--- Rajesh Tharyan <[email protected]> wrote:
> In some cases it may be preferable to throwing out the observations?
Maybe, but only based on some external evidence that that observation
is seriously flawed, and not on any automatic procedure identifying
observations as outliers. Extreme values may be extreme due to some
error in data collection (e.g. accidentily typing too many zeros), in
which case that observation can be rightly ignored, or it may just be
genuinly large, in which case a model may need to be adjusted to
explain it, but that observation should certainly not be discarded.
Choosing which scenario is true for which observation can only be done
using external information. As a result there is .ado file that can do
this for you. Also notice that all procedures discussed so far are
univariate, i.e. an outlier is defined by these procedures only by its
relative position in one variable, while many applications are
multivariate (i.e. regress) and this changes the definition of outlier.
Hope this helps,
Maarten
-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands
visiting address:
Buitenveldertselaan 3 (Metropolitan), room Z434
+31 20 5986715
http://home.fsw.vu.nl/m.buis/
-----------------------------------------
___________________________________________________________
Yahoo! Mail is the world's favourite email. Don't settle for less, sign up for
your free account today http://uk.rd.yahoo.com/evt=44106/*http://uk.docs.yahoo.com/mail/winter07.html
*
* 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/
