A-Level数学概率分布二项泊松正态

1. 概率分布概述 Overview of Probability Distributions

In A-Level Mathematics, a probability distribution describes how the probabilities of a random variable are distributed across its possible values. Understanding distributions is fundamental to statistical inference, hypothesis testing, and modelling real-world phenomena. A discrete random variable takes a countable set of values, each with an associated probability that sums to 1 across the entire sample space. 概率分布描述了随机变量取值与其对应概率之间的关系。理解概率分布是统计推断、假设检验和现实世界建模的基础。离散型随机变量取可数个值，每个值对应一个概率，所有概率之和为1。

The probability mass function (PMF) gives P(X = x) for each value x in the domain. For discrete distributions, we often use the notation X ~ Distribution(parameters) to specify the model, and we compute probabilities using either the PMF formula directly or cumulative distribution tables provided in the formula booklet. 概率质量函数(PMF)给出定义域中每个值x对应的P(X = x)。对于离散分布，我们通常使用符号X ~ Distribution(参数)来指定模型，并通过PMF公式或公式手册中的累积分布表来计算概率。

2. 二项分布 The Binomial Distribution

The binomial distribution models the number of successes in a fixed number n of independent Bernoulli trials, each with the same probability of success p. The conditions for a binomial model are： each trial has exactly two outcomes (success or failure), the probability p remains constant across trials, the n trials are independent, and the number of trials is fixed in advance. If X represents the number of successes in n trials, we write X ~ B(n, p). 二项分布用于描述在固定次数n的独立伯努利试验中成功的次数，每次试验具有相同的成功概率p。二项模型的条件是：每次试验只有两种结果，概率p在试验间保持不变，n次试验相互独立，且试验次数预先确定。若X表示n次试验中成功的次数，我们记作X ~ B(n, p)。

The probability of exactly r successes is given by P(X = r) = C(n, r) × p^r × (1-p)^(n-r), where C(n, r) = n!/(r!(n-r)!) is the binomial coefficient. The mean or expected value of a binomial random variable is E(X) = np, and the variance is Var(X) = np(1-p). These properties are frequently tested in A-Level exam questions, especially when combined with hypothesis testing or when using the normal approximation for large n. 恰好r次成功的概率由P(X = r) = C(n, r) × p^r × (1-p)^(n-r)给出，其中C(n, r) = n!/(r!(n-r)!)是二项式系数。二项随机变量的期望值为E(X) = np，方差为Var(X) = np(1-p)。这些性质在A-Level考试中经常出现，尤其是结合假设检验或在大样本下使用正态近似时。

A common exam question involves finding the most likely number of successes, known as the mode. For X ~ B(n, p), the mode satisfies (n+1)p – 1 ≤ mode ≤ (n+1)p. When (n+1)p is an integer, there are two modes： (n+1)p – 1 and (n+1)p. Another typical problem asks for the smallest sample size n required to achieve a certain probability of at least one success： solve 1 – (1-p)^n ≥ target probability. 常见的考试题型是求最可能的成功次数，即众数。对于X ~ B(n, p)，众数满足(n+1)p – 1 ≤ 众数 ≤ (n+1)p。当(n+1)p为整数时，存在两个众数：(n+1)p – 1和(n+1)p。另一个典型问题是求达到至少一次成功的某个概率所需的最小样本量n：解1 – (1-p)^n ≥ 目标概率。

3. 泊松分布 The Poisson Distribution

The Poisson distribution models the number of events occurring in a fixed interval of time or space, given that events occur independently at a constant average rate λ. For a Poisson random variable X ~ Po(λ), the probability of exactly r events is P(X = r) = (e^(-λ) × λ^r) / r!, for r = 0, 1, 2, … . Unlike the binomial distribution, the Poisson distribution has no fixed upper limit on the number of events： the sample space extends to infinity. 泊松分布用于描述在固定的时间或空间区间内事件发生的次数，假设事件以恒定平均速率λ独立发生。对于泊松随机变量X ~ Po(λ)，恰好发生r个事件的概率为P(X = r) = (e^(-λ) × λ^r) / r!，其中r = 0, 1, 2, …。与二项分布不同，泊松分布对事件发生次数没有固定上限：样本空间延伸至无穷大。

The key property of the Poisson distribution is that the mean and variance are both equal to λ. This equality provides a diagnostic check： if a dataset has sample mean ≈ sample variance, the Poisson model may be appropriate. The Poisson distribution is also additive： if X ~ Po(λ1) and Y ~ Po(λ2) are independent, then X + Y ~ Po(λ1 + λ2). This property is useful when modelling combined event counts from multiple independent sources. 泊松分布的关键性质是期望值和方差都等于λ。这种相等性提供了一个诊断检查：如果数据集的样本均值约等于样本方差，泊松模型可能是合适的。泊松分布还具有可加性：若X ~ Po(λ1)且Y ~ Po(λ2)相互独立，则X + Y ~ Po(λ1 + λ2)。这一性质在模拟多个独立来源的组合事件计数时非常有用。

4. 泊松近似二项分布 Poisson Approximation to the Binomial

When n is large and p is small, the binomial distribution B(n, p) can be approximated by the Poisson distribution Po(np). The standard rule of thumb is that the approximation is acceptable when n > 50 and np < 5, or alternatively when n is large and p < 0.1. This approximation simplifies calculations significantly, as evaluating binomial coefficients for large n is computationally intensive. 当n很大而p很小时，二项分布B(n, p)可以用泊松分布Po(np)来近似。标准的经验法则是当n > 50且np < 5时近似是可接受的，或者当n很大且p < 0.1时。这种近似可以显著简化计算，因为对大的n计算二项式系数计算量很大。

For example, if a factory produces components with a defect rate of 0.2% and a batch contains 1000 components, the number of defects X ~ B(1000, 0.002) is well approximated by X ~ Po(2). The approximation works because λ = np = 2 satisfies np < 5. In exam questions, you should state the approximation conditions explicitly and compare the exact binomial probability with the Poisson approximation to demonstrate understanding. 例如，若某工厂生产元件的缺陷率为0.2%，一批包含1000个元件，则缺陷数量X ~ B(1000, 0.002)可以用X ~ Po(2)很好地近似。这种近似成立是因为λ = np = 2满足np < 5。在考试题中，你应明确说明近似条件，并比较精确二项概率与泊松近似值以展示理解。

5. 正态分布 The Normal Distribution

The normal distribution is the most important continuous probability distribution in A-Level statistics. A normal random variable X ~ N(μ, σ^2) has mean μ and variance σ^2, with the familiar bell-shaped probability density function f(x) = (1/(σ√(2π))) × e^(-(x-μ)^2/(2σ^2)). The standard normal distribution Z ~ N(0, 1) is obtained by standardising： Z = (X – μ)/σ. This transformation is the foundation of all normal probability calculations in the A-Level syllabus. 正态分布是A-Level统计中最重要的连续概率分布。正态随机变量X ~ N(μ, σ^2)具有期望值μ和方差σ^2，其概率密度函数为熟悉的钟形曲线f(x) = (1/(σ√(2π))) × e^(-(x-μ)^2/(2σ^2))。标准正态分布Z ~ N(0, 1)通过标准化得到：Z = (X – μ)/σ。这一变换是A-Level课程中所有正态概率计算的基础。

Students must be proficient in using the standard normal distribution table to find probabilities. For a given z-value, the table gives Φ(z) = P(Z ≤ z), the cumulative probability up to z. Common operations include： finding P(a < X < b) = Φ((b-μ)/σ) - Φ((a-μ)/σ), finding the value x such that P(X > x) = α by solving (x-μ)/σ = z_α where Φ(z_α) = 1-α, and finding symmetric intervals P(μ – kσ < X < μ + kσ). The 68-95-99.7 empirical rule states that approximately 68%, 95%, and 99.7% of observations fall within 1, 2, and 3 standard deviations of the mean. 学生必须熟练使用标准正态分布表来求概率。对于给定的z值，分布表给出Φ(z) = P(Z ≤ z)，即z之前的累积概率。常见操作包括：求P(a < X < b) = Φ((b-μ)/σ) - Φ((a-μ)/σ)，通过解(x-μ)/σ = z_α求满足P(X > x) = α的x值（其中Φ(z_α)=1-α），以及求对称区间P(μ – kσ < X < μ + kσ)。68-95-99.7经验法则指出约68%、95%和99.7%的观测值分别落在均值的1个、2个和3个标准差范围内。

6. 正态近似二项分布 Normal Approximation to the Binomial

For large n, the binomial distribution B(n, p) can be approximated by the normal distribution N(np, np(1-p)). The condition for this approximation is that both np > 5 and n(1-p) > 5, ensuring the distribution is sufficiently symmetric and the tails extend far enough. A continuity correction of ±0.5 must be applied because we are approximating a discrete distribution with a continuous one. 对于大样本量n，二项分布B(n, p)可以用正态分布N(np, np(1-p))近似。近似的条件是np > 5且n(1-p) > 5，确保分布足够对称且尾部延伸足够远。必须应用±0.5的连续性校正，因为我们用连续分布近似离散分布。

For example, to approximate P(X ≥ 45) where X ~ B(100, 0.4), we compute using Y ~ N(40, 24). With continuity correction： P(X ≥ 45) ≈ P(Y > 44.5) = P(Z > (44.5-40)/√24) = P(Z > 0.9186) ≈ 1 – Φ(0.92). The continuity correction is a common source of marks in A-Level exam questions： students who forget it lose accuracy marks even when the rest of the calculation is correct. 例如，要近似P(X ≥ 45)，其中X ~ B(100, 0.4)，我们使用Y ~ N(40, 24)计算。应用连续性校正：P(X ≥ 45) ≈ P(Y > 44.5) = P(Z > (44.5-40)/√24) = P(Z > 0.9186) ≈ 1 – Φ(0.92)。连续性校正是A-Level考试中常见的得分点：忘记校正的学生即使其余计算正确也会丢失准确性分数。

7. 假设检验 Hypothesis Testing with Distributions

Hypothesis testing is a core application of probability distributions in A-Level Statistics. The structure involves： stating null and alternative hypotheses (H₀ and H₁), choosing a significance level α (commonly 5% or 1%), calculating the test statistic from sample data, finding the p-value or critical region, and drawing a conclusion in context. For binomial tests, the test statistic is the observed number of successes, and probabilities are computed directly from the binomial distribution or its normal approximation. 假设检验是A-Level统计中概率分布的核心应用。其结构包括：陈述原假设和备择假设（H₀和H₁），选择显著性水平α（通常为5%或1%），从样本数据计算检验统计量，求p值或临界域，并在实际背景中得出结论。对于二项检验，检验统计量是观测到的成功次数，概率直接从二项分布或其正态近似计算。

For one-tailed tests, the critical region lies entirely in one tail of the distribution. For a right-tailed test H₁: p > p₀, the critical region is X ≥ c where P(X ≥ c | H₀) ≤ α. For two-tailed tests H₁: p ≠ p₀, the critical region is split between both tails, each with probability α/2. A common exam pitfall is halving the significance level incorrectly： in a two-tailed test, compare the p-value against α (not α/2), but find the critical value such that each tail has probability α/2. Always state the conclusion in the context of the original problem. 对于单尾检验，临界域完全位于分布的一侧尾部。对于右尾检验H₁: p > p₀，临界域为X ≥ c，其中P(X ≥ c | H₀) ≤ α。对于双尾检验H₁: p ≠ p₀，临界域分布在两侧尾部，每侧概率为α/2。常见的考试陷阱是错误地减半显著性水平：在双尾检验中，将p值与α（而非α/2）比较，但求临界值时每侧尾部概率为α/2。始终在原始问题背景中陈述结论。

8. 解题示例 Worked Examples

A manufacturer claims that at most 5% of its products are defective. A random sample of 50 products is inspected and 5 defectives are found. Test the manufacturer’s claim at the 5% significance level. Let X ~ B(50, 0.05) under H₀, so we test H₀: p = 0.05 vs H₁: p > 0.05. P(X ≥ 5) = 1 – P(X ≤ 4). Using the binomial table or calculator： P(X ≤ 4) ≈ 0.896, so p-value = 0.104 > 0.05. We do not reject H₀： there is insufficient evidence to dispute the manufacturer’s claim at the 5% level. 某制造商声称其产品最多5%有缺陷。随机抽取50个产品检查，发现5个缺陷品。在5%显著性水平下检验制造商的说法。设H₀下X ~ B(50, 0.05)，检验H₀: p = 0.05 vs H₁: p > 0.05。P(X ≥ 5) = 1 – P(X ≤ 4)。使用二项分布表或计算器：P(X ≤ 4) ≈ 0.896，因此p值 = 0.104 > 0.05。我们不拒绝H₀：在5%水平上没有足够证据质疑制造商的说法。

For a Poisson example, a call centre receives an average of 12 calls per hour. Find the probability of receiving more than 15 calls in a given hour. Let X ~ Po(12). P(X > 15) = 1 – P(X ≤ 15). Using cumulative Poisson tables or the formula： P(X ≤ 15) ≈ 0.844, so P(X > 15) ≈ 0.156. This means there is approximately a 15.6% chance of exceeding 15 calls in any given hour. A normal approximation using Y ~ N(12, 12) with continuity correction gives P(X > 15) ≈ P(Y > 15.5) = P(Z > 1.01) ≈ 0.156, matching the exact Poisson calculation closely. 对于泊松示例，某呼叫中心平均每小时接到12个电话。求在某小时内接到超过15个电话的概率。设X ~ Po(12)。P(X > 15) = 1 – P(X ≤ 15)。使用累积泊松表或公式：P(X ≤ 15) ≈ 0.844，因此P(X > 15) ≈ 0.156。这意味着在任何给定的小时内，有大约15.6%的概率超过15个电话。使用正态近似Y ~ N(12, 12)并应用连续性校正：P(X > 15) ≈ P(Y > 15.5) = P(Z > 1.01) ≈ 0.156，与精确泊松计算密切匹配。

9. 考试技巧 Exam Tips

Always check the model assumptions before applying any distribution. For the binomial： are there exactly two outcomes? Is p constant? Are trials independent? For the Poisson： are events occurring randomly and independently at a constant average rate? For the normal： is the underlying variable continuous and approximately symmetric? Stating these checks explicitly in your answer demonstrates exam technique and can earn method marks even if the subsequent calculation contains an error. 在应用任何分布之前务必检查模型假设。对于二项分布：是否恰好有两种结果？p是否恒定？试验是否独立？对于泊松分布：事件是否以恒定平均速率随机且独立地发生？对于正态分布：基础变量是否是连续的且近似对称？在答案中明确陈述这些检查可以展示考试技巧，即使后续计算有误也能获得方法分。

When using the formula booklet, know exactly where to find binomial cumulative probabilities, Poisson cumulative probabilities, and the standard normal distribution table. The binomial tables are organised by n and p, typically covering n up to 20 or 30. For larger n, use the Poisson or normal approximation as appropriate. For the normal distribution table, values of z are given to 2 decimal places, with the first decimal down the left column and the second decimal across the top row. Practice interpolating between table values for z-values with 3 decimal places： linear interpolation is sometimes required in A-Level exams. 使用公式手册时，准确知道二项累积概率、泊松累积概率和标准正态分布表的位置。二项分布表按n和p排列，通常覆盖n到20或30。对于更大的n，适当使用泊松或正态近似。对于正态分布表，z值精确到2位小数，第一位小数在左列，第二位小数在顶行。练习对3位小数的z值在表值之间进行插值：A-Level考试中有时需要线性插值。

10. 关键术语 Key Bilingual Terms

📚 需要课程辅导或获取完整资源？

联系电话 / 微信：16621398022

A-Level数学 概率分布 二项 泊松 正态