StevenPro":p4n3fvob said:
Thanks for the offer! I know my previous test was limited and a statistical analysis wasnot possible, but I only have 12 tanks. If I dropped down to two treatments along with the positive and negative controls, that would give me three tanks per group.
By the way, does increasing the number of Xenia fragments per tank help? I am guessing no.
Well you have to drop treatments out or increase tank numbers, only 10 bucks for a 10 gallon tank
If you set up a nested ANOVA then you could look at more xenia fragments per tank. This would allow you to use two tanks per treatment, with at least 3 frags per tank (use 5). We have to look at your question; the question is what is the medication doing to the corals, given that your tanks are exactly the same, which they never are, but they are going to be close. The only thing that would cause varitation would be your Xenia.
Much of stats can be made in to smoke and mirrors, the absolute best way to test the data would be to have 5 smaller tanks with one xenia frag per tank, however we are dealing with the real world and not the world of stats. If you only have so many tanks you have two choices, reduce the treatments or nest the data. Nesting is vaild when you are looking at survival. Some might argue that its not the best way but its a vaild test.
They killer would be if you have a difference between tanks in the same treatment, if your tanks had differneces then you would be SOL. I am betting that you are going to get good results with nested data.
andy