Statalist


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

st: AW: How does -kwallis2- compute adjusted p-values for significance?


From   "Martin Weiss" <martin.weiss1@gmx.de>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: AW: How does -kwallis2- compute adjusted p-values for significance?
Date   Wed, 11 Feb 2009 14:39:07 +0100

<> 


"Again, this is one of the most important failure of Stata: there are
packages to do everthing, but few of them can be mentioned in papers."

Not quite sure what you are getting at here. Is R any better in this
respect? Anyway, you should not call this a failure of Stata. Everything
inside official Stata is backed up by references in the help files and even
more extensively in the manuals. Official Stata is not responsible for what
authors of user-written packages do or do not do, though...

HTH
Martin


-----Ursprüngliche Nachricht-----
Von: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] Im Auftrag von Tiago V.
Pereira
Gesendet: Mittwoch, 11. Februar 2009 12:14
An: statalist@hsphsun2.harvard.edu
Betreff: st: How does -kwallis2- compute adjusted p-values for significance?

Thank you Martin! I really do that often, more often than the necessary,
btw.

I have already mentioned here that this unecessary task might be avoided
once each stata package points out references within the help file to
everybody rapidly to find out the answer.

I use more than 40 Stata packages weekly, and in several instances I have
no idea about which tests are employed! Frenquently, it is easier to
compare results to known packages such as SPSS, GraphPad or even R! -
instead of opening codes and try to decipher each code.

Again, this is one of the most important failure of Stata: there are
packages to do everthing, but few of them can be mentioned in papers.

If you don´t know stata programming, there is no way to know which methods
are programmed in the package.

All the best,

Tiago Pereira




There are several ways to get at this information, one being to -trace-
the thing and scroll up to the point where it is created afterwards. -set
tracehilite- is very helpful in this exercise...

******
sysuse census, clear
set tracehilite Adjusted
set tr on
kwallis2 medage, by(region)
set tr off
*******



Next possibility: If you -log- the process, you can look at the log in a
text editor and search for "Adjusted"...

Or you look at kwallis2.ado, for instance with NJC`s -ssc d fedit- where
you will find lines 123-124:

*******
 local c1=(1-$S_level/100)/(2*(`maxGrp'-1))
 di in gr "(Adjusted p-value for significance is " %7.6f `c1' ")" _n
*******



If you want, create your own copy of kwallis2.ado and toss in a -pause- so
you can inspect the state of the world when line 123 is processed...





HTH
Martin


============================================================================
Dear Statalisters,

Does anyone know about which method  -kwallis2-  uses to compute adjusted
p-values for significance when the Kruskal-Wallis Test is statistically
significant?

Cheers!

Tiago







*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index