Approximating the Binomial Distribution with Poisson

Approximating the Binomial Distribution with Poisson
Suppose that X is binomial with n trials and a probability of success p.
We can approximate the binomial distribution with a Poisson distribution if the following conditions hold:
- np<5 or n(1−p)<5 (some textbooks say 10)
- n is very large (most textbooks say 20)
Then, the Poisson mean is approximately equal to the Binomial mean:
λ=np
And the formula becomes:

Example: Approximating Binomial with Poisson
A majestic bird can lay many, many eggs to feed the village. Suppose there are two possible outcomes for each egg: shiny or not shiny. (Shiny eggs make great frittatas!) We assume 100 eggs are laid in an identical and independent manner, and the probability of getting a shiny egg is constant at 4%.
(a) Can we approximate the Binomial distribution with Poisson?
Yes, because nq=0.04(100)=4<5.
(b) What is the probability of getting at least 3 shiny eggs? [Use the binomial formula]
P(X⩾3)
=1−P(X⩽2)
=1−P(X=2)−P(X=1)−P(X=0)
Using the binomial formula:
=1−[(100C2)(0.04)2(0.96)98]−[(100C1)(0.04)1(0.96)99]−[(100C0)(0.04)0(0.96)100]
≈1−0.14498−0.07029−0.01687
≈0.76786
(c) What is the probability of getting at least 3 shiny eggs? [Use the Poisson formula]
λ=4
P(X⩾3)
=1−P(X⩽2)
=1−P(X=2)−P(X=1)+P(X=0)
Using the binomial formula:
=1−2!e−4(4)2−1!e−4(4)1−0!e−4(4)0
=1−0.14653−0.07326−0.01832
=0.76189
Very close answers!