You can use -sampsi-, but you have to remember that the inputs must be
proportions, not odds or odds ratios. So first you have to convert your
odds ratio question to the relevant proportions. In your case, your
exposure prevalence is 13%. Suppose you wanted to detect a 3-fold
increase in the odds. The proportion which corresponds to "3-fold
increase in the odds" is .31 (notice that it is not just .13 x 3).
Then your -sampsi- command is:
. sampsi .13 .31, n1(139) n2(83)
The power for the above command is about 85%.
Be sure when you run this command to associate your sample
specifications, e.g., n1() and n2() with the appropriate proportion:
n1() refers the first proportion entered in the command and n2() refers
to the second one.
-p
______________________________________
Paul F. Visintainer, PhD
Department of Epidemiology and Biostatistics
School of Public Health
New York Medical College
PH: (914) 594-4804
FX: (914) 594-4853
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of stata_user
stata_user
Sent: Tuesday, January 30, 2007 4:41 PM
To: [email protected]
Subject: RE: st: RE: power calculation in a case control design
Thanks a lot for all the suggestions and help.
I could not answer the emails earlier, sorry about that.
The prevalence of exposure in the control group is
19 or 13%
It is a difficult situation compared to the situation where we have to
test
equal means or equal proportions
I will appreciate any further suggestions
Sami
>From: "Visintainer, Paul" <[email protected]>
>Reply-To: [email protected]
>To: <[email protected]>
>Subject: st: RE: power calculation in a case control design
>Date: Tue, 30 Jan 2007 11:06:45 -0500
>
>What's the (assumed) prevalence of exposure in your control group?
>
>______________________________________
>Paul F. Visintainer, PhD
>Professor and Chair
>Department of Epidemiology and Biostatistics
>School of Public Health
>New York Medical College
>PH: (914) 594-4804
>FX: (914) 594-4853
>
>-----Original Message-----
>From: [email protected]
>[mailto:[email protected]] On Behalf Of stata_user
>stata_user
>Sent: Friday, January 26, 2007 4:58 PM
>To: [email protected]
>Subject: st: power calculation in a case control design
>
>Dear All,
>
>
>
>I have a case control design with 83 cases and 139 controls.
>
> I am intersted in the power of two sided 5% (or 1 sided) test to
>detect
>an approximate relative risk of less than 4.
>
>Is there any possibility to have a solution using stata?
>
>Thanks a lot
>Sami
>
>_________________________________________________________________
>Get Hotmail, News, Sport and Entertainment from MSN on your mobile.
>http://www.msn.txt4content.com/
>
>*
>* 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/
_________________________________________________________________
Get Hotmail, News, Sport and Entertainment from MSN on your mobile.
http://www.msn.txt4content.com/
*
* 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/
