Is Weinberg's differential rule valid?

Weinberg's differential method, whereby the frequency of dizygotic twins in a population is obtained by doubling the number of opposite-sexed pairs, might not be valid. Data are presented indicating a higher frequency of same-sexed than opposite-sexed pairs among samples of dizygotic twins ascertained at birth, thus favoring the hypothesis that, contrary to the method's fundamental assumption, the sexes of the two twin zygotes be not independent.