H2 Maths Statistics Overview | Free Notes (Topic 6)

Study guideUpdated 21 Aug 2026
Marcus Pang
Reviewed by
Marcus Pang·Managing Director (Maths)

Free H2 Maths notes on Probability & Statistics: key formulas, exam techniques, and step-by-step solutions for distributions, testing, and regression.

Q: What does H2 Maths Statistics Notes: Probability & Statistics Overview cover?
A: Understand the H2 Maths statistics strand: counting, distributions, sampling, hypothesis tests, and regression, with pacing tips for promos and A-Levels.
How to use this guide
Section B supplies the statistical thinking that feeds subjects like Economics, Physics practicals, and university interviews. Use this roadmap to plan your sampling/hypothesis practice and to identify the exact MOE expectations before diving into the detailed notes.
  • Statistics is modelling uncertainty: Identify the random variable.
  • Each question has a model, condition, and conclusion: Check whether it is probability, distribution, sampling, testing, or regression.
  • Marks come from computation plus context: State what the number means in the scenario.

Concrete example: A hypothesis test answer is not finished after the p-value. You must say whether there is enough evidence, at the stated level, for the claim in the question context.

If you want weekly support on Paper 2 pacing, graphing-calculator routines, and method-mark working, see our H2 Maths tuition Singapore page.

Status: SEAB's current H2 Mathematics (9758) syllabus PDF is labelled for 2026. Paper 1 is Pure Mathematics only (100 marks). Probability and Statistics is assessed in Paper 2 Section B (60 marks) across Topics 6.1-6.6, with key 2026 exclusions, including no normal approximation to binomial and no correlation hypothesis tests.


Model-choice checkpoint

Before using the graphing calculator, decide what kind of statistical question you are answering. Most errors in Section B start when the working uses the right command on the wrong model.

Question wordingFirst model choiceWhat to write before calculatingCommon trap
"At least", "at most", "given that"Probability rules or a named random variable.Define the event or random variable clearly.Treating a conditional probability question like two independent events.
"Number of successes" in fixed trialsBinomial model, if the conditions fit.State the trial count, success probability, and what counts as success.Using a binomial command without checking independence or constant probability.
"Mean of a sample" or "evidence for a mean"Sampling or hypothesis-testing framework.Identify whether the random variable is an individual value or a sample mean.Mixing the population standard deviation with the spread of the sample mean.
"Predict", "estimate", or "linear relationship"Correlation and regression.State the explanatory variable and response variable.Extrapolating beyond the data range or reading correlation as causation.

Worked check: If a question says a machine produces defective items with probability 0.04 and asks for the chance that at most two out of twenty are defective, start by defining the count of defective items and checking a binomial model. If a later question asks whether the mean mass differs from a target value, switch to a sample-mean or hypothesis-testing frame instead of reusing the binomial setup.

1 | Topic 6 Overview

The entire statistics strand is grouped under Topic 66. We break it into six focused articles 6.1−6.66.1-6.6 that mirror the MOE sub-topics.

Sub-topicCore competenciesCommon pitfalls
6.1 ProbabilityCounting arguments, conditional probability, independence proofsMixing up mutually exclusive vs independent events; forgetting denominators in P(A∣B) P(A \mid B) .
6.2 Discrete random variablesProbability mass functions, expectation/variance, binomial X∼Bin⁡(n,p) X \sim \operatorname{Bin}(n, p) modelUsing incorrect parameters, neglecting scenario fit (e.g. independence assumptions).
6.3 Normal distributionStandardisation, properties of N(μ,σ2) \mathcal{N}(\mu, \sigma^2) , linear combinationsForgetting symmetry, mishandling variance scaling in Var⁡(aX+bY) \operatorname{Var}(aX + bY) .
6.4 SamplingXˉ \bar{X} as a random variable, E(Xˉ) E(\bar{X}) , Var⁡(Xˉ) \operatorname{Var}(\bar{X}) , CLT foundations (labelled “for teaching and learning only”)Applying CLT without stating “large n n ”; mixing sample SD with population σ \sigma .
6.5 Hypothesis testingFormulating H0/H1 H_0/H_1 for a population mean, critical regions, pp-valuesSwapping H0 H_0 and H1 H_1 ; misreporting conclusions in context; introducing excluded error terminology.
6.6 Correlation & regressionPMCC, least squares, transformationsExtrapolating beyond data range; interpreting correlation as causation; attempting excluded hypothesis tests.

2 | Sequencing Your Learning

  1. JC1 Term 1-2: Counting techniques (permutations/combinations), baseline probability, introduction to discrete random variables.
  2. JC1 Term 3: Binomial distribution applications, standard normal tables, first pass at sampling ideas.
  3. JC2 Term 1: Hypothesis testing workflow and CLT-powered approximations; practise linking statistics to real-life contexts (e.g. chemistry titration data).
  4. JC2 Term 2: Correlation/regression projects, statistical inference consolidation, exam drilling.

3 | Modelling Toolkit

  • Probability algebra: P(A∪B)=P(A)+P(B)−P(A∩B) P(A \cup B) = P(A) + P(B) - P(A \cap B) , P(Ac)=1−P(A) P(A^c) = 1 - P(A) , and the product rule for independent events.
  • Binomial mean/variance: E(X)=np E(X) = np , Var⁡(X)=np(1−p) \operatorname{Var}(X) = np(1 - p) .
  • Normal standardisation: Z=X−μσ Z = \dfrac{X - \mu}{\sigma} feeds the N(0,1) \mathcal{N}(0, 1) table; annotate diagrams to avoid sign mistakes.
  • Sampling variance: Var⁡(Xˉ)=σ2n \operatorname{Var}(\bar{X}) = \dfrac{\sigma^2}{n} ; when σ \sigma is unknown, use the unbiased sample variance s2 s^2 with divisor n−1 n - 1 .
  • Regression line: Use least squares to obtain y=a+bx y = a + bx ; check residuals before trusting predictions.

4 | Examination Strategy

  • Structure: Paper 1 is Pure Mathematics only. Paper 2 Section B (60 marks) contains the Probability & Statistics questions and demands interpretation, not just number crunching. Quote conclusions in plain English referencing the context (e.g. "There is sufficient evidence at the 5%5\% level that the mean tensile strength exceeds 4.2 kN \pu{4.2 kN} .").
  • Calculator readiness: Practise binomial/normal/PMCC commands on the GC; store sequences for quick re-computation.
  • Worked example workflow
    1. Sketch: tree diagram, distribution, or scatter plot.
    2. Compute with precise notation.
      e.g. P(X=3)=(103) 0.33 0.77 P\left(X = 3\right) = \binom{10}{3} \space 0.3^3 \space 0.7^7
    3. Conclude with units and wording tied to the scenario.

5 | What to Review from IP/O-Level

  • Additional Maths probability identities.
  • Sampling language from O-Level statistics (mean, variance formulations).
  • Linear graphs/transformations (to accelerate regression understanding).

Practice Quiz

Take a rapid pulse check across probability, distributions, sampling, and inference priorities for Section B.


6 | Next Steps

Dive into the detailed Topic 66 posts; each will provide:

  • Digestible concept summaries aligned with MOE's bullet list.
  • Step-by-step worked examples with calculator screenshots (where relevant).
  • Exam-style practice with annotated solutions.

Keep the entire roadmap handy via our H2 Maths notes hub, and start with H2 Maths Notes (JC 1-2): 6.1) Probability.


Frequently asked questions

Is Probability and Statistics in Paper 1 or Paper 2?
All of Topic 6 (Probability & Statistics) appears exclusively in Paper 2 Section B, which is worth 60 marks. Paper 1 is Pure Mathematics only.

Is the Central Limit Theorem (CLT) directly examinable?
Sub-topic 6.4 Sampling is labelled “for teaching and learning only” in the official syllabus, so the CLT itself is not directly tested as a standalone question. However, the sample-mean distribution Xˉ∼N(μ,σ2/n) \bar{X} \sim \mathcal{N}(\mu, \sigma^2/n) is used inside 6.5 Hypothesis Testing, so you must understand the ideas.

Can I use the graphing calculator for all statistics questions?
Yes. SEAB allows GC use throughout Paper 2, and most mark schemes include GC commands as valid working. However, you must still write down the distribution model, show the key probability statement, and interpret the answer in context.

Is normal approximation to the binomial tested in 2026?
No. The 2026 H2 Maths syllabus explicitly excludes the normal approximation to the binomial distribution. Treat binomial and normal as completely separate models in your exam preparation.


Sources