Message/Author |
|
Noa Cohen posted on Monday, May 11, 2015 - 7:03 am
|
|
|
I ran a CFA with 2 latent factors (with categorical variables). The fit indices are not good, but the loadings on each of the factors are good. I would love some suggestions how to have a better fit. |
|
|
You can try an EFA to see if your CFA is reasonable. You can look at modification indices to see where the misfit lies. General questions like this are more appropriate for a general discussion forum like SEMNET. |
|
|
Dear Dr Muthen, I am running a cross-lagged model, but the modelfit is really low (CFI aroud .45, TLI around .35, RMSEA around .25). The modification indices aren't of any help as there are no M.I.'s greater than 15. So I don't understand why the model fit is that low. Could it be because I have few participants (N=46)? My model does have fewer parameters than participants. Thank you for your help! Kind regards, Aurelie |
|
|
I don't think the test statistics are reliable at this low sample size and high number of paprameters (relative to N). |
|
|
Thank you for your quick reply. Would this mean that I can ignore the fit statistics? Or that I should avoid models with such high parameters? |
|
|
Generally, you should avoid models with more parameters than observations. |
|
Back to top |