Factors (Unobserved variables) measuring with summated items can be factorized like items in EFA or Not.
Please justify the above questions.
Morning Sivapalan!
Yes andefinitely now. If you work with SPSS than you can save super-variable as a regression made of items, construing it. But it is not about technology, it is about theory. After you checked that your items work with EFA (build a factor as predicted) than you should perform the CFA in AMOS. You can even just start with the CFA. After you check that your measuring model work good (model fit, convergent and discriminant validity of the proposed items, reliability etc) than you can simple aggregate your items to a super-variable.
The same applies for the second level constructs. You have to show in, e.g. AMOS the same statistics as for first order constructs. Here is how to do it: https://youtu.be/HBQPqj63Y7s
Sincerely,
Eugene
Hello,
yes indeed, you can fit second-order CFA if it is in accord with a theory.
See the picture below,
M.
Yes Factors (Unobserved variables) measuring with summated items can be factorized. these factors can be treated as manifest variables of latent construct in SEM through AMOS software provide you have saved the factor score (option is available in SPSS). Above all everything should be grounded in sound theory.
- Badges
- Science topic
- Similar topics
- Psychometrics
- Measurement