Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: Permutation tests of paired analyses


From   "Dietrich, Mary" <mary.dietrich@Vanderbilt.Edu>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: Permutation tests of paired analyses
Date   Wed, 7 Jun 2006 16:06:32 -0500

I am just beginning to use permutation tests and seem to have begun with
a situation that is not commonly covered in the documentation and online
examples. I want to do permutation tests of paired analyses - for the
sake of simplicity, paired t-tests. I have no trouble with permutation
tests on independent groups...

The portion of these permutations that I can not get around is the need
to specify some 'variable' on the "permute" command. There is no single
variable for paired analyses. Then I thought I needed to create a
program to save the resultant 't' - I did that (see below) and the
program works but I still seem to be missing a step. I still need some
type of variable for the "permute" command and it does not accept the
't' value that I thought I was saving from the previous program 'return'
statement. I must be overlooking some very basic concept in this
process.

In the following example, the variables, "RT_R" and "RT_L" comprise a
set of paired values.

program pttest, rclass
  1. ttest RT_R==RT_L
  2. return scalar t=r(t)
  3. end

. pttest

. permute ???? t=r(t), reps(1000)...

Thanks for any assistance!

--Mary

*
*   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/



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