Comparative Theories: Hypothesis Tests for Two Populations | 比较理论：两个总体的假设检验

📚 Comparative Theories: Hypothesis Tests for Two Populations | 比较理论：两个总体的假设检验

In Edexcel A-Level Mathematics, comparative theories bring together hypothesis testing, distributions and confidence intervals to answer one central question: is an observed difference between two groups statistically meaningful, or could it have arisen by chance? These methods appear in the Statistics component and are tested through real-world contexts such as medical trials, production quality and behavioural studies.

在 Edexcel A-Level 数学中，比较理论将假设检验、分布和置信区间结合起来，回答一个核心问题：两个组之间观察到的差异是否具有统计意义，还是可能由随机因素引起？这些方法出现在统计部分，并通过医学试验、生产质量和行为研究等实际背景进行考查。

1. The Comparative Mindset in Statistics | 统计中的比较思维

Comparison is at the heart of statistical inference. We rarely care about a single number in isolation; we usually want to know whether one treatment, one production line or one teaching method performs differently from another. A comparative test formalises this by setting up a null hypothesis of no difference and asking whether sample data provide enough evidence against it.

比较是统计推断的核心。我们很少孤立地关心一个数字；我们通常想知道一种处理、一条生产线或一种教学方法是否与另一种表现不同。比较检验通过设立“无差异”的原假设，并判断样本数据是否提供足够证据来拒绝它，从而将这个问题形式化。

The two main families are tests for means and tests for proportions. Within means, the key distinction is between independent samples, where the two groups have no natural pairing, and paired samples, where each observation in one group is matched to an observation in the other.

主要有两大类：均值检验和比例检验。在均值检验中，关键区别在于独立样本（两组没有自然配对）与配对样本（一组中的每个观测值与另一组中的观测值一一对应）。

2. Independent Samples vs Paired Samples | 独立样本与配对样本

Independent samples arise when two separate groups are selected and measured under different conditions, for example 30 patients given drug A and 30 given drug B. The sample sizes can differ, and the order of observations within each group does not matter.

独立样本出现在两个独立组被选中并在不同条件下测量的情况，例如 30 名患者服用药物 A，30 名服用药物 B。样本量可以不同，每组内观测值的顺序无关紧要。

Paired samples arise when the same individual or matched unit is measured twice, such as before and after a training programme, or when twins are split between two treatments. The comparison is based on the differences within each pair, which removes variation between individuals from the analysis.

配对样本出现在同一个体或匹配单元被测量两次的情况，例如培训前后，或者双胞胎被分到两种处理中。比较基于每对内部的差异，这从分析中消除了个体之间的变异。

3. Comparing Two Means: Known Population Variances | 比较两个均值：已知总体方差

When both population variances σ₁² and σ₂² are known, the sampling distribution of x̄₁ − x̄₂ is exactly normal. The test statistic is:

当两个总体方差 σ₁² 和 σ₂² 已知时，x̄₁ − x̄₂ 的抽样分布是精确正态的。检验统计量为：

z = (x̄₁ − x̄₂) / √(σ₁²/n₁ + σ₂²/n₂)

This situation is rare in practice, but it underpins the logic of all two-sample tests. At A-Level, it usually appears when a normal distribution is assumed and variances are stated, or when sample sizes are large enough for the central limit theorem to justify a normal approximation.

这种情况在实践中很少见，但它支撑了所有双样本检验的逻辑。在 A-Level 中，通常当假定正态分布并给出方差，或当样本量足够大、中心极限定理支持正态近似时，会出现这种情况。

4. Comparing Two Means: Unknown Equal Variances | 比较两个均值：未知但方差相等

More realistically, the population variances are unknown. If we are willing to assume that the two populations have the same variance, we combine the sample variances into a pooled estimate:

更现实的情况是总体方差

Published by TutorHao | A-Level Mathematics Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）