Hi Statalist,
Sorry if this is a stupid question- I am new to multiple mediators and can not find an easy way to get indirect effects and CIs for multiple mediators. Is possible to exponentiate direct and indirect effects coefficients from teffects command in sem and use as odds ratios or can path multiplication only owrk for the linear models? Is there an easy way to calculate CIs from gsem- like the previous post here I had no luck with gsem and nlcom. My mediators are inter-related, so the causal approach that I have used for a single mediator is probably not applicable from what I understand of Vanderweele. I would greatly appreciate any advice from smarter people
Thanks, Kylie
Sorry if this is a stupid question- I am new to multiple mediators and can not find an easy way to get indirect effects and CIs for multiple mediators. Is possible to exponentiate direct and indirect effects coefficients from teffects command in sem and use as odds ratios or can path multiplication only owrk for the linear models? Is there an easy way to calculate CIs from gsem- like the previous post here I had no luck with gsem and nlcom. My mediators are inter-related, so the causal approach that I have used for a single mediator is probably not applicable from what I understand of Vanderweele. I would greatly appreciate any advice from smarter people
Thanks, Kylie
