📚 Common Pitfalls in Year 13 CAIE Statistics and How to Correct Them | Year 13 CAIE 统计常见误区与纠正方法
In Year 13 CAIE Statistics, students often lose marks not because they cannot perform calculations, but because they misunderstand subtle conditions, apply wrong distributions, or misinterpret results. This guide walks you through the most common pitfalls seen across Paper 5 and Paper 6 topics—especially in probability, sampling, estimation, and hypothesis testing—and shows you exactly how to correct them.
在 Year 13 CAIE 统计中，学生丢分往往不是因为不会计算，而是因为对细微条件理解有误、选错分布或错误解读结果。本指南将逐一剖析试卷 5 和试卷 6 中最常见的误区——尤其涉及概率、抽样、估计和假设检验的内容，并告诉你如何精准纠正。
1. Misinterpreting the Conditions for a Poisson Distribution | 误解泊松分布的条件
A common mistake is applying the Poisson distribution to any count data without checking whether events occur independently and at a constant average rate within a fixed interval. For example, treating the number of cars passing a point per minute as Poisson without ensuring the rate is stable across minutes will lead to flawed models.
常见的错误是不检查事件是否在固定区间内独立发生且具有恒定的平均速率，就直接对任何计数数据使用泊松分布。例如，在未确保每分钟的速率稳定的情况下，把每分钟通过的车辆数当作泊松变量来建模，会得出错误的模型。
The correction: Always verify the three core conditions—events are independent, the rate λ is constant, and two events cannot occur at exactly the same instant. State these conditions explicitly before naming the distribution.
纠正方法：始终验证三个核心条件——事件相互独立、速率 λ 恒定、两个事件不可能在同一瞬间发生。在写出分布名称前，务必明确陈述这些条件。
2. Confusing Binomial and Poisson Approximations | 混淆二项分布与泊松分布的近似条件
Many students approximate a binomial distribution B(n, p) by Poisson Po(λ) whenever n is large, ignoring that p must also be small. They incorrectly use Poisson approximation for a binomial with n=200 and p=0.4, where the normal approximation is far more appropriate.
许多学生只要 n 较大，就把二项分布 B(n, p) 近似为泊松分布 Po(λ)，却忽略了 p 也必须很小这一条件。他们错误地对 n=200、p=0.4 的二项分布使用泊松近似，而这种情况用正态近似才合适得多。
Correction: Use Poisson approximation only when n is large, p is small, and np < 10 (commonly np ≤ 10). If np > 10 and n(1-p) > 10, use normal approximation with continuity correction instead.
纠正：仅当 n 大、p 小且 np < 10（通常 np ≤ 10）时，才使用泊松近似。如果 np > 10 且 n(1-p) > 10，则应改用带连续性校正的正态近似。
3. Forgetting Continuity Correction When Using Normal Approximations | 使用正态近似时忘记连续性校正
A classic pitfall: when approximating a discrete distribution (binomial or Poisson) by a normal distribution, candidates simply standardize the exact value without adjusting the interval. For instance, to find P(X = 12) in a binomial approximated by N(μ, σ²), they compute z = (12-μ)/σ instead of using the corrected interval from 11.5 to 12.5.
一个经典错误：用正态分布近似离散分布（二项或泊松）时，考生直接把精确值标准化而不调整区间。例如，对近似为正态 N(μ, σ²) 的二项分布求 P(X = 12)，他们算的是 z = (12-μ)/σ，而没有使用从 11.5 到 12.5 的修正区间。
Correction: Always apply a continuity correction by expanding the discrete value by ±0.5. For P(X ≤ k) use k + 0.5, for P(X ≥ k) use k – 0.5, and for exact probability use interval [k–0.5, k+0.5]. State the correction clearly in your working.
纠正：始终通过 ±0.5 的扩展来应用连续性校正。对于 P(X ≤ k)，用 k + 0.5；对于 P(X ≥ k)，用 k – 0.5；对于精确概率，用区间 [k–0.5, k+0.5]。在解题过程中要清晰写出校正步骤。
4. Misunderstanding the Sampling Distribution of the Mean | 误解样本均值的抽样分布
Students often incorrectly assume that the distribution of the sample mean is exactly normal for any sample size, or they forget to adjust the variance. A typical error is writing X̄ ~ N(μ, σ²) instead of X̄ ~ N(μ, σ²/n). This leads to wrong test statistics and confidence intervals.
学生经常错误地认为，无论样本量多大，样本均值的分布都是精确正态的，或者忘记调整方差。一个典型错误是把 X̄ 写成了 N(μ, σ²)，而不是 N(μ, σ²/n)。这会导致检验统计量和置信区间全错。
Correction: By the Central Limit Theorem, the sampling distribution of X̄ is approximately N(μ, σ²/n) for large n. If the population is normal, X̄ is exactly normal for any n. Always halve the variance divide by n and confirm whether the population standard deviation is known or estimated.
纠正：根据中心极限定理，当 n 大时，X̄ 的抽样分布近似为 N(μ, σ²/n)。如果总体本身是正态分布，则对任意 n，X̄ 都精确服从正态分布。务必把方差除以 n，并确认总体标准差是已知的还是用样本估计的。
5. Incorrect Use of the t-distribution versus the Normal | t 分布与正态分布的错误使用
A frequent error is using the normal distribution when the population variance is unknown and the sample size is small. In such cases, the t-distribution should be used, but students still read critical values from normal tables.
一个频繁错误是，当总体方差未知且样本量较小时，还在用正态分布。这种情况应该使用 t 分布，但学生仍从正态分布表中查临界值。
Correction: Use the t-distribution with (n – 1) degrees of freedom when the population standard deviation is estimated by the sample standard deviation s, especially when n < 30. As n increases, the t-distribution approaches the normal, but for small samples the difference in critical values is significant.
纠正：当总体标准差用样本标准差 s 估计时，尤其当 n < 30 时，要使用自由度为 (n – 1) 的 t 分布。随着 n 增大，t 分布趋近于正态分布，但在小样本中，临界值的差异非常显著。
6. Confusing Confidence Intervals with Prediction Intervals | 混淆置信区间与预测区间
Many candidates interpret a 95% confidence interval for the mean as an interval that will contain 95% of future individual observations. This is wrong because a confidence interval reflects uncertainty about the population mean, not about individual values.
许多考生把均值的 95% 置信区间解释为会包含未来单个观测值 95% 的区间。这是错误的，因为置信区间反映的是对总体均值的不确定性，而不是对个体取值的不确定性。
Correction: A 95% confidence interval for μ means that if we repeated the sampling process many times, 95% of such intervals would capture the true μ. For predicting an individual value, a prediction interval (typically wider) would be needed. Keep these interpretations distinct in exam responses.
纠正：μ 的 95% 置信区间意味着，如果反复多次抽样，那么这些区间中有 95% 会包含真实的 μ。要预测单个取值，需要使用预测区间（通常更宽）。在答题时务必将这两种解释区分清楚。
7. Misinterpreting p-values in Hypothesis Testing | 假设检验中对 p 值的错误解读
A p-value is often taken as the probability that the null hypothesis H₀ is true, which is incorrect. Students might say ‘p = 0.03, so H₀ has a 3% chance of being true.’ This misinterpretation loses marks in CAIE context questions.
p 值经常被当作原假设 H₀ 为真的概率，这是错误的。学生可能会说“p = 0.03，所以 H₀ 有 3% 的可能性为真”。这种曲解在 CAIE 情景题中会失分。
Correction: The p-value is the probability of observing a test statistic as extreme as, or more extreme than, the one obtained, given that H₀ is true. It does not give the probability that H₀ is true. Use p-value to compare with the significance level α: if p < α, reject H₀.
纠正：p 值是在 H₀ 为真的条件下，观察到与当前检验统计量一样极端甚至更极端结果的概率。它并不给出 H₀ 为真的概率。用 p 值与显著性水平 α 比较：若 p < α，则拒绝 H₀。
8. Type I and Type II Error Confusion | I 型错误与 II 型错误的混淆
Students frequently mix up Type I error (rejecting a true null hypothesis) and Type II error (failing to reject a false null hypothesis). In context, they might describe convicting an innocent person as a Type II error, which is actually a Type I error.
学生经常弄混 I 型错误（拒真错误）和 II 型错误（取伪错误）。在实际情境中，他们可能把“错判无辜者有罪”说成 II 型错误，但这其实是 I 型错误。
Correction: Memorise: Type I = rejecting H₀ when it is true (false positive), probability α. Type II = not rejecting H₀ when it is false (false negative), probability β. The power of a test is 1 – β. Relate each error to the consequences in the given scenario.
纠正：记住：I 型错误 = H₀ 为真时拒绝了它（假阳性），概率为 α。II 型错误 = H₀ 为假时没有拒绝它（假阴性），概率为 β。检验的功效为 1 – β。要结合给定情境将每种错误与后果联系起来。
9. Treating Dependent Variables as Independent in Variance Calculations | 在方差计算中将相关变量视为独立变量
When calculating Var(X + Y) or Var(X – Y), many students automatically use Var(X) + Var(Y), forgetting that this holds only when X and Y are independent. If the variables are dependent, the covariance term is missing, leading to incorrect variances and confidence intervals.
在计算 Var(X + Y) 或 Var(X – Y) 时，许多学生不假思索地使用 Var(X) + Var(Y)，忘记了这仅在 X 与 Y 独立时才成立。如果变量之间存在相关性，遗漏协方差项会导致方差和置信区间错误。
Correction: Use the general formula: Var(aX + bY) = a²Var(X) + b²Var(Y) + 2ab Cov(X, Y). For pairwise independent variables, covariance is zero, so the formula simplifies. In paired samples or repeated measures, dependence must be accounted for.
纠正：使用一般公式：Var(aX + bY) = a²Var(X) + b²Var(Y) + 2ab Cov(X, Y)。对于两两独立的变量，协方差为零，公式可简化。在配对样本或重复测量中，必须考虑依赖性。
10. Mishandling Paired Samples | 错误处理配对样本
When presented with paired data—such as before-and-after measurements—students sometimes use an independent two-sample t-test instead of a paired t-test. This ignores the natural pairing and reduces the power of the test, often leading to a wrong conclusion.
当遇到配对数据时——例如前后测量数据——学生有时会用独立双样本 t 检验而不是配对 t 检验。这忽略了自然配对关系，降低了检验功效，往往导致错误结论。
Correction: Recognise paired designs: each subject provides two observations. Compute the differences dᵢ, then perform a one-sample t-test on these differences, testing H₀: μ_d = 0. The test statistic is t = (d̄)/(s_d/√n) with n – 1 degrees of freedom. State clearly that a paired t-test is appropriate because the data are not independent.
纠正：识别配对设计：每个受试者提供两个观测值。计算差值 dᵢ，然后对这些差值进行单样本 t 检验，检验 H₀: μ_d = 0。检验统计量为 t = (d̄)/(s_d/√n)，自由度 n – 1。要明确说明采用配对 t 检验的原因是数据并非独立。
11. Using Population Variance Instead of Sample Variance in Estimation | 估计时使用总体方差而非样本方差
In estimation questions, when σ² is unknown, students sometimes incorrectly use σ² in their standard error formula instead of the unbiased estimator s². For a normal population, this leads to ignoring the extra uncertainty and incorrectly using z-values instead of t-values.
在估计问题中，当 σ² 未知时，学生有时会在标准误公式中错误地使用 σ²，而不是无偏估计量 s²。对于正态总体，这会忽略额外的不确定性，并错误地使用 z 值而不是 t 值。
Correction: Replace unknown population variance with the sample variance s² = Σ(x – x̄)²/(n – 1). Then the standard error of the mean is s/√n. If σ² is truly known, use z; otherwise, rely on the t-distribution. Remember that CAIE often provides raw data so you must compute s² using (n – 1) divisor.
纠正：用样本方差 s² = Σ(x – x̄)²/(n – 1) 替换未知的总体方差。于是均值的标准误为 s/√n。如果 σ² 确实已知，用 z；否则应借助 t 分布。记住，CAIE 常提供原始数据，所以你必须用除数 (n – 1) 来计算 s²。
12. Assuming the Normal Distribution for Small Samples without Justification | 在小样本中无理由地假定正态分布
In hypothesis tests or confidence intervals for the mean, students often assume the sampling distribution of the mean is normal even when the sample size is small and no information about the population distribution is given. This assumption is invalid unless the population is known to be normal.
在对均值进行假设检验或计算置信区间时，即使样本量很小且未给出总体分布信息，学生也常常假定样本均值的抽样分布是正态的。除非已知总体是正态分布，否则这种假定是无效的。
Correction: If n is small (e.g., n < 30) and the population distribution is unknown, a normal-based test cannot be justified unless you are told the population is normal. For large samples, the Central Limit Theorem supports approximate normality. Always check the question for statements like 'the population is normally distributed'.
纠正：如果 n 很小（如 n < 30）且总体分布未知，就不可随意基于正态进行检验，除非题中说明总体服从正态分布。对于大样本，中心极限定理支持近似正态性。始终检查题目中是否有“总体服从正态分布”之类的陈述。
Published by TutorHao | Statistics Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
屏轩国际教育cambridge primary/secondary checkpoint, cat4, ukiset,ukcat,igcse,alevel,PAT,STEP,MAT, ibdp,ap,ssat,sat,sat2课程辅导，国外大学本科硕士研究生博士课程论文辅导