People with Exactly Two Children and At Least One Daughter

[bup]

Do you have two daughters, or one son and one daughter?

Please only answer the poll if you have exactly two children, and at least one is a daughter. Thanks.

[Jack Batty]

I'm kinda wondering why you want to know this.

[bup]

Originally Posted by Jack Batty

I'm kinda wondering why you want to know this.

This thread.

[garygnu]

Can I answer on behalf of any of my cousins' families? All three have one boy and one girl.

[Skammer]

5 year old boy, 3 year old girl here.

If you want a bigger sampling, couldn't you just ask people to answer based on their oldest two kids? Pretending they never had any more after that.

[Voyager]

Two children, both girls.

[bup]

Originally Posted by Skammer

5 year old boy, 3 year old girl here.

If you want a bigger sampling, couldn't you just ask people to answer based on their oldest two kids? Pretending they never had any more after that.

I don't want to, in case there's some argument to be made that people with two older daughters, or one older daughter and one older son, would be more likely to vote- some weird variation on the illusory correlation problem. I don't know which way it would go, but no need to muddy the waters.

Originally Posted by garygnu

Can I answer on behalf of any of my cousins' families? All three have one boy and one girl.

Let's keep it simple. Besides, people only get one vote.

[Skammer]

With a sufficient sample, you sould find that 67% will answer one of each. (Right now it's 100%).

[ZipperJJ]

Can I answer for my parents? They've got two kids, and I'm their only daughter.

I've got no kids.

[Jack Batty]

I couldn't make heads nor tails out of the linked thread (and please, I'm begging you, don't try to explain it to me).

My boy and girl are fraternal twins if that has any bearing to the results.

[ToeJam]

What if I don't have kids but my Parents had 2 kids (including me), do you want me to answer for them?

[bup]

You know, I understand it shouldn't affect the results, but since the question this is based on says somebody has two kids, let's make all respondents live by the same rules. Otherwise, I could point out that I have two kids, but I am also one of exactly two children, and I could answer for that if I didn't have my own children, etc.

I want to KISS it, so that doubters have the least ammunition possible to explain why reality is wrong.

In other words, I'm politely requesting that people only answer for their own kids, if they have exactly two with at least one daughter. Everyone else sit on your hands.

[corkboard]

Two girls.

[Hedda Rosa]

one of each

[Giles]

Originally Posted by Jack Batty

My boy and girl are fraternal twins if that has any bearing to the results.

The only situation that would skew the results (apart from infanticide or abortion based on the child's gender) would be a family with just a pair of identical twins, since that makes BB or GG slightly more likely that BG or GB. But since identical twins are relatively rare, I don't think you're going to see an affect on the result.

And I didn't vote, because I have 4 children. (BBGB for those interested).

[BigT]

Originally Posted by bup

You know, I understand it shouldn't affect the results, but since the question this is based on says somebody has two kids, let's make all respondents live by the same rules. Otherwise, I could point out that I have two kids, but I am also one of exactly two children, and I could answer for that if I didn't have my own children, etc.

I want to KISS it, so that doubters have the least ammunition possible to explain why reality is wrong.

In other words, I'm politely requesting that people only answer for their own kids, if they have exactly two with at least one daughter. Everyone else sit on your hands.

I suspect your problem will be that the sample size will be rather small, then. First you're going to eliminate the child-free/childless. Then you eliminate anyone who has only one child. Then you eliminate anyone who has more than one child. Finally, you eliminate people who don't have a daughter

[Fuzzy Dunlop]

Originally Posted by BigT

I suspect your problem will be that the sample size will be rather small, then. First you're going to eliminate the child-free/childless. Then you eliminate anyone who has only one child. Then you eliminate anyone who has more than one child. Finally, you eliminate people who don't have a daughter

30 responses so far assuming they're all valid. Seems pretty good.

[bup]

Originally Posted by BigT

I suspect your problem will be that the sample size will be rather small, then. First you're going to eliminate the child-free/childless. Then you eliminate anyone who has only one child. Then you eliminate anyone who has more than one child. Finally, you eliminate people who don't have a daughter

Depends. If all I want to do is reject the null hypothesis (that the true split should be 50/50) I think I've already got enough (haven't run the numbers - will tonight).

If I want to say the real split should be within X% of 66.666% to 33.333%, well, I might need more.

[Giles]

It's currently about 35.5% to 64.5%, which is much closer to the 33.3/66.7 split than it is to the 50/50 split.

[Skammer]

It's currently 11-22. Math wins!

[ToeJam]

Originally Posted by bup

In other words, I'm politely requesting that people only answer for their own kids, if they have exactly two with at least one daughter. Everyone else sit on your hands.

Damn, forgiveness is easier than permission I suppose.
Scratch one of the "Yes" then. I answered for my parents.

[Greg Charles]

I have no children, but I voted on behalf of my brother who has two daughters.

[ToeJam]

Originally Posted by bup

Depends. If all I want to do is reject the null hypothesis (that the true split should be 50/50) I think I've already got enough (haven't run the numbers - will tonight).

I think (based on what I'm learning in Stats) that one should have an N of 100 to have a nice good sample size but I have no idea if that's really true or not, and I'm doubting it.

Also, I think you should also have a 2:1 ratio in this test of yours.

The 4 TRUE options should have been:

MM/MF/FM/FF (and thus getting 2:1:1 as your possible answers). It would only be 50% if you compared it as "Same Sex" vs. "Not-same sex" with each having a 50% ratio, at least that's the case if we were doing punnet squares and vastly simplifying the inheritance rations and such to 50% with 2 variables of the ration (XX*XY).

I doubt it's that easy and realistic in the real world.
But hey, that's a nice AP Bio question right there if you simplified it down that way!

[bup]

Originally Posted by ToeJam

I think (based on what I'm learning in Stats) that one should have an N of 100 to have a nice good sample size but I have no idea if that's really true or not, and I'm doubting it.

Meh. I think you're thinking of 30, at least for a t-distribution.

Also, I think you should also have a 2:1 ratio in this test of yours.

Of course. But if you look at the thread linked in post 3, there are those out there who believe the answer should be 50/50. I thought this poll might convince doubters that, even if they don't believe it's 1:2, at least it's not 1:1.

Rejecting a null hypothesis that it is 1:1 is easy, using the t-distribution. With the present breakout of 14:27, we get 65.9% that the other child is a son, and a standard error with 41 observations of 7.4%. So even right now we can say with 95% confidence that the real breakout in the full population is somewhere between 51% and 81%.

In other words, it ain't 50/50.

This does assume that the people who answered the poll have no expected bias from the population at large that has two kids, one of whom is a daughter.

[ToeJam]

fair enough.

[Rushgeekgirl]

I have only two children, both girls, but I've been pregnant four times in my life if that matters.

[Edward The Head]

I've got two little girls, I'm surrounded by estrogen.

[Jack Batty]

Oh yeah ... to further muddy the waters ... my boy and girl, the fraternal twins ... they have a half sister and half brother too.

[Ann Onimous]

Two children: one boy, one girl.

[bup]

Updating the stats - having the other child a boy is now at 63.4%, with a standard error of 5.7%.

[Leaffan]

One of each for me.

[L. G. Butts, Ph.D.]

Two girls.

It's good to see that this poll matches my math!

[JSexton]

Both my son and my daughter love counterintuitive things.

[Lord Mondegreen]

I have two children, both daughters.

And I love them with all my being.

[PunditLisa]

Two children, both daughters.

[IvoryTowerDenizen]

Nice results!