If we have a quantitative research confirming effect of one independent variable on one dependent variable in the presence of a moderating and a mediating variable and the construct of the scale/instrument/questionaire is correlates in such a way that all items are highly corelated to measure the same regressional relationship can we analyse cronbach alpha of all items together or do we have to do it seperately...? What makes more sense, suggestions are welcomed.
Regards
Sidra
Cronbach's alpha is a measure of reliability for a unidimensional scale, specifically, a set of items that measure a single latent variable ("factor") with equal factor loadings. It does not make sense (and is not interpretable) to compute alpha for items that are multidimensional (measure multiple latent variables/factors), even if the different factors are strongly correlated.
You should calculate one Cronbach's alpha for each of your separate scales.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS