Translation

The purpose of translating the questionnaire is to produce instruments that maintain an equivalence measurement across languages and/or dialects, and to achieve a functional and culturally appropriate version of the original instrument. Only then is it possible to achieve results that are comparable cross-nationally.

An accurate translation that conveys the context or conceptual meaning − rather than a literal translation − of the WG-SS is crucial.  Literal translation – even with back translation as a validation mechanism – is insufficient, since it can leave many inconsistencies or misinterpretations in place.  

For example, in one disability survey undertaken in Mumbai, India, a question put forward for consideration was “Does your child have difficulty walking around the house?” In native-English-speaking countries, “around the house” generally means within the house, for example walking from room to room. However, when this is interpreted into Marathi (an Indian language spoken predominantly by the Marathi people of Maharashtra), the phrase actually means “walking around the circumference of the house,” and would thus yield a high degree of confusion and prompt respondents to ask why their child would circle the outside of their house. 

Evidently, back translation failed to catch this error, since translating back from Marathi to English yielded the exact same English words but failed to catch the fact that in the two languages they were two different questions. After testing the translation, it was determined that the words for “across the room” in Marathi captured the intended concept without causing confusion among respondents.

Consequently, the WG applies an alternative protocol for the translation of its tools, which recommends that two individuals or two teams first review the WG-SS concepts before working separately to translate the questions and response options into the target language. The teams, and at least one reviewer, subsequently meet to review the translations and, comment on issues found or, by consensus, make recommend changes. An adjudicator (who may or may not be at the review meeting) ultimately decides whether to adopt the changes or recommendations, or to make alternative changes based on the findings. The reviewed translated document is then cognitively pretested. 

Therefore, translators should have (or be):

• Excellent knowledge of the source and target languages;
• Familiarity with the concepts upon which the questions are based;
• Awareness of when literal translation is insufficient;
• Bicultural rather than just bilingual.

Furthermore, it is equally important that the translation procedure is applied in both the questions and the response categories. Throughout the translation process, decisions made at every step are documented to inform designers and analysts on how the final translation was reached.

Since this process is vital, questions and response options should be of the highest quality. Furthermore, a great deal of effort and expense goes into administering a survey, the data from which can be used for many years by diverse stakeholders to help design and evaluate public policies. 

The translation of the disability domains differed among the countries in the Arab region implementing the WG-SS (See Table 1), with some terms also translated differently from the exact English terms.  For example:

• “Walking” − motion, mobility and movement
• “Seeing” − vision, looking and eyesight
• “Memory” − remembering, understanding and perception
• “Communicating” − understanding and pronunciation

Similarly, translation of the levels of difficulty showed a variety in the use of Arabic terms in some countries. Table 2 provides examples of some national translations of the four levels of difficulty.