I have studied few studies in which this is being done, probably to ensure the validity of the data; to increase its robustness, and also to analyse the model in a more comprehensive manner...
Statistically you cannot split your data for EFA and than for CFA. However you can do CFA after EFA.
Find the attached file for detail...
Article An Easy Approach to Exploratory Factor Analysis: Marketing P...
Thanks a lot Noor Ul Hadi for your article. But, I have seen a few studies and too published in some good journals having the data being split into two halves- one used for applying EFA and the other one for appllying CFA (based on the factors which emerged out of EFA). Can you please explain the probable logic behind doing so?
Mehra Sir, May be this link solve your purpose:
https://www.researchgate.net/post/Can_we_do_exploratory_and_confirmatory_factor_analysis_in_the_same_data_set2
To tell the truth, it is not just "not ok" but it is wrong.
EFA is meant for Pilot Study data while CFA is meant for field study data. Both are being carried out for different reasons. EFA is to determine the dimensionality of items measuring the construct while CFA is to validate the measurement model of a latent construct
- Badges
- Science topic