Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: statsby?

From   Roger Newson <[email protected]>
To   [email protected]
Subject   st: RE: statsby?
Date   Tue, 03 May 2005 18:19:11 +0100

An alternative possibility is to use -parmest- together with -somersd- to produce an output dataset (or resultsset) with 1 observation per Kendall's tau-a, which you can then list and/or plot and/or tabulate. This method has the advantage of producing confidence limits as well as P-values for Kendall's tau-a, and also the option of using a normalising and/or variance stabilising transformation, such as the arcsine or the hyperbolic arctangent. The -somersd- and -parmest- packages are both downloadable from SSC.

To find out more about confidence intervals for Kendall's tau-a and other rank statistics, type in Stata

findit params

and you will be signposted to my website, where you can download a preprint of Newson (2002). To find out more about -parmest- and resultssets in general, type

findit resultsset

and you will be signposted to my website, where you can download a copy of my Stata User Meeting presentation on resultssets.

I hope this helps.



Newson R. 2002. Parameters behind "nonparametric" statistics: Kendall's tau, Somers' D and median differences. The Stata Journal 2(1): 45-64.

At 16:43 03/05/2005, you wrote:

Dear Nick

Your solution is exactly what I needed.
A thousend thanks,

Nick Winter wrote:
In Stata 8, this should do the trick:
. statsby "ktau x y " TauB=r(tau_b) P=r(p) , by(z) saving(myfile)
In Stata 9, it would be:
statsby TauB=r(tau_b) P=r(p) , by(z) saving(myfile) : ktau x y

--Nick Winter
At 04:49 PM 5/3/2005 +0200, you wrote:

With STATA 8, I tried

_statsby "ktau x y" exp_list, by(v w z) saving(myfile)_

where for exp_list I tried _se and _b (I have to admit that the grammar of exp_list remains a riddle to me).

The error produced is r(301): last estimates not found.

Regards, Sascha

Nick Cox wrote:

In principle, -statsby- sounds a good solution.
However, we cannot comment on what went wrong without seeing your code. Also, the syntax of -statsby- differs between Stata 8 and Stata 9, so you would need to specify version as well.
Nick [email protected]
Sascha Robert

I have to do a bunch of Kendall's tau correlations. Now the output of

_ktau x y, by(z)_

is quite exaustive, and I am mostly interested in
- Kendall's tau-a
- Prob > �z�

Is there a way to direclty produce a table of the results, like

tau-a prob ...
a b ...
c d ...,

instead of producing text output?

Is _statsby_ a solution for my problem? It does not seem to work for my puposes, though.

*   For searches and help try:

***** sascha robert | mailto: [email protected] *****
*   For searches and help try:
Nicholas Winter                           607.255.8819 t
Assistant Professor                       607.255.4530 f
Department of Government              [email protected] e
308 White Hall   w
Cornell University
Ithaca, NY 14853-4601

*   For searches and help try:

***** sascha robert | mailto: [email protected] *****
*   For searches and help try:

Roger Newson
Lecturer in Medical Statistics
Department of Public Health Sciences
Division of Asthma, Allergy and Lung Biology
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: [email protected]

Opinions expressed are those of the author, not the institution.

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index