Conceptually, statistical significance is baffling, almost backward: it tells us the chance of observing the data given a particular theory, the theory that there is no effect. Really, we’d like to know the opposite, the probability of a particular theory being true, given the data.
