Well, that's the thing. You can't even prove that any factor has any effect at all. All I know is that from my anecdotal experience a lot of my male peers got into programming tangentially through gaming because they liked playing games and wanted to learn how to make games. And there's a fair amount of proof that young boys are more likely to play video games than young girls.
As for whether or not it'll rectify itself, I'm a little more skeptical. Girls are more likely to be casual gamers, and casual gamers are less likely to like games enough to want to build their own. Of course, this could easily be my projection bias showing, as I am a casual gamer who has never felt a strong passion to build my own game.
Right, and again, my case shows a difference even there: I went into software. But my brother went into carpentry and animal husbandry. Same family, same environment -- gaming, coding with dad, etc. He's just as smart and could've done tech if he wanted, but he had different interests.
That's an interesting point on casual gaming. However, it's also balanced by the fact that casual games are the only games you can easily create as a one man band or hobbyist.
If you're into puzzle games, say, you can probably put together your own with a bit of HTML/JS. Or use one of the puzzle game builders.
If you're into Call of Duty, though, you've got no chance of writing your own version of that until you've graduated and gotten a job somewhere. Even the ubiquitous modding community that used to be a path into development has diminished as games are increasingly locked down.
As for whether or not it'll rectify itself, I'm a little more skeptical. Girls are more likely to be casual gamers, and casual gamers are less likely to like games enough to want to build their own. Of course, this could easily be my projection bias showing, as I am a casual gamer who has never felt a strong passion to build my own game.