How many studies can we find where the same data was used to come to completely different conclusions? More than you might think.