In my study design, I originally set out to ustilise a multi-measure method, utilising two subscales from the MIPS perfectionism questionnaire and 2 subscales from the SMPS questionnaire, both of which contained a subcale for perfectionist concerns and strivings. As well as this, I've taken in 7 coping scores from the ACSI-28. I was planning on combining the perfectionism scales using Z scores to generate total perfectionist strivings and total perfectionist concerns however, didn't realise i would need to conduct a EFA and CFA to before combining the two standardised measures.
Is there any logic in only inputting one questionnaire Into my regression model, i.e if one questionnaire obtained higher cronbach alphas?
Usually researchers use either the EFA or the CFA method, using both doesn't really add that much information. Considering you already seem to be working under the hypothesis that all or most of these variables refer to perfectionism I'd recommend the CFA.
I'd also advise you to just run the factor analysis first before deciding to ignore the other questionnaires. Maybe you'll end up with two good factors that correspond with perfectionist strivings and perfectionist concerns, and you won't need to throw away data. If that's not the case you can still decide that using the original scale from one questionnaire is the best course of action, nothing gained nothing lost.
Hi Sidney,
if you plan to create a composite anyway than I don't think that you need a CFA beforehand, as you move from "measurement of a (or two) latent variables" to a composite/aggregate construct. It may be, however, that future reviewers will see this differently.
An alternative would be to stick with the original factors and hypothesize effects of the original perfectionism facets. This requires you to think about whether the facets have unique effects, or whether they represent an underlying second-order factor (which actually has an effect--and not the facets). An aggregate/composites avoids these questions. You get "an effect" without learning something important of the underlying mechanisms.
Best,
Holger
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology