Comment Grouping Versus non-grouping (Score 0) 981
Is it just me, or isn't it obvious why the probability changes? All you are doing is adding more parameters to what you are trying to actually determine. The most basic question you can do (not in this question, but overall) is simply: What is the odds of a child being a girl or boy. That's theoretically 50/50.
Now, each other question is subtly altering the grouping of what you are trying to determine the probability. It's just that english (or probably any language) obscures the fact that you are making those other elements important. The real trick of these is the fact that language is vague, and most people are missing that you are making necessary additions to the group you are trying to determine information about. You are in fact not just asking about the sex of the other child, but you are building a situation that makes the information about the date, or the order of the children, important. If the importance of those other elements disappears, then they are not bounding boxes to the situation and the results fall back to 50/50.