Introduction
A Bernoulli trial or Bernoulli experiment is a random experiment with exactly two mutually exclusive outcomes: “Success” and “Failure” with probabilities p and q = 1-p respectively. The probability of success remains constant every time the experiment is conducted. The distribution is named after Swiss scientist Jacob Bernoulli.
Bernoulli Trials
Trials of a random experiment are called Bernoulli trials if they satisfy:
- Each trial has exactly two possible outcomes (“Success” and “Failure”)
- The trials are independent (outcome of one trial doesn’t influence others)
- The probability of success remains constant from trial to trial
Examples of Bernoulli Trials
- Tossing a coin (“Heads” or “Tails”)
- Sex of newborn baby (“Male” or “Female”)
- Answer to true/false question (“True” or “False”)
- Product quality (“Defective” or “Non-defective”)
- Customer decision (“Purchase” or “No Purchase”)
Bernoulli Distribution Definition
Consider a random experiment with two possible outcomes: “Success” (S) with probability p and “Failure” (F) with probability q = 1-p.
The Bernoulli distribution describes the probability distribution of a random variable X representing the number of successes in a single Bernoulli trial.
Probability Mass Function (PMF)
$$P(X=x) = p^x q^{1-x}, \quad x=0,1; \quad 0<p,q<1; \quad q=1-p$$
where:
- X = number of successes (0 or 1)
- p = probability of success
- q = 1-p = probability of failure
Verification:
- $P(X=x) \geq 0$ for all x
- $P(X=0) + P(X=1) = q + p = 1$ ✓
Key Features of Bernoulli Distribution
Mean (Expected Value)
$$E(X) = \mu = p$$
Proof: $$E(X) = 0 \cdot P(X=0) + 1 \cdot P(X=1) = 0 \cdot q + 1 \cdot p = p$$
Variance
$$\text{Var}(X) = \sigma^2 = pq = p(1-p)$$
Proof: $$E(X^2) = 0^2 \cdot q + 1^2 \cdot p = p$$
$$\text{Var}(X) = E(X^2) - [E(X)]^2 = p - p^2 = p(1-p)$$
Standard Deviation
$$\sigma = \sqrt{pq} = \sqrt{p(1-p)}$$
Relationship to Other Distributions
- Binomial Distribution: Sum of n independent Bernoulli trials
- Geometric Distribution: Number of Bernoulli trials until first success
- Negative Binomial Distribution: Number of Bernoulli trials until r-th success
Bernoulli Distribution Examples
Example 1: Battery Defects
A battery manufacturing process produces batteries where 85% are non-defective. A battery is selected at random.
a) What is the probability distribution?
Let X = 1 if non-defective, 0 if defective
- p = 0.85 (probability of non-defective)
- q = 0.15 (probability of defective)
$$P(X=x) = 0.85^x \cdot 0.15^{1-x}, \quad x=0,1$$
| X | P(X=x) |
|---|---|
| 0 | 0.15 |
| 1 | 0.85 |
b) Probability that battery is defective:
$$P(X=0) = 0.85^0 \cdot 0.15^1 = 0.15$$
c) Mean and variance:
$$E(X) = p = 0.85$$
$$\text{Var}(X) = pq = 0.85 \times 0.15 = 0.1275$$
$$\sigma = \sqrt{0.1275} = 0.357$$
Example 2: Coin Toss
Toss a fair coin once. Let X = 1 if heads, 0 if tails.
Given: p = 0.5 (fair coin)
Probability distribution:
$$P(X=x) = 0.5^x \cdot 0.5^{1-x} = 0.5, \quad x=0,1$$
Properties:
$$E(X) = 0.5$$
$$\text{Var}(X) = 0.5 \times 0.5 = 0.25$$
$$\sigma = 0.5$$
Example 3: Customer Purchase Decision
A customer visiting a store has probability 0.3 of making a purchase. Let X = 1 if purchase, 0 if no purchase.
Distribution:
$$P(X=x) = 0.3^x \cdot 0.7^{1-x}$$
a) Probability of no purchase:
$$P(X=0) = 0.7$$
b) Probability of purchase:
$$P(X=1) = 0.3$$
c) Expected number of purchases:
$$E(X) = 0.3$$
On average, 30% of customers make a purchase.
d) Variance:
$$\text{Var}(X) = 0.3 \times 0.7 = 0.21$$
Example 4: Medical Test Result
A medical test for disease has 95% accuracy. Let X = 1 if positive, 0 if negative (assuming person has disease).
Given: p = 0.95
Probability distribution:
$$P(X=1) = 0.95, \quad P(X=0) = 0.05$$
Expected value of test outcome:
$$E(X) = 0.95$$
Variance:
$$\text{Var}(X) = 0.95 \times 0.05 = 0.0475$$
When to Use Bernoulli Distribution
✓ Binary outcomes (success/failure, yes/no, pass/fail) ✓ Single trial or observation ✓ Constant probability across trials ✓ Independent trials ✓ Basis for other distributions (Binomial, Geometric, Negative Binomial)
Properties Summary
| Property | Formula |
|---|---|
| PMF | $P(X=x) = p^x(1-p)^{1-x}$ |
| Mean | $\mu = p$ |
| Variance | $\sigma^2 = p(1-p)$ |
| Std Dev | $\sigma = \sqrt{p(1-p)}$ |
| Support | X ∈ {0, 1} |
| Notation | X ~ Ber(p) |
Applications
- Quality control (defective/non-defective)
- Medical testing (positive/negative result)
- Customer behavior (purchase/no purchase)
- Website visits (click/no click)
- Reliability testing (failure/success)
- Genetics (trait present/absent)