Skip to content

Help with Amelia

3 messages · Amit Patel, Mark Seeto, Alex Karner

#
Hi
I have used the amelia command from the Amelia R package. this gives me a number 
of imputed datasets. 

This may be a silly question, but i am not a statistician, but I am not sure how 
to combine these results to obtain the imputed dataset to usse for further 
statistical analysis. I have looked through the amelia and zelig manuals but 
still can not find the answer. This maybe because I dont understand the science 
behind it. Can anyone help

Thanks in advance
#
Amit Patel-7 wrote:
The details depend on what you are doing with the data.

Take the method you would use if there were no missing data, and apply it to
each imputed data set. Average the estimates of interest from each imputed
data set to get overall estimates. You then have to adjust the covariance
matrix of the estimates to account for variability due to imputation. This
is explained in various places, for example "Statistical Analysis with
Missing Data" by Little and Rubin, and "Regression Modeling Strategies" by
Harrell.

Frank Harrell's rms package can do this combining of imputations for you.
#
Take a look at package 'mitools' and 
http://faculty.washington.edu/tlumley/survey/svymi.html
for some examples in analyzing multiply-imputed survey data.