Edexcel Year 13 Statistics: Interdisciplinary Problem Drills | Edexcel 13年级统计：跨学科综合题型训练

📚 Edexcel Year 13 Statistics: Interdisciplinary Problem Drills | Edexcel 13年级统计：跨学科综合题型训练

In Edexcel Year 13 Statistics, exam questions increasingly blend core statistical methods with scenarios from physics, biology, economics, and engineering. Success depends on recognising the appropriate model, extracting parameters, and interpreting results in a subject-specific context.

在 Edexcel 13 年级统计中，考题越来越多地将核心统计方法与物理、生物、经济学和工程学等场景结合。能否成功取决于识别合适的模型、提取参数并在学科情境下解读结果。

This article provides a structured drill across eight interdisciplinary themes, illustrating how normal, Poisson, binomial, and regression models apply outside the mathematics classroom. Each section pairs an English explanation with a Chinese translation, followed by worked insights and common pitfalls.

本文围绕八个跨学科主题提供结构化训练，展示正态、泊松、二项和回归模型如何应用于数学课堂之外。每个小节均配有中英文解释、解题思路和常见陷阱。

---

1. Recognising Interdisciplinary Contexts | 识别跨学科背景

Interdisciplinary problems rarely announce the distribution. Instead, you meet a paragraph about engine temperatures, radioactive decay counts, or customer defaults. The first task is to translate the narrative into a statistical variable and choose a suitable model.

跨学科题目很少直接标明分布。相反，你会遇到关于引擎温度、放射性衰变计数或客户违约的描述。首要任务是将叙述转化为统计变量并选择合适的模型。

Key clues: repeated independent trials with constant success probability suggest a binomial distribution; events occurring randomly at a constant average rate point to Poisson; large samples of continuous symmetric measurements often follow a normal distribution. Contexts like ‘mass of apples from an orchard’ or ‘thickness of glass sheets’ usually invoke the normal curve.

关键线索：具有恒定成功概率的重复独立试验暗示二项分布；以恒定平均速率随机发生的事件指向泊松分布；大量连续对称的测量值通常服从正态分布。如“果园苹果的质量”或“玻璃板的厚度”通常会用到正态曲线。

A quick mapping table helps consolidate the connections between disciplines and statistical models:

一张快速映射表有助于巩固学科与统计模型之间的联系：

Discipline 学科	Typical Variable 典型变量	Common Model 常用模型
Mechanical Engineering 机械工程	Shaft diameter, breaking force	Normal N(μ, σ²)
Biology 生物学	Bacterial colony count per cm²	Poisson Po(λ)
Finance 金融	Number of loan defaults out of n	Binomial B(n, p)
Medicine 医学	Blood pressure reduction, survival count	Normal or Binomial
Environmental Science 环境科学	Annual CO₂ level vs. temperature anomaly	Bivariate regression
Psychology 心理学	Reaction time, IQ score	Normal N(μ, σ²)
Physics 物理学	Measurement error, half-life count	Normal or Poisson

Once the model is identified, the statistical engine starts: standardisation, plugging into probability functions, or setting up hypothesis tests.

一旦模型确定，统计引擎便可启动：标准化、代入概率函数或建立假设检验。

---

2. Mechanical Engineering & Normal Distributions | 机械工程与正态分布

A common Edexcel problem presents a machined component whose critical dimension X follows a normal distribution. For example, the diameter of a steel pin is known to be normally distributed with mean 10.00 mm and standard deviation 0.05 mm. A pin is acceptable only if its diameter lies between 9.92 mm and 10.08 mm.

Edexcel 常见问题给出一个机加工部件，其关键尺寸 X 服从正态分布。例如，钢销直径服从均值为 10.00 mm、标准差为 0.05 mm 的正态分布。仅当直径在 9.92 mm 到 10.08 mm 之间时销钉才合格。

We define the random variable X ~ N(10.00, 0.05²). The probability of acceptance is P(9.92 < X < 10.08). Standardising both limits yields z₁ = (9.92 − 10.00)/0.05 = −1.6 and z₂ = (10.08 − 10.00)/0.05 = 1.6.

定义随机变量 X ~ N(10.00, 0.05²)。合格概率为 P(9.92 < X < 10.08)。将两个界限标准化得到 z₁ = (9.92 − 10.00)/0.05 = −1.6，z₂ = (10.08 − 10.00)/0.05 = 1.6。

P(accept) = Φ(1.6) − Φ(−1.6) = 0.9452 − 0.0548 = 0.8904

Hence around 11% of pins are rejected. To improve quality, engineers may aim to reduce σ. If σ is halved, the z‑values double, and the rejection rate drops dramatically — a practical link between variability and cost.

因此约 11% 的销钉被拒收。为提高质量，工程师可努力减小 σ。若 σ 减半，z 值翻倍，拒收率将显著下降——这直观体现了变异性与成本之间的实际联系。

In a reverse problem, given a required rejection rate (e.g., 1%), you must find the target standard deviation. Set P(accept) = 0.99, work back to z = 2.5758, then solve 0.08/σ = 2.5758 to obtain σ ≈ 0.031 mm. Such ‘tolerance design’ tasks are classic interdisciplinary drills.

在反向问题中，给定目标拒收率（如 1%），你需要求出目标标准差。设接受概率为 0.99，反推 z = 2.5758，然后解方程 0.08/σ = 2.5758，得 σ ≈ 0.031 mm。这类“公差设计”任务正是经典的跨学科训练题。

---

3. Biological Populations & Poisson Distributions | 生物种群与泊松分布

In microbiology, the number of bacterial colonies appearing on a Petri dish per unit area often follows a Poisson distribution. Suppose the mean number of colonies per cm² is 4. The random variable Y ~ Po(4) models the colony count on a randomly chosen 1 cm² area.

在微生物学中，培养皿上单位面积内出现的细菌菌落数通常服从泊松分布。假设每 cm² 平均菌落数为 4。随机变量 Y ~ Po(4) 可用于描述任意选取的 1 cm² 区域上的菌落数。

P(Y = k) = e⁻⁴ × 4ᵏ / k! for k = 0, 1, 2, …

To find the probability of observing exactly 2 colonies, compute P(Y = 2) = e⁻⁴ × 4² / 2! = e⁻⁴ × 8 ≈ 0.1465. This elementary calculation forms the basis of dilution assays in biology.

要计算观察到恰好 2 个菌落的概率，计算 P(Y = 2) = e⁻⁴ × 4² / 2! = e⁻⁴ × 8 ≈ 0.1465。这一基础计算构成了生物稀释试验的基础。

Often the question extends to a larger area. If the mean is 4 per cm², then for a 5 cm² region the mean becomes λ = 20. For large λ, the Poisson can be approximated by a normal distribution N(20, 20). An Edexcel task might ask: ‘Find the approximate probability that the number of colonies on a 5 cm² plate exceeds 25.’ Using continuity correction:

题目通常会扩展到更大面积。如果每 cm² 均值为 4，则 5 cm² 区域的均值变为 λ = 20。对于较大的 λ，泊松分布可用正态分布 N(20, 20) 近似。Edexcel 题目可能会问：“求 5 cm² 培养皿上菌落数超过 25 的近似概率。”使用连续性校正：

P(Y > 25) ≈ P(Z > (25.5 − 20)/√20) = P(Z > 1.23) ≈ 0.1093

This blending of Poisson and normal approximations perfectly illustrates the interdisciplinary nature of Year 13 statistics, requiring seamless switching between models.

这种泊松与正态近似的混合完美体现了 13 年级统计的跨学科特性，要求在不同模型之间自如切换。

---

4. Financial Risk & Binomial Models | 金融风险与二项模型

Credit risk analysis often applies the binomial distribution. A bank issues 200 personal loans, and historically 4% of borrowers default. The number of defaults, D, can be modelled as D ~ B(200, 0.04).

信用风险分析常用二项分布。某银行发放 200 笔个人贷款，历史违约率为 4%。违约数量 D 可建模为 D ~ B(200, 0.04)。

Calculate the probability of exactly 5 defaults: P(D = 5) = ²⁰⁰C₅ (0.04)⁵ (0.96)¹⁹⁵. Although tedious by hand, binomial tables or Poisson approximation (λ = np = 8) can be used. With Poisson: P ≈ e⁻⁸ × 8⁵ / 5! ≈ 0.0916.

计算恰好有 5 笔违约的概率：P(D = 5) = ²⁰⁰C₅ (0.04)⁵ (0.96)¹⁹⁵。尽管手算繁琐，但可使用二项分布表或泊松近似（λ = np = 8）。采用泊松：P ≈ e⁻⁸ × 8⁵ / 5! ≈ 0.0916。

A more demanding task is hypothesis testing: the bank suspects the default rate has risen above 4%. Set up H₀: p = 0.04, H₁: p > 0.04. In a sample of 200, 13 defaults are observed. Under H₀, D ~ B(200, 0.04) with mean 8. The p‑value is P(D ≥ 13). Using a normal approximation with continuity correction:

更具挑战性的任务是假设检验：银行怀疑违约率已上升至 4% 以上。建立 H₀: p = 0.04，H₁: p > 0.04。在 200 个样本中观察到 13 笔违约。在 H₀ 下，D ~ B(200, 0.04)，均值为 8。p 值为 P(D ≥ 13)。使用带连续性校正的正态近似：

z = (12.5 − 8) / √(200 × 0.04 × 0.96) = 4.5 / √7.68 ≈ 1.624

P(Z > 1.624) ≈ 0.052. At the 5% significance level, this is borderline, prompting further investigation. Interdisciplinary finance questions reward clear contextual conclusions: ‘There is insufficient evidence at the 5% level to conclude the default rate has increased, but the result is close to the threshold.’

P(Z > 1.624) ≈ 0.052。在 5% 显著性水平下，这属于临界结果，需进一步调查。跨学科金融题目看重清晰的语境结论：“在 5% 水平下，没有足够证据表明违约率已上升，但结果接近临界值。”

---

5. Medical Trials & Hypothesis Testing | 医学试验与假设检验

Medical scenarios test your ability to design a hypothesis test and interpret it for a non‑statistical audience. Consider a trial evaluating a new drug that claims to reduce systolic blood pressure by at least 10 mmHg. A random sample of 36 patients yields a sample mean reduction of 9.2 mmHg with a sample standard deviation of 4.8 mmHg.

医学场景考查你能否设计假设检验并向非统计受众解读结果。假设一项试验评估一款声称能将收缩压至少降低 10 mmHg 的新药。随机抽取 36 名患者，样本平均降幅为 9.2 mmHg，样本标准差为 4.8 mmHg。

Because the sample size is large, the distribution of the sample mean is approximately normal. Test H₀: μ = 10 against H₁: μ < 10, using the test statistic:

由于样本量较大，样本均值的分布近似正态。检验 H₀: μ = 10 对 H₁: μ < 10，使用检验统计量：

z = (9.2 − 10) / (4.8/√36) = −0.8 / 0.8 = −1.0

The one‑tailed p‑value is P(Z < −1.0) = 0.1587, which is far above 0.05, so we do not reject H₀. The clinical conclusion: 'The trial did not provide statistically significant evidence that the mean reduction is less than 10 mmHg.'

单尾 p 值为 P(Z < −1.0) = 0.1587，远高于 0.05，因此我们不拒绝 H₀。临床结论为：“该试验并未提供具有统计学意义的证据表明平均降幅低于 10 mmHg。”

Edexcel questions often require a non‑technical sentence explaining the result, as well as the recognition that even a non‑significant result is valuable for patient safety monitoring. This bridges mathematical rigour and real‑world communication.

Edexcel 题目通常要求用非技术性语句解释结果，并认识到即便不显著的结果对患者安全监测也有价值。这在数学严谨性与现实沟通之间架起了桥梁。

---

6. Environmental Science & Correlation/Regression | 环境科学与相关回归

Environmental datasets invite bivariate analysis. A typical exercise provides annual mean CO₂ concentration (x, in ppm) and global temperature anomaly (y, in °C) for 20 years. Students calculate the product moment correlation coefficient (PMCC) to quantify the linear association.

环境数据集适合双变量分析。典型练习给出 20 年来的 CO₂ 年均浓度（x，单位 ppm）和全球温度异常（y，单位 °C）。学生计算积矩相关系数 (PMCC) 以量化线性关联。

If the calculated r = 0.78, a hypothesis test for ρ = 0 uses the test statistic t = r √(n−2) / √(1−r²). With n = 20, t = 0.78 × √18 / √(1−0.6084) ≈ 3.31 / 0.626 ≈ 5.29. Comparing against t₁₈ critical values (two‑tail, 5% ≈ 2.101), there is strong evidence to reject H₀, suggesting a significant positive correlation.

若计算得 r = 0.78，检验 ρ = 0 的统计量为 t = r √(n−2) / √(1−r²)。n = 20 时，t = 0.78 × √18 / √(1−0.6084) ≈ 3.31 / 0.626 ≈ 5.29。与 t₁₈ 临界值（双尾，5% ≈ 2.101）比较，有充分证据拒绝 H₀，表明存在显著正相关。

The regression line y = a + bx can then be used with caution to predict temperature anomaly for a given CO₂ level. Edexcel rewards statements about reliability: ‘Extrapolation beyond the observed CO₂ range is unreliable because the linear trend may not persist.’

然后可谨慎使用回归线 y = a + bx 预测给定 CO₂ 水平下的温度异常。Edexcel 看重关于可靠性的表述：“在观测到的 CO₂ 范围之外

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

更多咨询请联系16621398022（同微信）