Robust regression vs Robust standard error

[Jo Smith]

Hello,

What is the difference between Robust regression and Robust Standard Error? Is the first one is to control outliers and the second one is for heteroscedasticity?
Is the second one only is enough?
What is the Stata code for the Robust regression?

Thanks

[Christian Otchia]

See these Stata Tidbits by Micheal Norman Mitchell:
http://www.michaelnormanmitchell.com...rs-part-1.html
http://www.michaelnormanmitchell.com...rs-part-2.html

[Carlo Lazzaro]

Jo:
the difference you report is, in general terms, correct.
However, things are trickier with robust regression (-rreg- in Stata); one of the first question in daling with this tool (and with the so called robust methods at large) should be: robust with respect to what? Are you sure that your so called outliers are not consistent with the data generating process that underpins your sample?
Anyway, for more details on -rreg-, you may be interested in taking a look at: http://papers.ssrn.com/sol3/papers.c...act_id=1369144