Message/Author |
|
|
Hello, I'm fitting a growth mixture model on continuous indicators from four time points but have trouble arriving at a sounds solution. I keep getting non-positive latent variable covariance matrices and someone suggested to check residual variances. Could you tell me how I can free the residual variances of the indicator items across classes and/or across time in the growth mixture model? Thanks in advance! Charlotte |
|
|
We need to see your full output to give advice on this - send to Support along with your license number. |
|
|
Thanks, but would it also be possible to provide the syntax I can use to free the residual variances of the indicator items across classes and/or across time? Thanks in advance! |
|
|
I don't think letting residual variances vary across classes will solve this problem. They are already free across time as the default. |
|
Back to top |