Well, yes, you can control for gender bias, but then you're studying a completely different domain. You're selling short the unique complexities of both heterosexual and homosexual courtship.