|
|
Quantifying the effect of a continuou... |
|
Message/Author |
|
|
Hi I have no success finding an option I believed Mplus had: the ability to estimate and quantify the effect of a continuous moderator on a regression (or a correlation) In LCA, Mplus can estimate the effect of latent classes on associations between two variables. I would like to have the possibility to estimate moderation for a non-categorical variable, having (1) a quantification of its effect on the association between two variables (2) being able to develop a model with longitudinal data and three-wave data where the influence of the moderator is estimated at all three measurements within one model. Is that possible. If so, how do I proceed? Thanks for any help. Christopher Bratt |
|
|
I might add... I know of the option to estimate interaction effects even between two latents (XWITH). That is an alternative. But it is not quite what I was hoping for when testing the moderating effect of a continuous variable. Regards, Christopher |
|
|
Ex5.23 uses the continuous variable "pihat" to moderate a correlation in a "QTL" analysis. Moderating a regression slope gets back to the usual interactions between variables. |
|
Back to top |
|
|