Thanks to Kit Baum two new commands are available on the SSC.
The first, nonparmde, uses non-parametric estimators to calculate the
Minimum Detectable Effect (MDE) of a cluster-randomized experiment. My
colleague, Joel Middleton, and I wrote this command for cases when
random assignment of clusters is planned and the user has access to
either cluster totals or averages of a prior outcome variable. More
information about the estimators can be found in Middleton & Aronow
(2011).
The second command creates crossvalidated predicted values of an OLS
regression. It allows you to either have the command generate folds
automatically or specify a custom "folds" variable. Xvalols is
required if you plan to use the crossvalidation option in the
nonparmde command.
If there are any comments, questions, or suggestions, please do not
hesitate to contact me.
Thanks,
John Ternovski
Director of Analytics
Analyst Institute
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/faqs/resources/statalist-faq/
* http://www.ats.ucla.edu/stat/stata/
