How simply we can ensure the validity and reliability of certain scale in different language?
What I mostly have done is first to check the translation validity through translation- backward translation and again translation. This also could go to an expert panel to give their score as to the validity of the tool that could be measured statistically too. The usual way after this is to design a study to implement the tool on your population and then using some statistical analysis such as factor analysis or Rasch analysis, depending on the structure of the tool, could help you to measure validity and reliability. What I mean buy structure of the tool is whether it gives total score, does it measure one construct or more, it is uni dimension or multi dimension etc. Depending on the purpose of using the tool, you can decide to go for a smaller group and measure the validity and reliability in a specific population for example with a sample size of 100 or go for a bigger sample size that included diverse group of people for instance with or without diagnosis. It does all depend on the purpose of the tool and your purpose of using that. If you are planning to use it in clinic only and not for research, sometimes a cognitive analysis would be a good method to study the relevance and applicability of the tool in a clinical setting. I would say then it all depends on several factors.
What I mostly have done is first to check the translation validity through translation- backward translation and again translation. This also could go to an expert panel to give their score as to the validity of the tool that could be measured statistically too. The usual way after this is to design a study to implement the tool on your population and then using some statistical analysis such as factor analysis or Rasch analysis, depending on the structure of the tool, could help you to measure validity and reliability. What I mean buy structure of the tool is whether it gives total score, does it measure one construct or more, it is uni dimension or multi dimension etc. Depending on the purpose of using the tool, you can decide to go for a smaller group and measure the validity and reliability in a specific population for example with a sample size of 100 or go for a bigger sample size that included diverse group of people for instance with or without diagnosis. It does all depend on the purpose of the tool and your purpose of using that. If you are planning to use it in clinic only and not for research, sometimes a cognitive analysis would be a good method to study the relevance and applicability of the tool in a clinical setting. I would say then it all depends on several factors.
I believe the scenario sketched by Farzaneh is rigorous and comprehensive!
Dual panel is another way to check translation.
http://www.med.lu.se/bagadilico/news_archive/100118_translations
- Badges
- Science topic
- Similar topics
- Linguistics
- Philology
- Language