Comparing growth curve models

[Klara Hagspiel]

Hey,

I am currently working on my thesis and want to find out if there is an association between social support and negative affect. Both variables are measured over 7 time points, and I have already created for each variable a latent growth model. There is a non-linear trajectory.

Now I would like to put these two models together and see if there is a significant association between the variable's change trajectories. In other words, if social support goes up, does negative affect go down.

Is this possible within the lavaan package ?

Thanks for any help !

K.

[Yago Luksevicius de Moraes]

Hi,

Not sure if I understood it right, so feel free to correct me if I'm wrong, but you seem to be interested in a cross-lagged panel model.

In this case, instead of creating a LGM to each variable, you should make a single model with both variables together and each variable at time t would be caused (regressed) by itself and the other variable at time t-1.

However, I understood you estimated a latent intercept and latent slopes of a non-linear function. In this case, I've never seen anyone correlating intercepts and slopes of two different models. Maybe you could plot the line described by these functions and compare them.

Best,

Yago

[Mauricio Garnier-Villarreal]

Hi

I think the LGM is fine for this, you can put the 2 growth curves in the same model, and correlate the latent slpes. These will be the correlations between change over time, can be a little odd to interpret. But is is not an unusual model

Can also correlate linear, quadratic etc slopes between different growth models. Might see these as coupled growth models for example