|
|
 |
| Question re: interpreting LGC |
 
|
| Message/Author |
|
|
|
|
Hello,
I am building latent growth curve model in MPlus and first fit an intercept-only model, then added a linear growth effect. After conducting a Chi-square difference test, the addition of the linear growth effect yielded significantly better model fit. However, when looking at the mean of the latent linear growth effect, the Wald test is non-significant. How do I make sense of a better fitting model with the inclusion of a non-significant linear growth effect? |
|
|
|
|
| Your model must have significant slope variance so that people vary in their growth but on average it is zero. |
|
|
| Back to top |
|
|
