Birth order and gender prediction - a totally unscientific survey

When I was pregnant with A and S, my mother said to me, "I think they'll be girls.  If you look at your friends you'll see that the second baby tends to be the same as the first.  After that, anything goes, but the first and the second are normally the same."

Now, to be honest, I thought that was a load of nonsense.  But then I looked around.  I'm a girl, my sister's a girl, then there's my brother.  B's a boy, then there are two more boys.  My mum and her sister are girls.  My dad and his brothers are boys.  I've got L and then A and S.  My brother and sister-in-law have girls.  My colleagues fitted the pattern, my uni friends do, and so do most of the new friends I'm making up here. It goes on and on.

So I find myself wondering whether it's true.  So I've done what every committed social science and scientific researcher does, and I've googled it.  And I haven't, after an intensive five minutes, found any proper research.  So I thought I'd do some improper research instead.

How about you?  Does your family (parents, siblings, your own children) follow this theory?  Or is it, as I first thought, yet another one of those wonderful myths about gender prediction (gold rings, round bumps, speedy heartbeats) that we really should have stopped repeating in the fifteenth century.  Write a comment and let me know, and I'll do a whizzy spreadsheet.... oh, and please do pass it on, the more info I have the more "reliable" the survey will be!