(My apologies if you've received this before. I was somehow unsubscribed from Edstat without my knowledge and had to resubscribe, with some of my user options changing in the process.)

I have what seems to be a fairly simple and common problem: given an estimate of the probability of occurrence of an event and a desired level of confidence (CL), I'm interested in estimating the minimum sample size (Nmin) needed to be CL% confident of observing the event (based on the conventional "frequentist" concept of confidence). I've done some extensive simulations as so can predict what I need, down to probabilities of occurrence of 1e-5. However, I can't seem to match these to theory. I know how to estimate the Nmin needed to estimate a proportion within a specified tolerance, but can't see how to apply that to this problem. I've scanned quite a few stats texts and have done Google and Scirus searches but can't locate what I need. Can anyone suggest a lead or something to read?

Rich Strauss


============================================================= Dr. Richard E. Strauss (806) 742-2719 Biological Sciences (806) 742-2963 Fax Texas Tech University [EMAIL PROTECTED] Lubbock, TX 79409-3131 <http://www.biol.ttu.edu/Strauss/Strauss.html> =============================================================

.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to