Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • bootstrapped indirect effects through multiple mediators using sem

    Dear statalisters,

    I am trying to run a mediation analysis with 2 indepvars, 5 mediators and 1 depvar using SEM. I would like to extract the bootstrapped indirect effects of each indepvar through each mediator.

    my syntax sofar is:

    sem (indepvar1 indepvar2 mv1-mv5 -> depvar) (indepvar1-> mv1-mv5 depvar) (indepvar2 -> mv1-mv5 depvar), nocapslatent vce(bootstrap, reps(5000))

    estat teffects


    here I get the bootstrapped teffects, indireffect and direffects, but I would like to decompose the indireffects.

    I tried using nlcom as follows

    nlcom _b[mv1:indepvar1] *_b[depvar:mv1]

    ..and so forth for each coefficient pair.

    I understood that nlcom uses coefficients as constants, in fact the products I get have 95%ci rather than normal-based ci.
    I have read previous posts addressing similar questions, but since I am not familiar with matrices nor programming, I could not adapt some syntax I found to my case, for example the following syntax using sureg

    capture program drop bootmm
    program bootmm, rclass
    syntax [if] [in]
    sureg (read math)(write math)(science read write math) `if' `in'
    return scalar indread = [read]_b[math]*[science]_b[read]
    return scalar indwrite = [write]_b[math]*[science]_b[write]
    return scalar indtotal = [read]_b[math]*[science]_b[read]+[write]_b[math]*[science]_b[write]
    end

    bootstrap r(indread) r(indwrite) r(indtotal), bca reps(5000): bootmm

    So my question is: should I write a program, and in this case could anyone help please? or is there a different strategy I am missing?

    thank you in advance,
    Rudy
    Last edited by rudy rossi; 12 Mar 2020, 07:55.

  • #2
    Hi Rudy
    The medsem package might be useful for your purposes. It's a Monte Carlo (parametric) bootstrap approach that is faster than non-parametric bootstrapping. Perhaps it can cover your specific use case?

    In case it's relevant, Instats is offering a series of seminars on Stata's sem command including bootstrapped CIs for indirect effects and conditional indirect effects. All seminars are combined in a Structured Course in case it's of interest.

    Best wishes
    Mike

    Comment

    Working...
    X