📚 High-Frequency Exam Topics and Common Mistakes Analysis for Year 13 CIE Statistics | Year 13 CIE 统计：高频考点与易错题分析
Year 13 CIE Statistics (the S2 component of Cambridge International A Level Mathematics, usually examined as Paper 6) contains a small number of topic areas that reappear in almost every session: continuous random variables, linear combinations of random variables, the Poisson distribution, sampling and estimation, and hypothesis testing. Because the syllabus is narrow, the examiner can probe each topic in depth, and the marks are often lost not through ignorance of the method but through small, predictable slips. This article maps the high-frequency exam topics onto the errors candidates actually make, so that you can revise with a checklist of traps rather than a vague sense of "doing more questions".
Year 13 CIE 统计（剑桥国际 A Level 数学的 S2 模块，通常对应 Paper 6）虽然内容不多，但考点高度集中：连续随机变量、随机变量的线性组合、泊松分布、抽样与估计、假设检验。由于考纲范围较窄，考官往往会把每个考点挖得很深，而失分通常不是"不会做"，而是"小地方出错"。本文把高频考点与考生真实易错点逐一对应，希望你的复习不只是"多刷题"，而是带着一张清晰的"避坑清单"上场。
1. Continuous Random Variables and Probability Density Functions | 连续随机变量与概率密度函数
e of x, and the total area under the curve over its domain equals 1. In exam questions you are typically asked to find an unknown constant k by setting the definite integral of f(x) over its support equal to 1, and then to use that PDF to find probabilities, the median, the mode or the mean.
连续随机变量 X 由概率密度函数（PDF）f(x) 描述。它必须满足两个条件：对任意 x 有 f(x) ≥ 0，且曲线在定义域内的总面积等于 1。考试中常见的第一问是通过令 f(x) 在其有效区间上的定积分等于 1 来求未知常数 k，随后用这个 PDF 求概率、中位数、众数或均值。
∫₋∞^∞ f(x) dx = 1 and f(x) ≥ 0 for all x
imum on the support. Note that for a continuous variable P(X = a) = 0 for any single value a, so P(a < X < b) and P(a ≤ X ≤ b) are the same thing.
中位数 m 满足左尾面积等于 0.5，即 ∫₋∞ᵐ f(x) dx = 0.5；众数则是 f(x) 在有效区间上取最大值处的 x 值。还要注意：连续型变量中任意单点概率为零，即 P(X = a) = 0，因此 P(a < X < b) 与 P(a ≤ X ≤ b) 完全等价。
- Finding k: write down the correct limits of integration first, then integrate and solve.
求 k：先写清积分上下限，再积分并解方程。 - Verifying: after finding k, confirm the resulting f(x) is non-negative on the whole support.
验证：求出 k 后，务必确认所得 f(x) 在整个区间内非负。 - Typical error: ating over (−∞, ∞) when f(x) is defined only on a finite interval such as 0 < x < 4.
典型错误：当 f(x) 仅在如 0 < x < 4 的有限区间上定义时，却在整条实轴上积分。 - Typical error: solving ∫₋∞ᵐ f(x) dx = 0.5 and accepting a root that lies outside the range of X.
典型错误：解出 ∫₋∞ᵐ f(x) dx = 0.5 后，接受了落在 X 取值范围之外的根。
2. Cumulative Distribution Functions | 累积分布函数
The cumulative distribution function (CDF) is defined by F(x) = P(X ≤ x) = ∫₋∞ˣ b) = F(b) − F(a). The median can also be found by solving F(m) = 0.5, and quartiles by solving F(q) = 0.25 and F(q) = 0.75.
累积分布函数（CDF）定义为 F(x) = P(X ≤ x) = ∫₋∞ˣ f(t) dt。它在有效区间上由 0 单调递增到 1，是求区间概率最方便的工具，因为 P(a < X < b) = F(b) − F(a)。中位数也可通过解 F(m) = 0.5 得到，下、上四分位数则分别通过解 F(q) = 0.25 与 F(q) = 0.75 得到。
When the PDF is defined piecewise, the CDF must be built up piece by piece, carrying forward the probability already accumulated on the previous interval. Only then can you add the contribution of the current piece.
当 PDF 分段定义时，CDF 必须逐段构造，并把前一段已累积的概率带过来，再加上当前段的贡献，不能各段独立积分。
- Useful relationship: F' = f(x) wherever f is continuous, so differentiating a CDF is a fast route back to the PDF.
重要关系：在 f 连续处有 F'(x) = f(x)，对 CDF 求导是快速回到 PDF 的捷径。 - Boundary behaviour: F(x) = 0 below the support and F(x) = 1 above it; state these explicitly.
边界行为：在有效区间之外，下方 F(x) = 0、上方 F(x) = 1，答题时应明确写出。 - Typical error: omitting the constant of integration when integrating the PDF to obtain F.
典型错误：由 PDF 积分求 F 时漏掉积分常数，导致 F 在起点不为 0。 - Typical error: producing a CDF that is not monotonic, or that exceeds 1, a clear signal that an accumulation step was mishandled.
典型错误：得到的 CDF 非单调或超过 1，这往往说明累积步骤出错。
3. Median, Mode, Quartiles and Percentiles from a PDF or CDF | 由 PDF 或 CDF 求中位数、众数、四分位数
These "location" artiles come from solving F(x) = 0.5, 0.25 and 0.75; the mode comes from maximising f(x) on the support; and the interquartile range is the difference between the upper and lower quartiles.
这类"位置量"问题在 S2 试卷中出现频率极高，因为它同时考查积分、解方程和结果解释。中位数与四分位数来自解 F(x) = 0.5、0.25、0.75；众数来自在有效区间上使 f(x) 取最大；四分位距则是上四分位数与下四分位数之差。
F(m) = 0.5, F(Q₁) = 0.25, F(Q₃) = 0.75, IQR = Q₃ − Q₁
(x) to occur at a boundary rather than at a stationary point, so differentiating and setting f'(x) = 0 alone is not sufficient.
对于分段 PDF，众数必须在每一段上分别检查，包括各段的端点。f(x) 的最大值很可能出现在区间端点而非驻点，因此仅靠令 f'(x) = 0 求极值是不够的。
- Typical error: assuming the median equals the mean; they coincide only for symmetric distributions.
典型错误：默认中位数等于均值；只有分布对称时两者才相等。 - Typical error: giving a decimal median without checking it lies inside the support.
典型错误：求出小数形式的中位数却不检验它是否落在有效区间内。 - Typical error: forgetting that skewness can be inferred from the relative positions of mode, median and mean, and being unable to comment on it.
典型错误：忘记可以通过众数、中位数、均值的相对位置判断偏态，从而无法作答"评论形状"的小问。
4. Expectation and Variance of Continuous Random Variables | 连续随机变量的期望与方差
The expectation of a continuous random variable is the integral of x multiplied by its density, and the expectation of a function of X follows the same "law of the unconscious statistician" pattern. Variance is most safely computed from the difference between the second moment and the square of the first moment.
连续随机变量的期望等于 x 与其密度函数乘积的积分；而 X 的函数的期望遵循同样的"代入法"模式。方差最稳妥的算法是"二阶矩减去一阶矩的平方"。
E(X) = ∫ x f(x) dx Var(X) = E(X²) − [E(X)]²
When the question asks for E(g(X)) for some function such as X², 1/X or 3X + 2, you integrate g(x) f(x) over the support. There is no need to find the distribution of g(X) first, which is a common source of wasted time.
当题目要求 E(g(X))（例如 g(x) = x²、1/x 或 3x + 2）时，只需在有效区间上对 g(x) f(x) 积分，不必先求 g(X) 的分布，很多同学在这里浪费了大量时间。
- Check your limits: l limits are always the support of X, never ±∞ unless the density really is defined there.
检查上下限：积分限永远是 X 的有效区间，除非密度确实定义在全实轴上，否则不要写 ±∞。 - Sanity checks: a probability must lie between 0 and 1, and a variance must be positive; if either fails, re-read the integral.
合理性检验：概率必须介于 0 与 1 之间，方差必须为正；若不符，请重新检查积分。 - Typical error: computing E(X²) as [E(X)]², which silently forces the variance to zero.
典型错误：把 E(X²) 当成 [E(X)]²，这会让方差变成 0 而毫无察觉。 - Typical error: mixing up "mean of X" with "mean of the distribution of X²" in later parts of a multi-part question.
典型错误：在多小问题目中把 E(X) 与 E(X²) 混用，导致后续小问连锁出错。
5. Linear Combinations of Random Variables | 随机变量的线性组合
This topic is almost guaranteed to appear. For constants a and b, the expectation scales linearly and simply shifts, whereas the variance scales by the square of the multiplier and ignores the shift entirely. When two variables are independent, variances add, whether the variables are being added or subtracted.
这一考点几乎每次必考。对常数 a、b，期望按线性伸缩并随常数平移，而方差则按系数的平方伸缩，且完全忽略常数项。当两个变量相互独立时，无论相加还是相减，方差都是相加的。
E(aX + b) = aE(X) + b Var(aX + b) = a²Var(X)
E(X ± Y) = E(X) ± E(Y) Var(X ± Y) = Var(X) + Var(Y), if X and Y are independent
For normal variables the result is even cleaner: any linear combination of independent normal variables is itself normal, with mean and variance obtained from the rules above. That is why the standard normal tables can still be used even when the question involves several variables at once.
对于正态变量，结论更为漂亮：独立正态变量的任意线性组合仍服从正态分布，其均值与方差由上面两条规则给出。正因如此，即便题目同时涉及多个变量，仍然可以继续查标准正态分布表。
- Standard deviation trap: if the question asks for a standard deviation, take the square root only at the very end.
标准差陷阱：若题目要的是标准差，请只在最后一步开平方。 - Independence statement: always quote "assuming the times are independent" before adding variances.
独立性的说明：在相加方差之前，务必写出"假设各时间相互独立"。 - Typical error: writing Var(X − Y) = Var(X) − Var(Y), which can even produce a negative variance.
典型错误：写成 Var(X − Y) = Var(X) − Var(Y)，有时甚至得出负方差。 - Typical error: forgetting to square the coefficient, e.g. using 3σ instead of 9σ² for Var(3X).
典型错误：忘记把系数平方，例如把 Var(3X) 写成 3σ 而不是 9σ²。
6. The Normal Distribution and Standardisation | 正态分布与标准化
If X follows a normal distribution with mean μ and variance σ², written X ~ N(μ, σ²), then standardising turns any normal problem into a single lookup in the standard normal table. The inverse process, finding a value of x given a probability, is equally common in S2 questions.
若 X 服从均值为 μ、方差为 σ² 的正态分布，记作 X ~ N(μ, σ²)，则通过标准化可把任意正态问题化为一次标准正态查表。反向操作——已知概率求 x 值——在 S2 中同样常见。
Z = (X − μ) ÷ σ, where Z ~ N(0, 1)
er a. Sketch a bell curve, shade the correct region and decide whether the z value should be positive or negative before you use the table; this single habit prevents most sign errors.
反向正态问题通常给出如 P(X < a) = 0.9 的概率条件，要求反求 a。动手前先画一条钟形曲线、标出阴影区域，并判断 z 值应为正还是负——这个习惯能避免绝大多数符号错误。
- Parameter or square? In CIE notation N(μ, σ²) the second entry is the variance; some textbooks write N(μ, σ). Read carefully.
参数还是平方？CIE 记号 N(μ, σ²) 中第二项是方差；部分教材写作 N(μ, σ)，务必看清。 - Rounding: keep full accuracy through the working and round probabilities to 3 significant figures at the end.
取整：中间过程保留足够精度，概率最后取三位有效数字。 - Typical error: using σ where σ² is needed when combining normal variables in a later part.
典型错误：在后续小问中组合正态变量时，该用 σ² 的地方用了 σ。 - Typical error: shading the wrong tail in a "greater than" question and quoting a probability above 0.5 that should be below it.
典型错误：在"大于"型问题中阴影画错尾巴，导致本该小于 0.5 的概率报成大于 0.5。
7. Sampling, the Sample Mean and the Central Limit Theorem | 抽样、样本均值与中心极限定理
When a random sample of size n is taken from a population with mean μ and variance σ², the sample mean is itself a random variable with its own distribution. Its mean equals the population mean, and its variance is the population variance divided by n, so its standard deviation - the standard error - shrinks as the sample grows.
从均值为 μ、方差为 σ² 的总体中抽取容量为 n 的随机样本时，样本均值本身就是一个随机变量，具有自己的分布：其均值等于总体均值，其方差等于总体方差除以 n，因此其标准差（即标准误）随样本量增大而减小。
X̄ ~ N(μ, σ² ÷ n), standard error = σ ÷ √n
If the population is already normal, this result is exact for any sample size. If the population is not normal, the central limit theorem guarantees that the distribution of the sample mean becomes approximately normal once n is reasonably large, so the same working can still be used.
若总体本身服从正态分布，则上述结论对任意样本量都精确成立；若总体非正态，中心极限定理保证当 n 足够大时，样本均值的分布近似正态，因此仍可沿用同一套做法。
- Divide by n, not by n²: Var(X̄) = σ²/n, so the standard error is σ/√n.
除以 n 而非 n²：Var(X̄) = σ²/n，故标准误为 σ/√n。 - Individual versus mean: P(X < 200) uses the individual distribution, P(X̄ < 200) uses the sampling distribution.
个体与均值：P(X < 200) 用个体分布，P(X̄ < 200) 用样本均值的抽样分布。 - Typical error: quoting the central limit theorem for a total rather than a mean, or applying it when n is very small and the population is clearly skewed.
典型错误：对"总和"而非"均值"引用中心极限定理，或在 n 很小且总体明显偏斜时强行使用它。 - Typical error: forgetting the √ when converting variance of the mean into standard error.
典型错误：把样本均值的方差换算为标准误时漏掉根号。
8. Point Estimation and Unbiased Estimators | 点估计与无偏估计量
An estimator is unbiased if its expected value equals the true population parameter for every possible value of that parameter. The sample mean is the classic unbiased estimator of the population mean, and in the Cambridge treatment the unbiased estimator of the population variance uses the divisor n − 1 rather than n.
若估计量的期望值恒等于总体真参数，则称该估计量为无偏估计量。样本均值是总体均值的经典无偏估计量；在剑桥的处理方式中，总体方差的无偏估计量使用分母 n − 1 而不是 n。
Unbiased means E(θ̂) = θ, for example E(X̄) = μ
A useful way to justify unbiasedness in a written answer is to substitute the definition of the estimator into the expectation operator, apply linearity, and show that the result equals the parameter. Marks are awarded for this structure, not only for
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