Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Jacob Felson <felsonj@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | st: Selecting cases for optimal match with vector of means |
Date | Tue, 1 Mar 2011 22:35:31 -0500 |
I'm wondering if there is a Stata routine that could facilitate the following task. Students in my department were recently administered a standardized test from the Educational Testing Service (ETS). This test was administered in 141 other schools. ETS offers me the option of selecting a set of no fewer than 10 other schools with which to compare the scores of students in my department. Of course, I would like to select a comparison sample of schools that matches my school in important ways. I have data on average SAT scores, percent black and percent Hispanic. Not surprisingly, there are no schools that are exactly the same as mine on the above three metrics. But I am attempting to select a sample of schools in an optimal fashion in order to minimize the mismatch between the metrics in the comparison school and the metrics for my school. What might be the best way to go about doing this? Propensity score matching is out of the question because ETS standardized test scores are not available to me. I would like to select a sample that provides an optimal balance of bias and efficiency. Thank you for your attention to this, Sincerely, Jacob Felson Assistant Professor Department of Sociology William Paterson University * * 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/