Re: st: How do I obtain conf. intervals for differences between percentiles of two different distributions?

[Tirthankar Chakravarty <tirthankar.chakravarty@gmail.com>]

See advice previously given here:
http://www.stata.com/statalist/archive/2009-06/msg00376.html

This produces bootstrap confidence interval around the difference of
percentiles.

T

2010/4/14 Liliana Olarte <liliana.olart@gmail.com>:
> Dear Statalist :
>
> I am trying to estimate the wage gap between females and males by
> percentiles.  However, I would like to estimate conf. intervals for
> the difference between the percentiles of the two distributions. ¿How
> can I do this?  I know cendif command does not work since cendif is
> ment for computing confidence intervals, not for differences between
> percentiles, but for percentile differences.
>
> This is the procedure I did for percentile(75) :
>
> - centile log_hourly_wage if gender==1, centile(75) normal
> - centile log_hourly_wage if gender==2, centile(75) normal
>
> Results for gender==1
>                                                  --Normal, based on observed centiles --
> Variable               |         Obs    Percentile      Centile    [95% Conf. Interval]
> ------------------------------------------------------------------------------------------------------------------------------------
> loghourlywage   |       2956    75           8.334871      8.291912  8.377831
>
>
> Results for gender==2
>                                                  --Normal, based on observed centiles --
> Variable               |         Obs    Percentile      Centile    [95% Conf. Interval]
> ------------------------------------------------------------------------------------------------------------------------------------
> loghourlywage   |       1415     75         8.477972        8.412637    8.543307
>
>
> gap (75th centile) = 8.477972 - 8.334871=0.1431
>
> These are the results when I use cendif :
>
> Percent Pctl_Dif        Minimum Maximum
> 75      .97528172       .90964651       1.0403433
>
> As it can be seen from the results above, cendif is not the appropiate
> command to use here.
>
> Regards,
> Liliana
>
> *
> *   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/
>



-- 
To every ω-consistent recursive class κ of formulae there correspond
recursive class signs r, such that neither v Gen r nor Neg(v Gen r)
belongs to Flg(κ) (where v is the free variable of r).

*
*   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/