Unit 4: Probability & probability distributions
Business Statistics notes · PTU syllabus (BCOMGE 201-18)
On this page
Unit summary
Probability measures uncertainty — the chance of a product being defective, a customer buying, or a loan defaulting. This unit covers the approaches to probability, the addition and multiplication laws, conditional probability and Bayes' theorem, and the binomial, Poisson and normal distributions.
After this unit you can
- Calculate probabilities using the classical, relative frequency and axiomatic approaches
- Apply the addition and multiplication laws and conditional probability
- Apply Bayes' theorem
- Use the binomial, Poisson and normal distributions to solve business problems
PTU syllabus topics
- Approaches to calculating probability
- addition and multiplication laws
- conditional probability and Bayes' Theorem
- Binomial distribution
- Poisson distribution
- Normal distribution and its properties
Topic 1
Approaches to probability
- Experiment
- An activity with uncertain outcome — tossing a coin
- Sample space (S)
- Set of all possible outcomes — {H, T}
- Event
- A subset of the sample space
- Mutually exclusive events
- Cannot occur together
- Exhaustive events
- Together cover the sample space
- Independent events
- One does not affect the other
- Complement
- A′ — event does not occur
Classical (a priori)
P(A) = favourable outcomes ÷ total equally likely outcomes
P(head) = 1/2, P(six on a die) = 1/6
Relative frequency (empirical)
P(A) = limit of (times A occurred ÷ number of trials)
120 defective out of 2,000 bulbs gives 0.06
Subjective
Personal degree of belief
Chance of a new product succeeding
Axiomatic (Kolmogorov)
0 ≤ P(A) ≤ 1; P(S) = 1; P(A ∪ B) = P(A) + P(B) for exclusive events
Mathematical base of the subject
Range
0 ≤ P(A) ≤ 1
Sure and impossible event
P(S) = 1, P(∅) = 0
Complement
P(A′) = 1 − P(A)
Odds in favour
P(A) ÷ P(A′)
Topic 2
Laws of probability
Addition law (any events)
P(A ∪ B) = P(A) + P(B) − P(A ∩ B)
Addition law (mutually exclusive)
P(A ∪ B) = P(A) + P(B)
Multiplication law (any events)
P(A ∩ B) = P(A) × P(B
A)
Multiplication law (independent)
P(A ∩ B) = P(A) × P(B)
Conditional probability
P(B
A) = P(A ∩ B) ÷ P(A)
Three events
P(A ∪ B ∪ C) = ΣP − ΣP(pairs) + P(A ∩ B ∩ C)
Example
A card is drawn from 52 cards. P(king or heart) = 4/52 + 13/52 − 1/52 = 16/52 = 4/13. Two cards drawn without replacement: P(both kings) = (4/52) × (3/51) = 1/221.
Example
In a class, 60% study Statistics, 50% study Economics and 30% study both. P(Stats or Economics) = 0.6 + 0.5 − 0.3 = 0.8; P(Economics | Stats) = 0.3 ÷ 0.6 = 0.5.
Meaning
Cannot occur together
Occurrence of one does not change the other
P(A ∩ B)
0
P(A) × P(B)
Example
Head and tail on one toss
Two separate coin tosses
- 1Define the experiment and sample space
- 2List the favourable outcomes
- 3Decide whether events are exclusive, independent or conditional
- 4Apply the correct law
- 5Check that 0 ≤ P ≤ 1
Topic 3
Bayes' theorem
Bayes' theorem revises the probability of a cause after an outcome has been observed.
Total probability
P(B) = Σ P(Aᵢ) × P(B
Aᵢ)
Bayes' theorem
P(Aᵢ
B) = [P(Aᵢ) × P(B
Prior probability
P(Aᵢ) before the evidence
Posterior probability
P(Aᵢ
B) after the evidence
Example
Three machines make 50%, 30% and 20% of the output with 1%, 2% and 3% defective items. A defective item is found — which machine made it?
| Machine | P(Mᵢ) | P(Defective given Mᵢ) | Product | Posterior P(Mᵢ given D) |
|---|---|---|---|---|
| M₁ | 0.50 | 0.01 | 0.005 | 0.005 ÷ 0.017 = 0.294 |
| M₂ | 0.30 | 0.02 | 0.006 | 0.006 ÷ 0.017 = 0.353 |
| M₃ | 0.20 | 0.03 | 0.006 | 0.006 ÷ 0.017 = 0.353 |
| Total | 1.00 | 0.017 | 1.000 |
Example
P(defective) = 0.017. Machine M₂ or M₃ is each 35.3% likely to be the source, so quality checks should focus on them.
Topic 4
Binomial distribution
A binomial experiment has n independent trials, each with two outcomes (success p, failure q = 1 − p), and p is constant.
Probability of r successes
P(X = r) = nCr pʳ qⁿ⁻ʳ, r = 0, 1, …, n
Mean
np
Variance
npq
Standard deviation
square root of npq
Mode
Integer part of (n + 1)p
Symmetric when
p = 0.5
With p = 0.5 the bars are symmetric; the mean is np = 2.5
Example
A fair coin is tossed 5 times: P(exactly 3 heads) = 5C3 (0.5)⁵ = 10 ÷ 32 = 0.3125. A firm finds 10% of items defective: in a sample of 4, P(none defective) = (0.9)⁴ = 0.656; P(at least one) = 1 − 0.656 = 0.344.
- Fixed trials
- n is fixed in advance
- Two outcomes
- Success or failure
- Constant p
- Same probability every trial
- Independence
- Trials do not affect each other
- Use
- Quality control, surveys, yes-no responses
Exam tip
Choosing the distribution: a fixed number of trials with two outcomes and known p → binomial; rare events counted over time or space with a known average → Poisson (also when n is large and p small, with m = np); continuous measurements such as height or marks → normal, converted to Z = (X − μ) ÷ σ.
Topic 5
Poisson distribution
The Poisson distribution models the number of rare events in a fixed interval of time or space (calls per hour, defects per metre, accidents per day).
Probability
P(X = r) = e^(−λ) λʳ ÷ r!, r = 0, 1, 2, …
Mean and variance
Both equal λ
e
e = 2.71828; e⁻² = 0.1353
Approximation to the binomial
When n is large and p small, λ = np
P(0) = 0.135, P(1) = 0.271, P(2) = 0.271, P(3) = 0.180, P(4) = 0.090 — probabilities fall quickly for larger r
Example
A switchboard receives 2 calls a minute on average. P(no call) = e⁻² = 0.1353; P(exactly 3 calls) = e⁻² × 2³ ÷ 3! = 0.1353 × 8 ÷ 6 = 0.1804; P(at least one) = 1 − 0.1353 = 0.8647.
Trials
Fixed n
Unlimited, over an interval
Parameters
n and p
λ
Mean and variance
np and npq (variance smaller)
λ and λ (equal)
Use
Counting successes in n trials
Counting rare events in time or space
Topic 6
Normal distribution
The normal distribution is the bell-shaped continuous distribution used for heights, marks, errors and returns.
Density
f(x) = [1 ÷ (σ square root of 2π)] e^(−(x − μ)² ÷ 2σ²)
Parameters
Mean μ and standard deviation σ
Standard normal variable
Z = (X − μ) ÷ σ, with mean 0 and SD 1
Total area
1
Empirical rule
68.3% within μ ± 1σ, 95.4% within μ ± 2σ, 99.7% within μ ± 3σ
Symmetric about the mean; mean = median = mode; tails never touch the axis
- Shape
- Bell-shaped and symmetric
- Central values
- Mean = median = mode
- Maximum ordinate
- At x = μ
- Points of inflexion
- At μ ± σ
- Asymptotic
- Tails approach but never touch the axis
- Area
- Total 1; 50% on each side of the mean
- Linear combinations
- Of normal variables are normal
- Kurtosis
- Mesokurtic
Within ±2σ lies 95.4% and within ±3σ lies 99.7%
| Z value | Area between 0 and Z |
|---|---|
| 0.5 | 0.1915 |
| 1.0 | 0.3413 |
| 1.5 | 0.4332 |
| 1.96 | 0.4750 |
| 2.0 | 0.4772 |
| 2.58 | 0.4951 |
| 3.0 | 0.4987 |
- 1
Write the given μ and σ
- 2
Convert x to Z = (x − μ) ÷ σ
- 3
Sketch the curve and shade the required area
- 4
Read the area from the standard normal table
- 5
Add or subtract areas using symmetry (0.5 on each side)
- 6
State the answer as a probability or percentage
Example
Marks are normal with μ = 60 and σ = 10. P(X > 75): Z = 1.5, area between 0 and 1.5 is 0.4332, so P = 0.5 − 0.4332 = 0.0668 (6.7%). P(50 < X < 70): Z from −1 to +1, area = 2 × 0.3413 = 0.6826 (68.3%).
- Quality control
- Control limits at ±3σ
- Finance
- Returns and risk
- Sampling
- Distribution of sample means for large n
- Approximations
- Binomial and Poisson for large n
- Standard scores
- Compare performance on different scales
Key terms
- Probability
- Measure of the likelihood of an event, from 0 to 1
- Mutually exclusive events
- Events that cannot occur together
- Independent events
- Events where one does not affect the other
- Conditional probability
- Probability of an event given that another has occurred
- Bayes' theorem
- Rule for revising probabilities given new evidence
- Binomial distribution
- Distribution of the number of successes in n independent trials
- Poisson distribution
- Distribution of rare events over an interval
- Normal distribution
- Symmetric bell-shaped continuous distribution
- Standard normal variable
- Z = (X − μ) ÷ σ
Quick revision
- Classical P = favourable ÷ total; P(A′) = 1 − P(A).
- Addition: P(A ∪ B) = P(A) + P(B) − P(A ∩ B); multiplication: P(A ∩ B) = P(A) P(B | A).
- Bayes: posterior = prior × likelihood ÷ total probability.
- Binomial: mean np, variance npq; Poisson: mean = variance = λ.
- Normal: Z = (X − μ)/σ; 68–95–99.7 rule.
- Correlation r = √(byx bxy); regression lines meet at the means.
Important exam questions
Practice questions written to the PTU exam pattern for this unit's syllabus: short answers (Section A style) and long answers (Sections B and C style).
Short-answer questions
- Q1.Define probability. State its range.
- Q2.State the addition law of probability.
- Q3.Differentiate between mutually exclusive and independent events.
- Q4.Define conditional probability.
- Q5.State Bayes' theorem.
- Q6.Write the probability function of the binomial distribution and its mean and variance.
- Q7.When is the Poisson distribution used? What are its mean and variance?
- Q8.State four properties of the normal curve.
- Q9.What is a standard normal variable?
- Q10.A coin is tossed 3 times; find the probability of exactly two heads.
Long-answer questions
- Q1.Explain the approaches to probability with examples.
- Q2.Explain the addition and multiplication laws with examples.
- Q3.Explain conditional probability and Bayes' theorem with a numerical example.
- Q4.Explain the binomial distribution and its properties with an example.
- Q5.Explain the Poisson distribution and when it can be used as an approximation to the binomial.
- Q6.Explain the normal distribution and its properties; solve a problem using the standard normal table.
- Q7.Calculate the correlation coefficient and fit regression lines for the given data.
Stuck on this unit?
Message SBS on WhatsApp for help with Business Statistics, or to ask about studying B.Com at Synetic.
