|
![](http://www.statmodel.com/discussion/icons/csc/1px.gif) |
Model indirect for categorical covari... |
![Previous](http://www.statmodel.com/discussion/icons/csc/navgen_arrow_left.gif) ![](http://www.statmodel.com/discussion/icons/csc/1px.gif)
|
Message/Author |
|
Gareth posted on Tuesday, August 25, 2009 - 8:22 am
|
|
|
When using model indirect, I interpret the standardized estimate as one standard deviation change in Y for each standard deviation increase in X that is mediated by M. When X is categorical, I want to use STDY. However, I am unsure how to interpret. Is the estimate now one standard deviation change in Y, as X changes from 0 to 1, for each unit increase in M? |
|
|
It is the one standard deviation change in Y, as X changes from 0 to 1, that is mediated by M. |
|
Gareth posted on Saturday, August 29, 2009 - 5:46 am
|
|
|
If the outcome, mediator and covariate were all latent variables, could I use STDXY to interpret as "one SD change in Y per SD change in X that is mediated by M?" |
|
|
You should then use STD which standardized with respect to the latent variable variances. |
|
Back to top |
|
|