CIS 732 Final report User Interface InternationalizationJia Shen

--- CIS732 Final Project ---

User Interface Internationalization

Submitted to: Professor Murray Turoff

Submitted by: Jia Shen

Date: 12/17/2000

Table of Content

1.Why is “User Interface Internationalization"?

2.Interface globalization Issues and Levels

2.1 Language

3.Culture and design

3.1 What is Culture

3.2 Culture Meta models

3.3 Some known and tested culture models

3.4 Culture and metaphor

3.5 Culture and icons and symbols

4.Software Engineering for International Computer Product

4.1 Initiative

4.2 Requirement

4.3 Design

4.4 International Usability Evaluation

5.Global Web site

6.Future research topics

7.Conclusion

8.Reference

1.Why is “User Interface Internationalization"?

The answer is simple: Business Decision.

Ten years ago, user interface internationalization is not quite an issue as it is today. Computer products made from America were sold all over the world without interface internationalization questions being asked because there were not many local competitors out there anyway. Or the sales overseas was not significant at all. Things have changed dramatically in recent years. The graphs below are survey results of percentage of Internet users by geography:

Internet Users by Geographic Locations (1996 - 1998)

(Source:

1996
/ 1997
/ 1998

The yellow areas represent Internet users in North America in all the charts, and it is obvious even by glance that its percentage is shrinking each year. Internet users from outside North America are increasing significant as well as rapidly during recent years. The implication is that the demands for computer hardware and software and other related appliances are also increasing, exponentially. Vendors now understand entering a new market takes more than supplying a product that is just technically superior, aesthetically pleasing, or relatively inexpensive. ([Elisa M. del Galdo, 1996]) Each culture has its own needs and desires when it comes to products. Only those who address these needs can compete with local or national vendors and win the market. Computer product user interface internationalization is not a moral decision, but a business decision.

2.Interface globalization Issues and Levels

The first step in addressing internationalization issues is to make clear what are the issues that are being involved. Shneiderman [Shneiderman, 1998] says the issues include:

  • Characters, numerals, special characters and diacritical
  • Left-to-right versus right-to-left versus vertical input and reading
  • Date and time formats
  • Numeric and currency formats
  • Weights and measures
  • Telephone numbers and addresses
  • Names and titles
  • Social-security, national identification, and passport numbers
  • Capitalization and punctuation

Tony Fernandes [Fernandes, 1995] lists the following issues as a few:

  • Nationalism: What is considered an inherent part of a nation or culture and what is consider a threat to it.
  • Language: a language and its various dialect
  • Social context: in many languages of the world, who made a statement has a much bearing as what the statement was.
  • Time: Date and time format vary from country to country as well as calendars.
  • Currency: symbols used to denote money vary from locale to locale
  • Units of measure: both the metric system and English units are used in the world
  • Cultural values: notion of quality, normality, cleanliness and property.
  • Symbols: food, animals and everyday objects can has symbolic meanings that may convey unintended messages.
  • Esthetics: use of color, patterns, shapes and textures.

Below I am categorizing these issues into four categories with three levels of objectives. The following table lists them from surface to bottom level.

Internationalization Issues and levels

Objectivity Levels / Internationalization
Issues / Example / Current Research Examples

Comprehensibility
Usability
Desirability / Language / Product language localization / Unicode;
Machine Translation;
Microsoft knowledge base for common computer word translation
Institutional matters / Time zone, date format, currency, measurement
Environmental factors / Esthetics, Icons and symbols / ISO symbols for interface;
Microsoft knowledge base for international color use
Social conventions / Forms and values. / Culture model

As Del Galdo and Nielsen [Elisa Del Galdo and Jakob Nielsen, 1996] described, the three levels are:

  1. Comprehensibility: A computer interface that is capable of displaying the user's native language, character set, and notations, such as currency symbols.
  2. Usability: A computer interface that is understandable and usable in the user's native language.
  3. Desirability: A system that is able to produce systems that accommodate users' cultural characteristics.

Companies have started to recognize the need to address issues on the first two levels. Issues such as language translation or date and currency format have been considered in some products. Current research work is concentrating on automatic machine translation, using Unicode for programming [David A. Schmitt, 2000], or establishing knowledge base for product internationalization, such as Microsoft Developer's network which includes a database for common computer word translation. Further research is needed in understanding the deeper level of culture issues and their influence on computer product desirability. Let's first briefly discuss the issue of language, then the next part is devoted to discussion on culture and design.

2.1 Language

In the past, people dealt with computer were assumed to know English. Products rarely needed to be translated. Nowadays, with the proliferation of computer all over the world, people with all ages and culture background started to use computer. The assumption is no more valid.

Language is usually the first issue we encounter in product internationalization. Programming languages are designed that better support making international software easily. Languages such as Java, Visual C++, Visual Basic all have supporting features, such as character coding, and operating systems such as Windows are adding functions to better support internationalization.

Besides translation, there are some other issues in this area such as language ambiguity. Trash can in the United States is called waste basket in Great Britain and rubbish in New Zealand.

A billion means a thousand million in the US whereas in Great Britain it means a million. So language internationalization should also deals with these issues.

3.Culture and design

There is no denying that culture influences human-computer interaction. The task is how to study culture effectively to address the needs of users worldwide in computer product design and implementation. To do this, let us first look at what is culture, its layers of issues that could possibly influence product design, and then some well known culture models and their impact on computer product.

3.1 What is Culture

Our working definition for this article is: culture is learned behavior consisting of thoughts, feelings, and actions. [Hoft, 1996] There is no agreement to a specific definition of culture. In 1954, for example, A. L. Kroeber and C. Kluckhohn reported over 300 definition of culture (Culture: A Critical Review of Concepts and Definitions, New York: Random Hourse). The definition used here is simple and seems to be in agreement with much research on culture.

3.2 Culture Meta models

A culture model helps to identify levels of issues being involved in this complex problem by using international variables, or dimensions of culture. International variables are categories that organize cultural data. [Hoft, 1996]

Before we study any specific cultural models, let's look at cultural meta models that help us to understand how and where culture comes to influence our lives in a profound way. There are four meta modals of culture that were reviewed in [Hoft, 1996], which are

  • Objective and Subjective culture
  • The Iceberg Model
  • The Pyramid Model
  • The Onion Model

Overall these models are all telling us one thing: culture has multiple layers and what we can normally observe is only the most external layer which counts for about 10% of its total influence. There is much more profound influence of which we may not even be aware. The Iceberg Model and the pyramid model are most helpful to illustrate this point.

  1. The Iceberg Model:

The Iceberg model is a popular metamodel that is often used in cross-cultural research. Below is the graph that shows the models.

The Iceberg model

The analogy drawn in the Iceberg model is that just as 10 percent of an iceberg is visible above the surface of the water, only 10 percent of the cultural characteristics of a target audience is easily visible to an observer (us). It follows that just as the remaining 90 percent of our cultural characteristics are hidden from view and are therefore easier to ignore and more difficult to identify and study. The model identifies three metaphorical layers of culture:

  • Surface: visible, obvious rules such as number, currency, time and date formats.
  • Unspoken rules: obscured, need context of situation to understand the rules
  • Unconscious rules: rules out of conscious awareness and difficult to study.
  1. The Pyramid Model

Another model that is well known was developed by Geert Hofstede was called the pyramid model.

The Pyramid Model

Geert Hofstede introduces three layers of culture in Pyramid model.

  • Personality: specific to a person and is learned and inherited
  • Culture: specific to a group or category of people. It is learned not inherited.
  • Human Nature: common to all human beings. It is universal and is inherited, not learned.

These meta models provide us with a sense of which layer of culture we would like to look at to test international computer products.

3.3 Some known and tested culture models

The purpose of reviewing some known and tested culture models is to provide what has been known in this area. They may serve as starting point to develop your own culture model for your needs. These models are based on questionnaires, surveys, extensive interviews, focus groups, and years of experience and observation.

Four well known models of culture

Author / Focus of Culture / Variables identified
Edward T. Hall / Determining what releases the right response rather than what sends the right message / Speed of messages
High context/ low context
Space
Polychronic/Monochromic Time
Information flow
Action chains
David A. Victor / Determining the aspects of culture most likely to affect communication specifically in a business setting / Language: fluency /accents etc.
Environment and Technology
Social organization
Context
Authority conception
Nonverbal behavior
Temporal conception
Geert Hofstede / Determining the patterns of thinking, feeling, and acting that form a culture's mental programming / Power Distance
Collectivism vs. Individualism
Femininity vs. Masculinity
Uncertainty Avoidance
Long-term vs. Short-term
Fons Trompenaars / Determining the way in which a group of people solves problems / Universalism vs. Particularism
Collectivism vs. Individualism
Specific vs. Diffuse
Achievement vs. Ascription
Attitude to time
Attitude to environment

The international variables being identified in these models tell us how different people from different culture are. Current computer products are not sophisticated to support cultural diversity to this degree, and thus not achieving the "desirability" level in our hierarchy. Some of the variables are directly related to observations done by some study in comparing Japanese and American computer users. Below I am summarizing a few interesting examples given in [Masao Ito and Kumiyo Nakakoji, 1996]

When difficulty is encountered in using a system, whom to blame: users or designers?

  • Japanese tend to conclude it's their fault for not reading manuals carefully before hand.
  • Americans tend to blame the system and its designers for not being considerate.

Learning style:

  • Japanese: read manuals first very carefully before start using a system.
  • Americans: immediately start using a system to see "what will happen"

Sequence of operations:

  • Japanese: first identify the object, then the action on the object, such as (file; delete). This is said to stem from Japanese grammar, which is subject followed by object followed by verb
  • Americans: accustomed to first specify the action, then the object as English grammar is: Subject + Verb + Object

Thus Japanese users find the sequence of a typical GUI operation (e.g. first select an object then specify an action) more natural than American users.

Group collaboration

  • Japanese: parental management style, stress on teamwork, not individual contribution. Harmonious meeting.
  • Americans: Consultative management style, appreciate individualism, encourage different ideas

Reading habit:

  • Japanese: Last sentence is always topic sentence.
  • Westerners: Topic sentence can be found in each paragraph in English.

Another interesting observation is that it has been said that Japanese use face marks (emoticon) in on line communication much more than westerns. The reasons are twofold: Japanese, like Chinese, are used to look at pictograms, and second, Japanese tend to use facial expression and context to judge another person’s meaning. Saying everything up front is considered to be impolite. Using Edward T. Hall's culture model, Japanese is a high context culture where the context of conversation carries as much weight in the communication as the message itself, if not more; whereas countries such as German are low context and thus things are expected to be articulated clearly. But instead of using emoticons, such as , Japanese prefer to invent their own symbols: (^-^) for happiness. They also invented their own such as (^o^;>) for Excuse me. The triangular shape on the right apparently represents a protruding elbow and stems from the fact that an embarrassed or apologetic person will sometimes scratch the back of his or her head.

As we can see, culture model is effective in studying culture in a meaningful way for the purpose of computer product design and implementation. One has to develop his own culture model to accommodate his own needs. To this end, he must determine the layer of culture issue that he wants to study, and develop his model and then collect data to test the model. The goal for any culture model is to understand the difference in culture and thus base the design on the culture.

3.4 Culture and metaphor

Many computer interfaces adopt real world objects to help users understand the task and the system. Unfortunately the world is usually American real world. Though the idea is valid, users' real worlds vary from place to place yet user interfaces don't. [Tony Fernandes, 1995]

An interesting example is given in [Masao Ito and Kumiyo Nakakoji, 1996] where typewriter is being used as metaphor for word processing software in western world. Concepts and notions such as tab-stops and margins have been well known and naturally understood by those who have experience using real typewriter. However, it was not the case in Japan, where Japanese and Chinese character sets are being used and typewriters were rarely encountered in daily life except for professionals and English-major students. In Japan, they typically use rule-lined writing pads with 20*20 grids and one character was written in for each grid from top to bottom and right to left. Given these differences, an appropriate metaphor for Japanese word processing without influcence of American software would be ruled writing-pad. But Japanese still use the western typewriter metaphor because they got accustomed to it and changes to other metaphors may see awkward. So this is a case of "acclimatization" overwrites culture difference.

3.5 Culture and icons and symbols

As the saying goes, "a good picture worth a thousand words." Most of today's software use icons and symbols to communicate with users. Icons are signs that are familiar, concrete representations of objects or people. Symbols are more abstract and require specific instructions to learn. For international computer interface, the design of icons and symbols pose great challenge. Similar to metaphors, the problem here is what is meaningful and natural to one group may be ambiguous, unintelligible, or arbitrary to anther. [Aaron Marcus, 1996] Some well known examples include the design of Windows folder icon, garbage cans, and mailbox. The typical Windows folder icon looks like this:


Though the yellowish color and the tab suggest to western users that apparently it is a file folder that they use everyday, it is rarely seen in some other countries, China for example. The folder used in China is usually a rectangular one made from thick brown paper without tabs. So for Chinese users, they are forced to associate this unfamiliar object to folders and remember this association. It is definitely extra work which should not be the intention of using icons and symbols.

There are other issues that are associated with culture and interface design such as culture and color, culture and computer mediated communication, culture and groupware etc. As a complicated topic, the current study of culture for product design is far from fully-fledged. But there are companies who realized its importance and carried out research in this field. Hewlett-Packard once conducted contextual research for their Inkjet printer to enter international users' families. [Susan Dray and Deborah Mrazek, 1996] Though based on limited time and budget the study was carried out only in four America cities and two European cities with 20 families, the results were extraordinarily valuable for HP to see how their products were actually being used in their families. The data collected played a significant role later in the product design. The idea is a thorough study of people and an understanding of their cultural background is a prerequisite for a successful international product.

4.Software Engineering for International Computer Product

Interface design issue is part of computer product design and thus to do things right, a proper method needs to be proposed to incorporate internalization elements into the processes of software development cycle.