CAIE Year 13 Statistics: High-Frequency Topics and Common Pitfalls | CAIE 13年级统计：高频考点与易错题分析

📚 CAIE Year 13 Statistics: High-Frequency Topics and Common Pitfalls | CAIE 13年级统计：高频考点与易错题分析

Year 13 CAIE Statistics (Paper 6/7, Probability & Statistics 2) deepens your understanding of inference, distributions, and hypothesis testing. The exam regularly targets a handful of high-frequency topics: Poisson distribution and approximations, continuous random variables, linear combinations, sampling with the Central Limit Theorem, hypothesis tests for means (normal and Poisson), and chi-squared tests. Despite their recurrence, these topics also generate common errors, from misapplying continuity corrections to confusing distribution conditions. This article systematically reviews these key areas and the pitfalls that cost marks, helping you refine your technique and boost your confidence.

13年级CAIE统计（试卷6/7，概率与统计2）加深你对推断、分布和假设检验的理解。考试反复考察一批高频主题：泊松分布及其近似、连续随机变量、线性组合、中心极限定理下的抽样、均值假设检验（正态与泊松）以及卡方检验。尽管反复出现，这些主题也常引发典型错误，从误用连续性校正到混淆分布条件。本文系统梳理这些重点领域和导致失分的易错点，助你打磨技巧、增强信心。

1. Poisson Distribution and Approximations | 泊松分布与近似条件

The Poisson distribution models the number of events in a fixed interval, with probability mass function P(X = k) = (e^{-λ} λ^k) / k! for k = 0,1,2,…. It assumes events occur independently at a constant average rate λ. A classic pitfall is using the Poisson approximation to the binomial without checking that n is large and p is small—typically both n > 50 and np < 5 are expected. Applying it when p is moderate (e.g., p=0.3) can seriously distort probabilities. Similarly, the normal approximation to the Poisson requires λ > 15 and a continuity correction. Many candidates either forget the correction entirely or mistakenly apply x+0.5 instead of x-0.5 for a right-tail probability P(X ≥ x).

泊松分布对固定间隔内的事件次数建模，概率函数为 P(X = k) = (e^{-λ} λ^k) / k!，k = 0,1,2,…。它假设事件独立发生，且平均发生率 λ 恒定。经典易错点是在没有检查 n 大、p 小的情况下使用二项分布的泊松近似——通常要求 n > 50 且 np < 5。当 p 中等（如 p=0.3）时误用，会严重扭曲概率。同样，泊松分布的正态近似要求 λ > 15 并需连续性校正。许多考生要么完全忘记校正，要么误把右尾概率 P(X ≥ x) 用 x+0.5 而非 x-0.5。

Another frequent error concerns the variance of a Poisson variable. Since E(X) = Var(X) = λ, some students incorrectly treat the variance as λ² when standardising, or they use λ instead of √λ in the denominator. Always remember that for X ~ Po(λ), the standard deviation is √λ. In hypothesis tests, using the wrong variance leads to an invalid test statistic and lost accuracy marks.

另一常见错误涉及泊松变量的方差。由于 E(X) = Var(X) = λ，一些学生在标准化时误将方差当作 λ²，或在分母中用 λ 代替 √λ。务必记住，对于 X ~ Po(λ)，标准差为 √λ。在假设检验中，用错方差会导致无效的检验统计量并丢失准确性分数。

2. Continuous Random Variables: PDF and CDF | 连续随机变量：概率密度函数与累积分布函数

A probability density function (pdf) f(x) must satisfy f(x) ≥ 0 and ∫ f(x) dx = 1 over its entire domain. When f(x) is defined piecewise, students often forget to verify that the total area equals one by splitting the integral correctly. The cumulative distribution function (cdf) F(x) = P(X ≤ x) is found by integration; to obtain the median m, solve F(m) = 0.5. A common mistake is to set f(m) = 0.5 instead, completely misunderstanding the definition. Also, when the pdf is piecewise, the median may fall in a specific segment—always identify which piece before solving.

概率密度函数 f(x) 必须满足 f(x) ≥ 0 且 ∫ f(x) dx = 1 在整个定义域上。当 f(x) 分段定义时，学生常忘记通过正确拆分积分来验证总面积等于1。累积分布函数 F(x) = P(X ≤ x) 通过积分得到；求中位数 m 需解 F(m) = 0.5。一个常见错误是去解 f(m) = 0.5，完全误解了定义。此外，当 pdf 分段时，中位数可能落在特定区间——解题前务必确定哪一段。

Integration limits cause additional trouble. To find P(a < X < b), integrate f(x) from a to b, but many candidates use F(b) – F(a) incorrectly when the cdf is given piecewise. When finding E(X) or Var(X), using the wrong limits or forgetting the E(X²) formula for variance leads to arithmetic mistakes. If the distribution is symmetric, exploit symmetry to skip unnecessary work—this is often overlooked in the exam.

积分上下限带来额外麻烦。求 P(a < X < b) 时需从 a 到 b 积分 f(x)，但很多考生在 cdf 分段时错误使用 F(b) – F(a)。计算 E(X) 或 Var(X) 时，用错积分限或忘记计算方差需 E(X²) 会导致计算错误。如果分布对称，可利用对称性省去不必要的工作——这点在考试中常被忽略。

3. Linear Combinations of Random Variables | 随机变量的线性组合

For a linear combination aX + bY, the expectation is E(aX + bY) = aE(X) + bE(Y). The variance, however, involves squares: Var(aX + bY) = a²Var(X) + b²Var(Y) + 2ab Cov(X,Y). When X and Y are independent, covariance is zero. A famously persistent error is writing Var(aX + bY) = aVar(X) + bVar(Y), forgetting to square the coefficients. Another is mishandling Var(X – Y); since Var(X – Y) = Var(X) + Var(Y) for independent variables, students often wrongly subtract the variances.

对线性组合 aX + bY，期望为 E(aX + bY) = aE(X) + bE(Y)。而方差涉及平方：Var(aX + bY) = a²Var(X) + b²Var(Y) + 2ab Cov(X,Y)。当 X 与 Y 独立时，协方差为零。一个有名的顽固错误是写成 Var(aX + bY) = aVar(X) + bVar(Y)，忘记平方系数。另一个易错点是处理 Var(X – Y)；对独立变量有 Var(X – Y) = Var(X) + Var(Y)，学生却常错用减法。

In problems involving normal variables, the sum of independent normals is again normal. Some candidates fail to state this and instead try unnecessary transformations. When scaling a Poisson variable, note that 2X is NOT Poisson with mean 2λ, because the support would only contain even numbers. This distinction trips up many, especially in questions about the distribution of total number of events from multiple independent Poisson sources—you must sum the variables, not scale a single one.

在涉及正态变量的问题中，独立正态变量之和仍为正态。一些考生忘记声明这一点而进行不必要的变换。缩放泊松变量时注意，2X 不具有均值为 2λ 的泊松分布，因为它的取值只能是偶数。这一点难倒许多人，尤其在多个独立泊松源的事件总数问题中——你必须对变量求和，而非缩放单个变量。

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

Unbiased estimators of the population mean μ and variance σ² are the sample mean x̄ and s² = Σ(x – x̄)² / (n – 1). A common slip is using n instead of n-1 in the denominator of s², which yields a biased estimate. The Central Limit Theorem (CLT) states that for large n (usually n ≥ 30), the distribution of the sample mean is approximately N(μ, σ²/n), regardless of the population shape. The CLT applies to the mean, not to individual observations. A typical error is to claim that a single large sample makes the raw data normal, or to apply CLT when n < 30 without checking for normality.

总体均值 μ 和方差 σ² 的无偏估计量分别是样本均值 x̄ 和 s² = Σ(x – x̄)² / (n – 1)。常见失误是在 s² 分母中用 n 代替 n-1，得出有偏估计。中心极限定理指出，当 n 大时（通常 n ≥ 30），样本均值的分布近似为 N(μ, σ²/n)，无论总体形状如何。CLT 适用于样本均值，而非个体观测值。典型错误是宣称单个大样本使得原始数据正态，或在 n < 30 未检验正态性时滥用 CLT。

When standardising the sample mean, the standard error is σ/√n. If σ is unknown and n is large, we use s/√n. Mistaking the standard deviation of the sample (s) for the standard error of the mean (s/√n) is a frequent cause of incorrect z-values. Always remember to divide by √n. Examiners also want you to explicitly state that the sample size is sufficiently large to invoke the CLT, so make this justification clear in your solution.

标准化样本均值时，标准误是 σ/√n。

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

更多咨询请联系16621398022（同微信）