Description
The purpose of this paper is to understand online consumer behavior better by analyzing online consumers'
decision-making styles. In this research, an online consumer style inventory, which is suitable
for online businesses to measure online consumers' decision-making styles, has been developed in
Macau. The current studies of online consumer behavior have not considered the weights of the variables
that can affect online consumer behavior. This paper measures online consumers' decision-making styles
in Macau based on the weights of the corresponding inventory items.
Online consumer decision-making styles for enhanced understanding of Macau
online consumer behavior
Kin Meng Sam
a, *
, Chris Chatwin
b
a
Department of Accounting and Information Management, University of Macau, Taipa, Macau, China
b
School of Engineering and Informatics, University of Sussex, Brighton, United Kingdom
a r t i c l e i n f o
Article history:
Received 16 January 2013
Accepted 9 April 2014
Available online 19 March 2015
Keywords:
Factor score coef?cient
Online consumer behavior
Online consumer style inventory
Online consumers'decision-making styles
a b s t r a c t
The purpose of this paper is to understand online consumer behavior better by analyzing online con-
sumers' decision-making styles. In this research, an online consumer style inventory, which is suitable
for online businesses to measure online consumers' decision-making styles, has been developed in
Macau. The current studies of online consumer behavior have not considered the weights of the variables
that can affect online consumer behavior. This paper measures online consumers' decision-making styles
in Macau based on the weights of the corresponding inventory items.
© 2015, College of Management, National Cheng Kung University. Production and hosting by Elsevier
Taiwan LLC. All rights reserved.
1. Introduction
Decision making is more complex and important for consumers
than in the past (Hafstrom, Chae, & Chung, 1992; Lysonski,
Durvasula, & Zotos, 1996). The increased varieties of products and
the abundance of information through company advertisements
have broadened the choices for consumers. A consumer decision-
making style refers to a mental orientation describing how a con-
sumer makes choices (Durvasula, Lysonski, & Andrews, 1993).
Pro?ling consumer decision-making styles is very important to
marketers and advertisers (Lysonski et al., 1996). In order to deal
with the emergence of e-commerce activities, it is necessary to
consider online consumers' decision-making styles that in?uence
the willingness of online consumers to purchase products.
China's Internet penetration rate reached 40.1% during June 2012,
exceeding the world's average of 34.3% at the same point (Internet
World Stats, 2012). Academic researchers suggested that China's
cultural history of preferring face-to-face business interactions,
coupled with its restrictive regulatory environment, may hinder the
development of online shopping in China (Raven, Huang, & Kim,
2007). Nevertheless, China's online retail market has been growing
steadily with an increase of 45.9% in the number of online shopping
users in 2009, leading to a total of 156 million users (AbouTourism,
2010), compared to 154 million users in the USA who shopped on-
line in 2009 (Forrester Forecast, 2010); online shopping penetration
in China still appears to hold considerable growth potential.
Macau has been a prosperous Chinese city since 2004, after the
gambling industry was opened up to external investors; this caused
the gross domestic product for the ?rst quarter of 2011 to expand
by 21.5% (New Zealand Consulate-General Hong Kong, 2011).
Macau's living standards have risen by over 300% in just one dec-
adedit is now destined to become the richest territory in Asia
(excluding the Middle East) by gross domestic product per capita
(Zimbabwemetro, 2011). According to Taobao (2010), Macau is the
city with the fastest increase in online shopping of all the cities in
China. For this reason it is valuable to focus on the analysis of online
consumers in Macau.
1.1. Consumer style inventory
One of the ways to characterize consumer styles is consumer
characterization, focusing on cognitive and affective orientations
related to consumer decision-making (Sproles, 1985). Consumer
characterization is very promising as it deals with the cognitive
orientation of consumers in making decisions (Sproles & Kendall,
1986). Sproles and Kendall (1986) designed a 40-item Consumer
Style Inventory (CSI) model to measure decision-making styles of
consumers based on a sample of US youth. The applicability of the
CSI model has been investigated across several cultures such as
Korea (Hafstrom et al., 1992), New Zealand (Durvasula et al., 1993;
Lysonski et al., 1996), India (Lysonski et al., 1996), Greece (Lysonski
* Corresponding author. Department of Accounting and Information Manage-
ment, University of Macau, Avenida Padre Tom as Pereira Taipa, Macau, China.
E-mail address: [email protected] (K.M. Sam).
Peer review under responsibility of College of Management, National Cheng Kung
University.
HOSTED BY
Contents lists available at ScienceDirect
Asia Paci?c Management Review
j ournal homepage: www. el sevi er. com/ l ocat e/ apmrvhttp://dx.doi.org/10.1016/j.apmrv.2014.12.005
1029-3132/© 2015, College of Management, National Cheng Kung University. Production and hosting by Elsevier Taiwan LLC. All rights reserved.
Asia Paci?c Management Review 20 (2015) 100e107
et al., 1996), southwestern United States (Shim, 1996) and China
(Fan & Xiao, 1998). The 40 items used in the CSI model are grouped
into eight cognitive characteristics of consumer decision-making
style as follows:
Perfectionistic and high-quality conscious consumer (searches
for the best quality in products).
Brand conscious and price equals quality consumer (buys the
more expensive, well-known national brands).
Novelty and fashion-conscious consumer (attracted to innova-
tive products)
Recreational consumer (goes to shop just for the fun of it)
Price conscious consumer (has high consciousness for sales
prices and lower prices in general)
Impulsive and careless consumer (tends to appear unconcerned
about how much he or she spends)
Confused by over-choice consumer (tends to be easily confused
by too many brands and stores from which to choose)
Habitual and brand-loyal consumer (repetitively chooses the
same favorite brands and stores)
Due to the emergence of e-commerce activities, the CSI model
(Sproles & Kendall, 1986) should be modi?ed in order to ?t the E-
commerce environment. In this research, the newOnline Consumer
Style Inventory (O-CSI) model in the E-commerce environment is
developed and the factor scores of online consumers' decision-
making styles can be computed and analyzed in order for man-
agers toestimateonlineconsumer behavior accurately. This research
aims to facilitate the understanding of online consumer behavior.
The remainder of this paper is structured as follows. Section 2
describes the development of the O-CSI model, which addresses
online consumers' decision-making styles in the business-to-
consumer environment. Section 3 discusses the computation of
factor scores for the online consumers' O-CSI model. Finally, con-
clusions are presented in Section 4.
2. O-CSI model
Based on the inventory items of the CSI model (Sproles &
Kendall, 1986), the inventory items of the O-CSI model should
satisfy one of the following criteria: (1) include only those items
that can affect the decision of selecting the suitable products or
services directly; and (2) include items that are directly related to
an online shopping environment.
For (1), some inventory items of CSI are directly related to the de-
cision of selecting suitable products for consumers, e.g., “getting very
good quality is very important to me.” On the other hand, there are
some inventory items that are not directly related to the decision of
selecting suitable products for consumers, e.g., enjoy shopping for fun.
In order to allow online businesses to analyze consumer behavior in
their buying decisions, only those inventory items related to the
following ?ve noetic characteristics, which are directly related to the
decision of selecting suitable products or services, will be included in
the O-CSI model: (a) perfectionistic and high-quality conscious con-
sumer; (b) brand conscious and price equals quality consumer; (c)
noveltyandfashion-consciousconsumer; (d) priceconsciousandvalue
for money consumer; and (e) habitual and brand-loyal consumer.
For (2), the emergence of e-commerce activities has caused e-
retailers and online consumers to pay more attention to the
following facilities in business websites:
2.1. Privacy and security
Consumers are willing to pay a little more to make transactions
with online retailers that are more likely to protect their private
information (Teresa, 2012). The vast majority of US consumers have
concerns about their online privacy and security and are wary of
doing business with businesses they believe don't have adequate
protection in place (Leggatt, 2012). Businesses, government
agencies and consumer groups agree that privacy protection needs
to be drastically improved (Higgins, 2011). The reluctance to reveal
personal information is still prevalent among Chinese online
shoppers (Gong, Maddox, & Stump, 2012).
2.2. Self-service technologies and well-organized pages
Self-service technologies, offered by e-businesses to address
customer needs, can lead to factors that can cause positive re-
actions to the online shopping service. Companies that adopt self-
service capabilitiesdsuch as product search engine and commu-
nication channels for product enquiry or order trackingdincreased
customer satisfaction (65%) and customer retention (39%), which
are signi?cantly higher rates than companies focused on cost
reduction only (Boyd, 2007). An Internet Retailer Survey in
February 2010 showed that the No. 1 and No. 2 website design
priorities across all US Online Retailers are:
Well organized and updated home, category, and product pages
and
Excellent search engine optimization.
2.3. Social networking sites
The world's Internet users spend over 110 billion minutes on
social networking and blogging sites. These numbers translate into
22% of all of the time spent on the Internet (Nielsen, 2010a). The
average social networking visitor was spending almost 6 hours in
April 2010, versus 3 hours and 31 minutes in April 2009, an increase
of almost 70% in 1 year (Nielsen, 2010a). According to a survey
(Ramsey, 2010) in 2010, more than half of all marketers engaged in
some form of social media activity and about 60% of them planned
to increase their spending in 2011. After friends and family, the
number one driver for brand trust was online reviews and feedback
from the social networking (Nielsen, 2010b). As a direct result,
advertisers are moving from a more traditional broadcasting based
marketing relationship with online consumers to a more interac-
tive based marketing relationship, where consumers directly
engage with marketing messages and pass them along to their
friends via social networking sites (Gibs &Bruich, 2010). As a result,
the social networking sites can greatly affect online consumers'
buying decisions.
2.4. Customer reviews
According to Siwicki (2009), products with reviews have a 20.4%
lower return rate than products without reviews and the return
rate continues to decline as products get more reviews. Products
with more than 50 reviews have a 65% lower return rate than
products with no reviews. The situation resulted in substantial
annual savings. In addition, the sales are increased substantially on
items with positive reviews.
Based on the facilities of business websites mentioned above,
several inventory items are included in the O-CSI model. As a result,
a list of 20 inventory items of online consumers' decision-making
styles is identi?ed in Table 1. The inventory items 1e12 can affect
online consumers' decisions in selecting suitable products or ser-
vices. The inventory items 13e20 are directly related to the online
shopping environment. These are the basis for the O-CSI model.
K.M. Sam, C. Chatwin / Asia Paci?c Management Review 20 (2015) 100e107 101
2.5. Research methodology
Each noetic characteristic identi?ed in Table 1 was measured by
some related items, which in turn are measured on a ?ve-point
Likert scale, starting from strongly agree to strongly disagree. The
information was the design framework used for questionnaires for
each of the four product industriesdapparel, information tech-
nology (IT) products, jewelry and cardin order to ?nd out whether
the O-CSI model can be applied to those four product industries.
The web-based and mailed surveys were adopted to collect quan-
titative data in Macau. For the web-based questionnaire, it was
distributed to employees at different job levels in different in-
dustries such as manufacturing, marketing and service, information
technology, government institutions and others. The questionnaire
clearly stated the purpose of the study and asked for their partici-
pation in the study by clicking a hyperlink to the survey form. For
the mailed questionnaire, it was mainly distributed to the students
in universities. As an incentive, respondents from the web-based
questionnaire and mailed surveys were offered a fast-food restau-
rant coupon.
Of the 400 mailed questionnaires and 1000 electronic ques-
tionnaires, 823 completed questionnaires were analyzed, with a
response rate of 58.78%. Table 2 shows the demographics of the
respondents.
For the data analysis process, factor analysis was adopted to
analyze and design the structure of a new O-CSI model. The pur-
pose is to reduce and categorize the items into several factors for
the O-CSI model.
First of all, the dimensionality of the consumer styles inventory
was assessed by examining the factor solution. In order to obtain a
factor solution, a principal components factor analysis was used
with varimax rotation. In addition, the amount of variance
explained by the extracted factors (i.e., their eigenvalues) was noted
in order to determine which items have very high correlation
values and which can be eliminated. The above process was
repeated for the four different product industries.
The results show that the ?rst seven components have eigen-
values higher or close to 1 for all the four product industries. As a
result, there are seven extracted factors in the O-CSI model. Table 3
shows the total variance and the initial eigenvalues of components
for Apparel and IT item industries.
Table 1
Inventory items in the O-CSI model.
1. Getting very good quality is very important to me.
2. Once I ?nd a product or brand I like, I stick with it.
3. The well-known national brands are best for me.
4. The higher the price, the better its quality.
5. I prefer buying the best-selling brands.
6. I usually have one or more products of the very newest style.
7. Fashionable, attractive styling is very important to me.
8. I buy as much as possible at sale price.
9. The lower price products are usually my choice.
10. I look carefully to ?nd the best value for the money.
11. When buying products, portability is very important to me.
12. The smaller the product size, the more I prefer them.
13. When I go shopping online, privacy and security are very important.
14. It is very important for the websites to offer communication channels to me
for product enquiries and order tracking.
15. It is very important for the websites to offer a product searching service to
me.
16. It is perfect if the websites can offer me richness of information about
products.
17. It will be annoying to get a lot of animated effects on the business websites.
18. Design layout of business website is one of the important factors in making
buying decisions.
19. It is good if the websites can offer customer reviews on the products.
20. It is good if the websites can offer social networking facilities so that I can
share product comments with my friends.
Table 2
Demographics of the respondents.
Demographics Number Percent
Gender
Female 434 52.8
Male 389 47.2
Age
< 29 365 44.3
30e39 311 37.8
40e49 117 14.2
> 50 30 3.7
Job positions
Top and middle managers 117 14.2
Line managers 139 16.9
Frontline staff 521 63.3
Others 46 5.6
Industries
Manufacturing 60 7.3
Marketing and service 283 34.4
Information technology 114 13.9
Government agencies 216 26.2
Others 150 18.2
Table 3
Total variance explained.
Component Initial eigenvalues Extraction sums of
squared loadings
Total % of
variance
Cumulative % Total % of
variance
Cumulative %
Apparel
1 3.006 18.347 18.347 3.006 18.347 18.347
2 2.607 15.525 33.872 2.607 15.525 33.872
3 1.954 11.658 45.530 1.954 11.658 45.530
4 1.513 9.306 54.836 1.513 9.306 54.836
5 1.258 7.684 62.520 1.258 7.684 62.520
6 1.069 6.739 69.259 1.069 6.739 69.259
7 0.976 6.022 75.281 0.976 6.022 75.281
8 0.906 5.147 80.428
9 0.791 4.195 84.623
10 0.707 3.628 88.251
11 0.614 2.913 91.164
12 0.574 2.613 93.777
13 0.517 2.172 95.949
14 0.383 1.385 97.334
15 0.350 1.004 98.338
16 0.314 0.697 99.035
17 0.283 0.482 99.517
18 0.103 0.259 99.776
19 0.051 0.148 99.924
20 0.025 0.076 100.000
Information technology items
1 3.787 21.039 21.039 3.787 21.039 21.039
2 2.872 16.757 37.796 2.872 16.757 37.796
3 2.217 12.926 50.722 2.217 12.926 50.722
4 1.829 9.485 60.207 1.829 9.485 60.207
5 1.361 5.848 66.055 1.361 5.848 66.055
6 1.172 5.102 71.157 1.172 5.102 71.157
7 0.969 4.627 75.784 0.969 4.627 75.784
8 0.828 4.232 80.016
9 0.728 3.843 83.859
10 0.697 3.670 87.529
11 0.642 3.369 90.898
12 0.498 2.566 93.464
13 0.353 1.819 95.283
14 0.306 1.402 96.685
15 0.267 1.072 97.757
16 0.218 0.713 98.470
17 0.161 0.528 98.998
18 0.133 0.457 99.455
19 0.074 0.319 99.774
20 0.041 0.226 100.000
Extraction method: principal component analysis.
K.M. Sam, C. Chatwin / Asia Paci?c Management Review 20 (2015) 100e107 102
2.6. Con?rmation of the seven-factor model
According to Table 3, a seven-factor solution was supported by
the percentage of variance accounted for by each factor. The seven-
factor solution was chosen based on the scree test for determining
the number of factors to be retained in factor analysis. Based on the
results, all eigenvalues exceeded or were close to 1.0 and our
requirement that at least 75% of the cumulative variance be
explained by the set of retained factors. More importantly, the
seven factors con?rm the importance of the characteristics pro-
posed and the seven-factor consumer style model can be applied to
the four product industries. Table 4 shows the seven factors and the
factor loadings of their corresponding inventory items.
Factor 1. This factor measures high quality and becomes a
conscious consumer characteristic. Items loading on this factor
measure howimportant a consumer thinks quality is for the buying
decision and it becomes a consumer habit for buying the same
high-quality products.
Factor 2. This factor identi?es a brand conscious consumer
characteristic, brand conscious, price equals quality. It measures
consumers' orientation towards buying the more expensive, well-
known national brands.
Factor 3. This factor measures a novelty-fashion conscious
consumer characteristic. High scores on this characteristic indicate
that a consumer prefers a new product style to those old fashioned
styles.
Factor 4. This factor measures a price conscious characteristic. A
consumer having a high score on this factor is sensitive to product
price and prefers buying low price products.
Factor 5. This factor measures the product portability conscious
characteristic. Those consumers who prefer a smaller size product,
so that it is ?exible for carrying around, have a high score on this
factor.
Factor 6. This factor measures the website content conscious
characteristic. What do consumers think about the facilities, such
as privacy, security, searching tools, communication tools for
product enquiry and order tracking, availability and richness of
product information, customer review and social networking
capability offered by online shops? The answer to this question can
affect the score on this factor. A high score indicates that consumers
care very much about the website facilities, such that it can affect
the consumers' buying decision.
Factor 7. This factor measures the website interface conscious
characteristic. The design of the website is important to some
consumers. Is it better to offer animation effect on the business
website to attract consumers' attention? Some consumers don't
like animation effects on business websites. The reasons include:
(1) confusion about the information displayed on screen; (2) low
data transmission speed on the Internet. Is it better to offer graphics
display instead of text display on sensitive information or infor-
mation that is not easy to understand? The location of the web tools
on the website can also affect some consumers when they want to
get some services from the website. Consumers who have a high
score consider the web interface very important, so it can affect
their buying decisions signi?cantly.
Table 5 presents Cronbach a reliabilities for the set of inventory
items within each decision-making style. Since most of the a co-
ef?cients have values around or above 0.7, the items are said to have
at least acceptable internal consistency.
2.7. A pro?le of online consumer style
A pro?le of online consumer style can be developed for each
of the product types based on the highest loading item on
each consumer style. These data were calculated by referring to
the raw scores on the highest loading item for each consumer
style. This result yields scores of 1e5 for each product on each
consumer style. Table 6 presents the means of highest loading
items and the percentages of scoring very high to very lowon each
scale for the apparel industry. Based on Table 6, a pro?le of online
consumer style can be established for the apparel industry in
Table 7.
Table 4
Online consumer decision-making styles (style characteristics): seven factor model.
Loadings: Apparel IT item Jewelry Car
Factor 1 e high-quality, become buying habit conscious consumer
Getting very good quality is very important to me. 0.35 0.81 0.58 0.81
Once I ?nd a product or brand I like, I stick with it. 0.75 0.35 0.40 0.59
Factor 2 e brand conscious consumer
The well-known national brands are best for me. 0.78 0.62 0.77 0.67
The higher the price, the better its quality. 0.84 0.78 0.46 0.63
I prefer buying the best-selling brands. 0.77 0.82 0.73 0.84
Factor 3 e novelty-fashion conscious consumer
I usually have one or more products of the very newest style. 0.82 0.80 0.74 0.85
Fashionable, attractive styling is very important to me. 0.73 0.87 0.59 0.79
Factor 4 e price conscious consumer
I buy as much as possible at sale price. 0.81 0.80 0.71 0.61
The lower price products are usually my choice. 0.76 0.56 0.87 0.81
I look carefully to ?nd the best value for the money. 0.40 0.70 0.45 0.57
Factor 5 e product portability conscious consumer
When buying products, portability is very important to me. 0.85 0.86 0.80 0.83
The smaller the product size, the more I prefer them. 0.83 0.57 0.78 0.82
Factor 6 e website content conscious consumer
When I go shopping online, privacy and security are very important. 0.38 0.50 0.32 0.52
It is very important for the websites to offer communication channels to me for product enquiries and order tracking. 0.75 0.82 0.82 0.83
It is very important for the websites to offer a product searching service to me. 0.90 0.87 0.92 0.83
It is perfect if the websites can offer me richness of information about products. 0.86 0.84 0.85 0.79
It is good if the websites can offer customer reviews on the products. 0.88 0.91 0.82 0.76
It is good if the websites can offer social networking facilities so I can share product comments with my friends. 0.85 0.92 0.77 0.84
Factor 7 e website interface conscious consumer
It will be annoying to get a lot of animated effects on the business websites. 0.62 0.81 0.68 0.72
Design layout of business website is one of the important factors to make buying decisions. 0.89 0.86 0.96 0.90
K.M. Sam, C. Chatwin / Asia Paci?c Management Review 20 (2015) 100e107 103
3. Computation of factor scores for O-CSI model
The earliest factor scoring methods developed yield least-
squares regression weights, factor score coef?cients, which are
applied to the data to estimate the factor scores. These exact-
regression methods have been criticized on several grounds: (1) the
vulnerability of shrinkage effects; (2) using degrees for each item
(Schweiker, 1967); and (3) not meeting the requirements of validity,
univocality, and orthogonality for an orthogonal factor solution.
As an alternative to the exact-regression methods, several early
authors (Cattell, 1952; Thurstone, 1947) introduced the unit-
loading method, which requires the investigator to examine the
structure coef?cients, select the items that surpass an arbitrary
salience criterion (e.g., ± 0.40), and unit-weight those items
(converted to z-scores if measured on different scales) in accor-
dance with the signs of the structure coef?cients. The unit-loading
method became popular and was used to measure e-marketing
mix elements for online businesses (Sam & Chatwin, 2012)
because of its computational convenience and the added bene?t of
avoiding the shrinkage effects associated with the regression-
based scoring methods. Wackwitz and Horn (1971) also found
that the unit-loading method produced factor scores that were
more valid and orthogonal than the scores produced by the exact-
regression methods. However, the unit-loading method will pro-
duce inaccurate factor score estimates (Wackwitz & Horn, 1971).
The loadings (structure coef?cients) do not necessarily indicate
how the items must be weighted and combined to create scores
for individual factors. That information is speci?cally conveyed by
the factor score coef?cients.
Table 5
Reliability coef?cients for seven online consumer styles.
Consumer style characteristics Cronbach a for subscale of all items
Apparel Information
technology item
Jewelry Car
High-quality, becoming
buying habit conscious
0.65 0.67 0.64 0.66
Brand conscious 0.93 0.84 0.76 0.82
Novelty-fashion conscious 0.85 0.84 0.76 0.83
Price-value conscious 0.67 0.73 0.78 0.67
Portability conscious 0.84 0.81 0.81 0.80
Website content conscious 0.94 0.96 0.98 0.94
Website interface conscious 0.88 0.92 0.80 0.85
Table 6
Statistical analysis of highest loading item for the seven online consumer characteristics for the apparel industry.
Style characteristic Highest loading item Percentage score of highest loading item
Highest loading item Mean Very High (1) High (2) Med. (3) Low (4) Very Low (5)
High-quality, buying habit
conscious
Once I ?nd a product or brand I like, I stick with it. 2.6 15% 42% 22% 14% 7%
Brand conscious The well-known national brands are best for me. 3.3 7% 22% 25% 31% 15%
Novelty-fashion conscious I usually have one or more products of the very
newest style.
3.3 7% 20% 24% 30% 19%
Price conscious I buy as much as possible at sale price. 2.6 21% 30% 22% 22% 5%
Portability conscious When buying products, portability is very
important to me.
3.0 10% 21% 42% 13% 14%
Website content conscious It is very important for the websites to offer
a product searching service to me.
1.7 54% 34% 6% 3% 3%
Website interface conscious Design layout of business website is one of the
important factors in making buying decisions.
2.9 13% 31% 24% 16% 16%
Table 7
A pro?le of online consumer style.
Your Name: _____________________________________________________________________
Style characteristics Your score Group mean Verbal interpretation of your consumer style
Apparel
High-quality, buying habit conscious 3.2 2.6 You are average in demanding and buying the same high-quality products.
Brand conscious 2.8 3.3 You are average to above average in brand name consciousness.
Novelty-fashion conscious 2.1 3.3 You are high in novelty and fashion consciousness.
Price conscious 3.8 2.6 You are low in price consciousness e price matters little to you.
Portability conscious 1.2 3.0 You are very high in portability consciousness, demanding a great deal of product ?exibility.
Website content conscious 2.6 1.7 You are below average in website content consciousness, not considering any
facilities inside the websites very much.
Website interface conscious 4.5 2.9 You are very low in website interface consciousness, there is no effect of website
interface on your buying decisions.
Information technology items
High-quality, buying habit conscious 3.2 1.7 You are below average to low in demanding and buying the same high-quality products.
Brand conscious 2.8 2.7 You are average in brand name consciousness.
Novelty-fashion conscious 2.1 3.1 You are high in novelty and fashion consciousness.
Price conscious 3.8 2.8 You are below average to low in price consciousness e price matters little to you.
Portability conscious 1.2 2.4 You are very high in portability consciousness, demanding a great deal of product ?exibility.
Website content conscious 2.6 1.6 You are below average in website content consciousness, not considering any facilities
inside the websites very much.
Website interface conscious 4.5 3.0 You are very low in website interface consciousness, there is no effect of website
interface on your buying decisions.
K.M. Sam, C. Chatwin / Asia Paci?c Management Review 20 (2015) 100e107 104
From above, both unit-loading and exact-regression methods
have advantages and disadvantages. A unit-regression strategy was
created to combine their advantages by applying the same logic of
the unit-weighting scheme to the multidecimal factor score co-
ef?cients (Harris, 1985). The items with salient factor score co-
ef?cients could be unit-weighted in accordance with the signs of
their respective coef?cients and summed. Those items with non-
salient score coef?cients would receive weights of zero and would
hence not contribute to the estimation of the factor scores. The
unit-regression estimates would consequently be resistant to the
shrinkage effects and would not suffer from the validity problems
that af?ict the unit-loading estimates. With respect to the factor
interpretation, the unit-regression method would also possess the
bene?cial feature of representing the items as a whole rather than
fractions.
In order to support the superiority of the unit-regression
method, Grice and Harris (1998) extended Wackwitz and Horn's
methodology (1971) to include complex factor structures for
assessing shrinkage effects among the exact-regression, unit-
regression, and unit-loading methods. The results generally sup-
ported the unit-regression strategy.
Results from comparisons between the unit-regression and
popular unit-loading methods also supported the unit-regression
method. The unit-regression estimates were found to be more
valid and orthogonal than unit-loading estimates. As a result, the
unit-regression method is used to evaluate factor scores of the
online consumers' decision-making styles.
3.1. Factor score coef?cient matrices
Table 8 presents the factor score coef?cient matrices of the O-
CSI model for apparel and IT item industries. The weights of in-
ventory items for the seven decision-making styles shown in
Table 8 are used to evaluate the factor scores of the online con-
sumers' decision-making styles.
Based on the coef?cient matrices in Table 8, the following facts
can be deduced:
For a particular industry, different decision-making styles have
different weights for the same inventory items.
For the same online consumers' decision-making style, there are
different weights for the inventory items in different industries.
Table 8
Factor score coef?cient matrix for the apparel and information technology industries.
Component (Factor)
1 2 3 4 5 6 7
Apparel industry
Prod1_ans1 0.028 À0.113 À0.162 0.222 À0.021 0.404 0.105
Prod1_ans2 0.059 À0.094 À0.013 0.057 À0.092 0.583 À0.004
Prod1_ans3 0.012 0.492 À0.044 À0.294 0.053 0.108 0.156
Prod1_ans4 0.049 0.390 À0.001 0.017 0.066 À0.079 À0.146
Prod1_ans5 À0.081 0.376 0.158 0.084 À0.087 À0.166 À0.081
Prod1_ans6 À0.024 À0.014 À0.082 0.543 0.027 0.072 0.107
Prod1_ans7 À0.070 À0.019 À0.068 0.409 0.126 0.152 À0.007
Prod1_ans8 0.019 0.063 À0.010 À0.116 0.535 À0.135 0.169
Prod1_ans9 0.173 À0.043 0.162 0.074 0.496 0.022 À0.278
Prod1_ans10 À0.048 À0.057 À0.039 0.090 0.377 À0.070 0.121
Prod1_ans11 0.016 À0.068 0.486 0.073 0.034 À0.049 0.080
Prod1_ans12 À0.071 0.004 0.474 À0.009 À0.104 0.044 À0.012
Prod1_ans13 0.405 À0.033 0.027 0.025 0.065 À0.031 À0.244
Prod1_ans14 0.403 0.066 0.092 À0.230 0.009 0.204 À0.084
Prod1_ans15 0.582 À0.045 0.075 À0.002 0.084 0.019 À0.078
Prod1_ans16 0.473 0.098 0.071 À0.094 À0.028 À0.148 0.126
Prod1_ans17 À0.094 0.048 À0.073 À0.107 0.021 0.042 0.481
Prod1_ans18 À0.012 0.008 À0.011 0.004 À0.035 À0.019 0.835
Prod1_ans19 0.484 À0.126 0.013 À0.072 À0.011 0.182 À0.136
Prod1_ans20 0.451 0.102 À0.118 0.131 À0.337 À0.221 À0.037
Information technology items
Prod2_ans1 0.044 À0.066 À0.056 À0.087 0.148 0.385 0.037
Prod2_ans2 À0.009 0.022 0.038 À0.007 0.091 0.649 À0.178
Prod2_ans3 0.025 0.448 0.034 À0.196 0.033 À0.188 0.217
Prod2_ans4 0.034 0.430 À0.091 À0.066 À0.099 0.052 À0.152
Prod2_ans5 0.011 0.363 À0.008 0.050 À0.010 À0.077 À0.086
Prod2_ans6 À0.040 À0.070 0.009 0.556 À0.055 À0.020 0.041
Prod2_ans7 0.052 0.071 À0.016 0.347 0.072 À0.152 0.035
Prod2_ans8 0.090 0.039 À0.104 À0.367 0.533 0.070 À0.030
Prod2_ans9 À0.073 À0.176 0.008 0.250 0.490 0.141 0.038
Prod2_ans10 À0.097 0.130 À0.067 0.222 0.326 À0.151 0.241
Prod2_ans11 À0.013 À0.106 0.510 À0.052 0.015 À0.104 À0.021
Prod2_ans12 0.003 0.106 0.572 À0.043 À0.154 0.013 0.007
Prod2_ans13 0.354 0.051 0.238 À0.193 À0.117 À0.332 0.211
Prod2_ans14 0.467 À0.091 À0.091 À0.036 0.010 0.117 À0.100
Prod2_ans15 0.517 À0.079 À0.021 0.003 0.029 0.097 0.016
Prod2_ans16 0.349 0.025 À0.006 À0.008 0.103 À0.112 0.029
Prod2_ans17 0.171 À0.017 À0.045 À0.010 À0.006 0.022 0.404
Prod2_ans18 À0.087 0.046 À0.057 À0.055 À0.023 0.123 0.713
Prod2_ans19 0.415 À0.118 0.043 À0.124 À0.111 0.162 À0.172
Prod2_ans20 0.386 0.032 À0.166 0.119 À0.475 À0.187 À0.043
Extraction method: principal component analysis.
Rotation method: varimax with Kaiser normalization.
K.M. Sam, C. Chatwin / Asia Paci?c Management Review 20 (2015) 100e107 105
Having considered the two facts above, the unit-regression
factor-scoring formula for Factor 1 in the apparel industry is
shown in Equation (1).
where: Weight
(app, 1, j)
¼ the weight of the j
th
inventory item with
respect to factor 1 in apparel industry; and Item_Score
(j)
¼the score
point of the j
th
inventory item
The factor score of factor i is calculated by summing up the n
weighted item scores of factor i, based on the weight of each in-
ventory item of the factor i. Since only the items whose weights are
0.4 are involved in evaluating the factor score, those items whose
weights are < 0.4 are ignored, as indicated by the value zero in
Equation (1).
Taking an online company selling electrical appliances as an
example, the registered users need to de?ne their O-CSI pro?les
by answering an online survey about the inventory items of the
O-CSI model, their pro?les are then stored and their factor scores
of online consumers' decision-making styles can be computed
based on Equation (1). As customers register in the online busi-
ness website, the group means of factor scores and the weights of
the factor items corresponding to the online consumers' decision-
making styles will be changed. By comparing the updated factor
scores with the frequently updated group means, the online
consumers' decision-making styles can be analyzed such that
business managers can estimate online consumer behavior
accurately.
4. Conclusions
In this paper, a 20-item O-CSI model has been designed for
online consumers in Macau. The results showed that there are
seven decision-making styles identi?ed in the O-CSI model: (1)
high-quality, buying habit consciousness; (2) brand conscious-
ness; (2) novelty-fashion consciousness; (4) price consciousness;
(5) portability consciousness; (6) website content consciousness;
and (7) website interface consciousness. In addition, the O-CSI
model can be applied to different product items. In order to allow
managers to analyze the characteristics of online consumers
quantitatively, the factor scores of the decision-making styles are
evaluated. The accuracy of the factor scores can be ensured by
adopting the unit-regression method in which the weights of
inventory items for their corresponding decision-making styles
play a very important role. Although this research focuses on
Macau, the methodology can be applied to any market or loca-
tion. By using the suggested methodology, online businesses will
be able to become more competitive and increase their market
share.
Con?icts of interest
All contributing authors declare no con?icts of interest.
Acknowledgments
This research was supported by the Department of Accounting
and Information Management at the University of Macau. In
addition, the survey could be completed quickly due to the support
of my colleagues and my friends.
References
AbouTourism. (2010). Statisical Survey Report on Internet Development in China.http://aboutourism.wordpress.com/2010/03/31/statistical-survey-report-on-
internet-development-in-china Accessed 05.04.10.
Boyd, A. (2007). Customer Self-Service: How May You Help Me?.http://www.
ecommercetimes.com/story/57949.html Accessed 21.06.07.
Cattell, R. B. (1952). Factor analysis: An introduction and manual for the psychologist
and social scientist. New York: Harper and Brothers.
Durvasula, S., Lysonski, S., & Andrews, J. C. (1993). Cross-cultural generalizability of
a scale for pro?ling consumers' decision-making styles. Journal of Consumer
Affairs, 27(2), 55e65.
Fan, J. X., & Xiao, J. J. (1998). Consumer decision-making styles of young-adult
Chinese. Journal of Consumer Affairs, 32(2), 275e294.
Forrester Forecast. (2010). Forrester Forecast: Online retail sales will grow to $250
billion by 2014.http://techcrunch.com/2010/03/08/forrester-forecast-online-
retail-sales-will-grow-to-250-billion-by-2014/ Accessed 01.09.11.
Gibs, J., & Bruich, S. (2010). Advertising effectiveness: Understanding the value of a
social media impression. New York: The Nielsen Company.
Gong, W., Maddox, L. M., & Stump, R. L. (2012). Attitudes toward online shopping: a
comparison of online consumers in China and the US. International Journal of E-
Business Development, 2(1), 28e35.
Grice, J., & Harris, R. (1998). A comparison of regression and loading weights for
the computation of factor scores. Multivariate Behavioral Research, 33(2),
221e247.
Hafstrom, J. L., Chae, J. S., & Chung, Y. S. (1992). Consumer decision-making styles:
comparison between United States and Korean young consumers. Journal of
Consumer Affairs, 26(1), 146e158.
Harris, R. J. (1985). A primer of multivariate statistics (2nd ed.). Orlando: Academic
Press.
Higgins, J. K. (2011). Internet Privacy: Trust Big Business or Trust the Government?.http://www.ecommercetimes.com/rsstory/72032.html Accessed 10.03.11.
Internet World Stats. (2012). Internet Users in Asia 2012.http://www.
internetworldstats.com/stats3.htm Accessed 03.01.13.
Leggatt, H. (2012). Online privacy real concern for 90% of US Internet users. http://
www.bizreport.com/2012/02/90-percent-of-online-adults-worry-about-their-
online-privacy.html Accessed 14.02.12.
Lysonski, S., Durvasula, S., & Zotos, Y. (1996). Consumer decision-making styles: a
multi-country investigation. European Journal of Marketing, 30(12), 10e21.
New Zealand Consulate-General Hong Kong. (2011). Hong Kong Review.http://www.
mfat.govt.nz/downloads/media-and-publications/newsletters/hong-kong/
HongKong-Review-June-2011.pdf.
Nielsen, W. (2010a). Social Networks/Blogs Now Account for One in Every Four and a
Half Minutes Online.http://blog.nielsen.com/nielsenwire/global/social-media-
accounts-for-22-percent-of-time-online/ Accessed 15.06.10.
Nielsen, W. (2010b). Friending the Social Consumer.http://blog.nielsen.com/
nielsenwire/online_mobile/friending-the-social-consumer/ Accessed 16.06.10.
Ramsey, G. (2010). 10 best practices for success with social media. New York:
eMarketer.
Raven, P. V., Huang, X., & Kim, B. B. (2007). E-business in developing countries: a
comparison of China and India. International Journal of E-Business Research, 3(1),
91e110.
Sam, K. M., & Chatwin, C. R. (2012). Measuring e-marketing mix elements for online
business. International Journal of E-Entrepreneurship and Innovation, 3(3), 13e26.
Schweiker, R. F. (1967). Factor scores aren't sacred: comments on abuses of factor
scores. American Educational Research Journal, 4, 168e170.
Shim, S. (1996). Adolescent consumer decision-making styles: the consumer so-
cialization perspective. Psychology & Marketing, 13(6), 547e569.
Siwicki, B. (2009). Getting them to write.http://www.internetretailer.com/2009/12/
29/getting-them-to-write Accessed 29.12.09.
Sproles, G. B. (1985). From perfectionism to fadism: measuring consumers'
decision-making styles. In 30th Annual Meeting of the American Council on
Consumer Interests.
Sproles, G. B., & Kendall, E. L. (1986). A methodology for pro?ling consumers'
decision-making styles. Journal of Consumer Affairs, 20(4), 267e279.
Taobao. (2010). China Online Shopping Report 2009e2010.http://www.linkagepro.
com/articles/Internet-Marketing-20100523.htm Accessed 10.05.10.
Factor Score
ðapp;1Þ
¼
X
n
j¼1
Weight
ðapp;1;jÞ
*Item Score
ðjÞ
; Weight
ðapp;1;jÞ
0:4
0; Weight
ðapp;1;jÞ
The purpose of this paper is to understand online consumer behavior better by analyzing online consumers'
decision-making styles. In this research, an online consumer style inventory, which is suitable
for online businesses to measure online consumers' decision-making styles, has been developed in
Macau. The current studies of online consumer behavior have not considered the weights of the variables
that can affect online consumer behavior. This paper measures online consumers' decision-making styles
in Macau based on the weights of the corresponding inventory items.
Online consumer decision-making styles for enhanced understanding of Macau
online consumer behavior
Kin Meng Sam
a, *
, Chris Chatwin
b
a
Department of Accounting and Information Management, University of Macau, Taipa, Macau, China
b
School of Engineering and Informatics, University of Sussex, Brighton, United Kingdom
a r t i c l e i n f o
Article history:
Received 16 January 2013
Accepted 9 April 2014
Available online 19 March 2015
Keywords:
Factor score coef?cient
Online consumer behavior
Online consumer style inventory
Online consumers'decision-making styles
a b s t r a c t
The purpose of this paper is to understand online consumer behavior better by analyzing online con-
sumers' decision-making styles. In this research, an online consumer style inventory, which is suitable
for online businesses to measure online consumers' decision-making styles, has been developed in
Macau. The current studies of online consumer behavior have not considered the weights of the variables
that can affect online consumer behavior. This paper measures online consumers' decision-making styles
in Macau based on the weights of the corresponding inventory items.
© 2015, College of Management, National Cheng Kung University. Production and hosting by Elsevier
Taiwan LLC. All rights reserved.
1. Introduction
Decision making is more complex and important for consumers
than in the past (Hafstrom, Chae, & Chung, 1992; Lysonski,
Durvasula, & Zotos, 1996). The increased varieties of products and
the abundance of information through company advertisements
have broadened the choices for consumers. A consumer decision-
making style refers to a mental orientation describing how a con-
sumer makes choices (Durvasula, Lysonski, & Andrews, 1993).
Pro?ling consumer decision-making styles is very important to
marketers and advertisers (Lysonski et al., 1996). In order to deal
with the emergence of e-commerce activities, it is necessary to
consider online consumers' decision-making styles that in?uence
the willingness of online consumers to purchase products.
China's Internet penetration rate reached 40.1% during June 2012,
exceeding the world's average of 34.3% at the same point (Internet
World Stats, 2012). Academic researchers suggested that China's
cultural history of preferring face-to-face business interactions,
coupled with its restrictive regulatory environment, may hinder the
development of online shopping in China (Raven, Huang, & Kim,
2007). Nevertheless, China's online retail market has been growing
steadily with an increase of 45.9% in the number of online shopping
users in 2009, leading to a total of 156 million users (AbouTourism,
2010), compared to 154 million users in the USA who shopped on-
line in 2009 (Forrester Forecast, 2010); online shopping penetration
in China still appears to hold considerable growth potential.
Macau has been a prosperous Chinese city since 2004, after the
gambling industry was opened up to external investors; this caused
the gross domestic product for the ?rst quarter of 2011 to expand
by 21.5% (New Zealand Consulate-General Hong Kong, 2011).
Macau's living standards have risen by over 300% in just one dec-
adedit is now destined to become the richest territory in Asia
(excluding the Middle East) by gross domestic product per capita
(Zimbabwemetro, 2011). According to Taobao (2010), Macau is the
city with the fastest increase in online shopping of all the cities in
China. For this reason it is valuable to focus on the analysis of online
consumers in Macau.
1.1. Consumer style inventory
One of the ways to characterize consumer styles is consumer
characterization, focusing on cognitive and affective orientations
related to consumer decision-making (Sproles, 1985). Consumer
characterization is very promising as it deals with the cognitive
orientation of consumers in making decisions (Sproles & Kendall,
1986). Sproles and Kendall (1986) designed a 40-item Consumer
Style Inventory (CSI) model to measure decision-making styles of
consumers based on a sample of US youth. The applicability of the
CSI model has been investigated across several cultures such as
Korea (Hafstrom et al., 1992), New Zealand (Durvasula et al., 1993;
Lysonski et al., 1996), India (Lysonski et al., 1996), Greece (Lysonski
* Corresponding author. Department of Accounting and Information Manage-
ment, University of Macau, Avenida Padre Tom as Pereira Taipa, Macau, China.
E-mail address: [email protected] (K.M. Sam).
Peer review under responsibility of College of Management, National Cheng Kung
University.
HOSTED BY
Contents lists available at ScienceDirect
Asia Paci?c Management Review
j ournal homepage: www. el sevi er. com/ l ocat e/ apmrvhttp://dx.doi.org/10.1016/j.apmrv.2014.12.005
1029-3132/© 2015, College of Management, National Cheng Kung University. Production and hosting by Elsevier Taiwan LLC. All rights reserved.
Asia Paci?c Management Review 20 (2015) 100e107
et al., 1996), southwestern United States (Shim, 1996) and China
(Fan & Xiao, 1998). The 40 items used in the CSI model are grouped
into eight cognitive characteristics of consumer decision-making
style as follows:
Perfectionistic and high-quality conscious consumer (searches
for the best quality in products).
Brand conscious and price equals quality consumer (buys the
more expensive, well-known national brands).
Novelty and fashion-conscious consumer (attracted to innova-
tive products)
Recreational consumer (goes to shop just for the fun of it)
Price conscious consumer (has high consciousness for sales
prices and lower prices in general)
Impulsive and careless consumer (tends to appear unconcerned
about how much he or she spends)
Confused by over-choice consumer (tends to be easily confused
by too many brands and stores from which to choose)
Habitual and brand-loyal consumer (repetitively chooses the
same favorite brands and stores)
Due to the emergence of e-commerce activities, the CSI model
(Sproles & Kendall, 1986) should be modi?ed in order to ?t the E-
commerce environment. In this research, the newOnline Consumer
Style Inventory (O-CSI) model in the E-commerce environment is
developed and the factor scores of online consumers' decision-
making styles can be computed and analyzed in order for man-
agers toestimateonlineconsumer behavior accurately. This research
aims to facilitate the understanding of online consumer behavior.
The remainder of this paper is structured as follows. Section 2
describes the development of the O-CSI model, which addresses
online consumers' decision-making styles in the business-to-
consumer environment. Section 3 discusses the computation of
factor scores for the online consumers' O-CSI model. Finally, con-
clusions are presented in Section 4.
2. O-CSI model
Based on the inventory items of the CSI model (Sproles &
Kendall, 1986), the inventory items of the O-CSI model should
satisfy one of the following criteria: (1) include only those items
that can affect the decision of selecting the suitable products or
services directly; and (2) include items that are directly related to
an online shopping environment.
For (1), some inventory items of CSI are directly related to the de-
cision of selecting suitable products for consumers, e.g., “getting very
good quality is very important to me.” On the other hand, there are
some inventory items that are not directly related to the decision of
selecting suitable products for consumers, e.g., enjoy shopping for fun.
In order to allow online businesses to analyze consumer behavior in
their buying decisions, only those inventory items related to the
following ?ve noetic characteristics, which are directly related to the
decision of selecting suitable products or services, will be included in
the O-CSI model: (a) perfectionistic and high-quality conscious con-
sumer; (b) brand conscious and price equals quality consumer; (c)
noveltyandfashion-consciousconsumer; (d) priceconsciousandvalue
for money consumer; and (e) habitual and brand-loyal consumer.
For (2), the emergence of e-commerce activities has caused e-
retailers and online consumers to pay more attention to the
following facilities in business websites:
2.1. Privacy and security
Consumers are willing to pay a little more to make transactions
with online retailers that are more likely to protect their private
information (Teresa, 2012). The vast majority of US consumers have
concerns about their online privacy and security and are wary of
doing business with businesses they believe don't have adequate
protection in place (Leggatt, 2012). Businesses, government
agencies and consumer groups agree that privacy protection needs
to be drastically improved (Higgins, 2011). The reluctance to reveal
personal information is still prevalent among Chinese online
shoppers (Gong, Maddox, & Stump, 2012).
2.2. Self-service technologies and well-organized pages
Self-service technologies, offered by e-businesses to address
customer needs, can lead to factors that can cause positive re-
actions to the online shopping service. Companies that adopt self-
service capabilitiesdsuch as product search engine and commu-
nication channels for product enquiry or order trackingdincreased
customer satisfaction (65%) and customer retention (39%), which
are signi?cantly higher rates than companies focused on cost
reduction only (Boyd, 2007). An Internet Retailer Survey in
February 2010 showed that the No. 1 and No. 2 website design
priorities across all US Online Retailers are:
Well organized and updated home, category, and product pages
and
Excellent search engine optimization.
2.3. Social networking sites
The world's Internet users spend over 110 billion minutes on
social networking and blogging sites. These numbers translate into
22% of all of the time spent on the Internet (Nielsen, 2010a). The
average social networking visitor was spending almost 6 hours in
April 2010, versus 3 hours and 31 minutes in April 2009, an increase
of almost 70% in 1 year (Nielsen, 2010a). According to a survey
(Ramsey, 2010) in 2010, more than half of all marketers engaged in
some form of social media activity and about 60% of them planned
to increase their spending in 2011. After friends and family, the
number one driver for brand trust was online reviews and feedback
from the social networking (Nielsen, 2010b). As a direct result,
advertisers are moving from a more traditional broadcasting based
marketing relationship with online consumers to a more interac-
tive based marketing relationship, where consumers directly
engage with marketing messages and pass them along to their
friends via social networking sites (Gibs &Bruich, 2010). As a result,
the social networking sites can greatly affect online consumers'
buying decisions.
2.4. Customer reviews
According to Siwicki (2009), products with reviews have a 20.4%
lower return rate than products without reviews and the return
rate continues to decline as products get more reviews. Products
with more than 50 reviews have a 65% lower return rate than
products with no reviews. The situation resulted in substantial
annual savings. In addition, the sales are increased substantially on
items with positive reviews.
Based on the facilities of business websites mentioned above,
several inventory items are included in the O-CSI model. As a result,
a list of 20 inventory items of online consumers' decision-making
styles is identi?ed in Table 1. The inventory items 1e12 can affect
online consumers' decisions in selecting suitable products or ser-
vices. The inventory items 13e20 are directly related to the online
shopping environment. These are the basis for the O-CSI model.
K.M. Sam, C. Chatwin / Asia Paci?c Management Review 20 (2015) 100e107 101
2.5. Research methodology
Each noetic characteristic identi?ed in Table 1 was measured by
some related items, which in turn are measured on a ?ve-point
Likert scale, starting from strongly agree to strongly disagree. The
information was the design framework used for questionnaires for
each of the four product industriesdapparel, information tech-
nology (IT) products, jewelry and cardin order to ?nd out whether
the O-CSI model can be applied to those four product industries.
The web-based and mailed surveys were adopted to collect quan-
titative data in Macau. For the web-based questionnaire, it was
distributed to employees at different job levels in different in-
dustries such as manufacturing, marketing and service, information
technology, government institutions and others. The questionnaire
clearly stated the purpose of the study and asked for their partici-
pation in the study by clicking a hyperlink to the survey form. For
the mailed questionnaire, it was mainly distributed to the students
in universities. As an incentive, respondents from the web-based
questionnaire and mailed surveys were offered a fast-food restau-
rant coupon.
Of the 400 mailed questionnaires and 1000 electronic ques-
tionnaires, 823 completed questionnaires were analyzed, with a
response rate of 58.78%. Table 2 shows the demographics of the
respondents.
For the data analysis process, factor analysis was adopted to
analyze and design the structure of a new O-CSI model. The pur-
pose is to reduce and categorize the items into several factors for
the O-CSI model.
First of all, the dimensionality of the consumer styles inventory
was assessed by examining the factor solution. In order to obtain a
factor solution, a principal components factor analysis was used
with varimax rotation. In addition, the amount of variance
explained by the extracted factors (i.e., their eigenvalues) was noted
in order to determine which items have very high correlation
values and which can be eliminated. The above process was
repeated for the four different product industries.
The results show that the ?rst seven components have eigen-
values higher or close to 1 for all the four product industries. As a
result, there are seven extracted factors in the O-CSI model. Table 3
shows the total variance and the initial eigenvalues of components
for Apparel and IT item industries.
Table 1
Inventory items in the O-CSI model.
1. Getting very good quality is very important to me.
2. Once I ?nd a product or brand I like, I stick with it.
3. The well-known national brands are best for me.
4. The higher the price, the better its quality.
5. I prefer buying the best-selling brands.
6. I usually have one or more products of the very newest style.
7. Fashionable, attractive styling is very important to me.
8. I buy as much as possible at sale price.
9. The lower price products are usually my choice.
10. I look carefully to ?nd the best value for the money.
11. When buying products, portability is very important to me.
12. The smaller the product size, the more I prefer them.
13. When I go shopping online, privacy and security are very important.
14. It is very important for the websites to offer communication channels to me
for product enquiries and order tracking.
15. It is very important for the websites to offer a product searching service to
me.
16. It is perfect if the websites can offer me richness of information about
products.
17. It will be annoying to get a lot of animated effects on the business websites.
18. Design layout of business website is one of the important factors in making
buying decisions.
19. It is good if the websites can offer customer reviews on the products.
20. It is good if the websites can offer social networking facilities so that I can
share product comments with my friends.
Table 2
Demographics of the respondents.
Demographics Number Percent
Gender
Female 434 52.8
Male 389 47.2
Age
< 29 365 44.3
30e39 311 37.8
40e49 117 14.2
> 50 30 3.7
Job positions
Top and middle managers 117 14.2
Line managers 139 16.9
Frontline staff 521 63.3
Others 46 5.6
Industries
Manufacturing 60 7.3
Marketing and service 283 34.4
Information technology 114 13.9
Government agencies 216 26.2
Others 150 18.2
Table 3
Total variance explained.
Component Initial eigenvalues Extraction sums of
squared loadings
Total % of
variance
Cumulative % Total % of
variance
Cumulative %
Apparel
1 3.006 18.347 18.347 3.006 18.347 18.347
2 2.607 15.525 33.872 2.607 15.525 33.872
3 1.954 11.658 45.530 1.954 11.658 45.530
4 1.513 9.306 54.836 1.513 9.306 54.836
5 1.258 7.684 62.520 1.258 7.684 62.520
6 1.069 6.739 69.259 1.069 6.739 69.259
7 0.976 6.022 75.281 0.976 6.022 75.281
8 0.906 5.147 80.428
9 0.791 4.195 84.623
10 0.707 3.628 88.251
11 0.614 2.913 91.164
12 0.574 2.613 93.777
13 0.517 2.172 95.949
14 0.383 1.385 97.334
15 0.350 1.004 98.338
16 0.314 0.697 99.035
17 0.283 0.482 99.517
18 0.103 0.259 99.776
19 0.051 0.148 99.924
20 0.025 0.076 100.000
Information technology items
1 3.787 21.039 21.039 3.787 21.039 21.039
2 2.872 16.757 37.796 2.872 16.757 37.796
3 2.217 12.926 50.722 2.217 12.926 50.722
4 1.829 9.485 60.207 1.829 9.485 60.207
5 1.361 5.848 66.055 1.361 5.848 66.055
6 1.172 5.102 71.157 1.172 5.102 71.157
7 0.969 4.627 75.784 0.969 4.627 75.784
8 0.828 4.232 80.016
9 0.728 3.843 83.859
10 0.697 3.670 87.529
11 0.642 3.369 90.898
12 0.498 2.566 93.464
13 0.353 1.819 95.283
14 0.306 1.402 96.685
15 0.267 1.072 97.757
16 0.218 0.713 98.470
17 0.161 0.528 98.998
18 0.133 0.457 99.455
19 0.074 0.319 99.774
20 0.041 0.226 100.000
Extraction method: principal component analysis.
K.M. Sam, C. Chatwin / Asia Paci?c Management Review 20 (2015) 100e107 102
2.6. Con?rmation of the seven-factor model
According to Table 3, a seven-factor solution was supported by
the percentage of variance accounted for by each factor. The seven-
factor solution was chosen based on the scree test for determining
the number of factors to be retained in factor analysis. Based on the
results, all eigenvalues exceeded or were close to 1.0 and our
requirement that at least 75% of the cumulative variance be
explained by the set of retained factors. More importantly, the
seven factors con?rm the importance of the characteristics pro-
posed and the seven-factor consumer style model can be applied to
the four product industries. Table 4 shows the seven factors and the
factor loadings of their corresponding inventory items.
Factor 1. This factor measures high quality and becomes a
conscious consumer characteristic. Items loading on this factor
measure howimportant a consumer thinks quality is for the buying
decision and it becomes a consumer habit for buying the same
high-quality products.
Factor 2. This factor identi?es a brand conscious consumer
characteristic, brand conscious, price equals quality. It measures
consumers' orientation towards buying the more expensive, well-
known national brands.
Factor 3. This factor measures a novelty-fashion conscious
consumer characteristic. High scores on this characteristic indicate
that a consumer prefers a new product style to those old fashioned
styles.
Factor 4. This factor measures a price conscious characteristic. A
consumer having a high score on this factor is sensitive to product
price and prefers buying low price products.
Factor 5. This factor measures the product portability conscious
characteristic. Those consumers who prefer a smaller size product,
so that it is ?exible for carrying around, have a high score on this
factor.
Factor 6. This factor measures the website content conscious
characteristic. What do consumers think about the facilities, such
as privacy, security, searching tools, communication tools for
product enquiry and order tracking, availability and richness of
product information, customer review and social networking
capability offered by online shops? The answer to this question can
affect the score on this factor. A high score indicates that consumers
care very much about the website facilities, such that it can affect
the consumers' buying decision.
Factor 7. This factor measures the website interface conscious
characteristic. The design of the website is important to some
consumers. Is it better to offer animation effect on the business
website to attract consumers' attention? Some consumers don't
like animation effects on business websites. The reasons include:
(1) confusion about the information displayed on screen; (2) low
data transmission speed on the Internet. Is it better to offer graphics
display instead of text display on sensitive information or infor-
mation that is not easy to understand? The location of the web tools
on the website can also affect some consumers when they want to
get some services from the website. Consumers who have a high
score consider the web interface very important, so it can affect
their buying decisions signi?cantly.
Table 5 presents Cronbach a reliabilities for the set of inventory
items within each decision-making style. Since most of the a co-
ef?cients have values around or above 0.7, the items are said to have
at least acceptable internal consistency.
2.7. A pro?le of online consumer style
A pro?le of online consumer style can be developed for each
of the product types based on the highest loading item on
each consumer style. These data were calculated by referring to
the raw scores on the highest loading item for each consumer
style. This result yields scores of 1e5 for each product on each
consumer style. Table 6 presents the means of highest loading
items and the percentages of scoring very high to very lowon each
scale for the apparel industry. Based on Table 6, a pro?le of online
consumer style can be established for the apparel industry in
Table 7.
Table 4
Online consumer decision-making styles (style characteristics): seven factor model.
Loadings: Apparel IT item Jewelry Car
Factor 1 e high-quality, become buying habit conscious consumer
Getting very good quality is very important to me. 0.35 0.81 0.58 0.81
Once I ?nd a product or brand I like, I stick with it. 0.75 0.35 0.40 0.59
Factor 2 e brand conscious consumer
The well-known national brands are best for me. 0.78 0.62 0.77 0.67
The higher the price, the better its quality. 0.84 0.78 0.46 0.63
I prefer buying the best-selling brands. 0.77 0.82 0.73 0.84
Factor 3 e novelty-fashion conscious consumer
I usually have one or more products of the very newest style. 0.82 0.80 0.74 0.85
Fashionable, attractive styling is very important to me. 0.73 0.87 0.59 0.79
Factor 4 e price conscious consumer
I buy as much as possible at sale price. 0.81 0.80 0.71 0.61
The lower price products are usually my choice. 0.76 0.56 0.87 0.81
I look carefully to ?nd the best value for the money. 0.40 0.70 0.45 0.57
Factor 5 e product portability conscious consumer
When buying products, portability is very important to me. 0.85 0.86 0.80 0.83
The smaller the product size, the more I prefer them. 0.83 0.57 0.78 0.82
Factor 6 e website content conscious consumer
When I go shopping online, privacy and security are very important. 0.38 0.50 0.32 0.52
It is very important for the websites to offer communication channels to me for product enquiries and order tracking. 0.75 0.82 0.82 0.83
It is very important for the websites to offer a product searching service to me. 0.90 0.87 0.92 0.83
It is perfect if the websites can offer me richness of information about products. 0.86 0.84 0.85 0.79
It is good if the websites can offer customer reviews on the products. 0.88 0.91 0.82 0.76
It is good if the websites can offer social networking facilities so I can share product comments with my friends. 0.85 0.92 0.77 0.84
Factor 7 e website interface conscious consumer
It will be annoying to get a lot of animated effects on the business websites. 0.62 0.81 0.68 0.72
Design layout of business website is one of the important factors to make buying decisions. 0.89 0.86 0.96 0.90
K.M. Sam, C. Chatwin / Asia Paci?c Management Review 20 (2015) 100e107 103
3. Computation of factor scores for O-CSI model
The earliest factor scoring methods developed yield least-
squares regression weights, factor score coef?cients, which are
applied to the data to estimate the factor scores. These exact-
regression methods have been criticized on several grounds: (1) the
vulnerability of shrinkage effects; (2) using degrees for each item
(Schweiker, 1967); and (3) not meeting the requirements of validity,
univocality, and orthogonality for an orthogonal factor solution.
As an alternative to the exact-regression methods, several early
authors (Cattell, 1952; Thurstone, 1947) introduced the unit-
loading method, which requires the investigator to examine the
structure coef?cients, select the items that surpass an arbitrary
salience criterion (e.g., ± 0.40), and unit-weight those items
(converted to z-scores if measured on different scales) in accor-
dance with the signs of the structure coef?cients. The unit-loading
method became popular and was used to measure e-marketing
mix elements for online businesses (Sam & Chatwin, 2012)
because of its computational convenience and the added bene?t of
avoiding the shrinkage effects associated with the regression-
based scoring methods. Wackwitz and Horn (1971) also found
that the unit-loading method produced factor scores that were
more valid and orthogonal than the scores produced by the exact-
regression methods. However, the unit-loading method will pro-
duce inaccurate factor score estimates (Wackwitz & Horn, 1971).
The loadings (structure coef?cients) do not necessarily indicate
how the items must be weighted and combined to create scores
for individual factors. That information is speci?cally conveyed by
the factor score coef?cients.
Table 5
Reliability coef?cients for seven online consumer styles.
Consumer style characteristics Cronbach a for subscale of all items
Apparel Information
technology item
Jewelry Car
High-quality, becoming
buying habit conscious
0.65 0.67 0.64 0.66
Brand conscious 0.93 0.84 0.76 0.82
Novelty-fashion conscious 0.85 0.84 0.76 0.83
Price-value conscious 0.67 0.73 0.78 0.67
Portability conscious 0.84 0.81 0.81 0.80
Website content conscious 0.94 0.96 0.98 0.94
Website interface conscious 0.88 0.92 0.80 0.85
Table 6
Statistical analysis of highest loading item for the seven online consumer characteristics for the apparel industry.
Style characteristic Highest loading item Percentage score of highest loading item
Highest loading item Mean Very High (1) High (2) Med. (3) Low (4) Very Low (5)
High-quality, buying habit
conscious
Once I ?nd a product or brand I like, I stick with it. 2.6 15% 42% 22% 14% 7%
Brand conscious The well-known national brands are best for me. 3.3 7% 22% 25% 31% 15%
Novelty-fashion conscious I usually have one or more products of the very
newest style.
3.3 7% 20% 24% 30% 19%
Price conscious I buy as much as possible at sale price. 2.6 21% 30% 22% 22% 5%
Portability conscious When buying products, portability is very
important to me.
3.0 10% 21% 42% 13% 14%
Website content conscious It is very important for the websites to offer
a product searching service to me.
1.7 54% 34% 6% 3% 3%
Website interface conscious Design layout of business website is one of the
important factors in making buying decisions.
2.9 13% 31% 24% 16% 16%
Table 7
A pro?le of online consumer style.
Your Name: _____________________________________________________________________
Style characteristics Your score Group mean Verbal interpretation of your consumer style
Apparel
High-quality, buying habit conscious 3.2 2.6 You are average in demanding and buying the same high-quality products.
Brand conscious 2.8 3.3 You are average to above average in brand name consciousness.
Novelty-fashion conscious 2.1 3.3 You are high in novelty and fashion consciousness.
Price conscious 3.8 2.6 You are low in price consciousness e price matters little to you.
Portability conscious 1.2 3.0 You are very high in portability consciousness, demanding a great deal of product ?exibility.
Website content conscious 2.6 1.7 You are below average in website content consciousness, not considering any
facilities inside the websites very much.
Website interface conscious 4.5 2.9 You are very low in website interface consciousness, there is no effect of website
interface on your buying decisions.
Information technology items
High-quality, buying habit conscious 3.2 1.7 You are below average to low in demanding and buying the same high-quality products.
Brand conscious 2.8 2.7 You are average in brand name consciousness.
Novelty-fashion conscious 2.1 3.1 You are high in novelty and fashion consciousness.
Price conscious 3.8 2.8 You are below average to low in price consciousness e price matters little to you.
Portability conscious 1.2 2.4 You are very high in portability consciousness, demanding a great deal of product ?exibility.
Website content conscious 2.6 1.6 You are below average in website content consciousness, not considering any facilities
inside the websites very much.
Website interface conscious 4.5 3.0 You are very low in website interface consciousness, there is no effect of website
interface on your buying decisions.
K.M. Sam, C. Chatwin / Asia Paci?c Management Review 20 (2015) 100e107 104
From above, both unit-loading and exact-regression methods
have advantages and disadvantages. A unit-regression strategy was
created to combine their advantages by applying the same logic of
the unit-weighting scheme to the multidecimal factor score co-
ef?cients (Harris, 1985). The items with salient factor score co-
ef?cients could be unit-weighted in accordance with the signs of
their respective coef?cients and summed. Those items with non-
salient score coef?cients would receive weights of zero and would
hence not contribute to the estimation of the factor scores. The
unit-regression estimates would consequently be resistant to the
shrinkage effects and would not suffer from the validity problems
that af?ict the unit-loading estimates. With respect to the factor
interpretation, the unit-regression method would also possess the
bene?cial feature of representing the items as a whole rather than
fractions.
In order to support the superiority of the unit-regression
method, Grice and Harris (1998) extended Wackwitz and Horn's
methodology (1971) to include complex factor structures for
assessing shrinkage effects among the exact-regression, unit-
regression, and unit-loading methods. The results generally sup-
ported the unit-regression strategy.
Results from comparisons between the unit-regression and
popular unit-loading methods also supported the unit-regression
method. The unit-regression estimates were found to be more
valid and orthogonal than unit-loading estimates. As a result, the
unit-regression method is used to evaluate factor scores of the
online consumers' decision-making styles.
3.1. Factor score coef?cient matrices
Table 8 presents the factor score coef?cient matrices of the O-
CSI model for apparel and IT item industries. The weights of in-
ventory items for the seven decision-making styles shown in
Table 8 are used to evaluate the factor scores of the online con-
sumers' decision-making styles.
Based on the coef?cient matrices in Table 8, the following facts
can be deduced:
For a particular industry, different decision-making styles have
different weights for the same inventory items.
For the same online consumers' decision-making style, there are
different weights for the inventory items in different industries.
Table 8
Factor score coef?cient matrix for the apparel and information technology industries.
Component (Factor)
1 2 3 4 5 6 7
Apparel industry
Prod1_ans1 0.028 À0.113 À0.162 0.222 À0.021 0.404 0.105
Prod1_ans2 0.059 À0.094 À0.013 0.057 À0.092 0.583 À0.004
Prod1_ans3 0.012 0.492 À0.044 À0.294 0.053 0.108 0.156
Prod1_ans4 0.049 0.390 À0.001 0.017 0.066 À0.079 À0.146
Prod1_ans5 À0.081 0.376 0.158 0.084 À0.087 À0.166 À0.081
Prod1_ans6 À0.024 À0.014 À0.082 0.543 0.027 0.072 0.107
Prod1_ans7 À0.070 À0.019 À0.068 0.409 0.126 0.152 À0.007
Prod1_ans8 0.019 0.063 À0.010 À0.116 0.535 À0.135 0.169
Prod1_ans9 0.173 À0.043 0.162 0.074 0.496 0.022 À0.278
Prod1_ans10 À0.048 À0.057 À0.039 0.090 0.377 À0.070 0.121
Prod1_ans11 0.016 À0.068 0.486 0.073 0.034 À0.049 0.080
Prod1_ans12 À0.071 0.004 0.474 À0.009 À0.104 0.044 À0.012
Prod1_ans13 0.405 À0.033 0.027 0.025 0.065 À0.031 À0.244
Prod1_ans14 0.403 0.066 0.092 À0.230 0.009 0.204 À0.084
Prod1_ans15 0.582 À0.045 0.075 À0.002 0.084 0.019 À0.078
Prod1_ans16 0.473 0.098 0.071 À0.094 À0.028 À0.148 0.126
Prod1_ans17 À0.094 0.048 À0.073 À0.107 0.021 0.042 0.481
Prod1_ans18 À0.012 0.008 À0.011 0.004 À0.035 À0.019 0.835
Prod1_ans19 0.484 À0.126 0.013 À0.072 À0.011 0.182 À0.136
Prod1_ans20 0.451 0.102 À0.118 0.131 À0.337 À0.221 À0.037
Information technology items
Prod2_ans1 0.044 À0.066 À0.056 À0.087 0.148 0.385 0.037
Prod2_ans2 À0.009 0.022 0.038 À0.007 0.091 0.649 À0.178
Prod2_ans3 0.025 0.448 0.034 À0.196 0.033 À0.188 0.217
Prod2_ans4 0.034 0.430 À0.091 À0.066 À0.099 0.052 À0.152
Prod2_ans5 0.011 0.363 À0.008 0.050 À0.010 À0.077 À0.086
Prod2_ans6 À0.040 À0.070 0.009 0.556 À0.055 À0.020 0.041
Prod2_ans7 0.052 0.071 À0.016 0.347 0.072 À0.152 0.035
Prod2_ans8 0.090 0.039 À0.104 À0.367 0.533 0.070 À0.030
Prod2_ans9 À0.073 À0.176 0.008 0.250 0.490 0.141 0.038
Prod2_ans10 À0.097 0.130 À0.067 0.222 0.326 À0.151 0.241
Prod2_ans11 À0.013 À0.106 0.510 À0.052 0.015 À0.104 À0.021
Prod2_ans12 0.003 0.106 0.572 À0.043 À0.154 0.013 0.007
Prod2_ans13 0.354 0.051 0.238 À0.193 À0.117 À0.332 0.211
Prod2_ans14 0.467 À0.091 À0.091 À0.036 0.010 0.117 À0.100
Prod2_ans15 0.517 À0.079 À0.021 0.003 0.029 0.097 0.016
Prod2_ans16 0.349 0.025 À0.006 À0.008 0.103 À0.112 0.029
Prod2_ans17 0.171 À0.017 À0.045 À0.010 À0.006 0.022 0.404
Prod2_ans18 À0.087 0.046 À0.057 À0.055 À0.023 0.123 0.713
Prod2_ans19 0.415 À0.118 0.043 À0.124 À0.111 0.162 À0.172
Prod2_ans20 0.386 0.032 À0.166 0.119 À0.475 À0.187 À0.043
Extraction method: principal component analysis.
Rotation method: varimax with Kaiser normalization.
K.M. Sam, C. Chatwin / Asia Paci?c Management Review 20 (2015) 100e107 105
Having considered the two facts above, the unit-regression
factor-scoring formula for Factor 1 in the apparel industry is
shown in Equation (1).
where: Weight
(app, 1, j)
¼ the weight of the j
th
inventory item with
respect to factor 1 in apparel industry; and Item_Score
(j)
¼the score
point of the j
th
inventory item
The factor score of factor i is calculated by summing up the n
weighted item scores of factor i, based on the weight of each in-
ventory item of the factor i. Since only the items whose weights are
0.4 are involved in evaluating the factor score, those items whose
weights are < 0.4 are ignored, as indicated by the value zero in
Equation (1).
Taking an online company selling electrical appliances as an
example, the registered users need to de?ne their O-CSI pro?les
by answering an online survey about the inventory items of the
O-CSI model, their pro?les are then stored and their factor scores
of online consumers' decision-making styles can be computed
based on Equation (1). As customers register in the online busi-
ness website, the group means of factor scores and the weights of
the factor items corresponding to the online consumers' decision-
making styles will be changed. By comparing the updated factor
scores with the frequently updated group means, the online
consumers' decision-making styles can be analyzed such that
business managers can estimate online consumer behavior
accurately.
4. Conclusions
In this paper, a 20-item O-CSI model has been designed for
online consumers in Macau. The results showed that there are
seven decision-making styles identi?ed in the O-CSI model: (1)
high-quality, buying habit consciousness; (2) brand conscious-
ness; (2) novelty-fashion consciousness; (4) price consciousness;
(5) portability consciousness; (6) website content consciousness;
and (7) website interface consciousness. In addition, the O-CSI
model can be applied to different product items. In order to allow
managers to analyze the characteristics of online consumers
quantitatively, the factor scores of the decision-making styles are
evaluated. The accuracy of the factor scores can be ensured by
adopting the unit-regression method in which the weights of
inventory items for their corresponding decision-making styles
play a very important role. Although this research focuses on
Macau, the methodology can be applied to any market or loca-
tion. By using the suggested methodology, online businesses will
be able to become more competitive and increase their market
share.
Con?icts of interest
All contributing authors declare no con?icts of interest.
Acknowledgments
This research was supported by the Department of Accounting
and Information Management at the University of Macau. In
addition, the survey could be completed quickly due to the support
of my colleagues and my friends.
References
AbouTourism. (2010). Statisical Survey Report on Internet Development in China.http://aboutourism.wordpress.com/2010/03/31/statistical-survey-report-on-
internet-development-in-china Accessed 05.04.10.
Boyd, A. (2007). Customer Self-Service: How May You Help Me?.http://www.
ecommercetimes.com/story/57949.html Accessed 21.06.07.
Cattell, R. B. (1952). Factor analysis: An introduction and manual for the psychologist
and social scientist. New York: Harper and Brothers.
Durvasula, S., Lysonski, S., & Andrews, J. C. (1993). Cross-cultural generalizability of
a scale for pro?ling consumers' decision-making styles. Journal of Consumer
Affairs, 27(2), 55e65.
Fan, J. X., & Xiao, J. J. (1998). Consumer decision-making styles of young-adult
Chinese. Journal of Consumer Affairs, 32(2), 275e294.
Forrester Forecast. (2010). Forrester Forecast: Online retail sales will grow to $250
billion by 2014.http://techcrunch.com/2010/03/08/forrester-forecast-online-
retail-sales-will-grow-to-250-billion-by-2014/ Accessed 01.09.11.
Gibs, J., & Bruich, S. (2010). Advertising effectiveness: Understanding the value of a
social media impression. New York: The Nielsen Company.
Gong, W., Maddox, L. M., & Stump, R. L. (2012). Attitudes toward online shopping: a
comparison of online consumers in China and the US. International Journal of E-
Business Development, 2(1), 28e35.
Grice, J., & Harris, R. (1998). A comparison of regression and loading weights for
the computation of factor scores. Multivariate Behavioral Research, 33(2),
221e247.
Hafstrom, J. L., Chae, J. S., & Chung, Y. S. (1992). Consumer decision-making styles:
comparison between United States and Korean young consumers. Journal of
Consumer Affairs, 26(1), 146e158.
Harris, R. J. (1985). A primer of multivariate statistics (2nd ed.). Orlando: Academic
Press.
Higgins, J. K. (2011). Internet Privacy: Trust Big Business or Trust the Government?.http://www.ecommercetimes.com/rsstory/72032.html Accessed 10.03.11.
Internet World Stats. (2012). Internet Users in Asia 2012.http://www.
internetworldstats.com/stats3.htm Accessed 03.01.13.
Leggatt, H. (2012). Online privacy real concern for 90% of US Internet users. http://
www.bizreport.com/2012/02/90-percent-of-online-adults-worry-about-their-
online-privacy.html Accessed 14.02.12.
Lysonski, S., Durvasula, S., & Zotos, Y. (1996). Consumer decision-making styles: a
multi-country investigation. European Journal of Marketing, 30(12), 10e21.
New Zealand Consulate-General Hong Kong. (2011). Hong Kong Review.http://www.
mfat.govt.nz/downloads/media-and-publications/newsletters/hong-kong/
HongKong-Review-June-2011.pdf.
Nielsen, W. (2010a). Social Networks/Blogs Now Account for One in Every Four and a
Half Minutes Online.http://blog.nielsen.com/nielsenwire/global/social-media-
accounts-for-22-percent-of-time-online/ Accessed 15.06.10.
Nielsen, W. (2010b). Friending the Social Consumer.http://blog.nielsen.com/
nielsenwire/online_mobile/friending-the-social-consumer/ Accessed 16.06.10.
Ramsey, G. (2010). 10 best practices for success with social media. New York:
eMarketer.
Raven, P. V., Huang, X., & Kim, B. B. (2007). E-business in developing countries: a
comparison of China and India. International Journal of E-Business Research, 3(1),
91e110.
Sam, K. M., & Chatwin, C. R. (2012). Measuring e-marketing mix elements for online
business. International Journal of E-Entrepreneurship and Innovation, 3(3), 13e26.
Schweiker, R. F. (1967). Factor scores aren't sacred: comments on abuses of factor
scores. American Educational Research Journal, 4, 168e170.
Shim, S. (1996). Adolescent consumer decision-making styles: the consumer so-
cialization perspective. Psychology & Marketing, 13(6), 547e569.
Siwicki, B. (2009). Getting them to write.http://www.internetretailer.com/2009/12/
29/getting-them-to-write Accessed 29.12.09.
Sproles, G. B. (1985). From perfectionism to fadism: measuring consumers'
decision-making styles. In 30th Annual Meeting of the American Council on
Consumer Interests.
Sproles, G. B., & Kendall, E. L. (1986). A methodology for pro?ling consumers'
decision-making styles. Journal of Consumer Affairs, 20(4), 267e279.
Taobao. (2010). China Online Shopping Report 2009e2010.http://www.linkagepro.
com/articles/Internet-Marketing-20100523.htm Accessed 10.05.10.
Factor Score
ðapp;1Þ
¼
X
n
j¼1
Weight
ðapp;1;jÞ
*Item Score
ðjÞ
; Weight
ðapp;1;jÞ
0:4
0; Weight
ðapp;1;jÞ