st: A correlation matrix after multiple imputation

[Alan Acock <acock@mac.com>]

Some journals request a correlation matrix and vector of standard deviations. The American Psychological Association stresses this in their journals. When doing multiple imputations, I'm unclear how to proceed.
1. If I have 20 imputed datasets, I could calculate the correlations on the entire n*20 observations, but this would not be how the regressions parameters were estimated and would not be consistent with them.
2. Using a program such as Mplus I could generate a "population" dataset consistent with the final solution and calculate the correlation matrix on that. I suppose I could do tests of significance on these values (APA likes these to be reported) using some N value. I'm not clear if this would be justified.
3. Is there an easy way to obtain the 20 correlation matrices, one for each of the 20 imputed datasets and then somehow pulling these?
4. I've seen manuscripts where people say here is the correlation matrix (casewise) before the multiple imputation, but this seems to be inconsistent with the entire logic of doing multiple imputation.

Any other ideas would be appreciated.

--Alan Acock

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/