I have a dataset of 587 with 67 items. I try to conduct CFA with 15 factors. RMSEA is approximately .05, Chi Square/df ratio, is lower than 3. But fit indices are quite low like .70, .60. Factor loadings are all good and all of their t values are significant. The picture does not change so much when I test the 15 + 1 (overarching factor) model. I use SIMPLIS syntax with LISREL, with no special modification (such as covariance etc). I would be glad if anyone can help me interpret this situation. Answeres and references to relevant materials are very much appreciated.
Dear Cinar Kaya,
Please go through the link.I think it may help you.
http://davidakenny.net/cm/fit.htm
http://statwiki.kolobkreations.com/wiki/Confirmatory_Factor_Analysis
Few related research paper is attached here.Kindly find them.
Ramendra
Çınar meraba, bence bütün indeksleri alma çünkü örneklem büyüklüğünden etkilenebiliyor. Ayrıca modelin kabul edilip edilmemesi için modelin geneli incelenerek karar verilir. sen sadece senin modelin verdiği uyum iyiliklerini ver yeterli. Ya da ölçek maddelerin fazla olduğu için örneklem sayını arttırarak DFA yı yeniden yap.
Thanks for the answers. Eyüp hocam teşekkürler. I will try to handle the issue given that sources.
As to my knowledge, it might be a try to reduce the number of items if possible, and then remain the best items. In this case, it might be useful to run EFA first before going to CFA. Hope it is useful. All the Best
Check the link below for details
http://statwiki.kolobkreations.com/index.php?title=Structural_Equation_Modeling#Model_fit_again
Regards
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Structural Equation Modeling