potential of the online business.
Hence, we need to consider the emergence of the
dimension of desirability in EC website design.
Desirability emerged from the realization of the
need to have new measures of users’ experience
driven by emotional factors (Dillon 2001, Spiller
2004). Norman, an advocator of emotional design
discussed the notion of emotional design through
elements of visceral, behavioral and reflective
factors (2004). His views, parallel the view of
Englelsted as cited in Aboulafia and Bannon (2004)
who discussed three temporal categories of
emotions: affect, emotion, and sentiment. For EC
website emotional design for desirability, we view
visceral factors or affect to be more pertinent. In
support of this, Mahlke and Thüring (2007) studied
affect and emotion as important parts of the users’
experience with interactive systems, aiming to
consider emotional aspects in the interactive system
design process. While admitting that emotion cannot
be designed, they assert the importance of deriving a
method for recognizing users’ emotion from
emotional evaluation procedures. In addition, they
developed the users’ experience framework
illustrated in Figure 1 that clearly illustrates
subjective feelings as a component of emotional user
reaction.
Despite the gained recognition, the subject of
emotional appeal of websites or desirability is often
neglected as designers tend to pay more attention to
issues of usefulness and usability (Buchanan 2000)
due to the availability of established design
methodology that addresses aspects of usefulness
and usability.
Design method that enables the incorporation of
emotional design requirements is lacking. In
addition, numerous studies conducted on emotional
design tends to look at minimizing irrelevant
emotions related to usability such as confusion,
anger, anxiety and frustration (Norman 2002). It is
then necessary to seek for a suitable design method
to handle design requirements based on emotional
signatures of websites.
Figure 1: Components of Users Experience.
3 KANSEI ENGINEERING
Kansei is a Japanese term that is used to express
one’s impression towards artifact, situation and
surrounding. Deeply rooted in the Japanese culture,
Kansei can be indirectly translated as the mental
state where knowledge, feeling, and sentiment are
harmonized (Nagamachi, 2003) and is able to evoke
subjective pleasurable feelings from the interaction
with an artefact (Nagasawa, 2004).
Kansei Engineering (KE) which is founded by
Nagamachi in the seventies is a customer-oriented
technology to assimilate human Kansei within the
engineering realms of product design to achieve
consumer satisfaction and enjoyment. KE focus on
the identification of product Kansei that trigger and
mediate customers’ emotional response. The KE
process implements different techniques to link
product Kansei with product properties. In the
process, the chosen product domain is mapped from
both a semantic and physical perspective. In terms of
a design methodology, the approach of KE is to
organize design requirements around the emotions
that embody users’ expectations and interaction
(Spiller, 2004). KE has been successfully used to
incorporate the emotional appeal in the product
design ranging from physical consumer products to
IT artifacts. KE is a well accepted industrial design
method in Japan and Korea and is gaining
acceptance in Europe where KE is better known as
emotional design.
Emotional user
reactions
Subjective feelings
Motor expressions
Physiological reactions
Cognitive appraisals
Behavioural tendencies
Perception of non-
instrumental qualities
Aesthetics aspects
Visual aesthetics
Haptic quality
Acoustic quality
Symbolic aspects
Associative symbolism
Communicative symbolism
Motivational aspects
Consequences of the user
experience
Source: Mahlke &
Thüring (2007)
User experience
Perception of
instrumental
qualities
Usefulness
Utility
Usability
efficiency
controllability
helpfulness
learnability
System
User
Task
Human-
technology
Interaction
APPLYING KANSEI ENGINEERING TO DETERMINE EMOTIONAL SIGNATURE OF ONLINE CLOTHING
WEBSITES
143

4 RESEARCH METHODOLOGY
4.1 The Kansei Web Design Method
In our earlier work, we introduced the concept of
Kansei EC website to integrate subjective feelings,
impressions and subjective emotion or Kansei with
website design elements to add emotional appeal
(Anitawati and Nor Laila, 2006) (Figure 2).
Figure2: Kansei EC Websites.
To engineer the Kansei values into website
design, we proposed the Kansei EC website design
method (Figure 3).
Figure 3: Design Method for Building Kansei EC Website.
The method is divided into four levels: L1-
selection of website specimen, L2-preparation of
Kansei checklist, L3-determination of Kansei, and
L4- prototyping of Kansei website design. L1 is sub-
divided into procedure PI and PII. PI begins with the
collection of website specimens with visible
differences from existing specific domain to obtain
valid specimens. In P2, a new product concept is
established to produce website samples for eliciting
the emotional design requirements. In L2, the Kansei
checklist is established through the process of
synthesizing Kansei words (KW) that are directly
related to the product domain. KW can be an
adjective or a noun and are synthesized by language
experts. The Kansei checklist produced is in the
form of the Osgood Semantic Differential scale
which is used to measure Kansei in the next level. In
L3, the determination of Kansei is through two
processes. The first process is Kansei measurement
involving either experts or consumers. Participants
of Kansei measurement rate their impressions
towards product specimen based on the Kansei
checklist. Next, the participants’ evaluation is
analyzed and validated to interpret the Kansei
responses against design elements identified in L1.
Finally, the outcome from L3 will be used as the
emotional design requirements in L4, the design of
Kansei product.
4.2 Research Method
To illustrate the Kansei EC design method, we
performed process L1, L2 and L3 as prescribed in
the previous sub-section on e-clothing EC websites.
The visual description of the method is illustrated in
Figure 4.
Figure 4: Research Method.
Some of the design elements considered include
colour, style, menu and page orientation. The
participants’ Kansei and the website design elements
were then analyzed separately and then mapped to
produce the Kansei EC website design.
4.3 Research Instruments
One hundred and sixty three online youth clothing
EC websites were selected based on visible design
differences and analyzed based on rules of colour,
L1. Selection
of specimen
L2. Establishment
of checklist
L3. Determination of
Kansei
L4. Prototyping
Collection of
specimen from
existing product
Investigation
of design
elements
Classification
of attribute and
value
Determination
of Valid
specimen
Collection
of KW
Selection of
domain
specific KW
Development
of checklist
Kansei
Measurement
Analysis &
Validation
Kansei
web design
New product
concept
Determine
design
elements
Classify
physical traits
Build
samples
PI
PII
Kansei EC
website
KANSEI
ENGINEERING
Subjective
emotion
Design
element
Kansei EC website design
Identification of
Website design element
(element1, element2,
element3
,
element-n
)
Measurement of
visitor’s Kansei
(Kansei1, Kansei2,
Kansei3, …,
Kansei-n)
Analysis
element2
element3
Kanse
Ka
Ka
Ka
Elements
element1
Selection of
EC Domain
ICEIS 2008 - International Conference on Enterprise Information Systems
144

design elements, layout, page orientations and
typography. From the analysis thirty five website
specimens were finally used in the Kansei
evaluation. The Kansei checklist was organized in a
5-point Semantic Differential (SD) scale and
consists of 40 KW (Figure 5).
Subject ID: ____________ Sample No. : ______
5 4 3 2 1
Adorable Not Adorable
Appealing Not Appealing
Beautiful Not Beautiful
Boring Not Boring
Calm Not Calm
Figure 5: The Sample Kansei Checklist.
4.4 Participants
Participants of the Kansei evaluation are
undergraduates from a local university from four
faculties: information technology, engineering,
architecture and business management. From each
faculty, thirty students consisting of fifteen males
and fifteen females with prior experience as web
users were selected.
4.5 Procedure
Four Kansei evaluation sessions were held for each
faculty group. In each session a briefing was given
before the participants began their evaluation
exercise. The thirty five website specimens were
shown one by one from a PC controlled projector to
all participants in a systematic and controlled
manner. Participants were asked to rate their feelings
into the checklist according to the given scale and
were given three minutes to rate their feelings
towards each specimen. A ten minutes break is given
after the fifteenth website specimen was shown.
Each Kansei evaluation session took approximately
two hours to complete.
5 RESULTS AND DISCUSSION
5.1 Reliability of Kansei Measurement
The Cronbach's alpha of the Kansei checklist was
calculated to measure the internal consistency and
the value obtained is 0.9512, which is higher than
the common benchmark value of 0.7. This confirms
the reliability of the Kansei checklist. We further
check on the participants response towards the web
specimens in accordance with Kansei study by Hon
and Ho (2006) who determined the average and
range of the participants’ Kansei score. The plot
values are shown in Figure 6.
Kansei Responses vs. Sample No.
0,00
0,50
1,00
1,50
2,00
2,50
3,00
3,50
4,00
1357911131517192123252729313335
Samp l e N o.
A verage
Rang e
Figure 6: Averaged value and range of evaluation results
between subjects.
The Kansei scores for each specimen is well
distributed above and below the value 3, which is
the neutral response point. This indicates that
subjects are responding well to the specimens.
Hence, from both readings we could conclude that
participants are sensitive to Kansei value.
5.2 Cluster Analysis and Partial Least
Square (PLS)
The participants’ Kansei score were then analyzed
using cluster analysis and the Partial Least Squares
(PLS) method to identify the Kansei word clusters
(Appendix I) to uncover the implicit relations
between KW and website designs elements. The
cluster analysis was performed to identify semantic
space of Kansei words. The result of the PC loadings
for the first and second principal components from
the evaluation result is shown in Figure 7.
Figure 7: PC Loadings result.
-0.5 0.0 0.5 1.0
-0.6
-0.2
0.2 0.6
PC 1
Adorable
Appealing
Beautiful
Boring
Calm
Charming
Chic
Childish
Classic
Comfortable
Cool
Creative
Crowded
Cute
Elegant
Feminine
Fun
Futuristic
Gorgeous
Impressive
Interesting
Light
Lively
Lovely
Luxury
Masculine
Mystic
Natural
Neat
Old.fashioned
Plain
Pretty
Professional
Refres hing
Relaxing
Sexy
Simple
Sophisticated
Stylish
Surreal
PC 2
-0.5 0.0 0.5 1.0
-0.6
-0.2
0.2 0.6
PC 1
Adorable
Appealing
Beautiful
Boring
Calm
Charming
Chic
Childish
Classic
Comfortable
Cool
Creative
Crowded
Cute
Elegant
Feminine
Fun
Futuristic
Gorgeous
Impressive
Interesting
Light
Lively
Lovely
Luxury
Masculine
Mystic
Natural
Neat
Old.fashioned
Plain
Pretty
Professional
Refres hing
Relaxing
Sexy
Simple
Sophisticated
Stylish
Surreal
PC 2
APPLYING KANSEI ENGINEERING TO DETERMINE EMOTIONAL SIGNATURE OF ONLINE CLOTHING
WEBSITES
145

The PC loadings were used to obtain semantic
structure of KW. The KW that produced large
positive first PC loadings (x-axis) are "Elegant",
"Gorgeous", “Stylish", and so forth. The dense area
of the right hand side of the chart corresponds to
such KW. On the other hand, KW that produced
large negative PC loadings is "Boring" and "Old-
fashioned". We represent this PC as the axis of
"Attractiveness". We can expect that websites with a
higher score on this component is likely to have
higher sense of attraction and conversely. In the
second PC loadings (y-axis), KW that have positive
large loadings are "Masculine", and KW that have
negative PC loadings are "Feminine", and "Chic".
We represent this PC as the axis of "Masculine-
Feminine". We can expect that websites with a high
score on this component will tend to have high
characteristic of masculinity and conversely.
The PLS analysis reveal the relationship between
design elements and the Kansei cluster (Appendix
II). For example for the Kansei cluster simple, the
design elements related includes page orientation:
content; page colour: brown; logo location: centre,
main background colour: blue and so on. From the
PCA and PLS performed, we identified the Kansei
semantic space and relationships between Kansei
and web design elements. From the results of our
findings, we can conclude that Kansei structure of
online clothing website design has two components:
attractive and masculine-feminine. This provides us
with the empirical evidence that Kansei EC website
design method proposed can produce the intended
Kansei semantic space, similar to other Kansei
Engineering studies in other product design.
6 CONCLUSION AND FUTURE
WORK
Work on the application of KE in website design is
still at the infancy stage. The dimension of
desirability draws on the new paradigm of producing
desirable websites as opposed to current focus on
website functional usability and performance. Our
study aims to explore whether the Kansei web
design method can determine the emotional values
to generate emotional signatures of online clothing
websites. From our findings we are able to gain
evidence that the proposed Kansei EC website
design method can produce the intended Kansei
semantic space, similar to other Kansei Engineering
studies in other product design. In addition, our
results also showed that the presence of Kansei
appeal in existing EC website.
However, the adoption of KE is not risk free.
Since Kansei is highly dependent on the indigenous
characteristics of the cultural race, engineering
Kansei into EC websites may not produce globally
accepted features. Consideration on the universal
and localized Kansei features will be considered in
our future work.
ACKNOWLEDGEMENTS
The research was supported by grants from the
Ministry of Science, Technology and Innovation,
Malaysia, under the e-Science Fund research grant
scheme [Project Code: 01-01-01-SF0029].
REFERENCES
Aboulafia, A., and Bannon, L. J. (2004). Understanding
affect in design: an outline conceptual framework.
Theoretical Issues in Ergonomics Science, 5(1), 4-15.
Anitawati, M. L. and Nor Laila, M.N. (2006) Kansei
engineering concept in EC website. Proceedings of the
International Conference on Kansei Engineering and
Intelligent Systems 2006 (KEIS '06), Aizu-Wakamatsu,
Japan.
Buchanan, R. 2000. Good Design in the Digital Age.
AIGA Journal of Design for the Network Economy,
1(1), American Institute of Graphic Arts: pp: 1-5.
Dillon, A. (2001). Beyond Usability: Process, Outcome
and Affect in human computer interactions. Paper
presented as the Lazerow Lecture 2001, University of
Toronto, March 2001
Hon, Y.S. and Ho, K.L. (2005). Visual Design of Web
Page. Proceedings of CybErg 2005. The Fourth
International Cyberspace Conference on Ergonomics.
Johannesburg: International Ergonomics Association
Press
Li, Na and Ping Zhang (2005), Consumer Online
Shopping Behavior, in Customer Relationship
Management, Fjermestad, J. and Romano, N. (eds),
Series of Advances in Management Information
Systems, Zwass, V. (editor-in-chief), M.E. Sharpe
Publisher.
Nagamachi, M. 2003. The Story of Kansei Engineering (in
Japanese), Japanese Standards Association, 6. Tokyo.
Nagasawa, S. 2004. Present State of Kansei Engineering
in Japan. 2004 IEEE International Conference. Vol.1.
pp:333-338.
Norman, D. A. 2002. Emotional Design: Attractive
Things Work Better in Interactions: New Visions of
Human-Computer Interaction, IX: pp. 36-42.
Norman, D. A. 2004. Emotional Design: Why We Love
(or Hate) Everyday Things. New York: Basic Books.
Sigala, M., Mich, L. and Murphy, J. 2007. Exploring the
Use of Emotional Features at Romantic Destination
Websites. Proceedings of the International
ICEIS 2008 - International Conference on Enterprise Information Systems
146

Conference in Information and Communication
Technologies in Tourism 2007, Ljubljana, Slovenia,
2007 Springer Vienna.
Spillers, F. 2004. Emotion as a Cognitive Artifact and the
Design Implications for Products that are Perceived as
Pleasurable, Experience Dynamics.
Tractinsky, N, Katz, A.S. and Ikar, D. 2000. What is
Beautiful is Usable. Interacting with Computers, 13:
127-145
APPENDIX I
Kansei Words Cluster.
CLUSTER NAME KANSEI WORDS
PLEASING Adorable, Appealing, Beautiful, Calm, Charming, Comfortable, Cool, Elegant, Gorgeous, Impressive, Interesting,
Lively, Lovely, Pretty, Professional, Refreshing, Relaxing, Sexy, Stylish
OLD-STYLE Boring, Old-fashioned
ADVANCE Chic, Classic, Creative, Cute, Fun, Futuristic, Luxury, Sophisticated, Surreal
STRONG Masculine, Mystic
FEMININE Feminine
SIMPLE Simple
CHILDISH Childish
CROWDED Crowded
APPENDIX II
Part of Relations Between Kansei Responses and Design Elements.
Design Element PLEASING CHILDISH CROWDED FEMININE SIMPLE ADVANCE OLD-STYLE STRONG
Bg Color Dk Brown Lt Blue Grey Lt Blue Black
Bg Style Texture Texture Texture
Page Shape Sharp
Page Menu Shape Mix
Page Style Table None
Page Orientation Content Banner-Content Content Content
Dominant Item Advertisement Text
Page Color Grey Not Specific Pink Brown Grey White Black
Page Size Small Small
Page Border Existence Yes
Logo Location Center Right Center Center Center
Header Bg Color Grey Red Black Grey Blue Grey
Header Bg Picture Picture
Header Font Size Medium
Header Menu Link Style Button Button
… … … … … … … … …
APPLYING KANSEI ENGINEERING TO DETERMINE EMOTIONAL SIGNATURE OF ONLINE CLOTHING
WEBSITES
147