Hi Stas,
Do you suggest that even following Roger's suggestion by using
-somersd- package to compare confidence intervals for rank statistics,
such as median, I have to run a bootstrap to get correct comparison?
-ACC
2007/3/22, Stas Kolenikov <[email protected]>:
The issue might be quite tricky: median is not a smooth function of
the data, so many inference methods generally taken for granted do not
apply to it. Asymptotic normality of the median is derived in a rather
non-standard way and requires regularity conditions that are pretty
difficult to verify. If you are dealing with survey data, neither
linearization nor the jackknife will produce appropriate (consistent)
standard errors for the population quantiles, and one has to use the
balanced repeated replication or the even more complicated survey
bootstrap both of which require the data provider's replication
weights -- it is not at all that trivial to run those procedures on
the analyst's side without getting one's feet really, really wet.
See http://www.citeulike.org/user/ctacmo/article/1036970 if you can find it.
On 3/22/07, Ada C. Chen <[email protected]> wrote:
> > Is it possible to compare medians of a continuous variable between two
> > groups by using -pweight- instead of -fweight-? Does stata only allow
> > for fweight?
--
Stas Kolenikov
http://stas.kolenikov.name
*
* 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/
*
* 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/
