I have 414 data set for developing a motivation scale. I used EFA and CFA. So I split the data to two groups. 207 for EFA and 207 left for CFA. AFter I finished the analyze with SPSS for EFA and SPSS AMOS for CFA, I knew that some of questions are not valid in EFA while it is valid in CFA and vice-versa. what should I do with this? this is for my thesis project. should I just report it like that? and when I report it, some questions that are not valid in EFA and CFA at the same time. Which one should I call as valid? Thank you in advance.
What do you mean when you say some of the questions are not "valid" in EFA but are in CFA?
Thank you so much for the respon Sir David L Morgan
I think I did the wrong analytical steps. I re-did the AFA test analysis step for the CFA data. I should have only used the CFA data as data in the CFA analysis calculation and used the rotated component matrix value from the AFA test analysis results. I have analyzed the data set again and I found some connections in CFA model. Is the model still fit? Thanks again in advance.
In EFA, the closest comparison to CFA is a Maximum-Likelihood extraction with an oblique rotation (i.e., correlated factors).
And yes, you should run the CFA data entirely separately from the EFA data.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS