Calvert et al. (2003) put preadolescent children into a MUD (online game) with other children they did not know, and studied their interactions. The children's gender was apparent by their character in the game, so while the children did not know each other, they knew the gender of the other child they were interacting with. Girl-girl pairs tended to interact through written dialogue, while boy-boy pairs interacted more through action. In mixed pairs, girls increased their actions and boys increased the amount they wrote. Calvert et al conclude that in late childhood, children begin to moderate their behavior patterns in interactions with the opposite sex.
This split between action and verbal communication, if valid, suggests that males would tend to dominate an action-oriented environment such as online games, while females would tend to dominate a verbal-oriented environment such as blogs. Online gaming surveys seem to confirm the first part of this assumption;
Griffiths et al (2003) found that 85% of online gamers were male and
Project Massive's 2004 survey reported 90% males. However,
Huffaker and Calvert (2005) cite several studies that find that bloggers are evenly split between males and females, although their sampling method provided nearly twice as many female subjects as male. It is suggested that the "teen" group of bloggers is skewed female. Huffaker and Calvert also found high rates of expressed homosexuality among teen male bloggers (14%) but not among teen female bloggers (3%). Sexuality demographics for online gamers was not available.
The only online game I've ever really taken big part in was
Puzzle Pirates, which
BBC news called "highly successful" with women. However, I've noticed even on this friendly game that outing yourself as queer is usually a poor idea. In an online game, you have very little control over the feedback you get based on the information you provide. In a blog, you can disable or moderate comments, picking and choosing what reactions you'll receive. I think that the ability to exercise control over feedback may be an important part of the online/gender puzzle, but I'm not sure how it fits in.