A-Level Maths: Hypothesis Testing

Hypothesis testing is where A-Level statistics becomes an exercise in disciplined language: the marks hang on stating hypotheses correctly, interpreting p-values precisely, and concluding in context without overclaiming. This quiz trains that discipline directly.

The framework questions pin down the vocabulary the mark schemes police: the significance level as the probability of rejecting a true null hypothesis (the Type I error rate), the p-value as the probability of observing data at least as extreme under the null, and the exact meaning of "failing to reject" — insufficient evidence for the alternative, never proof that the null is true. That last distinction gets its own multi-select question because it is the single most common conclusion error on real scripts.

The setup questions cover hypothesis formulation: the one-tailed pair for a coin suspected of bias towards heads (H₁: p > 0.5), the null hypothesis for a correlation test (ρ = 0), and how a two-tailed test at 5% splits its significance level into 2.5% in each tail — the mechanical detail that decides critical regions.

The Normal-mean questions handle the distribution theory: the test statistic with standard error σ/√n in its denominator, and the distribution of the sample mean under H₀ for n = 25 and σ = 10, where the variance must shrink to 100/25 = 4. Decision logic appears through the critical-region question — a statistic falling inside it means rejection — and a binomial question that finds the critical region for B(20, 0.3) at the 5% level.

Each explanation states the definition and then applies it, modelling the two-part answers that written papers reward.

• State null and alternative hypotheses for one- and two-tailed tests
• Define significance levels and p-values precisely
• Interpret 'failing to reject' without overclaiming
• Use the distribution of the sample mean with standard error σ/√n
• Apply critical-region logic and find a binomial critical region

Topic scope follows section O (Statistical hypothesis testing) of the DfE's prescribed AS and A level mathematics subject content — 100% common across all exam boards: hypothesis language, testing a binomial proportion, testing a Normal mean via the sample mean's distribution, and testing correlation.