4.10 Binomial Distribution
1. Description of the Problem
The Binomial distribution describes the probability of a discrete random variable $X$. It applies to experiments with exactly two possible outcomes:
- SUCCESS (with constant probability $p$)
- FAILURE (with probability $1-p$)
We repeat the experiment $n$ independent times. The parameters are:
$p$ = probability of success in a single trial
The variable $X$ counts the number of successes. We say $X$ follows a binomial distribution, written $X \sim B(n,p)$. Since $n$ is the number of trials, $X$ takes the values:
EXAMPLE 1
We toss a die 5 times. Success is rolling a six. Here, $n=5$ and $p=\dfrac{1}{6}$. We can have 0, 1, 2, 3, 4, or 5 successes.
| $X$ | 0 | 1 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|---|---|
| $P(X=x)$ | 0.4019 | 0.4019 | 0.1608 | 0.0322 | 0.0032 | 0.0001 |
We can answer inequality questions using cumulative probabilities:
| Find the probability of: | Notation | Result |
|---|---|---|
| exactly 3 sixes | $P(X=3)$ | 0.0322 |
| at most 3 sixes | $P(X \le 3)$ | 0.9967 |
| less than 3 sixes | $P(X < 3)$ | 0.9645 |
| more than 3 sixes | $P(X > 3)$ | 0.0033 |
| at least 3 sixes | $P(X \ge 3)$ | 0.0355 |
2. The Binomial Formula
Remark: This formula helps conceptualize binomial probabilities.
- The probability of obtaining 5 sixes in a row is: $\left(\dfrac{1}{6}\right)^5$
- The probability of obtaining 0 sixes is: $\left(\dfrac{5}{6}\right)^5$
- The probability of obtaining exactly 2 sixes and 3 no-sixes is: $^5C_2 \left(\dfrac{1}{6}\right)^2 \left(\dfrac{5}{6}\right)^3$
Here, $^5C_2$ is the number of ways to arrange 2 sixes in 5 trials. In general, for $n$ trials with a success probability of $p$, $P(X=x)$ is:
3. Expected Value and Variance
For a random variable $X \sim B(n, p)$, the expected value (mean) and variance are:
$E(X) = 5\left(\dfrac{1}{6}\right) = \mathbf{\dfrac{5}{6}}$
$Var(X) = 5\left(\dfrac{1}{6}\right)\left(\dfrac{5}{6}\right) = \mathbf{\dfrac{25}{36}}$
$$ \begin{aligned} Var(X) &= E(X^2) - (E(X))^2 \\ E(X^2) &= Var(X) + (E(X))^2 \\ E(X^2) &= np(1-p) + (np)^2 \end{aligned} $$
EXAMPLE 2
A box contains 5 balls: 1 BLACK and 4 WHITE. We win if we select a BLACK ball. We play this game 10 times (with replacement). Find the probabilities.
This is $\mathbf{0.088}$.
[$P(X=4) = {^{10}C_4}(0.2)^4(0.8)^6 = 0.088$]
This cumulative probability is $\mathbf{0.967}$.
[$P(X \le 4) = P(X=0) + P(X=1) + P(X=2) + P(X=3) + P(X=4)$]
This is $\mathbf{0.893}$.
[Alternatively: $P(X \ge 1) = 1 - P(X=0) = 1 - 0.107 = 0.893$]
$E(X) = np = 10 \times 0.2 = \mathbf{2}$
$Var(X) = np(1-p) = 10 \times 0.2 \times 0.8 = \mathbf{1.6}$
EXAMPLE 3
Let $p=0.2$. Given $P(X=1) = 0.268$, find $n$.
Testing integer values shows that $n=10$ results in 0.268.
Hence, $\mathbf{n=10}$.
4. Mode of a Binomial DistributionHL ONLY
To find the mode (the outcome with the highest probability), first calculate $E(X) = np$. The mode is adjacent to the mean.
Say $n=20$ and $p=\dfrac{1}{6}$, meaning $E(X) = \dfrac{20}{6} \approx 3.33$.
Evaluate the probabilities for the nearest integers (3 and 4):
$P(X=3) = 0.237$
$P(X=4) = 0.202$
Since $0.237 > 0.202$, the highest probability is at $X=3$. The mode is 3.
Say $n=60$ and $p=\dfrac{1}{6}$, meaning $E(X) = \dfrac{60}{6} = 10$.
Evaluate the probabilities for adjacent integers (9, 10, 11):
$P(X=9) = 0.134$
$P(X=10) = 0.137$
$P(X=11) = 0.126$
The peak probability is at the mean. The mode is 10.
For $n=5$ and $p=\dfrac{1}{6}$, $E(X) = \dfrac{5}{6} \approx 0.833$. Check the adjacent integers 0 and 1:
$P(X=0) = 0.4019$
$P(X=1) = 0.4019$
The probabilities are identical. Therefore, there are two modes: 0 and 1.