Sampling with order and with/out replacement as random variable

52 Views Asked by At

A random sample of size $n$ is drawn from a lot of $N$ items, of which a fraction $p$ are defective. Let $X = $ number of defective items in the sample. Determine $f_X (x) = P(X = x)$ if

$\cdot$ The sample is ordered with replacement

$\cdot$ The sample is ordered without replacement

If is ordered without replacement we have $n(\Omega) = \frac{N!}{(N-n)!}$ and I'm not sure if the numerator is $$n(A) = \frac{Np!}{(Np-x)!} \cdot \frac{(N-Np)!}{(N-Np-(Np-x))!} $$ because we want $x$ to be defective, and the rest not defective, and $P(X = x) =\frac{n(A)}{n(\Omega)}$.

If is ordered with replacement we have $n(\Omega) = N^{n}$ and I'm not sure if the numerator is

$$n(A) = Np (Np-1) (Np-2) \cdots (Np-x)! $$

and $P(X = x) =\frac{n(A)}{n(\Omega)}$.

I would appreciate some help, I am a little confused, with order it becomes more complicated for me to think.

1

There are 1 best solutions below

0
On BEST ANSWER

Order does not affect the distribution of the total number of defectives in the sample. Replacement does.

Sampling without replacement, you have a hypergeometric distribution:

$$\mathbb P(X=x)= \frac{{Np \choose x}{N-Np \choose n-x}}{{N \choose n}}=\frac{(Np)!\,(N-Np)!\,n!\,(N-n)!}{x! \,(Np-x)! \, (n-x)!\, (N-Np-n+x)!}.$$

Sampling with replacement, you have a binomial distribution:

$$\mathbb P(X=x)= {n \choose x} p^x (1-p)^{n-x}.$$