On 12 Nov 2003 04:10:50 -0800, [EMAIL PROTECTED] (Cristian) wrote: > Hi, > > I'm performing a statistical analysis of diagnostic test data. I want > to calculate the 95% CI of the sensitivity and specificity for my > individual studies according to Diamond's formula, I have read that > this formula provides a very good aproximation to the exact method. > > The question is that I don't know how to calculate the CI using > Diamond's formula. Any idea???
I think that I have never heard of Diamond's formula. I don't find anything by googling. Can you give a reference? Zar's textbook, as the first one I looked in, provides the exact limits for a binomial proportion by making use of the F-distribution. Is Diamond trying to make some further assumptions about sensitivity and specificity, or is that an approximation that should be particularly easy to use? -- Rich Ulrich, [EMAIL PROTECTED] http://www.pitt.edu/~wpilib/index.html "Taxes are the price we pay for civilization." . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
