|
Message/Author |
|
jan mod posted on Wednesday, December 02, 2015 - 3:36 pm
|
|
|
Hi, Is there a possibility to do a mediation model with: binary mediators, binary outcome variable, two level model, logit link, multiple imputed data? I tried it with model indirect but it didn't work. I'm looking into "model constraint"? Depends on the estimator I suppose? Help is much appreciated! |
|
|
Model constraint would be the way to go. But I am not sure what the correct effect formulas would be. Perhaps defined via latent response variables. |
|
Back to top |
|
|