Nonignorable Missing Data PreviousNext
Mplus Discussion > Missing Data Modeling >
Message/Author
 Moh Yin Chang posted on Saturday, July 21, 2007 - 1:55 pm
Dear Dr. Muthen,

I would like to ask if the current Mplus can fit the Diggle-Kenward selection model and/or shared-parameter model for nonignorable missing data. If yes, is there a code example that you may show me?

Thank you.
 Bengt O. Muthen posted on Saturday, July 21, 2007 - 5:15 pm
By Diggle-Kenward, do you mean missingness that is a function of the variable which has missingness? If so, yes. If not, please send me the article describing the model.
 V X posted on Wednesday, October 10, 2007 - 3:02 am
Dr. Muthen, I am also interested in learning Mplus to fit Diggle-Kenward selection model and shared-parameter model for nonignorable missing data (that is, missing not at random). Would you provide some Mplus code examples?

Thank you.
 Bengt O. Muthen posted on Wednesday, October 10, 2007 - 3:49 pm
I think Diggle-Kenward consider missingness as a function of the (latent) response variables y - what you would have observed if if wasn't missing. You could use DATA MISSING to create binary missing data indicators and then regress those "u's" on the "y's" that have missingness by regular ON statements (y on u). I am not familiar with the term "shared-parameter model".
 Bengt O. Muthen posted on Wednesday, October 10, 2007 - 3:51 pm
P.S. For more non-ignorable missing data modeling, see also the Lecture 17 handout from my UCLA latent variable course at

http://www.gseis.ucla.edu/faculty/muthen/Handouts.htm
 V X posted on Friday, October 12, 2007 - 1:00 pm
Dr. Muthen, the Lecture 17 notes for
Educ 231E class helps me a lot to understand the "outcome based dependent selection model". Thank you.

Currently, I have one question about the path diagram of "growth mixture model with nonignorable missingness as a function of y".

I know a circle represents a latent variable. Would you intepreate that, what the meaning of putting a circle on y in the diagram under the contents of selection model?
 Bengt O. Muthen posted on Friday, October 12, 2007 - 3:09 pm
For individuals who have missing data on y, the y variable is a latent variable.
 Moh Yin Chang posted on Friday, November 30, 2007 - 8:55 am
Dr Muthen,
You said "for individuals who have missing data on y, the y variable is a latent variable". Do you mean that by creating a latent variable CY like the figure in slide 6 of your Lecture 17, we can fit a nonignorable missingness model with missing Y?
 Bengt O. Muthen posted on Friday, November 30, 2007 - 3:29 pm
Yes. But you can also do non-ignorable missingness modeling without cy (which is categorical) using the model of slide 3.

Neither is an easy thing to do.
 Bruckers Liesbeth posted on Wednesday, November 25, 2009 - 8:03 am
Dear Dr. Muthen,

I was wondering if the syntax for the models that you discuss in the slides

http://www.gseis.ucla.edu/faculty/muthen/Handouts.htm (Lecture 17)

is available.

Kind Regards,

Liesbeth
 Bengt O. Muthen posted on Wednesday, November 25, 2009 - 9:33 am
No, it is not. But a new missing data paper will be posted within short which discusses alternative models for non-ignorable missing data and you can then request the Mplus setups for those analyses.
 Tim Stump posted on Friday, April 20, 2012 - 3:01 pm
I have a cohort of type II diabetes adolescents with hemoglobin a1c collected at baseline (prior to high school graduation), 3, 6, 9, and 12 month time pts. We know that our a1c outcome does not satisfy MAR assumption because we could not get all chart review data from physicians offices after adolescents left home. Baseline a1c is not missing, but missing increases over time. The cohort is relatively small with 180 subjects, but would like to explore some of the models outlined in "Growth Modeling With Nonignorable Dropout: Alternative Analyses of the STAR*D Antidepressant Trial". Our goal is simply to model a1c over time and see if trajectory is different for a couple of binary covariates and if missing a1c influences trajectory. Would have you any suggestions as to which type of NMAR model would work better with our small sample?
 Bengt O. Muthen posted on Friday, April 20, 2012 - 8:35 pm
Pattern-mixture modeling is probably the easier to work with.
 Martin Blais posted on Thursday, May 29, 2014 - 1:27 pm
Dear Drs. Muthen,

I am looking for a way to model non-ignorable missing data in a LCA model.

Case is I have 3 variables, each representing the age at onset of a 3-stage process. As stages 2 and 3 are only possible if the previous one has been reached, missing data on stage 2 and/or stage 3 are non-ignorable. The missing data structure looks like a monotone missing pattern, except that they are not dropout: the missing data are informative that the next stage was not reached and I want to include this information in the model.

Structure of the database is :
s1 s2 s3
9 12 13
12 14 17
14 15 .
11 16 .
13 . .
17 . .



Do you have any advice on how to implement such a model in Mplus? The closest I found is the Diggle-Kenward selection model (Ex. 11.3), but there are no i, s, q components in what I model...

Many thanks
 Bengt O. Muthen posted on Thursday, May 29, 2014 - 3:47 pm
So you have an LCA based on 3 cont's variables. Does s1 predict missingness on s2 and s3 and s1 is always observed, so MAR? Regarding non-ignorable, are you saying that the values that would have been observed for s2 and s3 predict their missingness?
 Martin Blais posted on Friday, May 30, 2014 - 5:02 am
You are right on the MAR component. I might have erred regarding the non-ignorable.

Cont's var are age at three stages: s1 = self-awareness; s2 = self-identification; s3 = disclosure. They follow a sequence constraining values in a way that s3>s2>s1. Data are cross-sectional.

s1 is always observed (inclusion criteria). s2 can be observed or not (if stage s1 is completed and s2 has been reached). s3 can only be observed if s2 is observed AND s3 has been reached.

I want to model trajectories that take into account both the occurrence (yes or no) of the stages and, if the next stage has been reached, the age at which that happened.

Thanks for your help!
 Bengt O. Muthen posted on Saturday, May 31, 2014 - 2:16 pm
I think I would need to understand the setting better to help you and that goes beyond Mplus Discussion. You may want to ask on SEMNET. You want to make clear why LCA is of interest to you (why mixtures?) and why you want to model trajectories (trajectories of what?). Two comments: Selection modeling of missing data like Diggle-Kenward can be done without a growth model; survival modeling might be relevant given that you want to model age at which the events happened (perhaps multivariate survival; see the Masyn dissertation on our website).
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: