Our team was conducting assessment of functional independence. We came across a scale in English language which has been developed and validated. Our study population speaks Hindi, the major language in India. We had followed the steps of translation and cross-cultural validation mentioned by the original copyright holders of the scale. Different teams conducted forward and back-translation in a blinded manner. We compared the back-translated scale with original scale and made appropriate changes with mutual discussion. Then we used the translated tool among a scientifically selected sample. Now during publication, we are being told that we also needed to collect test-retest data on a subsample of participants as well as another subsample of inter-rater reliability data. It has been more than a year since our data collection finished. Do we need to enroll new participants to collect test-retest and inter-rater reliability data?
I think the reviewer is making unreasonably specific requests. Instead, I think it would be fair to ask that you give evidence for the overall reliability and validity of your translated instrument. The most common standard for assessing reliability is coefficient alpha, so I would calculate and report that. As for validity,
I would recommend a "construct validity framework" where you assess convergent, divergent, and predictive validity.
In your letter of reply, you can thank the reviewer for pointing out the need to report issues of reliability and validity for your translated scale, without getting into the specifics of the original request.
Must you will measure test-retest reliabilities.
Also you must calculate other psychometric properties.
@David L Morgan and Noor muhammad, thank you for your response.
Some related requests in this regard:
1. I have heard the term pschometric properties, but know in patches only. Whats are these properties?
2. I don't know much about construct validity framework. I am new to scale validation. How do we calculate these in Epi Info software?
3. Can you share a resource to help in step by step of design, conduct and analysis of scale validation?
Thanks again
Psychometrics is a broad field, and you should concentrate on what your reviewer requested: evidence of reliability and validity.
Construct validity is a general approach to scale validation. A simple web search should find you all the resources you require.
Cross-cultural adaptation of a research scale entails guaranteeing consistency in the content and construct validity between the original and target versions of the scale. Therefore, it is strongly recommended that a researcher ascertain their scale version mirrors similar measurement properties of the original scale. To that end, and after translating the source scale, they should evaluate the psychometric properties of their adapted scale by testing it on the level of items (i.e., item-to-total correlation and internal consistency) and the level of scale (i.e., reliability, construct validity, and responsiveness). As regards your query on the test-retest reliability, it should be incorporated and reported. If not possible, its absence should be indicated in the limitations of the study. You might find the guidelines by Beaton et al. (2000), referenced below, relevantly informative.
Beaton, D. E., Bombardier, C., Guillemin, F., & Ferraz, M. B. (2000). Guidelines for the process of cross-cultural adaptation of self-report measures: Spine, 25(24), 3186–3191. https://doi.org/10.1097/00007632-200012150-00014
Good luck,
- Badges
- Science topic
- Similar topics
- Education