I read with interest the 7/13/08 post re: correlation using data from multiple imputation datasets (-ice-)
It makes sense to me that with 2 continuous variables, a work-around would be to standardize the dependent and independent variables, then use the mim module to run the regression (taking advantage of the fact that the regression coefficient in a bivariate regression is the correlation if you standardize both the IV & DV). Is there a similar workaround if one wishes to instead correlate dichotomous and continuous variables (point-biserial correlation) or dichotomous variables (tetrachoric correlation)?
Thank you
Jessica
*
* 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/
