📚 Hypothesis Testing Essentials for IGCSE WJEC Maths | IGCSE WJEC 数学：假设检验 考点精讲
Hypothesis testing is a fundamental statistical method used to make inferences about population parameters based on sample data. In the IGCSE WJEC Maths specification, you will learn to perform hypothesis tests for a population proportion using the binomial distribution. This guide covers all key concepts, step-by-step procedures, and exam techniques to help you master this topic and secure top marks.
假设检验是一种基于样本数据对总体参数进行推断的基础统计方法。在 IGCSE WJEC 数学考纲中，你将学习如何使用二项分布对总体比例进行假设检验。本指南涵盖了所有关键概念、分步流程和考试技巧，帮助你掌握这个主题并取得高分。
1. Introduction to Hypothesis Testing | 假设检验简介
Hypothesis testing allows us to decide whether a claim about a population parameter is likely to be true, based on evidence from a random sample. In WJEC IGCSE, the focus is on testing a population proportion p (e.g., the probability of success in a binomial experiment). We collect data and compare the observed number of successes to what we would expect if the claim were true. If the observed result is very unlikely under the claim, we reject the claim.
假设检验允许我们基于随机样本的证据，判断关于总体参数的某个声明是否可能为真。在 WJEC IGCSE 中，重点是对总体比例 p（例如二项试验中成功的概率）进行检验。我们收集数据，并将观察到的成功次数与若声明为真时的预期结果进行比较。如果观察结果在该声明下非常不可能发生，我们就拒绝该声明。
2. Null and Alternative Hypotheses | 原假设与备择假设
tive hypothesis, denoted H₁ or Hₐ, is what we suspect might be true instead (e.g., p > 0.5, p < 0.5, or p ≠ 0.5). The test aims to find statistical evidence against H₀. We never 'prove' H₁; we only conclude whether there is enough evidence to reject H₀ in favour of H₁.
每个假设检验都以两个对立的陈述开始。原假设，记作 H₀，是默认假设（例如 p = 0.5）。备择假设，记作 H₁ 或 Hₐ，是我们怀疑可能成立的另一种情况（例如 p > 0.5、p < 0.5 或 p ≠ 0.5）。检验的目的是寻找反对 H₀ 的统计证据。我们永远无法“证明” H₁，只能得出结论是否有足够证据拒绝 H₀ 而支持 H₁。
In WJEC exams, you must state both hypotheses clearly using correct notation. The null hypothesis always contains an equality. The alternative hypothesis indicates the direction of interest: one-tailed (if we suspect a change in one direction) or two-tailed (if we are testing for any difference).
在 WJEC 考试中，你必须使用正确的符号清晰地陈述两个假设。原假设总是包含等号。备择假设则指明关注的方向：单尾（如果怀疑朝某一方向变化）或双尾（如果检验是否存在任何差异）。
3. Significance Level and Type I Error | 显著性水平与第一类错误
The significance level, denoted by α (alpha), is the probability of rejecting H₀ when it is actually true. This mistake is called a Type I error. Common choices for α are 5% (0.05) and 1% (0.01). If the p-value obtained from the data is less than α, or if the test statistic falls in the critical region, we reject H₀. A smaller α makes it harder to reject H₀ and requires stronger evidence.
显著性水平，记作 α（alpha），是当 H₀ 实际为真时拒绝它的概率。这种错误称为第一类错误。α 的常见选择为 5% (0.05) 和 1% (0.01)。如果从数据得到的 p 值小于 α，或者检验统计量落入临界区域，我们就拒绝 H₀。α 越小，拒绝 H₀ 就越困难，所需的证据也越强。
In IGCSE WJEC, the significance level is usually given in the question. You do not need to calculate Type II errors (failing to reject a false H₀), but understanding the risk of a Type I error helps you interpret the strength of your conclusion.
在 IGCSE WJEC 中，显著性水平通常会在题目中给出。你不需要计算第二类错误（未拒绝错误的 H₀），但理解第一类错误的风险有助于你解释结论的可靠性。
4. Test Statistic: The Number of Successes | 检验统计量：成功次数
For a hypothesis test about a population proportion, the test statistic X is the number of successes in a fixed number of independent trials n. Under H₀, X follows a binomial distribution: X ~ B(n, p₀), where p₀ is the hypothesised proportion. The observed number of successes from the sample is denoted by x. We evaluate how extreme this x is, assuming H₀ is true.
在对总体比例的假设检验中，检验统计量 X 是在固定的独立试验次数 n 中的成功次数。在 H₀ 下，X 服从二项分布：X ~ B(n, p₀)，其中 p₀ 是假设的比例。样本中观察到的成功次数记作 x。我们在假定 H₀ 为真的条件下，评估这个 x 有多么极端。
You will either find the probability of obtaining a result at least as extreme as x (the p-value) or determine the critical region of values that would lead to rejecting H₀. Both approaches are accepted in WJEC and lead to the same conclusion.
你可以求得到至少与 x 一样极端的结果的概率（p 值），或者确定会导致拒绝 H₀ 的临界值区域。这两种方法在 WJEC 中均可接受，并且会得出相同的结论。
5. The Binomial Distribution Model | 二项分布模型
ssible outcomes (success or failure). The probability of exactly r successes is given by P(X = r) = C(n, r) × pr × (1 – p)n – r on cumulative binomial probability tables or your calculator to find P(X ≤ k) or P(X ≥ k).
当试验次数 n 固定、每次试验相互独立且成功概率 p 相同、只有两种可能结果（成功或失败）时，适用二项分布。恰好有 r 次成功的概率由 P(X = r) = C(n, r) × pr × (1 – p)n – r 给出。在 IGCSE 假设检验中，你需要借助二项累积概率表或计算器来求 P(X ≤ k) 或 P(X ≥ k)。
Always check that the conditions for a binomial model are satisfied before proceeding. In exam questions, these conditions are usually stated or implied (e.g., 'independent spins of a fair spinner', 'random sample of bulbs').
在进行检验之前，务必检查二项模型的条件是否满足。在考试题目中，这些条件通常会说明或隐含（例如“公平转盘的独立旋转”、“灯泡的随机样本”）。
6. One-Tailed vs Two-Tailed Tests | 单尾与双尾检验
A one-tailed test is used when the alternative hypothesis specifies a direction: H₁: p > p₀ (right-tailed) or H₁: p < p₀ (left-tailed). The entire significance level α is placed in one tail of the distribution. A two-tailed test is used for H₁: p ≠ p₀, where we split α equally into both tails (α/2 in each) because we are testing for a difference in either direction.
当备择假设指明了方向时，使用单尾检验：H₁: p > p₀（右尾）或 H₁: p < p₀（左尾）。此时整个显著性水平 α 都放在分布的一个尾部。对于 H₁: p ≠ p₀，使用双尾检验，我们将 α 均分至两个尾部（每侧 α/2），因为我们检验的是任一方向的差异。
Choosing the correct test is vital. Read the wording carefully: 'increase', 'more than', 'higher' suggest a right-tailed test; 'decrease', 'less than', 'lower' suggest a left-tailed test; 'different', 'changed', 'not equal' indicate a two-tailed test.
选择正确的检验至关重要。仔细阅读措辞：“增加”、“多于”、“高于”提示右尾检验；“减少”、“少于”、“低于”提示左尾检验；“不同”、“改变”、“不等于”表明双尾检验。
7. Critical Region Method | 临界区域法
tailed test, we find the smallest integer c such that P(X ≥ c) ≤ α; the critical region is X ≥ c. For a left-tailed test, we find the largest integer d such that P(X ≤ d) ≤ α; the critical region is X ≤ d. In a two-tailed test, we find both lower and upper critical values, each corresponding to α/2 probability in that tail.
临界区域由所有会导致在给定显著性水平下拒绝 H₀ 的检验统计量 X 的值组成。对于右尾检验，我们找出最小的整数 c 使得 P(X ≥ c) ≤ α；临界区域为 X ≥ c。对于左尾检验，我们找出最大的整数 d 使得 P(X ≤ d) ≤ α；临界区域为 X ≤ d。在双尾检验中，我们分别找出左右两个临界值，每个对应其尾部 α/2 的概率。
If the observed number of successes x falls inside the critical region, we reject H₀. Otherwise, we do not reject H₀. This method is often quicker when using binomial tables, as you can compare x directly against the critical value.
如果观察到的成功次数 x 落入临界区域内，我们就拒绝 H₀。否则，我们不拒绝 H₀。使用二项分布表时，这种方法通常更快，因为你可以直接将 x 与临界值进行比较。
8. P-Value Method | P值法
₁: p < p₀, p-value = P(X ≤ x). For a two-tailed test, the p-value is twice the smaller tail probability: p-value = 2 × min{P(X ≤ x), P(X ≥ x)}. If the p-value ≤ α, we reject H₀.
p 值是在假定 H₀ 为真的条件下，得到至少与观察结果一样极端的结果的概率。对于 H₁: p > p₀，p 值 = P(X ≥ x)。对于 H₁: p < p₀，p 值 = P(X ≤ x)。对于双尾检验，p 值是较小尾部概率的两倍：p 值 = 2 × min{P(X ≤ x), P(X ≥ x)}。如果 p 值 ≤ α，则拒绝 H₀。
Reporting the p-value provides extra information about the strength of evidence. A very
Published by TutorHao | IGCSE Mathematics Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
或截图后用「扫一扫 → 相册」选择本图