Hello Statalisters,
I'm modeling a mediation analysis with all continuous variables (treatment:aces_score, mediator:violence_score, and outcome:mh_score) in Stata 17 but are including an interaction between the treatment variable (aces_score) and a categorical covariate (relationship quality:srqcat) so am using the gsem instead of sem procedure.
Here is my model:
gsem (mh_score <- violence_score c.aces_score##i.srqcat i.age_gr i.educat2 i.wquintile rural, regress) ///
(violence_score <- c.aces_score##i.srqcat i.age_gr i.educat2 i.wquintile rural, regress)
I'd like to be able to report on the %total effect mediated with 95% CI. Is there an easy way to get this from Stata?
Thank you so much,
Alison
.
I'm modeling a mediation analysis with all continuous variables (treatment:aces_score, mediator:violence_score, and outcome:mh_score) in Stata 17 but are including an interaction between the treatment variable (aces_score) and a categorical covariate (relationship quality:srqcat) so am using the gsem instead of sem procedure.
Here is my model:
gsem (mh_score <- violence_score c.aces_score##i.srqcat i.age_gr i.educat2 i.wquintile rural, regress) ///
(violence_score <- c.aces_score##i.srqcat i.age_gr i.educat2 i.wquintile rural, regress)
I'd like to be able to report on the %total effect mediated with 95% CI. Is there an easy way to get this from Stata?
Thank you so much,
Alison
.