Thank you to everyone that reply to my post.
I also suspect, as Joseph pointed out, that the
statement in the article is incorrect regarding the
standard deviation.
Is it possible to compute a sample size if all you
know is the geometric mean and 95%CI.
Specifically if the geometric mean of a control group
is 2.52 (95% CI 2.28, 2.78) and I want to detect a
difference of say 0.5 between the GM of cases and
control, can this be done without additional
information?
Thank you again,
Ricardo.
--- Joseph Coveney <[email protected]> wrote:
> From: Richard Williams wrote:
>
> I think you meant
>
> sampsi 100 110, alpha(0.025) power(0.8) sd1(15)
> sd2(15)
>
> At least, that is what gives me an N of 43; the
> original command gives Ns
> of 20.
>
> [Excerpted]
>
>
----------------------------------------------------------------------------
>
> Yep. Either that or
>
> -sampsi 100 110, alpha(0.0167) power(0.8) sd1(15)
> sd2(15)-,
>
> which gives n = 48 for three tests between pairs of
> the three groups.
>
> Sorry about that, and thanks to Richard for catching
> it. (Talk about
> peppering with errors . . .)
>
> Joseph Coveney
>
>
>
>
>
>
> *
> * 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/
>
_______________________________
Do you Yahoo!?
Declare Yourself - Register online to vote today!
http://vote.yahoo.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/
