Assessment Overview & Difficulty Verdict
The 2025 CCEA GCSE Statistics Higher Tier assessment consists of two 2-hour calculator papers (Unit 1 GST12 and Unit 2 GST22), totaling 200 marks. The paper provides a well-calibrated distribution of difficulty, rating at 3.4 / 5.0. Standard procedural tasks (such as estimating means from grouped tables, basic probability trees, and standard stem-and-leaf plots) provide an accessible base of grade C/B marks, while higher-grade differentiation (A/A*) is achieved through multi-step statistical distributions, chain-base geometric means, comparative pie chart radii, and nuanced statistical interpretations.
Where the Marks Lie
Major mark concentrations are found across four main pillars:
- Time Series & Index Numbers: 4-point centered moving averages, trend line forecasting, chain-base index calculations, and geometric average rate of change.
- Probability & Distributions: Conditional probability using Venn and tree diagrams (without replacement), empirical normal distribution bounds \((\mu \pm k\sigma)\), standardized \(z\)-scores, and binomial expansion \((p+q)^5\).
- Quality Control & Reasoning: Control charts with warning/action limits, evaluation of statistical stability, and identifying misrepresentation in charts.
- Data Presentation & Spread: Stepped cumulative frequency, continuous cumulative curves with box plots, and class-width effects on mean reliability.
Key Pitfalls & Examiner Advice
Common areas where candidates forfeit marks include:
- Failure to contextualise: Stating generic answers such as 'it increases' rather than explicit statistical explanations (e.g., 'for every £1 million increase in transfer spending, points increase by 4.0 on average').
- Conditional Probability Denominators: Forgetting to restrict the sample space to the given condition in Venn and tree diagrams (e.g., \(P(\text{Both Red} \mid \text{At least one Red})\)).
- Comparative Pie Chart Radii: Omitting the square-root relationship when sizing radii proportional to sample frequencies (\(r_1 / r_2 = \sqrt{f_1 / f_2}\)).
- Extrapolation Warnings: Using lines of best fit or trend lines far beyond the observed domain without acknowledging model invalidity.
Revision & Exam Strategy
Focus high-yield revision on multi-step modeling problems: practice setting up binomial expressions using given polynomial expansions, calculating geometric means of index numbers \(\sqrt[n]{\prod I_k}\), and memorising key properties of standard distributions. Always include units, clear formulas, and double-check graph axis scales.