|
Message/Author |
|
|
Hello, I have a latent growth curve model with 8 time points. This 8 measurements I took within 4 weeks two times per week with identical items. My question is if it is possible to inter-correlate the errors (what would be logical)? Or if I should not/can not do this in a LGCM? (Also error intercorrelation would give me a much better model fit: RESMA .085 to .06 CFI .82 to .91) Thank you for your answer. |
|
|
I believe residuals at adjacent time points are sometimes correlated. |
|
Back to top |
|
|