CASE STUDY: Translation Is Not Localization

Posted on

UX Research On The Albanian UI Of Windows 7, 8.1

The Education Ministry of Albania commissioned in 2010 a Usability study on the Albanian localization of Windows 7 and 8.1 in 2014. I had the pleasure of leading a team of 5, compiled of linguists and IT professionals. We organized the project with a balance on qualitative and quantitative user testing and input. Because the localization of the Microsoft flagship OS was new in the Albanian speaking regions, the qualitative data was very useful in better understanding the user’s mood. That said, I tried to keep the analyses as cold as possible, removing any national subjectivity from the process.   

Data Gathering

  • Interviews of government officials and private organizations’ employees. This was the group that should have had already the localized LIP in their work computers. It came out that most of them were still using the US version of the OS, with some refusing altogether to switch to Albanian.
  • Questionnaires of the above group and students of Tirana’s University. We compiled the questions with only binary answers, for better clarity and data classification.
  • Direct questioning of random people one Sunday in the Tirana’s main park. This was the only way to get any solid information about the usability of the new localization.

Sadly, a huge majority of the feedback was negative; very few used Windows in Albanian. This brought the question: Why?

The answer this the second part of the research was focused in:

Observation And Card Sorting

Testing and observation sessions were set up to see how intuitive the Albanian localization of Windows was, and to better understand why the users in general were reluctant the use the Albanian UI.

  • The users were observed in using Windows in their preferred language and in Albanian, and then asked:
  • with card sorting, to pair up the original English (or other) term with the localized one.
  • to identify the function of a localized term.
  • to identify the user journey only with the Albanian localization.

Not going to deep on the conclusions of the project, the Windows 7 localization was very poor, and often unusable by the majority of the subjects, and then, latter, Windows 8.1 was over-localized to the point that terminology was becoming ridiculous. The language used on 7 was based mainly in a northern Albanian dialect, and on the 8.1 that situation was corrected, but the process had brought other major cultural and linguistic problems.

Even this study was not commissioned by Microsoft; I think that our research had its impact, as the Albanian government used it to push Microsoft to do a better job, and the latter localizations of Windows show progress.