CROSS-CULTURAL DESIGN
3.1 Graphical User Interface
As discussed in previous sessions, how people interact with objects around them is determined by their past experiences with these objects (or similar objects) and their expectations of how things should work when they use the objects. A graphical user interface (GUI) is the graphical and visual representation of, and interaction with, programs, data, and objects on the display of a system. An important feature of GUIs is the ability for users to directly manipulate elements and information on the screen display. A GUI system is portrayed as an extension of the real world. Thus, the users’ experiences and expectations play a critical role in the usability of GUI systems. The same is true when it comes to design for users with different cultural backgrounds in that the cultural experiences and expectations of the target audience need to be considered throughout the development lifecycle.
A GUI system usually includes four basic elements:
window, icon, menu, and pointing device (WIMP). The WIMP interaction was first developed at Xerox PARC in the 1970s with the desktop metaphor on a personal computer. The WIMP interaction and its variations and extensions are widely used in computing technologies today, including desktop applications, Internet browsing, Web applications, and mobile computing.
3.1.1 Information Organization and Representation
As discussed earlier, differences in cognitive styles exist among people of different cultural backgrounds.
Those differences have implications on how information should be represented and organized on GUIs to accommodate different thinking styles due to cultural differences.
Oftentimes on a GUI the information is organized into hierarchical structures. For example, menus display
listings of choices or alternatives that users have at appropriate points while using the system or create a set of listings that guide a user from a series of general descriptors through increasingly specific categories on following listings until the lowest level listing is reached. Design guidelines and best practices have been developed for designing usable menus (e.g., ISO, 1997;
Galitz, 1997; Weinschenk et al., 1997). Those guidelines call for logical and meaningful organization of menu options. The menu options should be arranged into conventional or natural groups known to users and follow logical order. The groupings should be logical, distinctive, meaningful, and mutually exclusive.
Designing menu structures for people with different cognitive styles cross-culturally needs to accommodate the differences in the tendency of information organi- zation and represent the information accordingly. As Choong (1996) points out when representing informa- tion of a system on a GUI, Chinese users will bene- fit from a thematically organized information structure, whereas American users will benefit from a functionally organized structure. Nawaz et al. (2007) report similar results of cultural differences on the ways the Chinese and Danish group objects, functions, and concepts into categories. In the card-sorting tasks, the Danish sub- jects prefer to highlight category name by its physical attributes, whereas the Chinese subjects highlight the category by identifying the relation between different entities. The Chinese subjects also utilized more the- matic categories than the Danish subjects in the study by Nawaz et al. (2007). Kim et al. (2007) report sim- ilar findings with Korean and Dutch users interacting with menu structure designed for mobile phone inter- faces. The relational grouping participants (Koreans) were more likely to select and prefer the thematically grouped menu, whereas taxonomic grouping participants (Dutch) had the tendency to select and prefer the func- tionally grouped menu.
All user interfaces use some forms of metaphors to provide visual and conceptual representations of major user interaction objects and their associated actions. A well-chosen metaphor can be helpful when all or part of the interface includes functions or features that are new to the target users. Metaphors are used to help users connect what they do not know with what they have known. It is imperative that the metaphors are sim- ple, easily understood, and quickly learned. Metaphors should allow target users to easily relate to their real- world experiences. When designing cross-cultural user interfaces, the use of metaphors becomes a challenge as the real world changes from culture to culture. Many companies choose to localize metaphors in their user interfaces by only redesigning the objects or translat- ing the text in a certain metaphor. However, “translat- ing” a metaphor is not sufficient; the entire metaphor will need to be reevaluated and replaced to make the interface mapped to the target users’ cultural experi- ence (Evers, 1998). For example, for designing multi- cultural applications, Salgado et al. (2009) propose five
conceptual metaphors to accommodate users’ attitudes and cultural variables: located “at home” metaphor, tele- scope observer metaphor, close observer metaphor, for- eigner “with subtitles” metaphor, and foreigner “without subtitles” metaphor. Salgado et al. expect that those five conceptual metaphors will help designers think about their own cultural perspective and also help the design- ers to think about how such perspectives can be experi- enced in different ways.
3.1.2 Graphics, Symbols, and Icons
Icons are an essential part in any system with a GUI, that is, WIMP interaction. Icons are small pictorial symbols used on GUIs to represent certain capabilities of the system and to be animated for bringing forth these capabilities for use by the users. The benefits of using icons include to represent visual and spatial concepts, to save screen space, for immediate recognition, for better recall, to reduce user’s reading time, and to help products go global (Horton, 1994).
A number of researchers have written about design- ing icons for specific cultures or for international use.
Shen et al. (2007) report on Chinese Web design with cultural icons. Cultural issues in designing interna- tional biometric symbols are described in Rau and Liu (2010) and Choong et al. (2010). Pappachan and Ziefle (2008) discuss cultural influences on comprehensibility of icons. Kim and Lee (2005) report on cultural differ- ences in icon recognition on mobile phones. From these we can conclude that a good icon for cross-cultural use has the following properties:
• Mimics both the physical appearance and the function or action of the object it represents
• Clearly represents the state of the object if it is an object that can assume more than one state
• Uses only widely recognized conventions for color and shape
• Is not directional and can be used without rotation
• Is not culture bound
• No embedded text characters
• No culture-specific metaphors
3.1.3 Presentation, Navigation, and Layout There are existing guidelines (e.g., ISO, 1997) for designers to follow on how to develop usable presen- tation and layout of GUI components as well as the navigation among those components. The key is that the design has to match the user’s work flow, expe- rience, and expectation. In addition to following GUI design guidelines, there are considerations that need to be addressed in cross-cultural GUI design. For example, different cultures employ different format conventions and measurement systems that will affect the presenta- tion of such information on the GUIs. The arrangement of information on the screen affects how efficiently and comfortably people can scan, read, and find information.
The information should be laid out on the screen in a natural orientation for the target users. For example, in cultures where traditional Chinese form is used, such as Hong Kong and Taiwan, people tend to scan the screen “across the columns” following an “N” pattern. In countries where simplified Chinese is used, for example, China and Singapore, information is mostly printed in rows and read from left to right and top to bottom, that is, a “Z” pattern.
Language Issues The textual information on a GUI imposes significant challenges in cross-cultural design.
First, if English is the language of the base product, it is important to start with unambiguous language that minimizes the use of technological jargon, avoids abbre- viations, and uses plain, simple English. Second, the same language used in different locales can have dialect differences, including spelling, word usage, grammar, and pronunciation. Third, literal translation of terms often fails to accurately translate the meaning or concept underlying a word.
Linguistic differences should be taken into account, such as text directionality, linguistic boundaries, text wrappings, justifications, and punctuations. When pre- senting an ordered list or information, a single sequence or ordering of textual information should not be as- sumed, even if the languages share the same alphabetical system. Collating ideographic characters, such as Chi- nese Hanzi and Japanese Kanji, is more complex than sorting Latin characters. There are four different col- lating methods that GUI designers should be aware of and take into consideration: radicals, number of strokes, phonetic sequence, and frequency of use (Rau et al., 2010). Cross-cultural design also will need to take into account physical language variations such as direction- ality, hyphenation, stressing, fonts, sizes, orientation, layouts, spaces, wrapping, and justification. In Euro- pean languages, such as German, words tend to be long strings of characters. In contrast, Asian words tend to be shorter, but the characters are much more complex (such as Chinese) and will require more pixels to render clearly on a display screen. Cross-cultural GUI designers should be aware of the possible text expansion required both horizontally and vertically.
Conventional guidelines assume that people read left to right and thus use a left-to-right orientation for labeling text boxes, presenting text, scrolling text within a text box, and presenting a series of control buttons. However, designers should take note that some languages are read from right to left, for example, Arabic or traditional Chinese writing. When designing for such languages, the display features should accommodate the text direction as well as navigation flow.
There are some key linguistic differences to be con- sidered when translating a user interface from one lan- guage to another. For user interfaces, adequate screen space needs to be allocated for possible text expan- sion due to translation. For example, German words are usually longer than their counterpart English words.
Composite messages should be avoided, such as warn- ings with a word or words dynamically determined. The
CROSS-CULTURAL DESIGN 177 composite messages will not be translated well since
sentence structures could vary dramatically across lan- guages. The rules for word wrapping can also be very different from one language to another.
Sometimes, it is essential to support multiple lan- guages simultaneously. The product designers should avoid using national flags to toggle among languages or using words in one language for selecting among languages. Using national flags for language selection may be offensive to some users since there can be more than one country using the same language. For example, English is used in the United States, United Kingdom, Canada, and many other countries. Using words in one language as language-selecting options is also inappro- priate since the users will have to know that language to make the selection. For example, in an English user interface, a Chinese user will have a problem in picking out the wordChinesefrom a list consisting of language options written in English if the Chinese user does not understand English.
Scripts are a collection of characters and glyphs that represent a written version of a spoken language.
In many cases, a single script may serve to write tens or even hundreds of languages, for example, the Latin scripts. In other cases, only one language uses a particular script, for example, Hangul, which is used only for the Korean language. The writing systems for some languages may also use more than one script; for example, Japanese makes use of the Han (or Kanji), Hiragana, and Katakana scripts.
Written language can be bidirectional or unidirec- tional. Most languages are unidirectional. For example, English is written from left to right in a unidirectional fashion. Chinese is another example of unidirectional scripts, but the direction can be left to right, right to left, or top to bottom. A bidirectional script, such as Arabic, can be written from right to left and left to right (in certain situations, such as numbers) in the same context.
Some languages, such as Chinese, demand special consideration of alternative methods of text input.
Niu et al. (2010) of Nokia developed a new method
called Stroke++for Chinese character input on mobile phones with the goal of making keypad typing more accessible to novice mobile phone users. Chinese mobile phone input methods use either standardized phonetic notations, such as Pinyin and Zhuyin, or structural information about characters, such as Wubihua or Cangjie. Pinyin requires knowledge of Latinized Chinese Pinyin, which presents difficult barriers to elderly people and nonusers of Pinyin. Wubihua, or Stroke, another popular input method, accepts five distinguishing strokes only when the input sequence is the same as the standard writing order of the character.
These methods are difficult to learn because the standard varies between mobile phone manufacturers and there is significant variation in writing habits between different people. The new Stroke++method exploits the fact that the 600 most frequently used characters in Chinese (out of 20,000 total) can cover 92.9%.of the user’s needs in short messages, thus greatly reducing the required set of radicals. However, rather than defining rules to restrict the sequence of entry, this method allows users to input radicals in arbitrary order to form a character, the desired character being selected from a pull-down list of possibilities organized according to frequency of use. In addition, the keypad layout is designed according to Chinese characters’ square shape. The radicals are grouped to make the keypad meaningful. For example, the radicals (Metal), (Wood), (Water), (Fire), and (Earth) representing the five elements in traditional Chinese culture are arranged in a single line to help users to remember. and are put together in the middle to make the Chinese word for “woman.”
Table 5 provides guidelines for use of language.
Format Conventions and Measurement Systems Formats are an artifact and are specific to different locales. Numeric values can be represented in different ways. Separators are not always used; even if they are used, different locales use different symbols for separators and with different formats. For the same number, 1,234.56 is used in the United States and
Table 5 Guidelines for Use of Language
Supporting Research/
Category Guidelines Best Practice
Language Account for physical language variations such as directionality, hyphenation, stressing, fonts, sizes, orientation, layouts, spaces, wrapping, and justification.
Rau et al., 2010 Language Consider differences in sort order between different languages. Rau et al., 2010 Language Allow adequate screen space for possible text expansion due to translation.
Languages such as German can be especially problematic.
Rau et al., 2010 Language Avoid composite messages such as warnings with a word or words dynamically
determined.
Rau et al., 2010 Language Avoid using words from one language as language-selecting options, for
example, the word ‘‘China’’ to select Chinese language option.
Rau et al., 2010 Language Consider, for example, the directionality of languages in the design of text boxes. Rau et al., 2010 Language Carefully consider the need for alternative text input methods for languages such
as Chinese.
Rau et al., 2010
Table 6 Guidelines for Format Conventions and Measurements
Supporting Research/
Category Guidelines Best Practice
Format conventions and measurement systems
Be aware of the way in which numbers are represented in the target culture, including treatment of separators and glyph shapes.
Rau et al., 2010 Format conventions and
measurement systems
Consider the format conventions for currency, calendars, date and time, addresses, and telephone numbers for the target culture.
Rau et al., 2010 Format conventions and
measurement systems
Consider the measurement conventions of the target culture, e.g.
weights, dimensions, temperatures, and paper sizes.
Rau et al., 2010
Canada; “1.234,56 is used in Germany, Holland, and Italy; and 1.234,56 is used in France and Sweden. The numeric glyph shapes can be different as well.
Other format conventions that need to be taken into account include currency, calendars, date and time formats, names and addresses, and telephone numbers.
Research and practice in cross-cultural user interface design have focused on the internationalization and localization of display codes, such features as formats, colors, icons, and graphics (Liang and Plocher, 2003).
Designers need to be aware of different measure- ment conventions for different locales, for example, dimensions, weights, temperatures, and paper sizes.
Adequate accommodations need to be provided so that the product uses the appropriate measurements for the target locales. Table 6 provides guidelines for format conventions and measurements.
3.1.4 Color Coding and Affect
Color Associations and Common Safety Words There is no difference between cultures in the actual perception of colors since there are common physiolog- ical bases for color vision (Bonds, 1986). Further, in technical applications, there is considerable agreement between Asians and Americans about how common safety words are associated with colors (Courtney, 1986;
Luximon et al., 1998; Liang et al., 2000; Kaiser, 2002):
• Danger—red
• Go—green
• Hot—red
• Stop—red
• Safe—green
• Caution—yellow
However, a few color associations have been found to be similar in some cultures (Courtney, 1986; Liang et al. 2000; Kaiser, 2002) but different in others:
• Cold—white: Chinese, Japanese;
blue: American
• On—green: Chinese; red: American
The point is that most colors have at least some ambiguity associated with their meaning and should
be avoided for signaling important or safety-critical concepts unless they are combined with other coding such as text or icon or both. Designers should be aware of ambiguities and use color coding with caution.
Color and Affect What people feel about colors is more subject to cultural variation. Colors and the combination of colors have different meanings in different cultures. Many researchers have conducted studies on color and its impact for different product design.
Early studies on color codability demonstrated that people in different societies did not have the same array of colors to partition the color spectrum (Whorf, 1956).
Berlin and Kay (1969) argued that, if the mechanism underlying color perception is universal, there should be agreement on colors among those who speak different languages from different cultural environments in spite of variations in color vocabulary. They studied 20 languages and discovered meaningful regularities in the use of basic color terms which are names of color categories consisting of only one morpheme. They also noted an evolutionary progression in color terms in the sense that culturally simpler societies tended to have fewer basic color terms than culturally complex societies, for example, large-scale, industrial countries.
MacLaury’s (1991) work also demonstrates the effect of cultural factors on color coding. A comprehensive study of color naming has been presented by Russell et al.
(1997). Davies and Corbett (1997) studied speakers of English, Russian, and Setswana languages, which differ in their number of basic color terms and in how the blue-green region is categorized.
Prabhu and Harel (1999) studied users’ needs and preferences for digital imaging products in Japan and China. They found that Japanese men preferred single color fonts and simple fonts without emphasis on all the three lines of help, whereas Japanese women and Chinese men and women preferred multiple colors and highlighted or emphasized fonts. Also, Japanese preferred pastel colors for both the welcome screens and the interaction screens. Though Chinese men preferred Chinese colors, preference for women was mixed between Chinese and Japanese pastel colors.
Minocha et al. (2002) conducted informal observa- tions and analysis for the choice of colors on some e-finance sites in India and Taiwan. Three e-finance