Hi Statalist,
I am using the --cqreg-- command for censored quantile regression from Ken
Chay's Berkeley website and am having some interpreting the output.
In particular, is the bias-corrected estimate equal to the "observed"
estimate or "observed"-"bias"? The Horowitz Handbook of Econometrics Chapter
(V5, Chap52) would suggest the latter (esp example 3.1).
If this is correct, how can I extract these results? --eret li-- seems to
return results from the last repetition of the --qreg-- estimation in the
bootstrap repetitions.
I suspect this betrays a more general lack of familiarity with bootstrapping
in Stata, so perhaps expertise with --cqreg-- specifically is not needed to
answer this question. I have also experimented with --clad-- from Stata
Technical Bulletin 58 and am experiencing some of the same difficulties.
Sample code follows.
Many thanks,
Raymond
----- sample code using cqreg ------
sysuse nlsw88.dta
save fake_nlsw88, replace
gen fake_wage = max(wage,4)
sum fake_wage, detail
compress
save fake_nlsw88, replace
cqreg fake_wage grade age, quantile(0.6) ll(4) dots
------- end sample code -----------
------ output when running the above in Stata 10.0 for Windows ---
. cqreg fake_wage grade age, quantile(0.6) ll(4) dots
............................................................................
........................
Initial sample size = 2246
Final sample size = 2117
Bootstrap statistics
Variable | Reps Observed Bias Std. Err. [95% Conf. Interval]
---------+------------------------------------------------------------------
-
grade | 100 .8756039 -.0174292 .050704 .7749961 .9762118
(N)
| .7776431 .9591758
(P)
| .8009594 .9764307
(BC)
---------+------------------------------------------------------------------
-
age | 100 .0073803 .0124284 .0259035 -.0440179 .0587786
(N)
| -.0283723 .0686576
(P)
| -.0329161 .0533409
(BC)
---------+------------------------------------------------------------------
-
const | 100 -4.431019 -.210878 1.240352 -6.892147 -1.969892
(N)
| -7.371184 -2.613133
(P)
| -7.289414 -2.472435
(BC)
----------------------------------------------------------------------------
-
------- end output ------
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
