It is very important. In our field (management) we often do a lot of studies in the Asian context; where we borrow/develop scale from the United States. The common method is to use the Language A scale (English) to be translated into the native language (say Korean) for a pilot; and have an expert on both Korean-English back translate each item to the original English, and have an additional native English speaker/reader/writer judge the similarity between what was asked.

Or at least follow some variant of such a process. Be sure to discuss this as generally peer-reviewed articles can't just say "subjects can easily read these scales". What is "easy"? Easy according to whom? etc. As you can see, this is very problematic.

For an example, see:

Anderson, B. S., & Eshima, Y. (2013). The influence of firm age and intangible resources on the relationship between entrepreneurial orientation and firm growth among Japanese SMEs. Journal of Business Venturing, 28(3), 413-429. URL: http://www.sciencedirect.com/science/article/pii/S0883902611000863

Thank you ever so much for your reply, Dr. Sunny. I belong to Pakistan that has a unique socio-cultural setting. That was ruled as a colony for the British for about a century, as part of the Indian subcontinent. The medical students there, are enrolled in English medium schools since start, which is also the medium of instruction in medical schools. Therefore, most of the researchers use English versions of most of the scales. However, this might differ in the less educated strata of the society where we employ scales in native language. 

Answering your query, the readability here is measured by Flesch readability score and Flesch-Kincaid grade level. I usually do not translate those scales that have a good face validity, with a reading ease, suitable for 6-8 grade children in the states.

When we use English versions of the scales, some reviewers object to it, for not being cross culturally validated. So, I perform factor analysis for it. However, on the hand, sometimes the reviewers object to the validation analyses, since it has not been translated, leading to outright rejection of the manuscript.

Yes, although I share with your frustration on when reviewers reject manuscripts for these reasons, I also think factor analyses are conducted independent of cross-cultural linguistic issues with validity. I think it is just one thing to do to make it further in the review process.

And yes, I am quite familiar with the schooling/history - I am Bangladeshi, and hence formerly East Pakistani - ruled by the British as well, perhaps even longer :) In fact, I attended English medium schools in Bangladesh under the GCSE curriculum. Good luck!

Dear Ahmed Waqas,

Your interesting question addresses one of the key issues in doing research; namely, validity and reliability related to the interpretation of data from psychometric scales. Indubitably, translation of metrics into L1 is one of the ways of producing validity-related evidence that goes beyond the researcher's personal opinions and judgments and convinces the participants that the targeted scale is sufficiently meaningful.The main reason is that a scale's psychometric properties might be significantly different across cultural groups. Therefore, the items on the scale should be carefully examined and potentially rectified since they may not similarly represent the latent construct in participants' L1 culture. As such, translation, as you have rightly observed, can prevent misinterpretation of the construct effectively. Although translation alone does not guarantee that the scale's psychometric properties are comparable (see Tucker, et al., 2006), it is still a crucial preliminary step in a cross-cultural work.

Best regards,

R. Biria