I am interested to translate a universally accepted questionnaire into local language. What are the main parameters that i have to tested for validation and standardization?
The validation portion can be pretty tricky, considering there may be cultural effects. Comparing factor analysis on the local sample versus the reported factor analysis from the original is one step. If the questionnaire is older, the factor structure may have drifted, so finding a similar sample from a colleague who can administer the questionnaire in the original language is also helpful. As for translation, since the nuances of language can adversely impact interpretation, it is good to have someone translate into the language (say English --> Spanish) and then a separate, naive translator back translate it (Spanish --> English) and investigate how far the translation has drifted. Getting the two translators to then discuss differences and make edits where necessary will help create a more accurate representation of the original.
This answer might not address exactly what you are asking, but there is a lot of careful work beyond simple parameter testing to translate a questionnaire/scale/measurement instrument that must be taken into account.
Thank you... I have another doubt... Suppose we measuring the stress score... Using the questionnaire... How can i spit the score into 4 or 5 catagories and say mild, moderate, severe.. Like wise... Is it statistical technique?....
This is a question where the answer depends on the distribution of scores across the population. Once you translate your measurement device you need to normalize it to the population. In other words, you need to go out into the society and give it to a representative sample of the population. This will give you a baseline "stress score" as well as the expected distribution. You can then use that distribution to set what "high" versus "medium" versus "very low" actually means.
Note that my answer here is very brief and general. Scale creation, validation, norming, translation etc. is very easy to do incorrectly, thereby adding unknown biases to your outcomes and interpretation/scoring/etc. I would encourage you to partner with a graduate university to be sure the appropriate methods are followed. One option is to contact the original scale authors to see if they are interested or have partner universities that would be interested in helping to expand the languages their scale is used in. Cross-cultural validation is of interest to a lot of psychometricians and can result in publications and more widespread adoption of an instrument, making it worth their while to help you. Good luck.
Suppose i have normalise the data... but can we catagorize the score into high, low medium..?... using quartiles ?..
the first step, is translate. (i agree with Ivan en la complejidad del proceso)
1. translate the questionnaire of English to bengali (for example) for a person that lenguaje native is Bengali, but is expert in English
2. translate the questionnaire of English to bengali (for example) for a person that lenguaje native is English, but is expert in Bengali
3. To analize diferences
after the proposal regarding translations, perform an exploratory factor analysis
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics
- Survey Design
- Questionnaire