Open Journal of Applied Sciences, 2012, 2, 198-208
doi:10.4236/ojapps.2012.23030 Published Online September 2012 (http :/ /www.SciRP.org/journal/ojapps)
Kansei Engineering Applied to the Form Design of
Injection Molding Machines
Ming-Shyan Huang*, Hung-Cheng Tsai, Wei-Wen Lai
Department of Mechanical and Automation Engineering and Graduate Institute of Industrial Design,
National Kaohsiung First University of Science and Technology, Kaohsiung, Taiwan (China)
Received June 19, 2012; revised July 20, 2012; accepted July 30, 2012
ABSTRACT
This study investigated the relationship between a subject’s evaluation of injection molding machines (IMMs) and for-
mal design features using Kansei engineering. This investigation used 12 word pairs to evaluate the IMM configurations
and employed the semantic differential method to explore the perception of 60 interviewees of 12 examples. The re-
lationship between product featu re design and corresponding wo rds was derived by multiple regression an alysis. Factor
analysis reveals that the 12 examples can be categorized as two styles—advanced style and succinct style. For the ad-
vanced style, an IMM should use a rectangular form for the clamping-unit cover and a full-cover for the injection-unit.
For the succinct style, the IMM configuration should use a beveled form for the safety cover and a vertical rectangular
form for the clamping-unit cover. Quantitative data and suggested guidelines for the relationship between design fea-
tures and interviewee evaluations are useful to product designers when formulating design strategies.
Keywords: Category Classification; Kansei Engineering; Injection Molding Machines; Product Feature Design;
Semantic Differential Method
1. Introduction
The product development trend is changing to a user-
oriented development, i.e., user feelings are recognized
as invaluable during product development. Thus, designers
must consider user preferences and needs to increase
product success. Product configuration is linked with
user perceptions of products, and the resulting product is
a reflection of customer philosophy and preferences.
How to investigate the feelings and preferences of cus-
tomers about a product and then utilizing these feelings
and preferences as a reference when designing products
has become an important issue.
Product design is a process during which designers de-
liver a product design to customers via a product that
satisfies user feelings and preferences. However, opin-
ions of designers and users typically differ. Mental con-
cepts of users are often difficult to identify and, cones-
quently, a satisfactory product is difficult to design. In
particular, designing injection molding machines (IMMs)
often lacks user feedback and relies heavily on technical
specifications and imaginary targets of manufact urers.
Kansei engineering, founded by Nagamachi in 1970, is
a methodology for systematically exploring user percep-
tions about a product and translating these perceptions
into design parameters [1]. If consumer feelings can be
implemented in new product designs, consumers would
be satisfied with such products. The most important task
in Kansei engineering is to survey customers to identify
their preferences, or kansei, at the start of the product
development process. The scale of kansei words, con-
sisting of the semantic differential proper adjectives, is
needed for psychological measurements of customers. If
customer kansei is determined accurately, product de-
velopment will be extremely successful.
Kansei engineering has been applied when designing
automobiles, construction machinery, electrical home
appliances, office machines, and cosmetics [2-7]. Roy et
al. [8] utilized web-based cell phone images and ques-
tionnaires to gather data associated with emotional re-
sponses to existing products that can then be used when
designing new cell phones and pursuing added value in
the consumer marketplace. They developed a method
that can be used to reduce arbitrary decision making,
which is often applied during the product design process.
Chang and Wu [9] explored the mechanisms that domi-
nate user mental models for product form classification.
After analyzing real mobile phones, they concluded that
local features were the dominant mechanisms for classi-
fying a large number of similar products. Classification
based on real product samples can help when exploring
*Corresponding a uthor.

M.-S. HUANG ET AL. 199
the effects of various form factors on user visual percep-
tions. Hsiao et al. [10] developed a support model that
determines the psychological preferences of consumers
via a genetic algorithm. Hsu et al. [11] applied a semantic
differential method to examine the relationship between
subject evaluations of sample telephones and design ele-
ments.
IMMs are characterized with specialized assembling
modules and forms, which are ease of being categorized,
analyzed, and parameterized. Consequently, the form
images of IMMs are proper to be studied via Kansei en-
gineering. Up to the present time, although many studies
have used Kansei engineering during product design, few
have examined IMM designs. This study applied Kansei
engineering for IMM design, including the analysis of
factors affecting product style and the relationship be-
tween product features and its image. Positive and nega-
tive factors that affect product style were obtained via a
questionnaire. Most factors affecting designs were iden-
tified, and design criteria were generated.
2. Methodology
This investigation has the following three stages: 1) data
collection and analysis, including selecting examples of
identical IMM designs and experts choosing suitable
word pairs; 2) a qu estionnaire survey and factor analysis,
including designing a semantic differential scale to col-
lect interviewee opinions about word pairs to evaluate
IMM designs. The design factors corresponding design
features and the suggested design criteria were also gen-
erated; and 3) a design case study for IMMs is per-
formed to demonstrate the effectiveness of this proposed
method.
2.1. Part 1: Data Collection and Analysis
This stage comprises data collection and analysis. Forty
well-known sets of IMM design samples were first ob-
tained from websites and magazines for machine manu-
factures published in 2009. These samples were limited
to small or medium-sized machines with clamping forces
<750 tons. The pictures with the front view of these ma-
chines were 100 × 100 mm; the background, colors, lo-
gos, and model information were removed. Four expert
mechanical engineers with industrial design backgrounds
were invited to select identical IMM design examples.
Using the Kawakita Jiro (KJ) method, these experts clas-
sified samples based on their features and images. Table
1 shows the 12 examples, each representing the style in
one category selected by experts. These sets are example
IMM designs. The following section introduces the KJ
method and category classification.
Table 1. Twelve example IMM designs proposed by experts.
No. Example No.Example
S1
S2
S3 S4
S5 S6
S7
S8
S9
S10
S11
S12

M.-S. HUANG ET AL.
200
2.1.1. KJ Me thod
The KJ method, developed in 1965, achieves efficient
group communication and has become a very popular
decision-making method for expert groups. The KJ
method uses one card that states a single concept and
then categorizes cards into several groups. Experts are
then required to select one representative card from each
group [12].
2.1.2. Category Classification and Word Pai r Selection
Category classification is utilized to convert a verbal
product description into a detailed design. A literature
survey identified 342 word pairs used in design descrip-
tions [13-15]. Based on these word data, experts choose
appropriate adjectives to describe IMM designs. Adjec-
tives for describing IMM designs are used only when
more than 50% the experts agree. The selected best 12
word pairs were as follows: traditional-modern, complex-
simple, massive-compact, tacky-sleek, inferior-superior, ubi-
quitous-unique, imitative-innovative, rough-delicate, curvilin-
ear-foursquare, cheap-expensive, dirty-clean, and disagree-
able-agreeable (Table 2).
2.2. Part 2: Questionnaire Survey and Factor
Analysis
This section verifies the 12 word pairs for IMM designs.
The verification process involves conducting a question-
naire survey administered to experts to select appropriate
design attributes, and analyzing samples using word pair
data to generate suitable adjectives for describing IMM
configurations.
In designing the questionnaire, this study utilized a
seven-level semantic differential (SD) scale (Table 3) to
identify the degree to which a word pair describes an
Table 2. Word pairs describing the IMM configur ation.
Word pairs
Traditional-Modern
Complex-Simple
Massive-Compact
Tacky-Sleek
Inferior-Superior
Ubiquitous-Unique
Imitative-Innovative
Rough-Delicate
Curvilinear-Foursquare
Cheap-Expensive
Dirty-Clean
Disagreeable-Agreeable
Table 3. The IMM semantic differential table.
1234 5 6 7
Traditional □ □ □ □ □ □ □ Modern
Complex □ □ □ □ □ □ □ Simple
Massive □ □ □ □ □ □ □ Compact
Tacky □ □ □ □ □ □ □ Sleek
Inferior □ □ □ □ □ □ □ Superior
Ubiquitous □□ □ □ □ □ □ Unique
Imitative □ □ □ □ □ □ □ Innovative
Rough □ □ □ □ □ □ □ Delicate
Curvilinear □ □ □ □ □ □ □ Foursquare
Cheap □ □ □ □ □ □ □ Expensive
Dirty □ □ □ □ □ □ □ Clean
Disagreeable □ □ □ □ □ □ □ Agreeable
IMM design. To simply the style design problem, the
color factor is excluded in this study. Thus, 12 IMM
examples with different forms were all in gray and with-
out any logos. Interviewees evaluated each set of IMM
examples using the 12 word pairs. For instance, Nos. 1, 4,
and 7 represent full agreement to the left adjective, neu-
tral opinion, and full agreement to the right adjective,
respectively (Table 3).
To assess the 12 example IMM designs, this study in-
terviewed 60 experts of 38 males and 22 females from
two different groups—40 experts with work experience
in mechanical engineering, and 20 experts with work
experience in product design. The questionnaires for the
12 samples were given to these experts to establish the
relationship between form features and image percep-
tions of IMM designs. Experts were asked to express
their judgments by setting the levels on SD scales for all
word pairs in the questionnaires. From the average
judgment values of these experts, the significance calcu-
lated value exceeded 0.05 and it represents the opinions
from the two groups are identical and thus should be
combined in subsequent analyses.
To reveal the reliability and effectiveness of question-
naire survey, reliability analysis is conducted in this
study. In the reliability test, Cronbach  and Split-half
coefficients are calculated. A high value of reliability
coefficient means the questionnaire survey is stable,
namely, the consistency of survey results conducted in
different durations. For evaluating potential variables,

M.-S. HUANG ET AL. 201
assume that the observed value, Xo, and true value, Xt,
are different. The error, Xe, is defined as
,
X
eXtXo (1)
in which, the mean value of error, E(Xe), is assumed to
be zero. Thereby, the following statements are valid.



,EXo EXt (2)
22
XoXt Xe
2

  (3)
In which, 2
Xo

, Xt
2

, and Xe
2

are the variations of
observed value, true value, and error, respectively. The
reliability coefficient, r, is then formulated as
22 22
1
Xt XoXe Xo
r

  (4)
The r value in Equation (4) is highly related with the
ratio of 2
Xt

and 2
Xo

. A higher r value means that er-
ror between true value and observed value is small, i.e., a
high level of reliability in the observed data.
Factor analysis used in this study is to extract major
features among the SD results. This statistical method
used to describe variability among observed variables in
terms of a potentially lower number of unobserved vari-
ables called factors. Factor analysis searches for such
joint variations in respons e to unobserved latent variables.
The observed variables are modeled as linear combina-
tions of the potential factors, plus error terms. The in-
formation gained about the interdependencies between
observed variables can be used later to reduce the set of
variables in a dataset. Principal component analysis
(PCA) is related to factor analysis and performs a vari-
ance-maximizing rotation of the variable space. PCA
takes into account all variability in the variables, while
factor analysis estimates how much of the variability is
due to common factors. The two methods become essen-
tially equivalent if the error terms in the factor analysis
model can be assumed to all have the same variance. The
mathematical model of factor analysis is presented as
follows.
112 2
j
jj jmmj
Z
aFaFa FU  (5)
In which, Zj is the observed score of the jth variable
after standardization (mean value is 0 and standard
deviation is 1), Fi is the ith common factor, Uj is the
unique factor of the Zj variable, and aj1, aj2, ···, ajm are
the factor loadings. The common factors identified from
factor analysis are further used to in the following
multiple regression analysis.
The multiple regression analysis employed here is to
analyze the influential level of destructed form features
on factors, and the formula is listed below:
01
n
ijj
j
YX
it



 

 (6)
where Xji, j = 1, 2, ···, n are the observed values of the ith
variable, Yi is the ith response value, and 
0,

1,

2, ···,
 n are the coefficients of regression model.
2.3. Part 3: Case Design and Verification
This section verifies the feasibility of suggested guide-
lines for design features concluded from Parts 1 and 2,
whereas inferior designs out of the 12 samples were se-
lected for design modification in Photoshop graphical
software. During verification, the seven new design cases
are evaluated by the same 60 experts who are invited as
interviewee in Part 2. The mean value of each modified
design sample is compared with its original design sam-
ple to demonstrate the effectiveness. Meanwhile, t test
are used to evaluate whether the strategy of modification
on these seven samples is significant in promoting these
interviewee’s satisfactory level on product style. Gener-
ally, the verification attempts to ensure that the proposed
design guidelines can generate an ideal form design of
IMM, and provides further information on improving
design insufficiencies.
3. Survey Results and Discussion
This section discuses 1) style analysis; 2) product posi-
tioning analysis; and 3) the influence of design features
on user perceptions.
For reliability analysis, Cronbach’s alpha was 0.94;
thus, survey results are highly reliable and reproducible.
The recruited experts identified six sections of IMM de-
signs with 27 design features (Figure 1 and Table 4).
The six sections of IMM designs are safety-door window
(A), safety-door cover (B), clamping-unit window (C),
clamping-unit cover (D), injection-unit cover (E), and
machine base (F). Each section has four to five design
feature alternatives. For instance, the safety-door window
section contains design feature alternatives of square
(A1), horizontal rectangle (A2), vertical rectangle (A3),
polygon (A4), and round corner (A5). The safety-door
cover section contains design feature alternatives of square
(B1), rectangle (B2), vertical rectangle (B3), polygon
(B4), and bevel form (B5). The clamping-unit window
Figure 1. Description of IMM configurations.

M.-S. HUANG ET AL.
202
Table 4. Design feature analysis of the IMM configuration.
Section Design features
Safety-
door window
(A)
Square (A1) Horizontal rectangle (A2) Vertical rectangl e (A3)Polygon (A4 ) Round corner (A5)
Safety-
door cover (B)
Square (B1) Rectangle (B2) Vertical rectangle (B3)Polygon (B4) Beveled form (B5)
Clamping-unit
window (C)
Horizontal rectangle (C1) Vertical rectangle (C2) Square (C3) Irregular form (C4) No window (C5)
Clamping-unit
cover (D)
Square (D1) Arc (D2) Beveled form (D3) Rectangle (D4)
Injection-unit
cover (E)
Semi & round-corner cover (E1) Semi & irregular cover (E2)Small-area cover (E3) Full-cover (E4)
Machine base
(F)
Beveled-form cover linking wit h machine
base and clampi n g unit (F1) Rectangle-form cover linking with machine base and clampi n g unit (F2)
Partial cover linking with machine base
and clamping unit (F3) F4. No cover linking with machine base and clamping unit (F4)

M.-S. HUANG ET AL. 203
section contains design feature alternatives of horizontal
rectangle (C1), vertical rectangle (C2), square (C3), ir-
regular form (C4), and no window (C5). The clamping-
init cover section contains design feature alternatives of
square (D1), arc (D2), beveled form (D3), and rectangle
(D4). The injection-unit cover section has design feature
alternatives of semi & round-corner cover (E1), semi &
irregular cover (E2), small-area cover (E3), and full-
cover (E4). Finally, the machine base section has design
feature alternatives of beveled-form cover linking with
machine base and clamping unit (F1), rectangle-form
cover linking with machine base and clamping unit (F2),
partial cover linking with machine base and clamping
unit (F3), and no cover linking with machine base and
clamping unit (F4). Each IMM configuration was ana-
lyzed with its individual design features and encoded
with the defined codes. Table 5 shows feature analysis of
the IMM designs. These data were further used in multi-
ple regression analysis to identify the influence of each
feature on style and appeal.
Table 5. IMM feature analysis.
Example
Feature S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12
A1 1 0 0 0 0 0 0 0 0 0 0 1
A2 0 1 0 1 0 0 1 1 1 0 0 0
A3 0 0 0 0 0 0 0 0 0 0 1 0
A4 0 0 1 0 1 0 0 0 0 0 0 0
A5 0 0 0 0 0 1 0 0 0 1 0 0
B1 1 0 0 0 0 0 0 0 0 1 0 1
B2 0 1 0 1 0 0 1 1 1 0 0 0
B3 0 0 0 0 0 0 0 0 0 0 1 0
B4 0 0 1 0 1 0 0 0 0 0 0 0
B5 0 0 0 0 0 1 0 0 0 0 0 0
C1 1 0 0 0 0 0 0 0 1 0 0 0
C2 0 1 0 0 0 0 1 1 0 0 0 0
C3 0 0 0 0 0 0 0 0 0 0 1 0
C4 0 0 0 0 1 0 0 0 0 0 0 0
C5 0 0 1 1 0 1 0 0 0 1 0 1
D1 1 0 0 0 0 0 0 0 1 0 1 1
D2 0 0 0 1 0 0 0 1 0 1 0 0
D3 0 0 1 0 1 0 0 0 0 0 0 0
D4 0 1 0 0 0 1 1 0 0 0 0 0
E1 0 0 0 0 0 0 0 0 0 1 0 0
E2 1 0 1 1 0 0 1 0 1 0 0 1
E3 0 0 0 0 1 0 0 1 0 0 0 0
E4 0 1 0 0 0 1 0 0 0 0 1 0
F1 0 1 1 0 1 0 0 0 0 0 0 0
F2 0 0 0 1 0 1 1 1 1 1 0 0
F3 0 0 0 0 0 0 0 0 0 0 1 1
F4 1 0 0 0 0 0 0 0 0 0 0 0

M.-S. HUANG ET AL.
204
3.1. Style Analysis
Table 6 lists semantic differential valu es of the word pair
evaluations of the 12 examples (S1-S12); an average
>5.0 indicates a high degree of agreement to the right
adjective, that of 3.0 - 5.0 is neutral agreement between
the word pair, and that <3.0 indicates a high degree of
agreement to the left adjective. Analytical results are as
follows.
1) Examples with a high degree of agreement were S1
for “foursquare” and “clean,” S2 for “modern” and “su-
perior,” S6 for “clean,” S9 for “foursquare,” S10 for
“simple” and “clean,” S11 for “modern” and “four-
square,” and S12 for “foursquare”.
2) The only example with a disagreement was S12 for
“compact” only.
3) Images of “foursquare” and “clean” had the highest
average scores, indicating that most examples have these
styles.
Table 7 also lists factor analysis results for these 12
examples; the three highest scores are marked with solid
circles and the three lowest scores are marked empty
circles. Thus, S2, S6, S10, and S11 are most agreeable
described by various word pairs. Additionally, S8, S9,
and S12 have the lowest agree ment for word pairs. Nota-
bly, S1, S11, and S12 are examples of “foursquare and
S2, S3, and S10 are those of “curvilinear”. Since the
word “foursquare” describes product appearance and
without assessment, it is analyzed individually.
Table 6. Semantic differential values.
Styling High degree
(>5.0) Neutral degree (3.0 - 5.0) Reverse degree (<3.0) Average
Modern S2, S11 (16.7%) S1, S3, S4, S5, S6, S7, S8, S9, S10, S12 (83. 3% ) 4.33
Simple S10 (8.3%) S1, S2, S3, S4, S5, S6, S7, S8, S9, S11, S12 (91.7%) 4.27
Compact S1, S2, S3, S4, S5, S6, S7, S8, S9, S10, S11 (91.7%) S12 (8.3%) 3.70
Sleek S2 (8.3%) S1, S3, S4, S5, S6, S7, S8 , S9, S10, S11, S12 (91. 7 %) 4.17
Superior S2 (8.3%) S1, S3, S4, S5, S6, S7, S8, S9, S10, S11, S12 (91.7%) 4.00
Unique S1, S2, S3, S4, S5, S6, S7, S8, S9, S10, S11, S12 (100%) 4.01
Innovative S11 (8.3%) S1, S2, S3, S4, S5, S6, S7, S8, S9, S10, S12 (91.7%) 4.13
Delicate S1, S2, S3, S4, S5, S6, S7, S8, S9, S10, S11, S12 (100%) 4.12
Foursquare S1, S9, S11, S12 (3 3% ) S2, S3, S4, S5, S6, S7 , S8, S10 (67%) 4.77
Expensive S2 (8.3%) S1, S3, S4, S5, S6 , S 7 , S8, S9, S10, S11, S12 (91.7%) 4.09
Clean S1, S6, S10 (25%) S2, S3, S4, S5, S7, S8, S9, S11, S12 (75%) 4.62
Agreeable S1, S2, S3, S4, S5, S6, S7, S8, S9, S10, S11, S12 (100%) 4.34
Table 7. Factor analysis of 12 IMM designs.
Example
Styling S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12
Modern ● ○ ○ ● ● ○
Simple ● ○ ● ○ ● ○
Compact ○ ● ○ ● ● ○
Sleek ○ ● ● ○ ● ○
Superior ○ ●
● ○ ● ○
Unique ○ ● ○ ● ● ○
Innovative ● ○ ○ ● ● ○
Delicate
● ○ ○ ● ● ○
Foursquare ● ○ ○ ○ ● ●
Expensive ○ ● ● ○ ● ○
Clean ● ○ ● ○ ● ○
Agreeable ● ● ○ ● ○ ○
●: Highly relative; ○: Lowly relative.

M.-S. HUANG ET AL. 205
3.2. Product Positioning Analysis
This study utilized factor analysis and principal compo-
nent analysis (PCA) to investigate the word pair com-
monality shared by interviewees for the descriptions of
the 12 examples. The aim of factor analysis is to identify
the common factor among word pairs and the outcome is
used for cluster analysis to investigate the style common
to each cluster. Thus, the difference among clusters is re-
vealed. Through the Kaiser-Mayer-Olkin (KMO) meas-
ure and Bartlett evaluation, the KMO value was 0.714
and, thus, factor analysis was validated [16]. In factor analy-
sis, compact-massive and foursquare-curvilinear had little
commonality and with a low KMO value; thus, both were
deleted. The rema ining 10 word pairs were simpl ified into
two factor axes (Table 8). The first factor axis comprised
traditional-modern, cheap-expensive, ubiquitous-unique,
tacky-sleek, inferior-superior, imitative-innovative, and
rough-delicate. These seven sets of word pairs were then
defined as advanced style (accounting for 66.54% of
variance). The second factor axis comprised of dirty-
clean, massive-compact, and disagreeable-agreeable. These
three sets of word pairs were defined as succinct style
(accounting for 23.99% of variance). These categories
accounted for 93.55% of variance.
Figure 2 shows the scores for the first principal com-
ponent (advanced style) and second principal component
(succinct style). These two components constructed the
recognition space, which was divided into the following
four groups. 1) Group 1—advanced style and succinct
style (S6 and S10). These designs are highly advanced
Table 8. Factor analysis and classification of 12 IMM de-
signs.
Weighting value
Factor Word pairs
Factor 1 Factor 2
Modern-Traditional 0.98 0.05
Expensive-Cheap 0.97 –0.10
Unique-Ubiquitous 0.96 0.09
Sleek-Tacky 0.96 0.10
Superior-Inferior 0.96 0.07
Innovative-Imitative 0.93 0.07
Advanced style
Delicate-Rough 0.91 0.24
Clean-Dirty 0.09 0.95
Compact-Massive –0.20 0.93 Succinct style
Agreeable-Disagreeable 0.38 0.81
Eigenvalue 6.65 2.40
% of Variance 66.54 23.99
Cumulative explanation (%) 66.54 90.53
and succinct. 2) Group 2—succin ct style (S1, S3, S4, S7,
and S9). 3) Group 3—neither advanced style nor suc-
cinct style (S5, S8, and S12). Notably, S12 performed
worst. 4) Group 4—advanced style (S2 and S11). These
two high-quality machines made in Europe were catego-
rized as modern, sleek, and unique.
3.3. Influence of Features on User Perceptions
Table 9 lists th e significant features id entified by multip le
regression analysis. Interviewees believed the design
feature of a full-cover over the injection-unit (E4) is as-
sociated with an advanced style (Table 9(a)). Conversely,
small-area cover in the injection-unit (E3) is dirty, com-
plex, disagreeable, and negative to the succinct style and
should be avoided. Table 9(b) lists the positive and nega-
tive features for the advanced style and succinct style.
Suggestions for feature selection are summarized as follows.
1) For the advanced style, an IMM design should use a
rectangular form for the clamping-unit cover and a full-
cover for the injection-unit. The rectangular co ver linked
to the machine base and a ratio of 1 to 3 should be avoided.
2) For the succinct style, an IMM design should use a
beveled form as a safety cover and a vertical rectangle
for the clamping-unit cover. A horizontal rectangle as the
safety-door window, no window over the clamping unit,
a beveled form for the clamping-unit cover, small-area
cover for the injection- unit, an d no cover link ing with the
machine base should be avoided.
4. Case Redesign and Verification
To improve the positive images of IMM designs on the
advanced style or the succinct style, 7 IMM examples
(S3, S5, S7, S8, S9, S11 and S12) were selected for fur-
ther refining. Table 10 lists the new design cases
Figure 2. Principal component analysis of 12 IMM configu-
rations.

M.-S. HUANG ET AL.
206
Table 9. Analysis of designing features.
(a) Significant design factors versus emotional assessments.
Advanced style Positive features Negative features
Traditional-Modern E4 –
Cheap-Expensive E4, D2 F2
Ubiquitous-Unique E4 –
Tacky-Sleek E4, D2 –
Inferior-Superior E4 –
Imitative-Innovative E4 –
Succinct style Positive features Negative features
Dirty-Clean – E3, F3, A2, C5, D3
Massive-Compact – E3
Disagreeable-Agreeable C3, B5 E3, F3
(b) Significant design factors versus emotional assessments.
Factor Advanced style Succinct styl e
A. Safety-door window A5 A1
B. safety-door cover B3 B3
C. Clamping-unit window C3 C1
D. Clamping-unit cover D2 D4
E. Injection-unit cover E4 E2
F. Machine base F1 F4
(M1-M7) us ed f or ve rificatio n on suggested g u idelin es of
IMM design features which are listed in Table 9. In the
new design cases, negative features are considered to be
replaced with pos itive features in priority, as are listed in
Table 11. For example, M1 was redesigned by replacing
C5, F3 with C3, F4 in S12, respectively. Namely, the
original no-window style at the rear cover of clamping
unit was replaced with square window. Also, the original
design of machine base using partial cover linking with
machine base and clamping unit was replaced with no
cover. M2 was redesigned by replacing E3 with E4 in S5.
M3 was redesigned by replacing A2, F2 with A4, F1 in
S7, respectively. M4 was redesigned by replacing A2,
E3, F2 with A5, E4, F1 in S8, respectively. M5 was re-
designed by replacing C5, D3 with C3, D2 in S3, respec-
tively. M6 was redesigned by replacing F3 with F1 in
S11. M7 was redesigned by replacing A2, F2 with A1, F1
in S9, respectively. These seven design cases are again
surveyed for 12 imagery word-pairs by the same 60 ex-
perts who have assessed the original 12 samples before.
The results are further statistically analyzed through
t-test to assess whether the new design samples are valid
in improving image style whereas the t-value is com-
puted, i.e., by comparing to the ind ividu al origin al d esign
sample, the modified design is significant if the t-test is
less than 0.05. Table 11 depicts that the mean value of
every modified sample is significantly higher than the
corresponding original one, and these results indicate that
the design guidelines for IMM feature in Table 9 are
practical.
Table 10. List of seven-pair original and modified design cases.
Case Original CaseModified
S12 M1
Replaced features: no clamping-unit window (C5) and partial
cover linking with machine base and clamping unit (F3) Refined features: square clamping-unit window (C3) and no cover linking with machine base and clamping unit (F4)
S5 M2
Replaced features: small-area injection-unit cover (E3) Refined features: full injection-unit cover (E4)
S7
M3
Replaced features: horizontal rectangle safety-door window (A2) and
rectangle-form cover linking wit h machine base and clamping unit (F2) Refined features: polygon safety-do o r window (A4) and beveled-form cover linking with machine b ase and clamping unit (F1)

M.-S. HUANG ET AL. 207
S8 M4
Replaced features: horizontal rectangle safety-door window (A2),
small-area injection-unit cover (E3), and rec tangle-form cover
linking with m achine base and clampin g unit (F2)
Refined features: round corner safety-door window (A5), full
injection-unit cover (E4), and beveled-form cover linking
with machine base a nd clamping unit (F1)
S3 M5
Replaced features: no clamping-unit window (C5), and beveled
form clamping-unit cover ( D3) Refined features: square clamping-unit window (C3), and arc clamping-unit cover (D2)
S11 M6
Replaced features: partial cover linking with machine base
and clamping unit (F3) Refined features: beveled-form cover linking with machine base and clampi n g unit (F1)
S9
M7
Replaced features: horizontal rectangle safety-door window (A2),
and rectangle-form cover linking with machine base and clamping unit (F2)Refined features: square safety-door window (A1), and beveled-form cover linking with machine b ase and clamping unit (F1)
Table 11. Validation of modified design cases with t test.
Feature displacement Mean value
Case Original Modified Original Modified Significance
M1 C5, F3 C3, F4 3. 78 4.39 Yes
M2 E3 E4 4.23 5.03 Yes
M3 A2, F2 A4, F1 4.18 4.64 Yes
M4 A2, E3, F2 A5, E4, F1 4.18 4.76 Yes
M5 C5, D 3 C3, D2 4.21 4.70 Yes
M6 F3 F1 4.77 5.27 Yes
M7 A2, F2 A1, F1 4.16 4.63 Yes
5. Conclusions
This work applied the Kansei engineering approach to
investigate the relationship between images of IMM de-
signs and design features. The semantic difference method
was utilized to explore the recognitio n of 12 examples of
produ ct configu ration s from 60 interv iewees. In this stud y,
IMM design was deconstructed into six sections and 27
design features (Table 4). The configuration of each
IMM example was analyzed with its features. Configura-
tion analysis results were further used in multiple regres-
sion analysis to identify the influence of each configura-
tion on style appeal. The principal conclusions are as
follows.
1) Appropriate words describing an IMM design has
the following 12 word pairs traditional-modern, complex-
simple, massive-compact, tacky-sleek, inferior-superior,
ubiquitous-unique, imitative-innovative, rough-delicate, cur-
vilinear-foursquare, cheap-expensive, dirty-clean, and dis-
agreeable-agreeable.
2) Factor analysis results and principal component
analysis results suggest that the 12 word pairs can be
simplified as two factor axes. The first factor axis, de-
fined as advanced style, comprised traditional-modern,
cheap-expensive, ubiquitous-unique, tacky-sleek, inferior-
superior, imitative-innovative, and rough-delicate. The sec-
ond factor axis, defined as succinct style, co mp ri s ed d i r ty -

M.-S. HUANG ET AL.
208
clean, massive-compact, and agreeable-agreeable. These
two factors account fo r 93 .5 5 % of vari ance.
3) Multiple regression analysis for features indicates
that, for the advanced style, an IMM design should use
a rectangular form for the clamping-unit cover and a
full-cover for the injection-unit. A rectangular cover
linked to the machine base at a ratio of 1 to 3 should be
avoided.
4) For the succinct style, an IMM design should use a
beveled form for the safety cover and a vertical rectangle
for the clamping-unit cover. The following should be
avoided: a horizontal rectangular form for the safety-door
window; a clamping unit without a window; no cover for
the injection-unit, and no cover linked to the machine
base.
5) This study was conducted to support machinery
configuration design as a scientific meth od incorporating
human sensibilities into applied designs. Factor analysis
of appropriate product style design by comparative evalu-
ation using Kansei values can be utilized in generating
product design strat e gi es.
6. Acknowledgements
The authors would like to thank Fu Chun Shin Machin-
ery Manufacture Co. Ltd. for financial support on this
study. Ted Knoy is also appreciated for his editorial as-
sistance.
REFERENCES
[1] M. Nagamachi, “Kansei Engineering: A New Ergonomic
Consumer-Oriented Technology for Product Develop-
ment,” International Journal Industrial Ergonomics, Vol.
15, No. 1, 1995, pp. 3-11.
doi:10.1016/0169-8141(94)00052-5
[2] T. Jindo and K. Hirasago, “Application Studies to Car
Interior of Kansei Engineering,” International Journal
Industrial Ergonomics, Vol. 19, No. 2, 1997, pp. 105-114.
doi:10.1016/S0169-8141(96)00007-8
[3] K. Nakada, “Kansei Engineering Research on the Design
of Construction Machinery,” International Journal In-
dustrial Ergonomics, Vol. 19, No. 2, 1997, pp. 129-146.
doi:10.1016/S0169-8141(96)00009-1
[4] M. Nagamachi, “Kansei Engineering as a Powerful Con-
sumer-Oriented Technology for Product Development,”
Applied Ergonomics, Vol. 33, No. 3, 2002, pp. 289-294.
doi:10.1016/S0003-6870(02)00019-4
[5] S. Mondragón, P. Company and M. Vergara, “Semantic
Differential Applied to the Evaluation of Machine Tool
Design,” International Journal Industrial Ergonomics,
Vol. 35, No. 11, 2005, pp. 1021-1029.
doi:10.1016/j.ergon.2005.05.001
[6] E. A. Demirtas, S. Anagun and G. Koksal, “Determina-
tion of Optimal Product Styles by Ordinal Logistics Re-
gression versus Conjoint Analysis for Kitchen Faucets,”
International Journal Industrial Ergonomics, Vol. 39, No.
5, 2009, pp. 866-875.
doi:10.1016/j.ergon.2009.06.007
[7] M.-S. Huang, H. C. Tsai and T. H. Huang, “Applying
Kansei Engineering to Industrial Machinery Trade Show
Booth Design,” International Journal Industrial Ergo-
nomics, Vol. 41, No. 1, 2011, pp. 72-78.
doi:10.1016/j.ergon.2010.10.002
[8] R. Roy, M. Goatman and K. Khangura, “User-Centric
Design and Kansei Engineering,” CIRP Journal of Manu-
facturing Science and Technology, Vol. 1, No. 3, 2009,
pp. 172-178. doi:10.1016/j.cirpj.2008.10.007
[9] C. C. Chang and J. C. Wu, “The Underlying Factors
Dominating Categorical Perception of Product Form of
Mobile Phones,” International Journal Industrial Ergo-
nomics, Vol. 39, No. 5, 2009, pp. 667-680.
doi:10.1016/j.ergon.2009.02.012
[10] S. W. Hsiao, F. Y. Chiu and S. H. Lu, “Product-Form
Design Model Based on Genetic Algorithms,” Interna-
tional Journal Industrial Ergonomics, Vol. 40, No. 3, 2010,
pp. 237-246. doi:10.1016/j.ergon.2010.01.009
[11] S. H. Hsu, N. C. Chuang and C. C. Chang, “A Semantic
Differential Study of Designers’ and Users’ Product Form
Configuration,” International Journal Industrial Ergo-
nomics, Vol. 25, No. 4, 2000, pp. 375-391.
doi:10.1016/S0169-8141(99)00026-8
[12] K. Jiro, “KJ Method: A Scientific Approach to Problem
Solving,” Kawakita Research Institute, Tokyo, 2000.
[13] C. C. Tseng, “The Application of Kansei Engineering to
the Image Recognition Analysis and Computer-Aided
Logo Design for Banks,” Ph.D. Thesis, National Cheng
Kung University, Taiwan, 2003.
[14] C. H. Wang, “A Study on the Influence of the Features of
a Bicycle’s Frame Form on the Image Perception,” Mas-
ter Thesis, National Cheng Kung University, Taiwan,
2003.
[15] M. S. Shen, “A Study on the Combination of Multiple
Kansei Words,” Master Thesis, National Yunlin Univer-
sity of Science and Technology, Taiwan, 2004.
[16] J. C. Jones, “Design Methods,” Van Nostrand Reinhold,
London, 1992.