Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: RE: Extreme data points

From   "Sarah Edgington" <>
To   <>
Subject   st: RE: Extreme data points
Date   Tue, 7 Jun 2011 11:05:53 -0700

You asked this question yesterday.  Calling observations extreme data points
instead of outliers isn't going to magically change people's mind about the
advisability of what you want to do.  If you're going to outright ignore the
statistical advice you've gotten here that's your business but starting a
new thread asking the same question, as if you weren't aware of the advice
you've already been given, is downright disingenuous and rude.

More importantly, though, Austin Nichols already gave you an example of two
different methods for dropping variables with extreme values.  He didn't
give you the exact code you're asking for here but he did give you an
example that you could easily modify to do what you want to do.  You should
now do the work of figuring out what the example code does and modifying it
as necessary to implement it for your purposes (hint -help percentile-). I
would also highly recommend reading the rest of the thread and really
thinking about the advice you've been given.  It's entirely up to you
whether you follow it, but you should at least do yourself the favor of
reading it.


-----Original Message-----
[] On Behalf Of Achmed Aldai
Sent: Tuesday, June 07, 2011 3:56 AM
Subject: st: Extreme data points


how can I drop the top 0,5% and bottom 0,5% observations of a variable?
I have to do this for several variables.

It would be great if someone could give me an example.

Thank you very much
NEU: FreePhone - kostenlos mobil telefonieren!			
Jetzt informieren:
*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index