📚 High-Frequency Topics and Common Mistakes in Year 13 WJEC Statistics | Year 13 WJEC 统计：高频考点与易错题分析
Year 13 WJEC Statistics builds on the foundations of probability, data handling, and inference, introducing more advanced techniques such as the Central Limit Theorem, bivariate data analysis, hypothesis testing for correlation, and goodness-of-fit tests. In this article, we identify the most frequently examined topics and the errors candidates make year after year, so you can approach revision with clarity and confidence.
Year 13 WJEC 统计课程在概率、数据处理和统计推断的基础上，进一步引入了中心极限定理、双变量数据分析、相关系数假设检验以及拟合优度检验等更高级的内容。本文将梳理最高频的考点，并分析考生年复一年出现的常见错误，帮助你更有针对性地复习和提升。
1. Applying the Central Limit Theorem | 中心极限定理的运用
sampling distribution of the sample mean will be approximately normal, even if the population distribution is not normal. Its mean is the population mean μ, and its standard error is σ/√n.
中心极限定理指出，只要样本量足够大（通常 n ≥ 30），样本均值的抽样分布就近似服从正态分布，即便总体分布不是正态的。其均值为总体均值 μ，标准误差为 σ/√n。
A very common mistake is applying the CLT when the sample size is too small or when the population variance is unknown and no estimate is provided. Students also forget to state the distribution of the sample mean explicitly before calculating probabilities.
一个非常常见的错误是在样本量太小时就套用中心极限定理，或者在总体方差未知又未提供估计值时直接使用。学生也经常忽略在计算概率之前明确写出样本均值的分布。
In exam questions, you must often calculate P(X̄ > a) by standardising to Z = (X̄ - μ) / (σ/√n) and using normal tables. Be careful to adjust for continuity if the original variable is discrete, though this is rare at this level.
在考试题中，你需要通过标准化 Z = (X̄ - μ) / (σ/√n) 并查正态分布表来计算概率 P(X̄ > a)。如果原始变量是离散的，要注意进行连续性校正，不过这在当前级别较少涉及。
2. Confidence Intervals for Means and Proportions | 均值和比例的置信区间
For a population mean with known variance, the 95% confidence interval is x̄ ± 1.96 × σ/√n ndard deviation s, the t-distribution with n-1 degrees of freedom should be used, giving x̄ ± t × s/√n.
对于已知方差的总体均值，95% 置信区间为 x̄ ± 1.96 × σ/√n。当 σ 未知并用样本标准差 s 估计时，应使用自由度为 n-1 的 t 分布，即 x̄ ± t × s/√n。
Use of the normal distribution instead of the t-distribution when σ is estimated is a classic error. This mistake is penalised heavily because it shows a misunderstanding of the extra uncertainty introduced by estimating σ.
当 σ 被估计时仍使用正态分布而非 t 分布是一个典型错误。这个错误会被严重扣分，因为它显示出对估计 σ 所引入的额外不确定性缺乏理解。
For a population proportion p, the interval is p̂ ± z √[p̂(1-p̂)/n]. The condition for normality here is that np̂ and n(1-p̂ depending on the specification). Many candidates proceed without checking this condition.
对于总体比例 p，置信区间为 p̂ ± z √[p̂(1-p̂)/n]。这里要求 np̂ 和 n(1-p̂) 都大于 5（或 10，视考试大纲而定）。许多考生在未检查此条件的情况下直接计算。
3. Interpreting Confidence Intervals Correctly | 正确解读置信区间
A 95% confidence interval does not mean there is a 95% chance that the true parameter lies within the interval. Instead, it means that if we were to take many samples and compute a confidence interval from each, about 95% of those intervals would capture the true parameter.
95% 置信区间并不意味着真实参数有 95% 的概率落在该区间内。它的正确含义是：如果我们重复抽取很多样本，并为每个样本计算置信区间，那么大约 95% 的区间会包含真实的参数值。
Misinterpretation of confidence intervals is one of the most frequently tested ideas in WJEC Statistics exams. Always use wording such as "we are 95% confident that the interval ... captures the true mean" rather than probabilistic language about the parameter.
对置信区间的错误解读是 WJEC 统计考试中最常考的概念之一。始终使用诸如“我们有 95% 的置信度认为该区间……包含了真实均值”之类的表述，而不要用关于参数的概率性语言。
4. Product Moment Correlation Coefficient (PMCC) | 积矩相关系数（PMCC）
trength and direction of a linear relationship between two variables. Its formula is r = Sxy / √Sxx Syy, where Sxy = Σ(x - x̄)(y - ȳ). A common error is mixing up Syy and Sxx, or forgetting to square the denominator properly.
积矩相关系数 r 衡量两个变量之间线性关系的强度和方向。其公式为 r = Sxy / √Sxx Syy，其中 Sxy = Σ(x - x̄)(y - ȳ)。常见错误包括混淆 Syy 和 Sxx，或忘记对分母正确开平方。
Examiners frequently ask candidates to interpret an r value in context. Saying "there is a strong positive correlation" is insufficient; you must link it to the variables, for example, "as the temperature increases, the number of ice creams sold tends to increase."
考官经常要求考生在具体情境中解读 r 值。仅仅说“存在强正相关”是不够的；你必须联系变量，例如“随着温度升高，冰淇淋的销售量往往增加”。
5. Hypothesis Testing for Zero Correlation Using PMCC | 使用 PMCC 检验零相关
n correlation table with n-2 degrees of freedom. The null hypothesis is H₀: ρ = 0, and the alternative can be one-tailed or two-tailed.
为了检验总体相关系数 ρ 是否为零，我们使用检验统计量 r，并将其与皮尔逊相关系数表（自由度为 n-2）中的临界值进行比较。零假设为 H₀: ρ = 0，备择假设可以是单尾或双尾。
A frequent mistake is using the wrong degrees of freedom (n-1 instead of n-2) or applying a normal approximation when the sample size does not warrant it. WJEC tables provide exact critical values, and candidates should use them.
一个常见错误是使用错误的自由度（用 n-1 而不是 n-2），或者在样本量不够大时使用了正态近似。WJEC 的表格提供了精确的临界值，考生应该使用这些值。
Another pitfall is failing to state the conclusion in context. You must say something like "reject H₀, there is evidence of a linear relationship between ..." linking back to the variables studied.
另一个陷阱是没有结合情境陈述结论。你必须像这样说“拒绝 H₀，有证据表明……之间存在线性关系”，并将结论联系回所研究的变量。
6. Spearman’s Rank Correlation and Its Hypothesis Test | 斯皮尔曼秩相关及其假设检验
When data are not linear or are given as ranks, Spearman’s coefficient rs is used. It is calculated as 1 - (6Σd²) / [n(n²-1)], where d is the difference between paired ranks. It requires that there are no tied ranks, or a correction is needed.
当数据不是线性的或以排名形式给出时，使用斯皮尔曼系数 rs。计算公式为 1 - (6Σd²) / [n(n²-1)]，其中 d 是配对排名之差。要求没有相同排名，否则需要修正。
Errors arise when students apply the Pearson table for Spearman’s test. While the same critical values often work asymptotically, the correct approach is using the Spearman table if provided, or noting the approximation. Also, forgetting to rank data correctly (ascending/descending) is a common slip.
学生在斯皮尔曼检验中使用皮尔逊表会导致错误。虽然渐近时相同的临界值经常适用，但如果提供了斯皮尔曼表，正确的方法就是使用该表，或注明近似。此外，忘记正确排序（升序/降序）也是一个常见的疏漏。
Hypothesis tests for rs follow the same structure: state H₀ and H₁, find the test statistic, compare with critical value, and conclude in context. Many candidates lose marks by omitting the contextual conclusion.
rs 的假设检验遵循相同的结构：陈述 H₀ 和 H₁，求检验统计量，与临界值比较，并给出情境化的结论。许多考生因遗漏情境化结论而失分。
7. Chi-Squared Tests for Association and Goodness of Fit | 卡方检验：独立性检验与拟合优度
The chi-squared test is a high-frequency topic. For a contingency table (test of association), the test statistic is Σ(O - E)²/E, with degrees of freedom (r-1)(c-1). For goodness of fit, the df is (number of categories - 1 - number of estimated parameters).
卡方检验是高频考点。对于列联表（独立性检验），检验统计量为 Σ(O - E)²/E，自由度为 (r-1)(c-1)。对于拟合优度检验，自由度为（类别数 - 1 - 估计参数个数）。
/ grand total), using percentages instead of frequencies, and failing to combine categories when expected frequencies fall below 5. WJEC expects all expected frequencies to be at least 5; otherwise, you must merge neighbouring categories.
最常见的错误包括：计算期望频数出错（E = 行合计 × 列合计 / 总计），使用百分比而非频数，以及在期望频数低于 5 时未合并类别。WJEC 要求所有期望频数至少为 5；否则必须合并相邻类别。
For goodness of fit, knowing when to subtract an extra degree of freedom for estimated parameters (e.g. when fitting a binomial or Poisson distribution to data) is crucial. Many students forget this and overestimate the critical value, leading to incorrect conclusions.
对于拟合优度，知道何时为估计参数扣除额外的自由度（例如当用二项分布或泊松分布拟合数据时）至关重要。许多学生忘记这一点，从而高估了临界值，导致错误结论。
8. Combining Random Variables | 随机变量的组合
If the variables are normally distributed, any linear combination is also normally distributed. These results are frequently examined.
如果 X 和 Y 是独立随机变量，则 E(aX + bY) = aE(X) + bE(Y)，且 Var(aX + bY) = a²Var(X) + b²Var(Y)。如果变量服从正态分布，那么任何线性组合也服从正态分布。这些结论经常被考查。
A common error is dealing with the variance of aX - bY. Many forget that the variance is always a²Var(X) + b²Var(Y), regardless of the minus sign. Subtracting variables does not reduce variance; the coefficients are squared, so the sign disappears.
一个常见错误是处理 aX - bY 的方差。很多人忘记方差总是 a²Var(X) + b²Var(Y)，无论减号如何。变量相减并不会减少方差；系数被平方后，符号消失了。
Problems involving total weight, total time, or differences between measurements are classic. Always define your random variables clearly, write their distributions, and then compute the distribution of the combination before calculating probabilities.
涉及总重量、总时间或测量值之差的问题很经典。务必清晰地定义你的随机变量，写出它们的分布，然后计算组合的分布，再计算概率。
9. Sampling Methods and Bias | 抽样方法与偏差
WJEC frequently includes questions on simple random sampling, stratified sampling, cluster sampling, and quota sampling. You must be able to describe the procedure for each and identify advantages and disadvantages.
WJEC 经常考查简单随机抽样、分层抽样、整群抽样和配额抽样。你必须能够描述每种方法的步骤，并识别其优缺点。
Common mistakes include confusing stratified and quota sampling. In stratified sampling, items within each stratum are selected randomly using a sampling frame, while quota sampling involves non-random selection within strata based on convenience. The presence or absence of a sampling frame is the key differentiator.
常见错误包括混淆分层抽样和配额抽样。在分层抽样中，每个层内的个体是使用抽样框随机选择的；而配额抽样则是在层内根据便利性进行非随机选择。有无抽样框是关键区别。
Another pitfall is failing to recognise sources of bias: selection bias, non-response bias, and measurement bias. When asked to criticise a sampling method, always look for whether every member of the population had an equal chance of being chosen.
另一个陷阱是未能识别偏差来源：选择偏差、无响应偏差和测量偏差。当被要求评价某种抽样方法时，务必检查是否总体的每个成员都有平等的被选中的机会。
10. Interpreting Statistical Outputs and Diagrams | 统计输出与图表的解读
WJEC exams often provide computer output, residual plots, or scatter diagrams, and require interpretation. For residual plots, you should check for random scatter around zero; patterns indicate a poor model fit or non-linear relationship.
WJEC 考试经常会给出计算机输出、残差图或散点图，并要求解读。对于残差图，应检查是否围绕零值随机散布；如果有规律则表明模型拟合不佳或存在非线性关系。
With bivariate data, outliers and influential points can heavily affect the PMCC. An outlier may not always be an error; it could be a genuine extreme observation. Candidates must comment on its effect on r and the regression line.
在双变量数据中，离群值和有影响的点会严重影响 PMCC。离群值不一定总是错误，可能是真实的极端观测值。考生必须评论其对 r 和回归线的影响。
ent of determination, r², is also key. For example, if r = 0.8, then r² = 0.64, meaning 64% he variation in x. Many candidates state this percentage incorrectly or misinterpret it as "64% of the data points lie on the line", which is wrong.
解读决定系数 r² 也很关键。例如，若 r = 0.8，则 r² = 0.64，意味着 y 的变异中有 64% 可由 x 的变异解释。许多考生错误地表述这个百分比，或误解为“64% 的数据点落在直线上”，这是完全错误的。
11. Practical Issues in Data Collection and Experiment Design | 数据收集和实验设计中的实际问题
Questions on designing a statistical investigation often ask for identification of the population, sampling unit, and sampling frame. You must also consider how to reduce bias and ensure data are reliable.
关于设计统计调查的问题，经常要求识别总体、抽样单元和抽样框。你还必须考虑如何减少偏差并确保数据可靠。
Pilot studies are a common answer for checking questionnaire wording, estimating variability for sample size calculation, and testing logistics. Many students forget to mention the purpose of a pilot study in detail and just say "to check if it works".
试点研究是常见的答案，用于检查问卷措辞、估计变异程度以计算样本量以及测试后勤安排。许多学生忘记详细说明试点研究的目的，而只是说“看看是否可行”。
Randomisation, blocking, and blinding in experiments can be tested. The key is understanding how they eliminate confounding variables. A common error is confusing single-blind and double-blind designs.
实验中的随机化、区组化和盲法也可能被考查。关键在于理解它们如何消除混杂变量。一个常见错误是混淆单盲和双盲设计。
12. Error Avoidance Checklist | 避错清单总结
Before the exam, review this checklist: Have I used the correct distribution (normal vs t)? Have I checked conditions (np > 5, expected frequencies ≥ 5)? Have I stated hypotheses correctly and concluded in context? Have I interpreted confidence intervals with appropriate language? Have I combined variances correctly, adding rather than subtracting? Have I used the correct degrees of freedom?
考前请回顾这份清单：我是否使用了正确的分布（正态还是 t）？我是否检查了条件（np > 5，期望频数 ≥ 5）？我是否正确地陈述了假设并给出了情境化结论？我是否用恰当的语言解读了置信区间？我是否正确组合了方差，用加法而非减法？我是否使用了正确的自由度？
By focusing on these high-frequency topics and actively avoiding the common errors detailed above, you can significantly improve your performance. Practise past paper questions with these pitfalls in mind, and you will be well-prepared for the WJEC Year 13 Statistics exam.
专注于这些高频考点，并有意识地避免上述常见错误，你的成绩将得到显著提升。带着对这些陷阱的警觉去刷历年真题，你将充分准备好应对 WJEC Year 13 统计考试。
Published by TutorHao | Statistics Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
或截图后用「扫一扫 → 相册」选择本图