On Mon, 1 Oct 2001 12:23:28 +0200, "Bernhard Kuster" <[EMAIL PROTECTED]> wrote:
> Hi > > I am interessted in the question of optimal sample size in general, not for > a special statistical technique. (a) There was a notable 1974 article on "Believability when N=1" and here is an academic webpage on the subject. It includes good references. http://www-personal.umich.edu/~hinderer/scrdoutline.html (b) Richard Peto's writing may have been the impetus behind one or two "mega-studies." He pointed out that by having a clear-cut, randomized treatment ("treatment <aspirin?> given within x hours of a heart attack") and an unambiguous outcome ("still alive after 30 days"), it should be possible to combine the experience of hundreds of hospitals, and 10,000's of patients. And it was done. I think that the randomization was in there, and was possible because no one had great faith in the treatment. The huge N was a *necessary* sample size because of the small size of the expected effect as an odds ratio, and the small fraction of people who would be mortalities-who-might-be-saved. There you have the extremes of what is optimum for "sample size in general, not for a special technique." > > My questions: (1) What do I have to keep in mind if I compute optimal sample > size, what is relevant? (2) What are the classic studies and who has highly > influenced the subject? (3) What are the problems discussed right now by the > scientific community? (4) What are the relevant journals and is there some > information on the web? I think you might find some issues included in Robert P. Abelson, "Statistics as principled argument." I think I am trying to say, gently, that your basic question doesn't make very good sense to me; and it did not, to Dennis, either. "Optimal" is one problematic word. Another problem is that you seem to ask about all research, in all of the world.... It might be a clever way to attack 'sample size', but I think that hasn't been done. -- Rich Ulrich, [EMAIL PROTECTED] http://www.pitt.edu/~wpilib/index.html ================================================================= Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =================================================================
