Statistical hypothesis testing (A2 Unit 4: Applied Mathematics B) — WJEC A-Level Mathematics
Statistical hypothesis testing (A2 Unit 4: Applied Mathematics B) explained
Statistical hypothesis testing (A2 Unit 4: Applied Mathematics B) covers 3 specification points, including Understand and apply statistical hypothesis… and (The calculation of correlation coefficients….
Read the full explanation
Open a point below for its explanation and revision guidance.
Understand and apply statistical hypothesis testing to correlation coefficients as measures of how close data points lie to a straight line and be able to interpret a given correlation coefficient using a given p-value or critical value.
The Product Moment Correlation Coefficient (PMCC) r measures the strength and direction of linear association between bivariate normal variables, bounded by −1 ≤ r ≤ 1. To test whether an observed sample correlation r indicates a genuine linear association in the population, candidates formulate hypotheses about the population correlation coefficient ρ: H₀: ρ = 0 against H₁: ρ > 0, ρ < 0, or ρ ≠ 0. Candidates compare the sample r against tabulated critical values for sample size n at significance level α, or compare the provided p-value against α. If |r| exceeds the critical value (or p < α), H₀ is rejected in favour of significant linear correlation.
(The calculation of correlation coefficients is excluded.)
In WJEC A-Level Mathematics, manual calculation of the Product Moment Correlation Coefficient r from raw summation formulas (S_xx, S_yy, S_xy) is excluded. Instead, numerical values of r are supplied directly in examination questions or retrieved from calculator statistical functions. The assessable focus is interpretive: understanding that r measures linear association, testing the significance of r using critical value tables or given p-values, recognising that r close to zero does not rule out non-linear association, and evaluating how outliers or restricted sampling ranges can distort r.
Conduct a statistical hypothesis test for the mean of a Normal distribution with known, given or assumed variance, and interpret the results in context.
When testing a claim about the population mean μ of a Normal distribution X ~ N(μ, σ²) with known variance σ²: 1. State H₀: μ = μ₀ and H₁: μ < μ₀, μ > μ₀, or μ ≠ μ₀. 2. For a sample of size n, the sample mean follows X̄ ~ N(μ, σ²/n). 3. The test statistic is the standardised score Z = (X̄ − μ₀) / (σ / √n), or the observed sample mean X̄ is compared directly to critical bounds. 4. At significance level α, compare Z to standard normal critical values (e.g. ±1.645 for 5% one-tailed, ±1.960 for 5% two-tailed), or calculate the p-value. 5. Conclude: state whether to reject H₀ and provide a contextual conclusion.
Your focus
- Formulate null and alternative hypotheses for correlation tests using population parameter ρ.
- Evaluate sample correlation coefficients against critical values or p-values at given significance levels.
- Interpret test outcomes in practical context, avoiding claims of causal determination.
Show all 9 objectives
- Interpret provided PMCC values in terms of the direction and strength of linear association.
- Execute hypothesis tests on correlation using given r values and statistical tables.
- Explain why correlation coefficients close to zero do not preclude strong non-linear relationships.
- Formulate hypotheses for population mean tests using correct parameter notation μ.
- Calculate test statistics using the sampling distribution of the mean X̄ ~ N(μ, σ²/n).
- Execute one-tailed and two-tailed Normal mean tests and interpret findings in context.
Statistical hypothesis testing (A2 Unit 4: Applied Mathematics B) exam tips
Marking Points
- stating hypotheses in terms of the population correlation coefficient ρ: H₀: ρ = 0 and H₁: ρ > 0, ρ < 0, or ρ ≠ 0
- looking up or quoting the correct critical value for sample size n at the specified significance level
- comparing sample r to the critical value (or p-value to α) and stating a clear statistical decision
- writing a non-dogmatic conclusion in context regarding the linear correlation between the two variables
- interpreting a provided sample correlation coefficient r in terms of strength and direction
- applying critical values or p-values to conduct a hypothesis test without deriving r manually
- explaining that r measures only linear association and cannot detect curved relationships
- stating hypotheses correctly using the population mean μ: H₀: μ = μ₀ and H₁: μ < μ₀, > μ₀, or ≠ μ₀
- identifying the sampling distribution of the mean X̄ ~ N(μ, σ²/n) with standard error σ/√n
- calculating the Z test statistic Z = (X̄ − μ₀) / (σ / √n) or finding the critical region bounds
- comparing with critical values or significance level and stating a contextual conclusion
Examiner Tips
- 💡Always write the Greek letter ρ (rho) for hypotheses: H₀: ρ = 0, H₁: ρ ≠ 0 (or > 0 or < 0).
- 💡For two-tailed tests at significance level α, use the column for α/2 in one-tail tables or look up the two-tail column.
- 💡Use the given r value directly: do not attempt to calculate it from summary formulas.
- 💡Remember that r measures linear correlation only; a scatter plot with a parabolic curve can have r ≈ 0.
- 💡Remember that the variance of the sample mean is σ²/n, so the standard deviation of X̄ is σ/√n.
- 💡For two-tailed tests at 5%, the critical Z values are ±1.960; for 1%, they are ±2.576.
Common Mistakes
- stating hypotheses in terms of sample correlation r (e.g. H₀: r = 0) rather than population parameter ρ
- concluding that rejecting H₀ proves a causal relationship exists between the two variables
- wasting examination time computing S_xy, S_xx, and S_yy when r is provided or calculated via calculator
- believing r = 0 proves that the two variables are completely independent when a non-linear relationship exists
- using σ instead of the standard error σ/√n in the denominator of the test statistic Z
- stating hypotheses in terms of the sample mean x̄ instead of the population mean μ