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

  • How to obtain CIs for standardized direct/indirect effects when using SEM builder

    Hi all,

    When using the SEM builder there is an option to have indirect/direct effects reported for mediation analysis. Additionally, one may request these effects reported as standardized (e.g. estat teffects, standardize). Typically one would request the direct/indirect effects in the same format that they report the rest of the model. However, when you requests the effects to be standardize, stata no longer reports confidence intervals. Is there way to obtain CIs for standardized teffects outputs?

    I believe this is a problem that is relatively well known and I wanted to see if there was any new information out on how to best handle it besides manually standardizing each variable and running the model as such.

    Thank you in advance!