multi-level sequential mediation analysis

[Aiqing Ling]

Dear all,

I am working with Stata12 and I would like to conduct sequential mediation analysis on multi-level with boostrapping.

I used "ml_mediation" syntax to conduct multi-level mediation analysis with one mediator as following:
ml_mediation, dv(dv_var) mv(mv_var) iv(iv_var) l2id(sid) cv(ctrl_var)
quietly bootstrap r(ind_eff) r(dir_eff) r(tot_eff), ///
strata(sid) reps(5000): ml_mediation, dv(dv_var) iv(iv_var) ///
mv(mv_var) cv(ctrl_var) l2id(sid)

in which, "dv" indicates dependent variable, "mv" indicates mediator, "iv" indicates independent variable, "cv" indicates control variable, and "l2id" indicates subject ID. The bootstrapping helps to estimate the direct effect from IV to DV and indirect effect from IV -> MV -> DV.

For the sequential mediation analysis, there would be two mediators involved in sequence. Specifically, the indirect effect goes from IV -> MV1 -> MV2 -> DV while the direct effect still goes from IV to DV. I wonder anyone can help me with this on implementing the syntax.

Thank you all in advance.

Aiqing Ling

[Anika Peschl]

Dear Aiqing,

I am confronted with same difficulty. I would like to perform a multi-level sequentail mediation analysis in stata. Have you found a solution?

Maybe, anyone else could help as well? I would really appreciate your help!!

Thank you in advance!

Anika