Missingness correlated variables PreviousNext
Mplus Discussion > Missing Data Modeling >
 David Alarcon posted on Tuesday, August 02, 2016 - 3:45 pm
I wonder if I could on a Monte Carlo simulation generate a variable which would be correlated with the missingness of other variable as specified on the Model Missing.
Thank you,
 Bengt O. Muthen posted on Tuesday, August 02, 2016 - 6:15 pm
Yes, for example missing on a set of y's can be influenced by x's.
 David Alarcon posted on Tuesday, August 02, 2016 - 6:25 pm
Thanks but I would like to know how do to it for example in an montecarlo where y x z h:
Model Population:
y on x*.6;

Model missing:
y on z*.4;

I want to "h" be correlated with "y"-being missing. So how could I generate the missingness variable e.x. ymis and do something like: h with ymis*.5;

Thank you
 Bengt O. Muthen posted on Wednesday, August 03, 2016 - 10:49 am
You can do that by generating a missing data indicator. It's a bit involved, but to do so, you might be helped by Chapter 10 of our new book that we describe on our website. We go through several missing data Monte Carlo analyses there.

The book is once again available at the end of this week.
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