On 11 Jul 2003 07:15:38 -0700, [EMAIL PROTECTED] (Jason Owen) wrote: > Let me get this straight -- you want to create a confidence interval > for a mean that you *know* is equal to 53,000? > That is a very good point. Logically, I have a great deal trouble getting past it....
> Even so, your formula needs a finite population size correction factor. > That seems to be true, but in a very peculiar way. What is the CI showing confidence about, once the exact mean is known? - the variance/ inaccuracy of the sampling method? There is already a claim for finite and known variance. I think I conclude that the problem has claims that are self-contracting. That answer certainly would not have a CI with width based on N= 120, where 120 was never sampled.... Nor should the CI exclude the KNOWN mean. [ snip, problem details .] -- Rich Ulrich, [EMAIL PROTECTED] http://www.pitt.edu/~wpilib/index.html "Taxes are the price we pay for civilization." Justice Holmes. . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
