I have been reading the article entitled, "Some Statistical Techniques for Bio-Efficacy Trials" (http://www.iasri.res.in/ebook/ebadat/6-Other%20Useful%20Techniques/6-Statistical%20Methods%20in%20Bioefficicacy%20Trials.pdf).

I am conducting a mortality trial with a non-uniform population. It appears that the Sun-Shepherd formula is best suited for corrected efficacy. I would be very grateful for feedback concerning the controls:

1. As we are normalizing to the controls, do I normalize the controls with the formula? Or do I leave the controls as is when conducting statistical comparisons? 2. When determining the value of the controls do I use the average of the controls or do I randomly select a control value?