|
![](http://www.statmodel.com/discussion/icons/csc/1px.gif) |
Autoregressive model with count data |
![Previous](http://www.statmodel.com/discussion/icons/csc/navgen_arrow_left.gif) ![](http://www.statmodel.com/discussion/icons/csc/1px.gif)
|
Message/Author |
|
|
Dear Prof Muthén, When running an autoregressive model with (zero inflated) count data, I get the message that covariances (among different variables at the same time points) are not allowed. Does this imply I can interpret an autoregressive model without covariances or should I rather specify a completely different model for count data? Many thanks! |
|
|
Use a factor to make two count DVs have a residual correlation. f BY y1-y2; f@1; This gives you the res cov in the y2 loading. |
|
Back to top |
|
|