EVALUATION OF CROSS-CULTURAL WEB INFORMATION
SYSTEM DESIGN GUIDELINES
Gatis Vitols, Irina Arhipova
Faculty of Information technologies, Latvia University of Agriculture, Liela 2, Jelgava, Latvia
Yukako Hirata
Hewlett-Packard Japan, Ltd., 1-2-1, Kamiogi, Suginami-ku, Tokyo, Japan
Keywords: Cross-cultural design, Web information systems, Usability, Accessibility.
Abstract: This article summarizes, analyzes and evaluates the selected set of guidelines provided for the cross-cultural
web information system design. Globalization affects various fields of human activities such as
communication, business, travelling and others. In a business field, many companies search a way to expand
their services globally. One of the main support tools for a global business is an information system that can
function online and instantly serve information to people from various parts of the world. Such information
system is often called a cross-cultural web information system. In this paper hypothesis is brought forward
that part of the guidelines provided for cross-cultural web information system designers are not appropriate
for application in the existing or in future developed projects. Guidelines are selected based on certain
criteria. Evaluations of most visited web information systems by users from Japan and Latvia are performed
by researchers from both countries. Evaluation results compliance to the summarized guidelines is
reviewed.
1 INTRODUCTION
Cross-cultural information system (IS) design is a
significant issue for IS developers. It is recognized
that the design process of cross-cultural IS does not
only involve just a translation of contents to an
appropriate language, but also includes many other
design tasks (Shaoyi, 2001).
People from various cultures think and act
differently. These actions can affect people
interaction with IS, including web information
systems (WIS). At the same time, there is another
point of view that WIS design can shape up users
behaviour, such as thinking ways, actions and values
that have been fundamental to their own culture
(Rau et al., 2008). Partly it can be explained with the
rapidly increasing number of WIS users from
different countries and cultures, which use their
culture characteristics to interact with the existing
WIS and developing new ones (Aykin, 2004).
Published researches (Rau et al., 2008) have
already shown that if a WIS that have been
developed for users in a particular culture is later
without improvements introduced to those from the
other cultures, usability and accessibility of this WIS
decrease.
In the field of cross-cultural IS development
various guidelines and practices have been
introduced by several authors. Some of those
guidelines give design recommendations for WIS
designers, and many are acquired from researches
based on examination of cultural dimensions
introduced by works of Hofsted (Hofsted and
Hofsted, 2005).
However, there are criticisms against Hofsted
works coming in past years, such as performed by
Oshlyansky (Oshlyansky, 2006). Oshlyansky argues,
that cultures keep mixing and modifying with an
impact of various factors. For example, Chinese
users have accepted internationally used colour
associations, such as yellow for attention, however
previously some researches did not show such
tendency (Rau et al., 2008).
According to such criticism, the hypothesis can
be brought forward that part of the guidelines
provided for WIS designers could not be appropriate
275
Vitols G., Arhipova I. and Hirata Y..
EVALUATION OF CROSS-CULTURAL WEB INFORMATION SYSTEM DESIGN GUIDELINES.
DOI: 10.5220/0003471602750280
In Proceedings of the 13th International Conference on Enterprise Information Systems (ICEIS-2011), pages 275-280
ISBN: 978-989-8425-56-0
Copyright
c
 2011 SCITEPRESS (Science and Technology Publications, Lda.)

for applications in the existing or in future
developed projects.
The aim of this article is to summarize, analyze
and evaluate the set of guidelines provided for cross-
cultural web information system designers and
examine cross-cultural WIS design trends. To reach
the aim, following tasks are brought forward:
1. Analyze previous researches and other types of
literature that relate to cross-cultural WIS design
and summarize the main existing guidelines for
designers.
2. Select two culture groups and evaluate the
existing WIS page designs compared with those
summarized guidelines.
3. Summarize evaluation results and review
guidelines compliance to the results.
2 MATERIALS AND METHODS
When summarizing and analysing various literature
sources, multiple WIS design guidelines can be
identified. While some of the guidelines are general
and give issues for considerations, others provide
precise recommendations for WIS designers.
The guidelines for this study were selected based
on certain criteria. Only those guidelines were
selected that could be evaluated with selected
content analysis methods. These methods include the
analysis of WIS pages source code, examination of
page rendering on web browser and execution of
automation tools for support of examination, such as
page loading speed measurers. Scientific articles and
web designer community websites were chosen as
main sources for collection of the guidelines. From
the analysis of selected literature, it can be seen that
designer's community websites mention only few
main guidelines that are also mentioned in selected
scientific papers, such as by Marcus and Rau
(Marcus and Rau, 2009).
In this research following guidelines are brought
forward and are ordered in groups (Table 1). The
groups include guidelines that address page layout,
text content, graphical content and accessibility.
To evaluate those guidelines, two different culture
groups are selected. Users from countries such as
Latvia and Japan are chosen. Latvian users are
chosen as Europe Union countries representative,
while the users from Japan are chosen because they
are separated from Latvian users by distance and
have different characteristics and preferences than
those from European countries (Okazaki, 2004).
Some parts of the research were performed in Japan
and others Latvia.
Table 1: Summarized cross-cultural WIS design
guidelines.
Guideline
Layout
1. For content layout, liquid layout is recommended,
because same text in different languages can use different
space (Choong, 2001; Marcus and Rau, 2009; Reece et
al.
,
2010
)
.
2. Navigation menu should be placed according to users
reading direction. (Reece et al., 2010) As a universal
solution, top horizontal position is recommended (Arno,
2010
)
.
Text content
3. It is recommended to use universal text encoding
system, such as UTF-8 for supporting various characters
(Arno, 2010; Andrew, 2010).
4. Selection of appropriate font families has to be taken
into account. Font families and their loading priorities has
to be defined using Cascade Style Sheets (CSS) (Marcus
and Rau
,
2009
;
Andrew
,
2010
;
Arno
,
2010
)
.
5. Language reading direction has to be considered. Some
countries read from left to right, some right to left and
some top to bottom (Marcus and Rau, 2009; Reece et al.,
2010
)
.
6. Designers should be aware when using alphabetic
numbering for displaying ordered lists, because some
countries have different sequence of letters. Alternative
markin
g
for lists is recommended
(
Marcus and Rau
,
Graphical content
7. Colours can have cultural meanings (Marcus and Rau,
2009; Choong, 2001). White, grey, black, blue and
yellow are most used cross-cultural interface colours
(
Kondratova and Goldfarb
,
2007
)
. Variations of blue
8. Text should not be included in design of icons (Marcus
and Rau, 2009).
9. Publication of non ethical images is not recommended
(Reece et al., 2010). In some cases WIS can be blocked
by legal authorities if the content is offensive and
considered non ethical b
y
users from certain culture
10. Usage of appropriate and understandable icons is
recommended. For example, for e-mail function, image of
an envelope is recommended (Choong, 2001; Reece et
al.
,
2010
)
.
Accessibility related
11. Users use various input and output devices to interact
with IS. Main input device is keyboard. Keyboards are
different in type and layout throughout the world
(
Choon
g
, 2001
)
. International Domain Names
(
IDN
)
can
12. Inclusion of appropriate META keywords in
documents can improve web information system rating at
local search engines (Arno, 2010).
13. Placing a web server in appropriate country can
improve faster access to web information system (Liao et
al., 2010; Andrew, 2010).
14. There is a need to take into consideration network
bandwidth of countries. Multiple versions of WIS should
be developed, including version for low bandwidth (Liao
et al., 2010; Andrew, 2010
)
.
Content analysis was chosen as methodology,
which is also used in various researches that relate to
ICEIS 2011 - 13th International Conference on Enterprise Information Systems
276

identification of cross-cultural WIS design
differences (Okazaki, 2004).
One hundred most visited WIS from each culture
were selected on the first week of September, 2010.
To acquire such information, statistics gathered by
Alexa Internet Company services were used. The
statistics provide a list of most accessed WIS by
certain country, which is chosen as main criteria for
WIS selection.
Each WIS from the acquired list was executed on
web browser and examined. Only the main page of
each WIS was examined. First of all, page source
and CSS file codes of each page were manually
evaluated to find necessary information about page
encoding, META tags, colour codes, font families
and type of layout. Information which was not
possible to gather from the manual examination of
source code was gathered by examination of page
visual rendering on a web browser.
To evaluate certain guidelines, additional tools
were used, such as online WHOIS protocol service
provided by DomainTools.com that allows examine
domains and determine web information system
server location. Also Link Vendor SEO tool named
“Website speed check” was used to determine size
of each page. In the data collection process, part of
data was coded with number 0 meaning non-
affirmative or 1 meaning affirmative, and the other
part as text string.
The results of evaluation were recorded and
processed in spreadsheet documents. Most accessed
Japanese user WIS page evaluation data were
gathered in Japan from September to November,
2010, but most accessed Latvian WIS page
evaluation data in Latvia from October 2010 to
January 2011. Table 2 shows one record of the
evaluation results.
In total, 200 WIS main pages were examined in
this study. From 200 WIS pages, 2% of pages which
are preferred by users from Japan and 1% which are
preferred by users from Latvia were not available for
examination.
3 RESULTS AND DISCUSSIONS
The examination of 1. and 2. layout guidelines
(Table 1) show the following results.
The results acquired from the evaluation of pages
preferred by users from Japan show that 61% of
pages use liquid layout, while 37% have fixed one.
The results acquired from the evaluation of pages
preferred by users from Latvia show that 62% use
liquid layout, while 37% have fixed layout.
Table 2: One record of gathered data from the WIS main
page evaluation process.
Criteria Result
URL of the WIS jugem.jp
IDN is used as main URL 0
Are META keywords present? 1
Is UTF-8 character encoding
used?
0
Which alternative character
encoding used?
EUC-JP
Are CSS styles used for content
presentation?
1
First priority font family Meiryo
Reading direction Left to right
Main colour for design Blue
Are icons with text available? 0
Icon for email Not available
Position of navigation menu Horizontal
Is liquid layout used? 0
Is alphabet list used? 0
Area of web server location Japan
Size of main page 70.68KB
Is non ethical images published? 0
Main language Japanese
Layout examination shows that liquid layout is
more widely used and preferred by WIS designers. It
can be related not only to the fact that same text in
different languages takes additional space, but also
to increasing development of various portable
devices with different screen sizes. So WIS pages
need to have ability to be stretched and adjusted to
both smaller and bigger size displays.
The results acquired from the evaluation of pages
preferred by users from Japan show that 64% of
pages have horizontal navigation menu, while 2% do
not have menu, 22% used vertical menu and 10%
used mixed menu, which includes combination of a
horizontal and a vertical menu. The results acquired
from the evaluation of pages preferred by users from
Latvia show that 59% have horizontal navigation
menu, while 7% do not have menu, 24% used
vertical menu and 9% used mixed menu. Even it is
recommended that a navigation menu should be
placed according to users' reading direction, for
example a menu should be placed in the right side of
a WIS page in Japanese case, however there were no
WIS pages met which menu was placed on the right
side which is one of the primary side where to start
reading from in Japanese.
The examination of 3.–6. text content guidelines
(Table 1) show the following results.
The evaluation of pages preferred by users from
Japan show that 57% of pages use UTF-8 encoding,
while 41% use encodings such as Shift_JIS and
EUC-JP. The evaluation of pages preferred by users
from Latvia show that 85% use UTF-8 encoding,
while 14% use encodings such as ISO-8859-1,
EVALUATION OF CROSS-CULTURAL WEB INFORMATION SYSTEM DESIGN GUIDELINES
277

Windows-1251 or Windows-1257. It can be seen
that in Latvian case WIS designers have in most
cases implemented UTF-8 encoding that can support
various character encoding. It can be related to the
fact that many of these WIS pages are bilingual and
have to display information in Latvian as the
primary language and the secondary most used
language which is Russian. Japanese WIS pages
have less UTF-8 application cases and still the
Japanese language character encoding scripts such
as Shift-JIS and EUC-JP are widespread applied.
The evaluation of pages preferred by users from
Japan show that 91% of pages have font families
defined in priority using CSS, while 7% do not use
font family priorities. Most used font families in
pages preferred by Japanese users are MS PGothic,
Hiragino and Osaka. The results acquired from the
evaluation of pages preferred by users from Latvia
show that 99% of pages have font families defined
in priority using CSS and most used font families are
Arial, Verdana and Tahoma. It can be seen that
developers from Japan and Latvia have adopted WIS
page styling with CSS and only in few cases it can
be seen that font families are coded with the help of
for example Hypertext Markup Language (HTML).
There is a seen font-family usage difference. While
Latvian WIS pages use widely known font families,
such as Arial and Tahoma, Japanese use less known
font families, such as Osaka and Hiragino.
Japanese traditional writing direction is top to
bottom and right to left, called tategaki, but Latvian
writing direction is left to right. The evaluation of
pages preferred by users from Japan show that none
of the pages have reading direction from top to
bottom and right to left, while the evaluation of
pages preferred by users from Latvia show that 99%
use reading direction from left to right.
The evaluation of pages preferred by users from
Japan show that 98% of pages do not use alphabetic
numbering for lists or do not have lists, but the
evaluation of pages preferred by users from Latvia
show similar results that is 99% of pages do not use
alphabetic numbering for lists or do not have lists. In
both cultures, for pages containing unordered lists,
various styles of bullets as markers are used, but for
ordered lists, Arabic numerals instead of alphabetic
numbering are chosen.
The examination of 7.–10. graphical content
guidelines (Table 1) show the following results.
The evaluation of pages preferred by users from
Japan show that 49% use variations of blue as a
main colour for design, followed by variations of
grey colour in 29% of cases. 20% use red, yellow,
orange, pink, white and green as main colour. The
results acquired from the evaluation of pages
preferred by users from Latvia show that 46% of
pages use variations of blue as a main colour for
design, followed by variations of grey in 31% and
orange in 13% of pages. 9% use red, black, white,
green and yellow colours. It can be seen that main
colour preferences are similar between both cultures
and that blue colour is one of the most used colours
in evaluated WIS pages.
The evaluation of pages preferred by users from
Japan show that in 88% of pages icons do not
include text or icons are not used on the page, while
10% had icons with texts. As for users from Latvia
case, evaluation show that 98% icons do not include
text or icons are not used on the page, while 1% had
icons with texts. Also examination of WIS pages
shows that icons are rarely used as developers give
more preference to buttons with text explanation.
The evaluation of non ethical images is very
relative and time consuming task. In this research
simplified evaluation has been performed, where
non ethical images are considered those which
contain pornography related content. The results
acquired from the evaluation of pages preferred by
users from Japan show that 9% of pages include non
ethical images, while 89% did not contain non
ethical images. The evaluation of pages preferred by
users from Latvia show that 12% include non ethical
images, while 86% did not contain non ethical
images.
For this research evaluation of e-mail icon has
been performed. In case icon is used, envelope as a
cross-cultural icon is recommended for this purpose.
The results of the evaluation of pages preferred by
users from Japan show that in 12% of pages
envelope was used as an icon for marking e-mail,
50% use plain text to mark e-mail, while 36% of
pages did not had e-mail function. Users from Latvia
case show that in 4% of pages envelope was used as
an icon for marking e-mail, 37% use plain text to
mark e-mail, while 57% did not had e-mail function.
There was no case met when for email icon other
image than envelope was used. It is concluded that
WIS developers have taken into account such
recommendation.
The examination of 11.–14. accessibility related
guidelines (Table 1) show the following results.
The evaluation of pages preferred by users from
Japan show that from selected WIS pages there was
no page which used IDN for a primary domain, but
in case of users from Latvia, the evaluation show
that there was one case when IDN for a primary
domain is used. Even IDN is popularised, available
for registration and advised for application, it is seen
ICEIS 2011 - 13th International Conference on Enterprise Information Systems
278

that WIS developers for primary domain names use
Latin characters. In Latvian case when IDN was
used, WIS page still had alternative Latin character
domain name.
The results acquired from the evaluation of pages
preferred by users from Japan show that in 66% of
pages META keywords are used, while in 32%
META keywords are not used. The evaluation of
pages preferred by users from Latvia show that in
48% of pages META keywords are used, while in
51% of pages META keywords are not used. It is
seen that META keywords are used only in half of
overall evaluated pages. META keywords are
mainly used for WIS page indexing by search
engines. So this could be explained by the actions of
developers who realize that the search engine
indexing strategies are changing and META
keyword impact on search results is decreasing.
The evaluation of pages preferred by users from
Japan shows that in 61% of cases, servers for storage
of pages are located in Japan, but in 47% of cases
outside of Japan. The evaluation of pages preferred
by users from Latvia show that in 38% of cases,
servers for storage of pages are located in Latvia, but
in 61% of cases outside of Latvia.
The results acquired from the evaluation of pages
preferred by Japanese and Latvian users show that
none of the WIS offers multiple speed versions,
although the creation of multiple versions is advised
by guidelines. Also none of the tested pages had a
size over 350KB. An average size of the page
preferred by users from Japan is 60,7KB, but for
users from Latvia is 60,3KB. Figure 1 shows
evaluated guideline compliance to the evaluation
results in percents.
Figure 1: Guideline compliance to the evaluation results.
4 CONCLUSIONS
The guidelines selected for this study relate to WIS
design and with the results gathered from the
experiments show certain trends for developers.
It is seen that many of the recommendations
included in the guidelines are applied in practice and
have good compliance with the experiments results.
However, designers of published WIS do not or
partly follow certain recommendations.
This study found that almost one third of
summarized guidelines have compliance to the
results beneath 50% (Figure 1).
This result can be directly linked to happening
synthesis of cultures with the de facto practices of
WIS development. One of clear examples is
preferences by users from Japan. Although the
guidelines recommend taking into consideration
reading direction of certain culture, which would be
starting from the right side and reading top to bottom
for Japanese users, there was no case found where
such content layout direction is applied. IDN
application cases of are also low in numbers.
Developers are also not focusing on creation of
multiple bandwidth versions of their WIS and files
storage on country local servers, even it is
recommended by the guidelines. It is a seen trend
that with the development of global networking
developers will be less focusing on matters relating
network bandwidth, even in a cross-culturally used
WIS pages.
This study shows that developers should be
cautious with published guidelines straightforward
application in existing and future WIS projects.
The previously brought forward hypothesis
appears to be true. It can be concluded that users
from various cultures are accepting and using
widespread applied information displaying practices
and adjusting their cultural preferences.
Still, for more successful design of WIS, further
researches are required with local users from
selected cultures.
There is also a need to do a further researches
with the literature that include guideline groups that
were not evaluated in this study and are addressing
issues such as typography, verbal style and aesthetic
of WIS pages.
ACKNOWLEDGEMENTS
Funding support for this research provided by
Europe Social Fund program “Support for doctoral
studies in LUA”, agreement 2009/0180/1DP/
1.1.2.1.2/09/IPIA/VIAA/017.
EVALUATION OF CROSS-CULTURAL WEB INFORMATION SYSTEM DESIGN GUIDELINES
279

REFERENCES
Alexa Internet. (2010). Alexa - Top Sites for Countries
and Cities. Retrieved September 23, 2010, from
http://www.alexa.com/topsites/countries
Andrew, P. (2010). Tips and Thoughts on Cross-Cultural
Global Web Design. Retrieved October 30, 2010, from
http://speckyboy.com/2010/06/30/tips-and-thoughts-
on-cross-cultural-global-web-design/
Arno, C. (2010). How to Design Websites That Appeal
Across Cultural Divides. Retrieved November 11,
2010, from http://tutorialblog.org/how-to-design-
websites-across-cultural-divides/
Aykin, N. (2004). Overview: Where to Start and What to
Consider. In Aykin, N. (Ed.), Usability and
Internationalization of Information Technology (pp. 3-
21). Boca Raton: CRC Press.
Choong, Y. (2001). Cross-cultural Issues in Human-
Computer Interaction. In Karwowski, W. (Ed.),
International Encyclopedia of Ergonomics and Human
Factors (Vol. 1, Part 5, pp. 644-647). London: Taylor
& Francis.
Hofstede, G. and Hofstede, G.J. (2005). Culture and
Organizations: Software of the Mind. New York:
McGraw-Hill.
Kondratova, I. and Goldfarb, I. (2007). Colour Your
Website: Use of Colours on the Web. In Aykin, N.
(Ed.), Usability and Internationalization Part II
(LNCS Vol. 4560, pp. 123-132). New York: Springer-
Verlag.
Liao, H., Guo, Y., Savoy, A. and Salvendy, G. (2010).
Content Preparation Guidelines for the Web and
Information Appliances: Cross-Cultural Comparisons.
Boca Raton: CRC Press.
Marcus, A. and Rau, P.R. (2009). International and
Intercultural User Interfaces. In Stephanidis, C. (Ed.),
The Universal Access Handbook (pp. 156-166). Boca
Raton: CRC Press.
Okazaki, S. (2004). Does Culture Matter?: Identifying
Cross-national Dimensions in Japanese Multinationals'
Product-based Websites. Electronic Markets, 14(1),
58-69.
Oshlyansky, L., Cairns, P. and Thimbleby, H. (2006). A
cautionary tale: Hofstede’s VSM revisited. In
Proceedings of the 20th BCS HCI Group Conference
(Vol. 2, pp. 11-15). London: British HCI Group.
Rau, P.R., Gao, Q. and Liang, S.M. (2008). Good
computing systems for everyone – how on earth?
Cultural aspects. Behaviour & Information
Technology, 27(4), 287-292.
Reece, G., Nesbitt, K., Gillard, P. and Donovan, M.
(2010). Identifying Cultural Design Requirements for
an Australian Indigenous Website. In Proceedings of
the Eleventh Australasian Conference on User
Interface (Vol. 106, pp. 89-97). Darlinghurst:
Australian Computer Society, Inc.
Shaoyi, H. (2001). Interplay of Language and Culture in
Global E-commerce: A Comparison of Five
Companies' Multilingual Website. In Proceedings of
the 19th annual international conference on Computer
documentation (Vol. 1, pp. 83-88) New York: ACM.
ICEIS 2011 - 13th International Conference on Enterprise Information Systems
280