|
![](http://www.statmodel.com/discussion/icons/csc/1px.gif) |
Nested models in growth modelling |
![Previous](http://www.statmodel.com/discussion/icons/csc/navgen_arrow_left.gif) ![](http://www.statmodel.com/discussion/icons/csc/1px.gif)
|
Message/Author |
|
Julia posted on Wednesday, March 01, 2017 - 12:20 am
|
|
|
I have a question about nested models in the growth modelling. The linear growth model is nested within the quadratic growth model. So we can compare them with chi-square difference? What about the latent basis model? Is it nested within the linear growth model? If not, how to compare the fit o the latent basis model and the linear growth model? Thank you very much in advance. |
|
|
If the two models differ in terms of variances being zero versus free, then the chi-square difference test is not ok. So if the quadratic model has a free variance for the quadratic term, then you can't test the linear model against the quadratic. I would use BIC. So use that same principle for your latent basis question. |
|
Back to top |
|
|