Hi all, I have 11 raters that I am comparing to a gold standard test.

I would like to run a test that randomly selects 5 (also 7 and 9) raters for several iterations and gives me the sensitivity and specificity that I could expect, at each cutoff. i.e. what is the accuracy of 5 random raters with 1, 2, 3, 4 and 5 raters correct?

I only have SPSS to work with.

Perhaps it is very simple and I'm missing something obvious?