I want to compare 4 different methods of bladder cancer induction and I have to calculate sample size. How many animals I have to used to get appropriate statistical power?
What is the variability in the response of your animals to the drug you will apply? How accurate do you need to be? If drug A induces bladder cancer in 15% of animals and drug B induces bladder cancer in 32% of animals is this difference large enough to matter? What if I change the values to 15% and 15.02%? What is the expected success rate? The answer will be very different if only 1 in 300 animals get the right cancer versus 9 animals in 10.
It is difficult to tell. You need to know the variability in the response to the drug. I would suggest to perform some tests initially to urothelial cell cultures prior to animals.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size