I remember from college statistics that blocking can be used, especially in agriculture. I would like to block but use binomial distributed data. Is this possible? For example: I have 100 embryos. I want to treat half of them one way and half another way and then freeze them. Upon thawing I want to use two different methods and then use blocking to determine any differences in the two treatments. (2 X 2) I seem to remember doing this with ANOVA with continuous data, but what about binomial data. My results will be percent to develop after 4 days of culture.

I know I could do the first experiment and then do the second and analyze each with a Fisher's Exact, but I would like to use less embryos by blocking.

Sorry if my terminology is a little off. It has been a while since I have been involved in statistics. Thanks!