The strongest evidence is that a priori there is not reason to expect populations that have historically been geographically separate to have the same distribution of IQ.
Generally, one never expects two imperfectly related continuous variables to be exactly identical. This tells us nothing about how peaked about 0 our prior distribution should be. In other words, the existence of a difference is no guarantee of existence of a significant difference.
Generally, one never expects two imperfectly related continuous variables to be exactly identical. This tells us nothing about how peaked about 0 our prior distribution should be. In other words, the existence of a difference is no guarantee of existence of a significant difference.