I am new to statisics, but as the only MS access programmer in our startup company; have been asked to calculate the sample size for a study. Since VC funding dried up there is no money to retain a statistician. Any help would be greatly appreciated.
== Background ==
1) We have a Testing-device (T) designed to evaluate if an individual fallopian tube in a Rabbit; is Open or Closed. We are comparing the same tube to the Gold Standard test (G), to confirm if the same tube is actually Open or Closed.
2) From one past study we have observed that of <10> fallopian tubes evaluated, <9 times when we said the tube was Open, the gold standard confirmed it was Open. There was 1 time when the Testing device said the tube was Open, the gold standard said it was Closed>
== Question ==
1) How many rabbits do we need so that we can say: <Every time our testing device says a tube is Open, it is really Open>. That way Vets will only need to do the Gold Standard test, when our device says a tube is <Closed>.
2) How would the number of rabbits change if wanted to make this statement with 90%, 95%, 98% confidence.
3) Could you please let me how you go about calculating this, what would be the appropriate statistical test to use, and how would you phrase the statement?
Thanks so much.