Reply #6: that's the problem with studies where you can only rely on correlational

[liontamer]

data. You can probably find lots of statistically significant correlations in their raw data. An interesting study would be to take two groups of pregnant women and give them the drugs randomly, and then see if their kids are gay. Of course that's one study that will never get funded.
And even an animal model of the study would be difficult, because it would be difficult to deduce how an animal identified.