I don't think you understood what you're responding to as it was not talking about grammatical gender, which is why it mentioned the use of statistical methods. The claim was that men and women do write in different dialects although they are less pronounced than in other languages.