We present clean experimental evidence that a methodological confound was introduced by Andreoni and Miller (2002) that leads to diametrically opposed conclusions regarding comparisons of preferences between categories of fellow human beings distinguished by gender or age. Our study is a warning...