Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: st: power calculations


From   "Edgar Munoz" <munozedg@hotmail.com>
To   <statalist@hsphsun2.harvard.edu>
Subject   RE: st: power calculations
Date   Fri, 5 Feb 2010 11:15:55 -0600

Also this could be useful...

QUANTO - Version 1.2 - January 2007 - Developed by:
Jim Gauderman, Ph.D. and John Morrison, M.S.
Department of Preventive Medicine
University of Southern California
e-mail: jimg@usc.edu, jmorr@usc.edu
Updates and Information: http://hydra.usc.edu/gxe

Thanks. Edgar

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Neil Shephard
Sent: Friday, February 05, 2010 3:32 AM
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: power calculations

On Thu, Feb 4, 2010 at 8:38 PM, louise hornstrup
<louisehornstrup@hotmail.com> wrote:
> Dear listers,
>
> I am doing two diffent studies, using Stata 11. The first is a
cross-sectional study (n=42,298), the second a case-control study
(n1=4,851/n2=4,851).
>
> I am looking at a specific genotype with a frequency of 0.6% in both
studies (variable of interest), the endpoint is IHD (Ischemic Heart
Disease).  I have some difficulties with the power calculations. I want to
estimate the minimal value (odds ratio) that can be detected with 80% power
for both studies...But what is the right way to do this in Stata? As far as
I can tell the sampsi command can+determine power or sample size - not
minimal detectable effect size?
>

Its not a Stata solution but you may also find the excellent Genetic
Power Calculator by Shaun Purcell and colleagues of use...

Purcell S, Cherny SS, Sham PC. (2003) Genetic Power Calculator:
design of linkage and association genetic mapping studies of complex
traits. Bioinformatics, 19(1):149-150.


http://pngu.mgh.harvard.edu/~purcell/gpc/

You can of course script (in Perl, Python [your choice of scripting
language]) repeated queries to the server for a range of parameters to
give a better view of the power of your sample than a single point
estimate.

Although of course knowing the power your sample has to detect
different effect sizes isn't particularly enlightening and is kind of
back to front as you should estimate your effect size from published
studies/pilot studies and recruit an appropriately sized cohort to
detect the estimated effect.  If you've already done your study and
not found any association calculating the effect size your sample has
sufficient power to detect is not a good idea. A couple of references
on this are below, but there are more out there.


Hoenig J.M., Heisey D.M. (2001) The Abuse of Power: The Pervasive
Fallacy of Power Calculations for Data Analysis. The American
Statistician 55:19-24


Levine M, Ensom MH (2001) Post hoc power analysis: an idea whose time
has passed? Pharmacotherapy 21.4:405-409

Neil


-- 
"... no scientific worker has a fixed level of significance at which
from year to year, and in all circumstances, he rejects hypotheses; he
rather gives his mind to each particular case in the light of his
evidence and his ideas." - Sir Ronald A. Fisher (1956)

Email - nshephard@gmail.com
Website - http://slack.ser.man.ac.uk/
Photos - http://www.flickr.com/photos/slackline/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index