I am combining different scales in my questionnaire. These scales have been tested for their validation and their reliability. 1st. Do I still need to examine discriminant validity? 2nd. The questionnaire is long and I need to eliminate items, is it possible to do that based on my experience and the review of other experts in the field? Thanks
Deleting items from a scale can affect its reliability and validity, so you would need to reassess those standards on the revised scale, once you collect it. Of course, that can be risky.
As for discriminant validity, if you are worried that some of your scales essentially measure the same thing, you might be better off shortening your questionnaires by dropping one of those overlapping measures.
Dear Carine
You will remove the items only when you administer the scale on a representative sample through discrimination power and difficulty value. After eliminating items you will administer the scale again then go for validity and reliability again.
I totally agree with Prof. Morgan. Don't simply delete the items. Run the EFA (reliability & validity test) first then delete the items with low reliability and factor loading.
- Badges
- Science method
- Similar topics
- Mathematical Sciences
- Statistics
- Applied Statistics
- Survey Design