next up previous
Next: Point estimation Up: Contents Previous: Bivariate distributions

Sampling distributions

Activity 5.1   Continuing Activities 3.8 and 3.11,show that the mean and variance of the binomial distribution in (which is the same as the sum of $ n$ independent Bernoulli trials) has mean $ n\pi$ and variance $ n\pi(1-\pi)$.
$ \blacksquare$

Answer 5.1   From the formulae, the proportion of successes, which is the mean of the $ n$ Bernoulli trials has mean $ \pi$ and variance $ \pi(1-\pi)/n$. So the sum of $ n$ Bernoulli trials has mean $ n\pi$ and variance $ n\pi(1-\pi)$.
$ \blacksquare$

Activity 5.2   From the results in 5.1, by fixing $ \lambda=n\pi$, and allowing $ n$ to increase, and $ \pi$ to get small, discover that the mean and variance of a Poisson distribution are both $ \lambda$.
$ \blacksquare$

Answer 5.2   The mean is $ n\pi=\lambda$. The variance is $ n\pi(1-\pi)=\lambda(1-\pi)$. This tends to $ \lambda$ as $ \pi$ tends to 0.
$ \blacksquare$



M.Knott 2002-09-12