WJEC Year 13 Statistics: In-Depth Past Paper Analysis | 历年真题深度解析

📚 WJEC Year 13 Statistics: In-Depth Past Paper Analysis | 历年真题深度解析

WJEC Year 13 Statistics requires not only a solid understanding of core concepts but also the ability to apply them under timed conditions. By analysing past papers deeply, we uncover patterns in question style, common pitfalls, and efficient solution strategies. This article walks you through key topics with worked examples and annotations that reflect actual exam demands.

WJEC 十三年级统计考试不仅要求扎实掌握核心概念，还要求能在限时条件下灵活应用。通过深度分析历年真题，我们可以发现题型规律、常见陷阱和高效解题策略。本文将带你梳理关键主题，配有符合实际考试要求的工作示例和注释。

1. Exam Overview and Mark Allocation | 考试概览与分值分配

Year 13 Statistics (typically Unit 3 or A2 unit) is assessed through a written paper lasting around 1 hour 45 minutes, carrying 80–100 marks. Questions often mix pure statistical computation with interpretation and comment. Marks are allocated for method (M), accuracy (A), and sometimes for concluding in context (E).

十三年级统计（通常为单元3或A2单元）通过约1小时45分钟的笔试考核，总分80–100分。试题经常将纯统计计算与解释、评论相结合。分数分布于方法分（M）、准确度分（A），有时还涉及在情境中下结论（E）。

Typical topics tested include: discrete and continuous random variables, expectation algebra, Poisson and normal distributions, sampling distributions, confidence intervals, hypothesis tests for means and proportions, chi-squared tests, and regression/correlation. The balance of marks tends to favour probability distributions, confidence intervals, and hypothesis testing.

常考主题包括：离散与连续随机变量、期望的代数运算、泊松分布与正态分布、抽样分布、置信区间、均值和比例的假设检验、卡方检验以及回归与相关。分值权重倾向于概率分布、置信区间和假设检验。

Topic	Approx. marks
Probability & discrete distributions	15–20
Poisson & normal distributions	18–25
Confidence intervals & hypothesis tests	25–30
Chi-squared & regression	15–20

主题与分值：概率与离散分布约15–20分，泊松与正态分布18–25分，置信区间与假设检验25–30分，卡方与回归15–20分。

2. Probability and Counting Techniques | 概率与计数技巧

Combinations, permutations, and the product rule appear frequently in paper beginnings to set up probability distributions. A typical WJEC question might ask: “A bag contains 5 red, 3 blue and 2 green balls. Three balls are drawn without replacement. Find the probability that exactly two are red.”

组合、排列和乘法原理经常出现在试卷开头，用于构建概率分布。典型的WJEC考题可能问：“袋中有5个红球、3个蓝球和2个绿球。不放回地抽取3个球。求恰好两个红球的概率。”

Always define the sample space carefully. For without replacement, use combinations: total ways to choose 3 from 10 is ¹⁰C₃ = 120. Favorable ways: ⁵C₂ × ⁵C₁ = 10 × 5 = 50. The probability is 50/120 = 5/12. Some candidates mistakenly treat it as with replacement or overlook the ‘other’ selection.

必须仔细定义样本空间。不放回时使用组合：从10个中选3个的总数为¹⁰C₃ = 120。有利方式：选2个红球有⁵C₂ = 10，再选1个非红球有⁵C₁ = 5，共50。概率为50/120 = 5/12。有些考生错误地当作有放回处理，或忽略“其他”的选择。

Conditional probability also features heavily. Past paper example: “Given that the first ball drawn is red, find the probability the second is also red.” Apply the conditional formula directly. Use tree diagrams to visualise – they often attract method marks even if arithmetic slips.

条件概率也大量出现。真题示例：“已知第一个抽出的球是红的，求第二个也是红的概率。”直接应用条件概率公式。使用树状图来可视化——即使计算有误，树图通常也能获得方法分。

P(B|A) = P(A ∩ B)/P(A)

3. Discrete Random Variables and Expectation | 离散随机变量与期望

Expectation and variance formulas are essential: E(X) = Σx·P(X=x), Var(X) = E(X²) – [E(X)]². Past papers test linear combinations thoroughly: E(aX + b) = aE(X) + b, Var(aX + b) = a²Var(X). A common extension is E(X+Y) = E(X)+E(Y), and Var(X+Y) = Var(X)+Var(Y) if independent.

期望和方差公式至关重要：E(X) = Σx·P(X=x)，Var(X) = E(X²) – [E(X)]²。历年真题深入考查线性组合：E(aX + b) = aE(X) + b，Var(aX + b) = a²Var(X)。常见延伸：E(X+Y) = E(X)+E(Y)，且若独立则Var(X+Y) = Var(X)+Var(Y)。

Worked example: The probability distribution of X is given. Find E(2X–3) and Var(2X–3). Compute E(X) first, then apply the transformations. Marks are awarded for recognising that the constant ‘–3’ does not affect variance.

工作示例：给定X的概率分布，求E(2X–3)和Var(2X–3)。先计算E(X)，再应用变换。认识到常数’–3’不影响方差，此点可得方法分。

Some questions embed ‘find the constant k’ before calculating probabilities, relying on Σ P(X=x) = 1. Always verify your k by summing the probabilities again – a quick sanity check prevents cascading errors.

有些问题在计算概率前需要“求常数k”，依靠Σ P(X=x) = 1。求出k后一定要将概率重新求和验证——这个快速检查能避免连锁错误。

4. Binomial Distribution | 二项分布

Recognition of a binomial setting is a key skill: fixed number of trials, two outcomes, constant probability, independence. WJEC questions often embed binomial random variables within real-life contexts: defective items, responses to a survey.

识别二项分布场景是一项关键技能：固定试验次数、两种结果、概率恒定、各次试验独立。WJEC题目常将二项随机变量嵌入现实场景：次品数量、问卷调查回应等。

Use the notation X ~ B(n, p). Probability mass function: P(X = r) = ⁿCᵣ pʳ (1–p)ⁿ⁻ʳ. The exam expects efficient use of the formula booklet or cumulative tables. A classic error is mixing up n and p: p is the probability of the event being counted.

使用符号 X ~ B(n, p)。概率质量函数：P(X = r) = ⁿCᵣ pʳ (1–p)ⁿ⁻ʳ。考试要求熟练运用公式手册或累积概率表。经典错误是混淆n和p：p是被计数事件的概率。

For large n, the binomial can be approximated by a Poisson or normal distribution. The paper often asks: “State which distribution you would use to approximate…” and then “Justify your choice.” Justification must mention the conditions: n is large, p is small and np is moderate → Poisson; np > 5 and n(1–p) > 5 → normal (with continuity correction).

对于大n值，二项分布可用泊松或正态分布近似。真题常问：“你会用哪种分布来近似…”然后“说明你的选择理由。”理由必须提及条件：n大、p小且np适中→泊松；np>5且n(1–p)>5→正态（带连续性校正）。

Past paper nuance: when using normal approximation, they frequently test continuity correction. P(X ≥ r) becomes P(X > r–0.5). Failing to apply the half-unit adjustment loses accuracy marks.

真题细微之处：使用正态近似时经常测试连续性校正。P(X ≥ r) 变为 P(X > r–0.5)。未进行半单位调整会丢失精度分。

5. Poisson Distribution | 泊松分布

Poisson processes model random events in a fixed interval (time, area, volume). Parameter λ represents the mean number of occurrences. A classic WJEC scenario: “Cars arrive at a toll booth at an average rate of 4 per minute. Find the probability that in a 2-minute interval, at most 3 cars arrive.” Here λ = 8.

泊松过程对固定区间（时间、面积、体积）内的随机事件建模。参数λ表示平均发生次数。WJEC经典场景：“汽车以每分钟4辆的平均速率到达收费站。求在2分钟间隔内至多有3辆车到达的概率。”此处λ = 8。

Scaling λ to match the interval is crucial. If rate per minute is μ, then over t minutes λ = μt. Common mistake: forgetting to multiply. The calculation then uses tables or formula: P(X ≤ 3) = Σ (e⁻⁸ × 8ˣ)/x! for x = 0 to 3.

将λ调整到匹配时间间隔至关重要。若每分钟速率为μ，则t分钟内的λ = μt。常见错误：忘记相乘。然后使用查表或公式计算：P(X ≤ 3) = Σ (e⁻⁸ × 8ˣ)/x!，x从0到3。

Sum of independent Poisson variables: if X ~ Po(λ₁) and Y ~ Po(λ₂) independently, then X + Y ~ Po(λ₁ + λ₂). Past papers often combine two sources and ask for the probability the total exceeds a threshold, immediately testing this addition property.

独立泊松变量之和：若X ~ Po(λ₁)，Y ~ Po(λ₂)且独立，则X + Y ~ Po(λ₁ + λ₂)。真题常将两个来源合并，要求计算总量超过某阈值的概率，直接考查这一可加性。

When the question provides rates ‘per metre’ for flaws in fabric, and asks about a 5-metre length, compute λ for the length. Remember: if flaws occur at 0.3 per metre, λ for 5 m = 1.5. Clear reasoning lines in the solution help secure method marks.

当题目给出织物瑕疵“每米”的发生率，并问5米长度的情况时，计算该长度对应的λ。记住：若瑕疵率为每米0.3，则5米的λ = 1.5。解题时清晰的推理步骤有助于拿到方法分。

6. Normal Distribution and Approximations | 正态分布与近似

The normal distribution N(μ, σ²) is the backbone of many topics. Standardizing: Z = (X – μ)/σ. WJEC questions often start with “Given X ~ N(50, 4²), find P(X < 45)." Use the standard normal table for Φ(z) carefully, noting that tables may give lower tail or upper tail probabilities.

正态分布N(μ, σ²)是许多主题的基石。标准化：Z = (X – μ)/σ。WJEC题目常以“已知X ~ N(50, 4²), 求P(X < 45)”开始。使用标准正态分布表时要小心，注意表格给出的是下尾还是上尾概率。

Finding an unknown mean or standard deviation is a favoured exam question. Example: “X ~ N(μ, 3²). Given that P(X > 20) = 0.025, find μ.” Set up (20 – μ)/3 = z-value such that Φ(z)=0.975, hence z = 1.96. Then solve. Many students mix up the sign or misread the table.

求未知均值或标准差是偏爱的题型。示例：“X ~ N(μ, 3²)。已知P(X > 20) = 0.025，求μ。”建立方程 (20 – μ)/3 = z值，使得Φ(z)=0.975，因此z = 1.96，然后求解。许多学生混淆符号或读错表。

Normal approximation to binomial or Poisson carries a continuity correction. For binomial, if X ~ B(n, p), then X ≈ N(np, np(1–p)). WJEC requires the correction unless otherwise stated. Write P(X ≤ r) ≈ P(Y < r + 0.5) where Y ~ N(np, np(1–p)). This small addition can be worth 1–2 marks.

二项或泊松的正态近似需进行连续性校正。对于二项分布，若X ~ B(n, p)，则X ≈ N(np, np(1–p))。WJEC除非另有说明，否则要求使用校正。记为P(X ≤ r) ≈ P(Y < r + 0.5)，其中Y ~ N(np, np(1–p))。这个小小的加项可值1–2分。

P(X ≤ r) ≈ Φ((r + 0.5 – np)/√(np(1–p)))

7. Sampling and Central Limit Theorem | 抽样与中心极限定理

When sampling from any distribution with mean μ and variance σ², the sample mean x̄ has distribution: E(x̄) = μ, Var(x̄) = σ²/n. If the population is normal, x̄ is exactly normal; otherwise, the Central Limit Theorem (CLT) states x̄ is approximately normal for large n (typically n ≥ 30).

从均值为μ、方差为σ²的任意分布中抽样时，样本均值x̄的分布为：E(x̄) = μ，Var(x̄) = σ²/n。若总体服从正态分布，则x̄精确服从正态；否则，中心极限定理（CLT）指出当样本量较大（通常n ≥ 30）时，x̄近似服从正态分布。

WJEC frequently asks: “State the distribution of the sample mean” and then tests confidence intervals or hypothesis tests based on that distribution. For unknown population variance, the t-distribution is used when n is small; however, at Year 13, large-sample z-procedures are more common unless specified.

WJEC常问：“写出样本均值的分布”，然后基于该分布考查置信区间或假设检验。当总体方差未知且样本量较小时，使用t分布；但十三年级更常见大样本z过程，除非另有说明。

Example from past papers: “A random sample of 50 items from a population with μ = 20 and σ = 5 is taken. Find the probability that the sample mean exceeds 21.” Standard error = 5/√50 ≈ 0.7071. Then z = (21–20)/0.7071 ≈ 1.414, giving P(Z > 1.414) ≈ 0.0786.

真题示例：“从μ = 20、σ = 5的总体中随机抽取50个样本。求样本均值超过21的概率。”标准误 = 5/√50 ≈ 0.7071。接着z = (21–20)/0.7071 ≈ 1.414，则P(Z > 1.414) ≈ 0.0786。

8. Confidence Intervals | 置信区间

Confidence interval for a population mean with known variance: x̄ ± z* × σ/√n, where z* is the critical value (e.g. 1.96 for 95%). The interval interpretation is frequently examined: “We are 95% confident that the interval captures the true mean.” Avoid saying “probability that μ is in the interval”.

总体方差已知时均值的置信区间：x̄ ± z* × σ/√n，其中z*为临界值（如95%对应1.96）。区间解释是常考点：“我们有95%的把握认为该区间包含真实均值。”避免表述为“μ落在区间内的概率”。

For proportions, the confidence interval uses p̂ ± z* × √(p̂(1–p̂)/n). WJEC markers expect the check that n is sufficiently large (n p̂ ≥ 5, n(1–p̂) ≥ 5) before constructing the interval. Omitting this check can cost a mark.

关于比例，置信区间为p̂ ± z* × √(p̂(1–p̂)/n)。WJEC阅卷人期望在构建区间前核实样本量足够大（n p̂ ≥ 5, n(1–p̂) ≥ 5）。忽略此检查可能失分。

A challenging twist: “Find the smallest sample size such that a 95% confidence interval has width at most 0.1.” Set the margin of error z* × σ/√n ≤ 0.05 (since width = 2 × margin). Solving for n gives the required size. Be ready to round up to the next integer.

一个有难度的变体：“求使95%置信区间宽度最多为0.1的最小样本量。”令误差边际z* × σ/√n ≤ 0.05（因宽度=2×误差边际）。求解n得到所需样本量，记住要向上取整。

n ≥ (1.96σ / 0.05)²

9. Hypothesis Testing for Means and Proportions | 均值和比例的假设检验

Set up null hypothesis H₀: μ = μ₀, alternative H₁: μ ≠ μ₀ (or one-tailed). Compute test statistic z = (x̄ – μ₀)/(σ/√n). Compare with critical value or find p-value. Conclude in context: “There is sufficient evidence at the 5% level to reject H₀ and suggest the mean has changed.”

设定原假设H₀: μ = μ₀，备择假设H₁: μ ≠ μ₀（或单尾）。计算检验统计量z = (x̄ – μ₀)/(σ/√n)。与临界值比较或求p值。在情境中下结论：“在5%显著性水平下，有足够证据拒绝H₀，表明均值已改变。”

For proportions, the test statistic is z = (p̂ – p₀)/√(p₀(1–p₀)/n). Note the denominator uses the hypothesized p₀, not p̂. This distinction is often missed by students, leading to an incorrect standard error.

对于比例，检验统计量为z = (p̂ – p₀)/√(p₀(1–p₀)/n)。注意分母使用假设的p₀而非p̂。这一区别常被学生忽略，导致标准误错误。

Type I and Type II errors: A Type I error occurs if H₀ is true but we reject it (false positive). Its probability is α. Type II error is failing to reject H₀ when H₁ is true. Questions may ask to calculate P(Type II error) for a specific alternative mean – this is a mark differentiator.

第一类错误和第二类错误：若H₀为真却拒绝H₀（假阳性），发生第一类错误，其概率为α。第二类错误是当H₁为真时未能拒绝H₀。可能要求计算特定备择均值下的P(第二类错误)——这是拉开分数差距的题目。

Worked example: H₀: μ=100, σ=15, n=25, x̄=105, H₁: μ>100. z = (105-100)/(15/5)=1.67. Critical value at 5% is 1.645. Since 1.67 > 1.645, reject H₀. Some papers require the exact p-value: P(Z > 1.67) = 0.0475.

工作示例：H₀: μ=100, σ=15, n=25, x̄=105, H₁: μ>100。z = (105-100)/(15/5)=1.67。5%临界值为1.645。由于1.67 > 1.645，拒绝H₀。有些试卷要求精确p值：P(Z > 1.67)=0.0475。

10. Chi-Squared Tests | 卡方检验

Chi-squared tests assess goodness of fit or independence. Test statistic: χ² = Σ (O – E)²/E. WJEC papers typically include a table of observed frequencies, and you must calculate expected frequencies based on the null hypothesis. Degrees of freedom must be stated.

卡方检验评估拟合优度或独立性。检验统计量：χ² = Σ (O – E)²/E。WJEC试卷通常提供一张观测频数表，需要根据原假设计算期望频数。自由度必须明确写出。

For goodness of fit: If no parameters are estimated, df = categories – 1. If parameters are estimated from the data (e.g. fitting a binomial with estimated p), then df = categories – 1 – number of estimated parameters. Candidates often lose marks by giving incorrect df.

拟合优度检验：若未估计参数，自由度 = 类别数 – 1。若从数据中估计了参数（如拟合二项分布并估计p值），则自由度 = 类别数 – 1 – 所估计参数的个数。考生常因给出错误的自由度而失分。

Test for independence in a contingency table: df = (rows–1)×(columns–1). The null hypothesis is that the variables are independent. A high χ² value relative to critical value suggests association. Always combine cells if any expected frequency falls below 5 – WJEC rewards stating this requirement explicitly.

列联表独立性检验：自由度 = (行数–1)×(列数–1)。原假设为变量独立。相对于临界值较高的χ²值表明存在关联。若任何期望频数低于5，必须合并单元格——WJEC对明确陈述此要求给予奖励分。

χ² = Σ (O – E)²/E, with Yates’ correction if df=1

χ² = Σ (O – E)²/E，若自由度为1需使用Yates校正

11. Correlation and Regression | 相关与回归

Product moment correlation coefficient r measures linear association. Interpretation of r² (coefficient of determination) is routinely required: “r² = 0.85 means 85% of the variation in y can be explained by the linear relationship with x.” This contextual comment scores highly.

积矩相关系数r衡量线性关联。对r²（决定系数）的解释是常规要求：“r² = 0.85意味着y中85%的变异可由与x的线性关系解释。”这种情境评论能得高分。

Least squares regression line: y = a + bx, where b = Sxy/Sxx and a = ȳ – b x̄. WJEC often provides summary statistics ∑x, ∑y, ∑x², ∑y², ∑xy, and you must compute the coefficients accurately to 3 significant figures.

最小二乘回归线：y = a + bx，其中b = Sxy/Sxx，a = ȳ – b x̄。WJEC常提供汇总统计量∑x, ∑y, ∑x², ∑y², ∑xy，你必须准确计算出系数并保留三位有效数字。

A common pitfall is using incorrect formula for Sxy when coding is involved. If variables are coded as u = (x – h)/k, then recalculating summary stats for u is safer. Always check your regression line by seeing if it passes through (x̄, ȳ).

常见陷阱是编码后使用了错误的Sxy公式。若变量经编码u = (x – h)/k，重新计算u的汇总统计量更为稳妥。始终通过检查回归线是否通过(x̄, ȳ)来验证。

Confidence intervals for the slope or prediction intervals may also appear. They rely on standard error of the slope and t-distribution. Even if the full interval is not required, a question might ask “Explain why the slope is significantly different from zero,” linking back to hypothesis testing.

斜率的置信区间或预测区间也可能出现。它们依赖于斜率的标准误和t分布。即使不要求完整区间，也可能问“解释为什么斜率显著不为零”，这又回到了假设检验。

12. Common Pitfalls and Exam Tips | 常见陷阱与考试技巧

Many marks are lost through misinterpretation. Read the question twice: underline ‘exactly’, ‘at least’, ‘more than’, ‘between’. In probability, translate ‘at most 2’ to P(X ≤ 2), ‘at least 3’ to 1 – P(X ≤ 2). Use diagram or distribution notation to avoid slips.

许多分数因误解题意而丢失。仔细读题：划出‘恰好’、‘至少’、‘多于’、‘介于’。在概率中，将‘至多2个’转化为P(X ≤ 2)，‘至少3个’转化为1 – P(X ≤ 2)。使用图示或分布符号以避免失误。

Write hypotheses with population parameters, not sample statistics. H₀: μ = 100, not x̄ = 100. Include units where applicable. In contextual conclusions, avoid vague phrases like “enough evidence” unless you explicitly state the significance level.

书写假设时要用总体参数而非样本统计量。H₀: μ = 100，而非x̄ = 100。适当时机要带上单位。在情境结论中，避免使用模糊词汇，如“足够证据”，除非明确说明显著性水平。

Time management: tackle the paper in order but mark questions you find difficult and return later. Show substitution into formulas clearly; even if the final answer is wrong, you can secure method marks. Use the formula booklet efficiently – know exactly where each formula is.

时间管理：按顺序答题，但标记出难题稍后返回。清晰地展示代入公式的过程；即使最终答案错误，也能获得方法分。高效使用公式

Published by TutorHao | Year 13 统计 Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

WJEC Year 13 Statistics: In-Depth Past Paper Analysis | 历年真题深度解析