st: RE: Signtest (statistics question), thanks and a further question

["Guillaume Frechette" <gurst1@hotmail.com>]

Thanks to the three guys that responded. Nick Cox wrote:

> -ranksum- works quite well in this situation.
My understanding (which I think is confirmed by Roger Newson's answer -- see 
below) was that since I have paired observations Ranksum would be 
inapropriate.

Roger, should I understand from your answer that there are tests taking into 
consideration what I am talking about but that 1) I have to read your thesis 
to learn about it, and 2) that it wouldn't be trivial to code? We can move 
this discussion to direct email if that is more appropriate.

Best,

g


> As it happens, the sign test tests for a zero value for the difference 
> between 2 conditional probabilities
>
> Pr(x1>x2 | x1!=x2) - Pr(x1<x2 | x1!=x2)
>
> which is, in a way, a special case of Somers' D, which is also a difference 
> between 2 conditional probabilities. The Somers' D calculated by my 
> -somersd- package (downloadable from SSC) is like the two-sample t-test, 
> whereas the difference tested by the sign test is like the paired t-test.
>
> The difference between the 2 corresponding unconditional probabilities
>
> Pr(x1>x2) - Pr(x1<x2)
>
> is similarly a paired version of Kendall's tau-a.
>
> More on Somers' D, Kendall's tau-a and median differences can be found in 
> my Stata Journal insert on the subject (Newson, 2002).
>
> As it happens, I did some work on paired (or more general within-group) 
> versions of Kendall's tau-a in my DPhil thesis (Newson, 1987). I have 
> always wanted to generalise the -somersd- package to the within-group case, 
> and to the case of survival analysis, but have always thought that I would 
> need DLLs to do this properly. Not surprisingly, I am looking forward to 
> hearing what Bobby Gutierrez and Nguyen Chinh have to say on DLLs at the 
> 9th UK Stata User Meeting.
>
> Best wishes
>
> Roger
>
> References
>
> Newson RB. An analysis of cinematographic cell division data using 
> U-statistics. DPhil dissertation. Brighton, UK: Sussex University; 1987.
>
> Newson R. Parameters behind "nonparametric" statistics: Kendall's tau, 
> Somers' D and median differences. The Stata Journal 2002; 2(1): 45-64.
>
>
> --
> Roger Newson
> Lecturer in Medical Statistics
> Department of Public Health Sciences
> King's College London
> 5th Floor, Capital House
> 42 Weston Street
> London SE1 3QD
> United Kingdom
>
> Tel: 020 7848 6648 International +44 20 7848 6648
> Fax: 020 7848 6620 International +44 20 7848 6620
>   or 020 7848 6605 International +44 20 7848 6605
> Email: roger.newson@kcl.ac.uk
>
> Opinions expressed are those of the author, not the institution.
>
> *
> *   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/
_________________________________________________________________
Add photos to your messages with MSN 8. Get 2 months FREE*.  
http://join.msn.com/?page=features/featuredemail

*
*   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/