PCA - Parallel Analysis
Hi team,
In JASP, to perform Principle Component Analysis, there are three options: Parallel analysis, Eigenvalue, or manual.
For the Parallel analysis option: it can be either "based on principle components or based on factors with seed 1234". what are these tests?
Is Parallel analysis based on Factors still PCA test? or is it a factor analysis?
Comments
this is just to choose the number of components/factors, not how the actual analysis is conducted. So while I'm not an expert in this I suppose if you're doing PCA it makes more sense to choose the number of components with the parallel analysis based on PCA (using all the item variances) rather than FA (using shared variance only).
But this is just to choose the number of components: whatever you choose, the final analysis will be PCA (using all item variances), whether the number of components was chosen based on all or shared variance
Hi patc3,
Thanks for the explanation, appreciate it. I am doing my analysis with selecting Parallel analysis "based on FA" when performing Exploratory Factor Analysis. sounds reasonable since it is just to choose the number of factors.
Thanks!