I've made an experiment with two products used at several dose rates. I did an ANCOVA with the model Y=Dose Treatment Dose*Treatment.
I've found the interaction as significant. So the slopes of the dose responses are different from a treatment to another.
Now I would like to compare the slopes (given in the model) and compute a ratio with a confidence intervalle. I've found something on the net with the Fieller's theorem that I could run in Excel, but I would prefer to have the result by SAS.