Page 1 of 1

Missing values in the imputation model

Posted: Mon Nov 11, 2013 1:47 pm
by lillauslillosa
Hello,
I have a problem with the imputation of the data. I am trying to impute birth weight values using gender, survival status, educational level of the parents and mother’s perception of size as suggested by Channon in "Multilevel multiple imputation of missing birth weights in developing countries: Analysing neonatal and post-neonatal mortality."

However I have missing values for the mother’s perception of size variable.

Any suggestion about how I can deal with this issue?

Thank you so much.

Re: Missing values in the imputation model

Posted: Mon Nov 11, 2013 5:39 pm
by Harvey Goldstein
This is what realcom impute is designed to do. Please read the manual for examples and further details.
Do come back if you get stuck
Harvey Goldstein

Re: Missing values in the imputation model

Posted: Mon Nov 11, 2013 6:41 pm
by lillauslillosa
Dear Prof. Goldstein,

Thank you for your response.
I am not sure which manual you are referring to because I have read the RealcomImputeStata manual by Prof. Bartlett and the REALCOM-IMPUTE Software for Multilevel Multiple Imputation with Mixed Response Types by Prof. Carpenter, you, and Prof. Kenward but I have not found anything about how to approach the missing values if one of the variables I should include to impute the missing values of the response variable has missing values too. In this case the mother’s perception of size, which is one of the explanatory variables in the imputation model for birth weight, has missing values too.

I am sorry for this, but if you could guide me I would appreciate it a lot.

Best,
Liliana