I am struggling to find references for testing structural validity of a scale on non-clinical samples first, before running an EFA and CFA on people with personality disorders. I was told this is standard practice when developing a scale but I am just not sure about the research procedure. Any advice would be great.
Thanks.
Developing new scales undergo several process. Let say that the items are build through adequate process i.e. theoretical background and content validity is considered. Now the recommend approach is to use EFA on pilot data i.e. 30-50 responses (depends). Most researchers recommend that the first test on newly established items will be performed on separate sample to validate the items, then the validated items/new questionnaire will be used for proper data collection. Hope the comment will help. Let me know about further question
Regards
Hi. I believe you would find the discussion in this article useful. You don't have to test the structural validity of a scale on non-clinical samples first, in my opinion. Many scales has been developed in clinical samples and later it has been examined whether they are generalizable to general population samples, useful in epidemiological Research (e.g. to find normative scores and so on....).
Best of Luck.
jkv
http://onlinelibrary.wiley.com/doi/10.1348/0144665031752934/full
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running