s-news
[Top] [All Lists]

Re: testing equality of pairwise correlation across subsamples

To: phlasky@earthlink.net
Subject: Re: testing equality of pairwise correlation across subsamples
From: Patrick Burns <pburns@pburns.seanet.com>
Date: Thu, 22 Jan 2004 17:59:27 +0000
Cc: s-news@lists.biostat.wustl.edu
References: <410-220041422171545703@earthlink.net>
User-agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-GB; rv:1.0.1) Gecko/20020823 Netscape/7.0
Well, I can give a definitive answer to one of your questions:
No, the permutation test functions at burns-stat.com do not
handle ANOVA designs.

If an effect is zero, then the distribution (under some mild
assumptions) will be the same if you permute the labels for
that effect.  So if you are testing Size, then you can permute
the Size labels, but not permute anything that might have an
effect like Style, Family, and Subject (if I understand your data
properly).

And as Tim rightly pointed out yesterday, you do want to think
about whether or not the permutation has implications that may
not be warranted.  But again if I understand your problem, I
think you should be okay.

Patrick Burns

Burns Statistics
patrick@burns-stat.com
+44 (0)20 8525 0696
http://www.burns-stat.com
(home of S Poetry and "A Guide for the Unwilling S User")

Paul Lasky wrote:

How do I generalize permutation testing?

 Following up on the permutation testing question:

  I have a 3 factors (e.g. Size, Style, and  Family) and each factor has 3
levels
(e.g  Size: large, mid, and small). These are NOT nested in subjects.  Each
subject
is tested at all 3^3=27  levels.  Now how do I do permutation testing  for
this "repeated measures" designed experiment ?
 Is the S+ software available at the "S-Poetry" site  capable of handling
this simple
ANOVA design; does anyone have an idea of how to accomplish the permutation
testing?

 Paul H. Lasky
B & P Consultants



[Original Message]
From: Stahel, Christof <stahel_2@cob.osu.edu>
To: <s-news@lists.biostat.wustl.edu>
Date: 1/21/2004 7:49:47 AM
Subject: [S] testing equality of pairwise correlation across subsamples

Hi all,

I know this is not a specific S+ question, but I thought someone knows
of an easy solution. If I were to test pairwise calculated correlations
over two subsamples, how do I test whether they are equal.

To be more precise, let x = cbind(x1,x2) be a matrix of dim (T,2). I
would like to test if cor(x[1:n,1],x[1:n,2]) =
cor(x[n+1:T,1],x[n+1:T,2]).

Any suggestions are very welcome.
Thanks
Christof


--------------------------------------------------------------------
This message was distributed by s-news@lists.biostat.wustl.edu.  To
unsubscribe send e-mail to s-news-request@lists.biostat.wustl.edu with
the BODY of the message:  unsubscribe s-news



--------------------------------------------------------------------
This message was distributed by s-news@lists.biostat.wustl.edu.  To
unsubscribe send e-mail to s-news-request@lists.biostat.wustl.edu with
the BODY of the message:  unsubscribe s-news






<Prev in Thread] Current Thread [Next in Thread>