Geography at Multiple Levels
The foundation for a multilingual approach lies in employee data. The starting point is information about the country and specific location. Based on the country, the language in which an employee is invited is determined. In countries where multiple languages are spoken, further refinement can be made based on location. To achieve this tiered application, it is useful for the language determination rules to be not black and white, but rather in a sequential order. This provides the most flexibility and makes it easy to adapt to specific customer situations. An example of this is inviting employees at the headquarters in English while addressing the rest of the employees in the local language.
Tailored to Individual Employees
The multilingual approach can be further refined by recording whether an employee switches to another language during a survey. This "preferred language" is used in future research invitations to address employees more frequently in the correct language.
Translation and Localization
Numerical results pose no problem in a multilingual approach. For qualitative results (textual explanations), it is advisable to translate and store them in a continuous process alongside the original results. In dashboards and analysis tools, you can display the original and translated comments together. This allows for the analysis of all comments in one language at a global level, while the original language comments can be referred to at the local level.
Exotic Languages
When multiple languages are used within an organization, "exotic" languages often come into play. In this context, languages are considered exotic when they have their own character set, such as in Japanese or Chinese. These character sets need to be supported at all (technical) levels. The best way to achieve this is by using Unicode, the international standard for encoding graphical characters and symbols. This way, you can process every language seamlessly. If Unicode support is missing in one of the levels of the ‘feedback stack’ (database, web application, reporting environment), survey questions will not be displayed correctly, and textual feedback will be stored inaccurately.
The UTF-8 encoding we use allows for standardized 'text-to-speech' conversions, enabling texts to be read aloud. This ensures we meet international standards for accessibility and inclusivity.
Right-to-Left Writing Direction
Not all languages are read from left to right. In Hebrew, Arabic, and all derived languages, you read from right to left. A good online research platform ensures not only that texts can be read and aligned from right to left but also that the "blocks" of questions and answer categories are swapped accordingly.
Implementing multilingual research is much more than conducting a survey in multiple languages. A comprehensive approach requires a solid foundation and a systematic approach. It requires more effort to set up initially, but once properly established, you can benefit from all the advantages of communicating in employees' native languages, such as increased engagement, higher response rates, and higher-quality feedback.
Multilingual Approach Checklist:
Flexibly defining language rules
Language switching in surveys
Capturing individuals' preferred language
Supporting right-to-left languages
Supporting exotic languages
Automatically translating provided comments
Making original and translated comments available in reports