I am studying the influence of 7 variables on a DV. 7 questionnaires has been adapted and one for the DV has to be established. So is it okay for me to run a CFA on all the questionnaires? or only EFA on the newly established scale and CFA for the rest? Ive read that EFA is only for newly established scale or do i proceed with actual data collection and then run CFA for the entire scales?
If you already have hypotheses about the number of factors and the loading pattern (which variables measure which factors), EFA may not be needed and you can proceed straight to CFA.
Christian Geiser, Hi thank you for replying! yes i already have hypothesis that all the factors influence the single DV, I understand if they already have a model or theory supporting the hypothesis then i can bring this directly to CFA. Btw, the scales did go through translation as well.
1. How about if have one factor that required me to establish a new scale, can i run Efa only for this scale or do i also proceed this to CFA?
2.So does this mean that i would only know the validity of my scale when i run SEM to study the predictive factors?