> The null hypothesis has to be that both approaches are equally effective. There simply isn't any evidence to suggest otherwise. Sure, I think that is in the

> The null hypothesis has to be that both approaches are equally effective. There simply isn't any evidence to suggest otherwise. Sure, I think that is in the ballpark of an acceptable null hypothesis, for a given experiment. I'm not sure one can identify a null hypothesis that is universal among all experiments, but this captures the idea, sure.