Alternative starting values for boots... PreviousNext
Mplus Discussion > Structural Equation Modeling >
 Lucas LaFreniere posted on Monday, August 22, 2016 - 9:00 am
Hello Drs. Muthen, I'm attempting to analyze a mediation model with three latent variables and one observed variable in four timepoint chain (pre-->mediator-->post-->follow-up). There is missing data on all variables and I'm using FIML to account for it. I'm using bootstrapping to test the indirect effects, but I've noticed that when I run a bootstrapping model twice I get the exact same estimates. Is there any way to change the starting values (or seed) of my bootstrap to draw *different* bootstraps on different runs? Thank you kindly, Luc
 Bengt O. Muthen posted on Monday, August 22, 2016 - 1:00 pm
The parameter estimates are not affected by the bootstrapping, only the SEs and the CIs.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message