The way to stratify and cluster on the same variable is to use the
option -funtype(wcluster)- in addition to the -cluster(id)-.
The option -cluster(id)- tells -somersd- that we are sampling clusters
from a population of clusters, rather than sampling observations from a
population of observations. The option -funtype()- specifies whether we
are comparing within clusters (indicated by -funtype(wcluster)-),
between clusters (indicated by -funtype(bcluster)-), or both (indicated
by -funtype(vonmises)-). The default of the -funtype()- option is
-funtype(bcluster)-. So, if you try to specify -wstrata(id) cluster(id)-
without specifying -funtype(wcluster)-, then -somersd- assumes that you
are asking to restrict yourself to comparisons both between AND within
IDs, which of course is an empty set of comparisons.
The -funtype()- option is short for "functional type". Somers' D is a
ratio of means if you specify -funtype(wcluster)-, a ratio of Hoeffding
functionals or U-statistics if you specify -funtype(bcluster)-, and a
ratio of Von Mises functionals if you specify -funtype(vonmises)-. See
Newson (2006) for further references.
I hope this helps.
Best wishes
Roger
References
Newson R. Confidence intervals for rank statistics: Somers' D and
extensions. The Stata Journal 2006; 6(3): 309-334. Download
pre-publication draft from
http://www.imperial.ac.uk/nhli/r.newson/papers.htm
or buy a proper version from
http://www.stata-journal.com/article.html?article=snp15_6
Roger B Newson
Lecturer in Medical Statistics
Respiratory Epidemiology and Public Health Group
National Heart and Lung Institute
Imperial College London
Royal Brompton campus
Room 33, Emmanuel Kaye Building
1B Manresa Road
London SW3 6LR
UNITED KINGDOM
Tel: +44 (0)20 7352 8121 ext 3381
Fax: +44 (0)20 7351 8322
Email: [email protected]
Web page: www.imperial.ac.uk/nhli/r.newson/
Departmental Web page:
http://www1.imperial.ac.uk/medicine/about/divisions/nhli/respiration/pop
genetics/reph/
Opinions expressed are those of the author, not of the institution.
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Feiveson,
Alan H. (JSC-SK311)
Sent: 09 May 2008 17:11
To: [email protected]
Subject: st: somersd: clustering with wstrata ?
Hi - I wish to compare two variables within repeated observations on
subjects using -somersd-. If I use the "wstrata" option, would the
standard error also be adjusted to take care of the non-independent
observations pertaining to the same subject? Apparently I can't use
"cluster" and "wstrata" on the same variable. I don't think I want to
use "cluster" alone if I want the comparisons to be within each subject.
Al Feiveson
. somersd z6a age,wstrata(id)
Symmetric 95% CI
------------------------------------------------------------------------
------
| Jackknife
z6a | Coef. Std. Err. z P>|z| [95% Conf.
Interval]
-------------+----------------------------------------------------------
------
age | .3608484 .0250542 14.40 0.000 .311743
.4099537
. somersd z6a age,wstrata(id) cluster(id)
Symmetric 95% CI
(Std. Err. adjusted for 403 clusters
in id)
------------------------------------------------------------------------
------
| Jackknife
z6a | Coef. Std. Err. z P>|z| [95% Conf.
Interval]
-------------+----------------------------------------------------------
------
age | (dropped)
------------------------------------------------------------------------
------
. somersd z6a age,cluster(id)
Symmetric 95% CI
(Std. Err. adjusted for 403 clusters
in id)
------------------------------------------------------------------------
------
| Jackknife
z6a | Coef. Std. Err. z P>|z| [95% Conf.
Interval]
-------------+----------------------------------------------------------
------
age | .4528894 .0239833 18.88 0.000 .4058829
.4998958
------------------------------------------------------------------------
------
*
* 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/
*
* 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/
Privileged, confidential or patient identifiable information may be contained in this message. This information is meant only for the use of the intended recipients. If you are not the intended recipient, or if the message has been addressed to you in error, do not read, disclose, reproduce, distribute, disseminate or otherwise use this transmission. Instead, please notify the sender by reply e-mail, and then destroy all copies of the message and any attachments.
*
* 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/
