Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | "Martin Weiss" <martin.weiss1@gmx.de> |
To | <statalist@hsphsun2.harvard.edu> |
Subject | AW: st: AW: Run Regression for each quintile separately |
Date | Mon, 8 Mar 2010 11:47:12 +0100 |
<> -sort-ing and -list-ing was just to show you the results returned by Stata. I am reluctant to endorse your research approach, but technically you can condition on the quantiles as in: ************* sysuse auto, clear xtile quant = mpg, nq(5) forv i=1/5{ reg price weight length mpg if quant==`i' estimates sto quantile`i' } estimates table quantile1 quantile2 /* */ quantile3 quantile4 quantile5, /* */ style(oneline) ************* HTH Martin -----Ursprüngliche Nachricht----- Von: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] Im Auftrag von Stevie Kourama Gesendet: Montag, 8. März 2010 11:35 An: statalist@hsphsun2.harvard.edu Betreff: Re: st: AW: Run Regression for each quintile separately Martin, Thanks. I got the command. Please clarify a bit, 1. tile quant = mpg, nquantiles(5) this command will generate the quantile, is it ok? 2. what is this command for then "l mpg quant, sepby(quant) h(20)" sorry can you come again on how to run those regression. I managed to make the quantiles, sort, but now how to run the regressions separately. Thanks very much for the valuable assistance. Stephen On Mon, Mar 8, 2010 at 1:10 PM, Martin Weiss <martin.weiss1@gmx.de> wrote: > > <> > > For the quintiles, you may like -xtile-: > > > ************* > sysuse auto, clear > xtile quant = mpg, nquantiles(5) > sort quant > l mpg quant, sepby(quant) h(20) > ************* > > > > HTH > Martin > > -----Ursprüngliche Nachricht----- > Von: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] Im Auftrag von Stevie Kourama > Gesendet: Montag, 8. März 2010 10:52 > An: statalist@hsphsun2.harvard.edu > Betreff: st: Run Regression for each quintile separately > > I have the following model > > P_(d,k,q)=α_q+β_q P_k+μ_(d,k,q) > > Where d = a district, q = a quintile, and k is a region. The left hand > side variable is the average participation rate for a given district > and quintile, the right hand side variable is the average > participation rate at the regional level. > > Now I would like to estimate the regression separately for each > quintile using OLS. the intuition is to observe the variations in > departmental participation across the country and to understand how > increased coverage affects the participation of different population > groups. > I have the household budget survey data for two periods, 2000/01 and 2007. > > How do we control for fixed effects? > > Please assist on the appropriate stata commands. > > > > > > -- > Stevie Kourama > > * > * 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/ > > > * > * 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/ > -- Stevie Kourama * * 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/ * * 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/