Verwijderd schreef op vrijdag 10 december 2004 @ 22:21:
[...]
Niets over ten minste of oudste dus de meest simpele variant van dit soort vragen
logica: mogelijkheden zd,dz,zz (1/3 leidt tot 2e jongen)
Bayes
P(ZZ|Z)=P(ZZ+Z)/P(Z)=(1/4)/(3/4)=1/3
P(ZZ+Z) = deelverzameling van families met 2 kinderen die EN 2 zonen EN 'n zoon hebben (alleen zz)
P(Z)= families met 2 kinderen die 'n zoon hebben (zd dz en zz)
2e vraag= onvoorwaardelijke kans dus gewoon 1/2
Idd buitengewoon klein verschil in de 2 vragen. Alleen verschil in selectie. Hoe dan de kans op een zoon voor het 2e kind daardoor kan veranderen snap ik eerlijk gezegd niet helemaal.
Normaal zit bij dit soort vragen de clou (net als bij 16) er in dat bij de ene vraag wordt gevraagd naar de kans op 2 jongens en in de andere naar de kans dat het 2e kind een jongen is als het 1e kind al een jongen is. Dit laatste is idd een groot verschil alhoewel op beide NWQ fora veel mensen zijn die dat hetzelfde vinden. Het verschil in deze 2 engelse vragen is voor mijn gevoel veel kleiner. Ben benieuwd naar de officiele berekening in het boek.
Ben benieuwd naar het antwoord uit het boek, kun je dat posten?
OK, ik zal het hele verhaal posten.
Ik zal het hele voorbeeld citeren en niet vertalen (wat ik overigens wel aan het doen was).
Omdat er geen wiskunde symbolen mogelijk zijn heb ik de woorden subset, intersect,omega, etc er voor in de plaats gezet.
Consider all families with two children and assume that boys and girls are equally likely.
Thus the sample space may be denoted schematically by 4 points:
OMEGA={(b,b), (b,g), (g,b), (g,g)}
where b=boy and g=girl; the order in each pair is the order of birth; and the 4 points have probability of ¼ each.
If a family is chosen at random from OMEGA and found to have a boy in it what is the probability that it has another boy; namely that it is of the type (b,b)? A quickie answer might be ½ if you jumped to the conclusion from the equal likelihood of the sexes. This is a mistake induced by a misplaced “relative clause” for the conditional probability in question. Here is the detailed explanation.
Let us put
A ={omega|there is a boy in omega}
B={omega|there are two boys in omega}
Then B subset A and so A intersect B=B, thus
P(B|A)=P(B)/P(A)=1/4:3/4=1/3.
This is the correct answer to the question.
But now let us ask a similar sounding but really a different question. If a child is chosen at random from these families and is found to be a boy, what is the probability that the other child in his family is also a boy?
This time the appropiate representation of the sample space should be
OMEGA*={g
g, g
b, b
g, b
b}
where the sample points are not families but the children of these families, and g
g = a girl who has a sister, g
b = a girl who has a brother, etc.
Now we have
A*={ omega*|omega* is a boy}
B*={omega*|omega* has a brother}
so that
A* intersect B* ={omega*|omega*=b
b}
Therefore
P(B*|A*)= P(A* intersect B*)/P(A*)=1/4:1/2=1/2
This is a wonderful and by no means artificial illustration of the importance of understanding ”
what we are sampling” in statistics.
Einde citaat.
Ik zou aan de laatste regel nog willen toevoegen:
en dan komt als lastigste de interpretatie van de resultaten.