WJEC Year 13 Statistics: Interdisciplinary Problem-Solving Practice | WJEC Year 13 统计：跨学科综合题型训练

📚 WJEC Year 13 Statistics: Interdisciplinary Problem-Solving Practice | WJEC Year 13 统计：跨学科综合题型训练

WJEC Year 13 Statistics demands more than just calculation skills—it requires you to apply the correct technique in unfamiliar, real-world scenarios. This article walks you through interdisciplinary problem-solving practice, covering hypothesis tests, confidence intervals, regression, and chi-squared methods, all set in contexts from biology to economics. Every section pairs an English explanation with its Chinese equivalent so you can sharpen both your conceptual understanding and your exam technique.

WJEC Year 13 统计不仅考察计算能力，更要求你在陌生、真实的情境中选出正确的统计方法。本文将带你进行跨学科综合题型训练，涵盖假设检验、置信区间、回归分析、卡方检验等，全部融入从生物学到经济学的不同背景。每个部分都以英文和中文成对呈现，帮助你同步强化概念理解与考试技巧。

1. The Power of Statistics Across Disciplines | 统计学的跨学科力量

Statistics is the universal language of uncertainty. Whether you are evaluating a new drug, predicting economic trends, or assessing environmental change, the same core methods—hypothesis testing, estimation, and modelling—form the backbone of scientific decision-making. WJEC Year 13 explicitly rewards candidates who can transfer statistical reasoning from one field to another.

统计学是描述不确定性的通用语言。无论你在评估一种新药、预测经济趋势还是分析环境变化，相同的核心方法——假设检验、估计与建模——构成了科学决策的支柱。WJEC Year 13 明确奖励那些能够在不同领域之间迁移统计推理的考生。

The exam will often embed a familiar test, such as a binomial test or a chi-squared test, in an unfamiliar narrative. Your first task is to identify the statistical framework hidden in the context. Is the data categorical or numerical? Is there a known population parameter? Is the sample large or small? Answering these questions unlocks the correct approach.

考试经常会将二项检验或卡方检验等熟悉的测试嵌入一个陌生的叙述中。你的首要任务是从情境中识别隐藏的统计框架。数据是分类的还是数值型的？是否存在已知的总体参数？样本量大还是小？回答这些问题就能找到正确的方法。

---

2. Core Statistical Methods for Year 13 | Year 13 核心统计方法

Before tackling integrated problems, you must have at your fingertips the most commonly tested procedures: binomial hypothesis tests, Poisson goodness-of-fit, normal distribution probability calculations, t-tests for means, chi-squared tests for independence and goodness-of-fit, product moment correlation coefficient (PMCC), and least squares regression. The WJEC specification also expects you to understand confidence intervals for population means and proportions, as well as the distinction between one-tailed and two-tailed tests.

在攻克综合问题之前，你必须熟练掌握最常见的检验程序：二项假设检验、泊松拟合优度检验、正态分布概率计算、均值的 t 检验、独立性与拟合优度的卡方检验、积矩相关系数（PMCC）以及最小二乘回归。WJEC 考纲还要求你理解总体均值与比例的置信区间，并分清单尾与双尾检验的区别。

Key notation includes H₀ and H₁ for null and alternative hypotheses, α for significance level, p-value, X ~ B(n, p) for binomial, X ~ Po(λ) for Poisson, and X ~ N(μ, σ²) for normal distributions. When dealing with sample means, the central limit theorem justifies using the normal distribution for large samples, while small samples require the Student t-distribution with appropriate degrees of freedom.

关键符号包括用于原假设与备择假设的 H₀ 和 H₁、显著性水平 α、p 值、二项分布 X ~ B(n, p)、泊松分布 X ~ Po(λ) 以及正态分布 X ~ N(μ, σ²)。在处理样本均值时，中心极限定理为大样本使用正态分布提供了依据，而小样本则需要结合适当自由度的学生 t 分布。

---

3. Biology: Binomial Hypothesis Test for Drug Efficacy | 生物学：药物有效性的二项假设检验

A biotech company claims that a new antiviral drug is effective for at least 80% of patients. In a clinical trial, 48 out of 60 patients recover. Test the claim at the 5% significance level, using a binomial model.

一家生物技术公司声称其新型抗病毒药物对至少 80% 的患者有效。在一项临床试验中，60 名患者中有 48 人康复。请在 5% 的显著性水平下使用二项模型检验这一声明。

We let p represent the true proportion of patients who recover. The hypotheses are H₀: p = 0.8 and H₁: p < 0.8. This is a one-tailed test because the concern is whether the true percentage is actually lower than claimed.

令 p 表示患者康复的真实比例。假设为 H₀: p = 0.8，H₁: p < 0.8。这是一个单尾检验，因为我们关心的是真实百分比是否低于声明值。

Under H₀, X ~ B(60, 0.8). The observed number of successes is 48. The p-value is P(X ≤ 48). Using binomial cumulative tables or a calculator, we find P(X ≤ 48) = 0.1441 (hypothetical value). Since 0.1441 > 0.05, we do not reject H₀. There is insufficient evidence to say the drug is effective in less than 80% of patients. The sample result is consistent with the company’s statement.

在 H₀ 下，X ~ B(60, 0.8)。观测到的成功次数为 48。p 值 = P(X ≤ 48)。查阅二项累积分布表或使用计算器得到 P(X ≤ 48) = 0.1441。由于 0.1441 > 0.05，我们不拒绝 H₀。没有充分证据表明该药物对少于 80% 的患者有效。样本结果与公司的声明一致。

p-value = P(X ≤ 48) for X ~ B(60, 0.8) ⇒ 0.1441 > 0.05, do not reject H₀

Always remember to interpret the conclusion in the original context: the data do not contradict the biotech company’s claim, so the drug may indeed be effective for at least 80% of patients.

一定要记得把结论放回原始情境中解释：数据没有与生物技术公司的声明相矛盾，因此该药物确实可能对至少 80% 的患者有效。

---

4. Medicine: Poisson Goodness-of-Fit for Emergency Admissions | 医学：急诊入院人数的泊松拟合优度检验

An A&E department believes that the number of emergency admissions per hour follows a Poisson distribution with mean 2.5. Over a random sample of 50 one-hour intervals, the following frequencies were observed. Using a χ² goodness-of-fit test at the 5% significance level, assess whether the Poisson model is appropriate.

某急诊科认为每小时急诊入院人数服从均值为 2.5 的泊松分布。在 50 个随机选取的一小时间隔中，观测到如下频数。请使用 5% 显著性水平的 χ² 拟合优度检验，评估该泊松模型是否恰当。

Admissions per hour (x)	0	1	2	3	4 or more
Observed frequency	4	10	16	12	8

We first calculate the expected probabilities using Po(2.5) and multiply by 50 to obtain expected frequencies. For categories where expected frequency is below 5, we combine adjacent groups to satisfy the test conditions.

首先使用 Po(2.5) 计算期望概率，并乘以 50 得到期望频数。对于期望频数低于 5 的类别，我们合并相邻组以满足检验条件。

After combining categories, we compute the test statistic χ² = Σ (O − E)² / E. Suppose the calculated value is 3.88 and the degrees of freedom after combining are 3. The critical value from the χ² table at the 5% level with 3 d.f. is 7.815. Since 3.88 < 7.815, we do not reject H₀. The Poisson model with mean 2.5 provides an adequate fit to the data.

合并类别后，我们计算检验统计量 χ² = Σ (O − E)² / E。假设计算结果为 3.88，合并后自由度为 3。查阅 χ² 表，5% 水平下自由度为 3 的临界值为 7.815。由于 3.88 < 7.815，我们不拒绝 H₀。均值为 2.5 的泊松模型对数据有足够好的拟合。

χ² = Σ (O − E)² / E, d.f. = number of categories − 1 − estimated parameters

Key reminder: The number of estimated parameters is subtracted from the degrees of freedom. Here we estimated one parameter (λ), so if we had 4 categories after pooling, d.f. = 4 − 1 − 1 = 2—always check your formula booklet.

重要提醒：自由度中需要减去已估计参数的个数。这里我们估计了一个参数（λ），因此若合并后有 4 个类别，自由度 = 4 − 1 − 1 = 2 ——一定要核对你的公式手册。

---

5. Psychology: Two-Sample t-Test for Cognitive Therapy | 心理学：认知疗法的双样本 t 检验

A psychologist investigates whether a new cognitive therapy (Group A) is more effective at reducing anxiety scores than a standard treatment (Group B). Independent random samples give the following summary statistics. Assuming equal population variances, test at the 1% level whether the new therapy yields a lower mean anxiety score.

一位心理学家研究新型认知疗法（A 组）是否比标准疗法（B 组）能更有效地降低焦虑得分。独立随机样本给出如下汇总统计量。假设总体方差相等，在 1% 水平下检验新疗法是否带来更低的平均焦虑得分。

Group	n	Sample mean (x̄)	Sample standard deviation (s)
A (new therapy)	15	18.4	4.1
B (standard)	15	22.6	5.2

The hypotheses are H₀: μₐ = μ_b, H₁: μₐ < μ_b, a one-tailed test. We calculate the pooled variance s²ₚ = [(nₐ−1)s²ₐ + (n_b−1)s²_b] / (nₐ + n_b − 2). Substituting values: s²ₚ = [14 × 4.1² + 14 × 5.2²] / 28 = (14×16.81 + 14×27.04) / 28 = (235.34 + 378.56)/28 = 21.89. Then standard error SE = √[s²ₚ(1/nₐ + 1/n_b)] = √[21.89 × (2/15)] = √2.9187 ≈ 1.708.

假设为 H₀: μₐ = μ_b，H₁: μₐ < μ_b，单尾检验。我们计算合并方差 s²ₚ = [(15−1)×4.1² + (15−1)×5.2²] / (15+15−2) = (14×16.81 + 14×27.04)/28 = (235.34 + 378.56)/28 = 21.89。然后标准误 SE = √[21.89 × (1/15 + 1/15)] = √2.9187 ≈ 1.708。

The test statistic t = (x̄ₐ − x̄_b) / SE = (18.4 − 22.6) / 1.708 = −2.459. Degrees of freedom = 28. The critical value for a one-tailed test at α = 0.01 with 28 d.f. is approximately 2.467. Since the calculated t = −2.459 > −2.467 (or |t| = 2.459 < 2.467), we do not reject H₀ at the 1% level. There is insufficient evidence to claim the new therapy significantly reduces anxiety compared with the standard treatment at this strict significance level.

检验统计量 t = (18.4 − 22.6) / 1.708 = −2.459。自由度为 28。α = 0.01 单尾检验在 28 自由度下的临界值约为 2.467。由于计算得到的 t = −2.459 > −2.467（即 |t| = 2.459 < 2.467），在 1% 水平下我们不拒绝 H₀。没有足够证据表明，在这一严格显著性水平下，新疗法比标准疗法显著降低了焦虑。

t = (x̄ₐ − x̄_b) / SE, SE = √[s²ₚ(1/nₐ + 1/n_b)]

Always check the equality of variances assumption. WJEC questions may provide variance values or expect you to use the pooled estimate. Note that if the sample sizes are very different or if variances are specified as unequal, you would use an unpooled approach with adjusted degrees of freedom.

务必检查方差齐性的假设。WJEC 试题可能直接给出方差数值，或期望你使用合并估计。注意如果样本量差异较大或明确说明方差不齐，则应使用非合并方法并调整自由度。

---

6. Economics: Correlation and Regression for GDP and Unemployment | 经济学：GDP 与失业的相关与回归

An economist collects data on GDP growth (%) and unemployment rate (%) for 10 countries. The product moment correlation coefficient (PMCC) is calculated to be r = −0.584. Test at the 5% significance level whether there is evidence of

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

更多咨询请联系16621398022（同微信）