Revision Notes
Cambridge International AS & A Level Mathematics 9709: Probability & Statistics 2 – Revision Notes
Condensed Cambridge 9709 Paper 6 revision notes: Poisson, linear combinations, PDFs, confidence intervals, hypothesis tests and a quick self-test.
- Subject
- Mathematics
- Level
- A LEVELS
- Topic
- Probability & Statistics 2
- Author
- Marlbridge Academic Team
- Updated
- Reviewed by
- Sajawal Zahid (what this means)
Aligned to Cambridge A Level Mathematics (9709), 2026-2027. Official specification .
Syllabus page (what it covers and how it is assessed): Cambridge A Level Mathematics.
Syllabus points this page covers
9709
- 6.1 The Poisson distribution
- 6.2 Linear combinations of random variables
- 6.3 Continuous random variables
- 6.4 Sampling and estimation
- 6.5 Hypothesis tests
This page covers all 5 subtopics recorded here in topic 6, Probability & Statistics 2.
Found an error? Report a correction.
Need help with this topic? Request a free trial class for A Level Mathematics (9709).
These notes condense topic 6, Probability & Statistics 2, of the Cambridge International AS & A Level Mathematics 9709 syllabus for exams in 2026 and 2027 (Version 4): sections 6.1 to 6.5. All of it is examined on Paper 6 (1 hour 15 minutes, 50 marks, 20% of the A Level; A Level only, and not combinable with Paper 4). Probability & Statistics 1 and the Pure Mathematics 3 calculus are assumed. For full explanations and worked examples, use the Probability & Statistics 2 study guide.
Other links: Probability & Statistics 2 practice questions, the Probability & Statistics 1 revision notes, the A Level Mathematics hub, the printable 9709 checklist and the free A Level diagnostic.
Formulas
| Result | Formula | In MF19? |
|---|---|---|
| Poisson Po(λ) | P(X = r) = e^(−λ) λʳ / r!; mean λ; variance λ | Yes |
| Linear function | E(aX + b) = aE(X) + b; Var(aX + b) = a²Var(X) | No |
| Sum of variables | E(aX + bY) = aE(X) + bE(Y) | No |
| Variance of sum (independent) | Var(aX + bY) = a²Var(X) + b²Var(Y) | No |
| Continuous mean | E(X) = ∫ x f(x) dx | Yes |
| Continuous variance | Var(X) = ∫ x² f(x) dx − {E(X)}² | Yes |
| Unbiased mean | x̄ = Σx / n | Yes |
| Unbiased variance | s² = (1/(n − 1))(Σx² − (Σx)²/n) | Yes |
| Central Limit Theorem | X̄ ~ N(μ, σ²/n) approximately | Yes |
| Sample proportion | approximately N(p, p(1 − p)/n) | Yes |
| CI for a mean | x̄ ± z σ/√n (or s/√n, large sample) | No |
| CI for a proportion | p̂ ± z √(p̂(1 − p̂)/n) | No |
6.1 The Poisson distribution
Conditions for Po(λ). Events occur singly, independently, at random, at a constant mean rate. Scale λ with the interval: rate 0.8 per metre → 2.4 in 3 metres.
Approximations.
| From | To | When (syllabus guide) | Continuity correction? |
|---|---|---|---|
| B(n, p) | Po(np) | n > 50 and np < 5, approximately | No |
| Po(λ) | N(λ, λ) | λ > 15, approximately | Yes |
| B(n, p) | N(np, npq) | np > 5 and nq > 5 | Yes |
Worked reminder (normal approximation). X ~ Po(18). Find P(X ≤ 12).
λ = 18 > 15, so use N(18, 18)
P(X ≤ 12) ≈ P(Y < 12.5)
z = (12.5 − 18)/√18 = −1.296
P = 1 − Φ(1.296) = 0.0974 (3 s.f.)
Sums. Independent X ~ Po(λ₁) and Y ~ Po(λ₂) give X + Y ~ Po(λ₁ + λ₂). Only sums: 2X is not Poisson.
6.2 Linear combinations
Method in steps.
- Write the combination in symbols first (T = X₁ + X₂ + Y, or D = 2B − A).
- Mean: combine the means with the same coefficients.
- Variance: square each coefficient, then add (independent variables).
- State the distribution (normal if the parts are normal).
- Standardise and use Φ.
Must-know distinction.
| X₁ + X₂ + X₃ (three separate items) | 3X (one item, tripled) |
|---|---|
| mean 3μ | mean 3μ |
| variance 3σ² | variance 9σ² |
Small reminder: if Var(X) = 3, then Var(5 − 2X) = (−2)² × 3 = 12. The constant 5 does not change the variance.
6.3 Continuous random variables
Checklist for a PDF on one interval.
- f(x) ≥ 0 everywhere on the interval.
- Total area = 1 (use it to find k).
- P(a < X < b) = area between a and b.
- E(X) and Var(X) from the MF19 integrals.
- Median m: area from the lower limit up to m = 1/2. Quartiles: 1/4 and 3/4.
Worked reminder. f(x) = (3/8)x² for 0 ≤ x ≤ 2.
area: (3/8) × [x³/3]₀² = (3/8)(8/3) = 1 ✓
E(X) = (3/8) ∫₀² x³ dx = (3/8)(4) = 1.5
E(X²) = (3/8) ∫₀² x⁴ dx = (3/8)(32/5) = 2.4
Var(X) = 2.4 − 1.5² = 0.15
median: (3/8)(m³/3) = 1/2 → m³ = 4 → m = 1.59 (3 s.f.)
For an infinite domain, e.g. x ≥ 2, evaluate the improper integral as a limit: terms such as 1/x³ → 0 as x → ∞.
The cumulative distribution function is not required explicitly; integrate f(x) directly.
6.4 Sampling and estimation
Unbiased, in simple terms. Each sample gives a different estimate, but the method is right on average. That is why s² divides by n − 1: dividing by n would, on average, underestimate the population variance.
Why randomness matters. A random sample gives every member an equal chance of selection, so results are not biased. A method is unsatisfactory when part of the population cannot be chosen (one time, one place, volunteers only).
Distribution of X̄.
| Population | n | X̄ |
|---|---|---|
| Normal | any | exactly N(μ, σ²/n) |
| Not normal or unknown | large | approximately N(μ, σ²/n) by the CLT |
| Not normal or unknown | small | no result available |
Confidence interval, method in steps.
- Find x̄ (or p̂) and, if needed, s² with divisor n − 1.
- Pick z from the critical values table: 90% → 1.645, 95% → 1.960, 98% → 2.326, 99% → 2.576.
- Standard error: σ/√n, s/√n, or √(p̂(1 − p̂)/n).
- Interval = estimate ± z × standard error, to 3 s.f.
- Interpret: about 95% of intervals built this way contain the true value.
Worked reminder (sample size). σ = 5. How large must n be for a 95% interval for μ to have width less than 2?
width = 2 × 1.96 × 5/√n < 2 → √n > 9.8 → n > 96.04
smallest n = 97
6.5 Hypothesis tests
Method in steps (every test).
- State H₀ and H₁ in terms of the parameter (p, λ or μ), not the sample value.
- State the distribution under H₀.
- Calculate the tail probability (or the z-value).
- Compare with the significance level (or critical value). For a two-tailed test, compare the tail with half the level.
- Conclude in context, without overstating: “evidence that…” or “insufficient evidence that…”.
Which test?
| Situation | Test statistic |
|---|---|
| Single observation, binomial | Direct P(X ≤ x) or P(X ≥ x); normal approximation if n large |
| Single observation, Poisson | Direct tail probability; normal approximation if λ large |
| Population mean, σ known, normal population | z = (x̄ − μ)/(σ/√n) |
| Population mean, large sample | z = (x̄ − μ)/(s/√n) |
Finding a discrete rejection region. H₀: p = 0.4, H₁: p > 0.4, n = 15, 5% level. Under H₀, P(X ≥ 9) = 0.0950 and P(X ≥ 10) = 0.0338. The first tail below 0.05 starts at 10, so the rejection region is X ≥ 10 and P(Type I error) = 0.0338.
Errors.
| H₀ true | H₀ false | |
|---|---|---|
| Reject H₀ | Type I error | Correct |
| Accept H₀ | Correct | Type II error |
- P(Type I) = P(in rejection region | H₀ true). For a normal test it equals the significance level; for binomial or Poisson it is the actual tail probability of the rejection region.
- P(Type II) = P(in acceptance region | a stated alternative value is true).
Quick self-test
- X ~ Po(4). Find P(X = 4).
- X ~ Po(1.5) and Y ~ Po(2.3) are independent. Find P(X + Y = 0).
- Var(X) = 3. Find Var(5 − 2X).
- X ~ N(20, 9) and Y ~ N(15, 16) are independent. State the distribution of X − Y.
- f(x) = kx for 0 ≤ x ≤ 2. Find k and E(X).
- Find unbiased estimates of μ and σ² from the sample 4, 7, 9, 10, 5.
- A random sample of 25 is taken from N(60, 10²). State the distribution of X̄.
- What value of z is used for a 99% confidence interval?
- Define a Type I error.
- Which distribution approximates B(100, 0.03)? Justify it.
- Which distribution approximates Po(20)?
Answers
- e^(−4) × 4⁴/4! = 0.195 (3 s.f.).
- X + Y ~ Po(3.8), so P = e^(−3.8) = 0.0224 (3 s.f.).
- (−2)² × 3 = 12.
- X − Y ~ N(5, 25): mean 20 − 15, variance 9 + 16.
- 2k = 1, so k = 1/2; E(X) = ∫₀² x²/2 dx = 4/3.
- x̄ = 35/5 = 7; s² = (9 + 0 + 4 + 9 + 4)/4 = 6.5.
- X̄ ~ N(60, 4), since 100/25 = 4. It is exactly normal because the population is normal.
- 2.576.
- Rejecting H₀ when H₀ is true.
- Po(3): n = 100 > 50 and np = 3 < 5.
- N(20, 20), with a continuity correction, because λ > 15.
Where marks are usually lost
- λ not rescaled when the interval changes.
- Normal approximation to a Poisson used without a continuity correction.
- Variances subtracted for a difference of two variables.
- nX used instead of X₁ + … + Xₙ for separate items.
- k found, but E(X) then worked out with f(x) missing the k.
- A median equation set equal to 0.5 over the wrong limits.
- Biased variance (divisor n) given when the unbiased estimate was asked for.
- Hypotheses written in terms of x̄ or the observed count instead of μ, p or λ.
- Two-tailed test compared with the full significance level in one tail.
- Type II error probability found under H₀ instead of the stated alternative value.
Frequently asked questions
When can a binomial distribution be approximated by a Poisson distribution? The 9709 syllabus guide is n > 50 and np < 5, approximately. Then use Po(np), with no continuity correction.
When can a Poisson distribution be approximated by a normal distribution? When λ > 15, approximately. Use N(λ, λ) with a continuity correction.
What is the difference between X₁ + X₂ + X₃ and 3X? Both have mean 3μ, but three separate items give variance 3σ², while one item tripled gives variance 9σ².
Why does the unbiased variance estimate divide by n − 1? Dividing by n would, on average, underestimate the population variance. Dividing by n − 1 gives an estimate that is right on average, which is what unbiased means.
Which z-values are used for common confidence intervals? From the critical values table: 90% uses 1.645, 95% uses 1.960, 98% uses 2.326 and 99% uses 2.576.
Official syllabus
Cambridge International AS & A Level Mathematics 9709 syllabus, for exams in 2026 and 2027 (Version 4), Cambridge University Press & Assessment. Topic 6, Probability & Statistics 2 (for Paper 6): sections 6.1 to 6.5.
Get free revision emails (optional)
Occasional emails with practice questions, worked explanations and links to free resources for the qualification and subjects you choose. No spam, and you can unsubscribe from any email. The free tools on this site never need an email.
Related resources
-
Study guides
Cambridge International AS & A Level Mathematics 9709: Circular measure – Study Guide
Radians, arc length, sector area, segments and tangent problems taught step by step with worked examples, for Cambridge 9709 Pure Mathematics 1.
Mathematics · Cambridge · A Level
Read guide -
Study guides
Cambridge International AS & A Level Mathematics 9709: Probability & Statistics 2 – Study Guide
Study guide for Cambridge 9709 Probability & Statistics 2 (Paper 6): Poisson, linear combinations, PDFs, estimation and hypothesis tests, fully worked.
Mathematics · Cambridge · A Level
Read guide -
Practice questions
Cambridge International AS & A Level Mathematics 9709: Probability & Statistics 2 – Practice Questions
12 original Cambridge 9709 Paper 6 practice questions with mark-by-mark answers on Poisson, PDFs, confidence intervals and hypothesis tests.
Mathematics · Cambridge · A Level
Try questions
Related articles
-
exam preparation
Where IGCSE Mathematics marks are lost early
The first weeks of an IGCSE Mathematics course rarely go wrong on difficulty. They go wrong on method, command words, rounding and units — four habits that cost marks a student had already earned.
24 August 2026
-
teaching
A Level tuition in Dubai: Cambridge or Pearson Edexcel International A Level?
Schools in Dubai and Abu Dhabi teach either Cambridge International AS & A Level or Pearson Edexcel International A Level. How the two are structured, why it matters for tuition, and how online classes from Lahore work.
9 October 2026
Studying this with a teacher
Working through Mathematics A LEVELS?
This page is free and stays free. If you would rather be taught it, Marlbridge runs Mathematics classes one-to-one and in small groups of up to 15, online in your own time zone. The first trial class is free. WhatsApp replies within an hour (8am–11pm Pakistan time, every day); email the same day.
Cambridge Mathematics teachers at Marlbridge