I'm working on a problem from a textbook (self-study) (Casella and Berger).
The Problem
Given these facts/assumptions...
- ⅓ of human twins are identical (one-egg) and ⅔ are fraternal (two-egg).
- identical twins are the same sex with sexes being equally likely.
- among fraternal twins, ¼ are female, ¼ are male, and ½ are male/female.
- among all US births, 1/90 are twins.
Let:
$$ \begin{align} A\ &=\ \{Births\ which\ Result\ in\ Twin\ Females\} \\ B\ &=\ \{Births\ which\ Result\ in\ Identical\ Twins\} \\ C\ &=\ \{Births\ which\ Result\ in\ Twins\} \end{align} $$
and find P($A \cap B \cap C$), i.e. the probability of identical twin females.
My attempt to solve
The first thing I tried to do was translate the above facts into probability statements that can be used to calculate P($A \cap B \cap C$).
- ⅓ of human twins are identical (one-egg) and ⅔ are fraternal (two-egg).
Given a set of twins, a fraction of them are either identical or fraternal:
$$ \begin{align} P(identical\ |\ twins) &= \frac{1}{3} \\ P(fraternal\ |\ twins) &= \frac{2}{3} \\ \end{align} $$
- identical twins are the same sex with sexes being equally likely.
$$ \begin{align} P(female\ |\ identical\ twins)\ = P(male\ |\ identical\ twins) = \frac{1}{2} \end{align} $$
- among fraternal twins, ¼ are female, ¼ are male, and ½ are male/female.
$$ \begin{align} P(female\ |\ fraternal\ twins) = P(male\ |\ fraternal\ twins) &= \frac{1}{4} \\ P(male/female\ |\ fraternal\ twins) &= \frac{1}{2} \end{align} $$
- among all US births, 1/90 are twins.
$$ P(twins) = \frac{1}{90} $$
Stuck on this one... do $A$, $B$ and $C$ have to be calculated individually in order to calculate P($A \cap B \cap C$)?