Work in the market research space? New findings have revealed that the results of surveys, questionnaires and market research can be skewed and manipulated by the translation of the terms used in questions.
According to a new study that was published in the Journal of Consumer Research, survey results of multilingual research may be biased if the participants do not know the translated terms. This is the outcome of a research conducted by Bert Weijters and Maggie Geuens of the Ghent University and Vlerick Business School and Hans Baumgartner of the Pennsylvania State University.
Weijters, Gueuens and Baumgarter believe that “Consumers are influenced by the specific labels used to mark the endpoints of a survey response scale. This is particularly important in multilingual research. If the response category labels used in different languages are not equivalent, this could bias survey results.” In many surveys, participants have to indicate whether they agree or disagree with a certain statement by using response scales. These scales often have category labels that for example say “strongly (dis)agree” or “completely (dis)agree.”
In the study English and French speaking consumers from the United States, the United Kingdom, Canada and France were used as test subjects. It appeared that the endorsement rates for the endpoints of the abovementioned agreement scales depended on the familiarity of the labels. If the associated labels were often used in everyday language (such as “completely agree” or “tout à fait d’accord”), response categories got more responses then when these scaled has less commonly used lables (“extremely agree” or “extrêmement d’accord”). This was true for both the English and the French language.
Other research has shown that the self-reported awareness of the cholesterol level of different types of food was a lot stronger when the Dutch equivalent of “completely disagree” to “completely agree” was used then when the researchers chose a scale that used “strongly disagree” and “strongly agree.”
These two studies reveal the importance of using the right labels when creating survey lists for research. This is why the authors conclude: “Survey researchers should pay more attention to the labels assigned to response categories on rating scales and make sure that the response category labels used in different languages are equivalent in terms of familiarity. Differences in the category labels used in different languages may lead to differences in responses resulting from poor translation.”
The findings are interesting and not something entirely new to us. Through providing translation services for the market research sector we have been asked by a couple of clients as part of our quality assurance (QA) process to conduct back translations for them. What’s a back translation? Check out the Wikipedia page for a detailed outline but in short, it is a translation of a translated text back into the language of the original text, made without reference to the original text.
Why did they want this as part of the QA? They reasoned it was necessary to carry out back translations to ensure meaning was 100% correct within the questions used when carrying out market research surveys or questionnaires. Through experience they had found that sometimes it was necessary to re-translate or tweak language to ensure the questions were interpreted in the correct way and consequently answered in the correct way.
The fascinating world of language and translation!