Hi, all, I'm running multiple imputation to handle missing data and I'm running into a problem. I can generate the MI data sets in both amelia and the mi package (they look fine), but I can't figure out how to get pooled results. The examples from the mi package, zelig, etc., all seem to go right to something like a regression, though all I want are the mean and SE for all the variables in the data set, and the cov and for among them. How would I request those basic descriptives (using mi, zelig, or any other package). Thanks! Mike
getting basic descriptive stats off multiple imputation data
3 messages · Michael Parent, Weidong Gu
Mike, You can retrieve each of imputed data sets and use Rubin's rule for combined analysis. I am not sure how to do combined analysis of cov, but mean and SE would be estimiable. For mi package to get individual copies of imputed data ?mi.completed HTH Weidong Gu
On Tue, Oct 18, 2011 at 5:13 PM, Michael Parent <michael.parent at ufl.edu> wrote:
Hi, all, I'm running multiple imputation to handle missing data and I'm running into a problem. I can generate the MI data sets in both amelia and the mi package (they look fine), but I can't figure out how to get pooled results. The examples from the mi package, zelig, etc., all seem to go right to something like a regression, though all I want are the mean and SE for all the variables in the data set, and the cov and for among them. How would I request those basic descriptives (using mi, zelig, or any other package). Thanks! Mike
______________________________________________ R-help at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Hi, Thanks! So, you're saying I should output the descriptives from the individual data files, and manually do the stats to get the combined estimates (or, I can use the files in mplus or spss to run it in those programs)? Is there no automated process in any of the mi or mi-related packages that automates that pooling? Thanks, Mike
On Oct 18, 2011, at 8:20 PM, Weidong Gu wrote:
Mike, You can retrieve each of imputed data sets and use Rubin's rule for combined analysis. I am not sure how to do combined analysis of cov, but mean and SE would be estimiable. For mi package to get individual copies of imputed data ?mi.completed HTH Weidong Gu On Tue, Oct 18, 2011 at 5:13 PM, Michael Parent <michael.parent at ufl.edu> wrote:
Hi, all, I'm running multiple imputation to handle missing data and I'm running into a problem. I can generate the MI data sets in both amelia and the mi package (they look fine), but I can't figure out how to get pooled results. The examples from the mi package, zelig, etc., all seem to go right to something like a regression, though all I want are the mean and SE for all the variables in the data set, and the cov and for among them. How would I request those basic descriptives (using mi, zelig, or any other package). Thanks! Mike
______________________________________________ R-help at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.