Ask Eric: Questions & Answers

August 16, 2007 – submitted by Joann Schissel of West Des Moines, IA

Question: I think I understand the concept and differences of globalization, localization, and internationalization. Can you please explain this formula? G11N=I18N+L10N. Also, do you have any recent statistics / research about English / non-English uses among internet visitors?

Eric's response: Here is an answer from Apala Lahiri Chavan. She is head of Asian operations at HFI and a leading light in cross cultural issues.

* * * * * * * *

I18N Internationalization: A development process of creating a base design to support various human languages and to behave in a culturally-sensitive manner.

L10N Localization: The process of making changes to a internationalized product to make it usable and viable in a particular market.

Technical L10N: Technical aspects of adapting a product to a local market, e.g. double-byte conversion, OS support, etc.

National L10N: Making the product behaviors / designs appropriate for a national setting, e.g. written language, punctuation, formats, etc.

Cultural L10N: Beyond functionality and correct language – cultural specific values, tastes, and conventions need to be addressed.

G11N Globalization: The design, development and creation of fully localized products to meet the cultural and linguistic requirements of a particular market.

G11N =I18N +L10N
What this means is that a truly "globalized" offering must at one level have a "base" interface that is internationalized (meets the I18N requirements) and at another level has localized offerings for the countries where a completely local offering is needed (meets the L10N requirements). So, for those countries where there is no localized offering, the "base" interface being internationalized makes it possible for users to interact and use the interface without being offended (since the interface would be a "neutral" interface from a cultural perspective).

July 23, 2007 – submitted by Julius Oyeleke of Windhoek, Namibia

Question: Hi Eric. Please I would like to ask you about ATM usability.

1) What are the new enhancement that need to be incorporated into the ATM system (Automatic Teller Machine).
2) What is probable outcome of using cognitive work through in ATM research usability.
3) sample of what academic research on ATM should look like.

Thank you for the the assistance. Julius

Eric's response: The ATM machines have been having more and more functionality added to them. The challenge is that with increased functionality comes increased complexity. So the usability work gets much tougher. At the same time ATMs are progressively being moved into more and more emerging markets and pushed to the "bottom of the pyramid" (poor users). We have often found that poor users LOVE the ATMs because they can avoid the denigrating attitude of bank employees. So that is a plus. But then you have to design for the more limited skills of this population.

A cognitive walkthrough of ATM operation is useful. Though I would be certain to also do usability testing. Just in case YOUR cognitive ideas don't match the actual users.

Research on ATMs should follow the guidelines of the publication they are planned for. For example you might see the American Psychological Association's style manual. Without other input, they probably have the best structure for academic papers.

May 30, 2007 – submitted by Maja Mkinga of Dar Es Salaam, Tanzania

Question: I want to design a try out model of a user interface for blind people. Could you please tell me the important principles to consider when designing one?

Eric's response: If you are designing a Web interface for use with a screen reader, then first look at the W3C accessibility guidelines. These are basic design and programming conventions that make it possible for a screen reader to access the interface. However, designing for a screen reader by no means ensures a usable interface. This takes real attention to the design. You can start by being sure to test the interface with a screen reader. Even just using one yourself will help.

If you are designing in a different environment you will be designing a non-visual interface. This probably does NOT mean a Braille interface (actually only a small fraction of blind users can read Braille). Instead, you will probably have a voice interface and apply the extensive literature on voice applications.

March 12, 2007 – submitted by Harikrishna VP of Kerala, India

Question: I have a doubt regarding, what date format should be followed for a international software product. Is it YYYY-MM-DD or DD-MM-YYYY. Can you please help me out. Thanks in advance.

Eric's response: Sure, I'll help.

When you DISPLAY a date (in display only situations) then you can use 12 SEP 2008 and feel very safe.

If the data is to be entered, or entered and displayed, it is harder. One very good method is to provide drop down fields in the '12 SEP 2006' format. This is good for a mouse-primary interface, but can be too slow for heavy data entry.

If dropdowns are not advisable then it is best to adapt to the local convention of the user. Microsoft does a fine job of doing this adaptation. So the USA uses MMDDYYYY and Europe DDMMYYYY (etc).

The YYYY-MM-DD is easy to sort inside the computer, but the only real value for users is that it tends to equally confuse everyone. I would avoid this "international" format as, after all, we really don't have any international users. Everyone operates from their own cultural context.

January 30, 2007 – submitted by Carol Hall of Connecticut, USA

Question: Hi, I'm an instructional designer interested in how companies are using MS Word readability/grade level scoring. I realize these are not a guarantee of understandability, however, it's quick and easy to do. Is it valuable to do? What grade level works in an insurance company for example? Thanks.

Eric's response: Calculations of reading grade level (RGL) are VERY valuable when you need to point out to someone that their writing is awful. I remember analyzing some legal documents and finding that you needed four doctorates to read them comfortably. It made an impression all the way to the executive level.

The RGL calculations are not precise. To REALLY know how difficult the text is to read you need to consider how common a word is. The RGL calculations assume a short word is easier, but if the short word is uncommon it might be much harder then a long common word.

In terms of target RGL, it is a question of your user population. A call-center population may need a lower RGL target than the lawyers in the legal department, or the underwriters.

(Sorry I have to do math...)
To set the right target find the distribution of RGL for the target users. Then find the RGL that can be read comfortably by the 5th-percentile reader (reads less well then 90% of the other users). That is generally a good target. The exact target can then be adjusted depending on context. If proper communication is essential, and communication failure could be life threatening, you might go for a much lower RGL (e.g., for prescription instructions). If you WANT to confuse people you can go much higher (e.g., in writing service agreements).