There would have to be a pretty large range between the test and the control to make such assumptions. I wouldn't bet the farm on 100 samples of anything.