home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky sci.math:17603 rec.puzzles:8197
- Newsgroups: sci.math,rec.puzzles
- Path: sparky!uunet!gatech!asuvax!ncar!noao!stsci!scivax!zellner
- From: zellner@stsci.edu
- Subject: Re: Marilyn Vos Savant's error?
- Message-ID: <1993Jan2.215958.1@stsci.edu>
- Lines: 91
- Sender: news@stsci.edu
- Organization: Space Telescope Science Institute
- References: <1gj5grINNk05@crcnis1.unl.edu> <1992Dec15.012404.24027@galois.mit.edu> <1992Dec15.052211.24395@CSD-NewsHost.Stanford.EDU> <1hvp6gINN9np@chnews.intel.com> <1992Dec31.203934.1@stsci.edu> <2B45F42A.3954@news.service.uci.edu>
- Distribution: na
- Date: Sun, 3 Jan 1993 02:59:58 GMT
-
- Steve White (srw@horus.ps.uci.edu) writes:
-
- > You forgot to label your table. You wanted it to be
-
- > Older Younger
- > ----- -------
- > Boy - Boy
- > Boy - Girl
- > Girl - Boy
- > Girl - Girl
-
- Correct.
-
- > Then your logic seems to hold. But if I label the table:
-
- > Kid Kid
- > Over there Elsewhere
- > ----- -------
- > Boy - Boy
- > Boy - Girl
- > Girl - Boy
- > Girl - Girl
-
- Yes, you could of course do that, but it would be a different problem, and
- free of the apparent paradox: Given that at least one of the kids is a boy,
- the probability of two boys depends on your knowledge of whether the one you
- see is the elder, the younger or simply "one of them."
-
- > Then case A gives probability 1/2 and case B gives 1/3.
- > In other words, this argument fails. Neither wording is the
- > same as "at least one of them is a boy".
-
- Sorry, you lost me there. If the man says "one of them is over there", and I
- look and see a boy, isn't that the same as "at least one of them is a boy"?
-
- > The correct answer is 1/2 for A and B.
-
- Try the following experiment. Toss a penny and a nickel some large number of
- times, or just make up a table with equal frequencies of H and T for each,
- e.g.,
-
- Penny Nickel
- ----- -------
- H T
- H H
- T H
- T T
-
- with equal frequency. Let the penny represent the older child, the nickel
- the younger child, and heads or tails the sex of each. Now go through and
- cross out all T-T combinations, that is, keep only those combinations in
- which at least one head (at least one boy) occurred. Now ask the following
- questions:
-
- 1. What is the probability of two heads, when the Penny is a head?
- 2. What is the probability of two heads, when the Nickel is a head?
-
- Obviously the probability is 1/2 in either case. But if you ask
-
- 3. What is the probability of two heads, when AT LEAST ONE COIN (of either
- species) is a head?
-
- Clearly that's 1/3. For a man to say "one of my two children is a boy" versus
- "the elder of my two children is a boy" samples two different statistical
- ensembles, with different results.
-
- Now then: Suppose a woman gets pregnant out of wedlock, and decides to give
- up the child for adoption straight from the delivery room, without even knowing
- its sex. A year later, she does it again. But in later years the rascal who
- kept getting her pregnant had a change of heart and married her and settled down
- to be a good husband and father, and they began to wonder about those two
- childen.
-
- Information about adopted children is tightly guarded in some jurisdictions, so
- they hired a private detective to see what he could find out. Suppose after
- some investigation he reported:
-
- "I've found the older child, and it's a boy."
-
- Or instead,
-
- "I've found one of the children, and it's a boy, but the records I've been
- able to peek at so far don't give its age, so I don't know whether it's
- the older or the younger."
-
- In each of the two cases, what's the probability that both children are boys?
- It seems intuitive that the age of the child that he found wouldn't matter, but
- mathematically it does. That's the paradox.
-
- Cheers, Ben
-
-