Translating Surveys Perfectly Will Make Them Perfect -- But It Doesn't Actually Matter!

Translating Surveys Perfectly Will Make Them Perfect -- But It Doesn't Actually Matter!

When translating consumer or business surveys into other languages, there are several approaches you might take. All have varying degrees of cost and complexity. For example, you can have a machine do it (uh–don’t do that) or a person. You can have one person translate or multiple. Two specific approaches recently tested by a team of researchers in Germany (reported in the Journal of Survey Statistics and Methodology) are back translation and team translation.

Back translation involves “translating an already translated document back to the original source language” via two independent translators. One person translates an English survey into Spanish, for example. When done, the second person translates the Spanish version back into English. Differences between the two English versions are then used to make adjustments and finalize the Spanish translation.

Team translation involves translators working together or in parallel on the same translation and then reconciling any differences. For example, two people translate an English survey into Spanish, and then they meet to review and reconcile the differences in how they translated it into Spanish.

Which one is better for surveys? According to the researchers, team translation gives measurably more accurate translations, so that would seem to be the preferred choice. But then they compared the two versions in terms of how they performed in the field, and there were no differences in measurement accuracy. In other words, even though back translation was less perfect than team translation, it worked just as well in achieving what it was supposed to do.

At Versta Research we always use team translation, so we are getting more perfect results in that sense. But this new research effectively points to the resiliency of survey respondents. Even if survey wording isn’t perfect, people who take surveys can generally figure out what we are trying to ask. When you think about it, that’s kind of amazing and encouraging for the outcomes we always hope to achieve.

Joe Hopper, Ph.D.

要查看或添加评论,请登录

Joe Hopper的更多文章

社区洞察