📚 Sums of Independent Random Variables | 独立随机变量的和
In A-Level Statistics, understanding how to add independent random variables is absolutely crucial. Whether you are dealing with normal, Poisson, binomial, or any other distribution, knowing the rules for expectations and variances of sums allows you to model real-world combined quantities. This article unpacks the key results for Edexcel Mathematics, covering linear combinations of independent random variables, sums of specific distributions, and the central limit theorem.
在A-Level统计学中，理解如何对独立随机变量进行加和是绝对关键的。无论你处理的是正态、泊松、二项还是其他分布，掌握和的期望与方差规则都能帮助你为现实世界中的组合量建立模型。本文针对Edexcel数学考试，拆解了独立随机变量的线性组合、特定分布的加和以及中心极限定理等重要结果。
1. Key Concepts: Independence and Sums | 关键概念：独立性与和
When we say two random variables X and Y are independent, the value taken by X has no influence on the value taken by Y. This assumption lets us simplify the variance of their sum and guarantees that certain results hold – for example, the convolution of their probability distributions.
当我们说两个随机变量X和Y独立时，意味着X的取值不会影响Y的取值。这个假设允许我们简化它们和的方差，并确保某些结论成立——例如，它们概率分布的卷积。
The “sum” here may be a simple X + Y or a more general linear combination such as aX + bY, where a and b are constants. Mastering these cases opens the door to dealing with totals, differences, averages and sample behaviour.
这里的“和”可以是简单的X + Y，也可以是像aX + bY这样更一般的线性组合，其中a与b为常数。掌握这些情形就能应对总量、差值、均值以及样本行为。
2. Expectation of a Linear Combination | 线性组合的期望
The expectation operator is always linear, even if the variables are dependent. For any constants a and b:
期望算子始终是线性的，即使变量之间存在依赖关系。对任意常数a与b：
E(aX + bY) = aE(X) + bE(Y)
ree independent random variables X, Y and Z, then E(2X - 3Y + 5Z) = 2E(X) - 3E(Y) + 5E(Z).
这一结果自然地推广到两个以上的变量。例如，若有三个独立随机变量X、Y和Z，则E(2X - 3Y + 5Z) = 2E(X) - 3E(Y) + 5E(Z)。
No independence is required for this property, so it is the safest starting point when working with combined quantities. In exam questions, you will often be given E(X) and E(Y) directly, or you will deduce them from the distribution parameters.
此性质不需要独立性假设，因此它是处理组合量时最安全的出发点。在考试题中，通常会直接给出E(X)和E(Y)，或者要求你从分布参数推导出来。
3. Variance of a Linear Combination (Independent Case) | 线性组合的方差（独立情况）
Unlike expectation, variance behaves multiplicatively with constants, and the independence condition is vital. For independent random variables X and Y:
与期望不同，方差对常数具有乘法效应，而且独立性条件至关重要。对于独立的随机变量X和Y：
Var(aX + bY) = a²Var(X) + b²Var(Y)
stants are squared. If you have a difference, the formula still adds the variances: Var(aX – ar(Y). The minus sign disappears because covariance is zero. This is a very common pitfall – students often incorrectly subtract variances.
注意常数都要平方。如果是求差，方差公式仍然是相加：Var(aX – bY) = a²Var(X) + b²Var(Y)。减号消失是因为协方差为零。这是非常常见的陷阱——学生经常错误地减去方差。
For n independent random variables all with the same variance σ², the variance of the total is nσ². Hence the standard deviation of the sum grows with √n. This property underpins sampling distributions.
对于n个独立且方差同为σ²的随机变量，总和的方差为nσ²。因此总和的标准差随√n增长。这一性质是抽样分布的基础。
4. Sums of Independent Normal Distributions | 独立正态分布的和
If X ∼ N(μ₁, σ₁²) and Y ∼ N(μ₂, σ₂²) are independent, then any linear combination also follows a normal distribution. In particular:
如果X ∼ N(μ₁, σ₁²) 与 Y ∼ N(μ₂, σ₂²) 独立，那么它们的任何线性组合也服从正态分布，具体有：
X + Y ∼ N(μ₁ + μ₂, σ₁² + σ₂²)
aX + bY ∼ N(aμ₁ + bμ₂, a²σ₁² + b²σ₂²)
This result is exceptionally useful in quality control, measurement errors and A-Level problems where combined lengths, weights or volumes are analysed. You must never add standard deviations; always add variances.
这一结果在质量控制、测量误差以及涉及长度、重量或体积组合的A-Level题目中格外有用。千万不能直接加标准差；一定要先加方差。
The difference of independent normals is also normal: X – Y ∼ N(μ₁ – μ₂, σ₁² + σ₂²). Again, the variances add, so the difference of two precise measurements can be less precise than each individually.
独立正态变量的差也服从正态分布：X – Y ∼ N(μ₁ – μ₂, σ₁² + σ₂²)。同样，方差相加，因此两个精确测量值的差可能比各自更不精确。
5. Sums of Independent Poisson Distributions | 独立泊松分布的和
If X ∼ Po(λ₁) and Y ∼ Po(λ₂) are independent, then the total number of events occurring in non-overlapping intervals or two independent streams can be modelled by:
如果X ∼ Po(λ₁)与Y ∼ Po(λ₂)独立，那么发生在不重叠时段或两个独立流中的事件总数可用下式建模：
X + Y ∼ Po(λ₁ + λ₂)
The parameter λ adds simply, preserving the Poisson shape. This property makes the Poisson distribution additive and ideal for counting processes such as calls arriving at a switchboard or defects in fabric.
参数λ可以直接相加，而泊松形状保持不变。这个性质使泊松分布具备可加性，非常适合对电话呼入、布料疵点等计数过程建模。
Be careful: this additivity holds only for independent Poisson variables. If the parameters are given as rates, multiply by the corresponding time or area first. For instance, if X ∼ Po(3 per hour) observed for 2 hours, treat it as Po(6).
注意：该可加性仅对独立的泊松变量成立。如果参数以速率给出，要先将速率乘以相应的时间或面积。例如，X ∼ Po(每小时3次) 观察2小时，则视为Po(6)。
6. Sums of Independent Binomial Distributions | 独立二项分布的和
When you have independent binomial variables with the same probability of success p, their sum is also binomial:
当你面对成功概率p相同的独立二项变量时，它们的和也服从二项分布：
X ∼ B(n₁, p), Y ∼ B(n₂, p) ⇒ X + Y ∼ B(n₁ + n₂, p)
The total number of trials adds up while p remains unchanged. This result arises naturally if you pool together two independent binomial experiments with the same success probability.
试验总次数相加，而p保持不变。如果你将两个成功概率相同的独立二项试验合并，这个结果便自然成立。
If the probabilities are different, the sum is not binomial. In that case you cannot use a simple closed form; instead you would need to consider all possible outcomes and their probabilities, or approximate using a normal distribution if the conditions allow.
如果概率不同，总和就不再是二项分布。此时不能使用简单的解析形式；你需要考虑所有可能的结果及其概率，或在条件允许时用正态分布近似。
7. The Distribution of the Sample Mean | 样本均值的分布
Let X₁, X₂, … pendent observations from a distribution with mean μ and variance σ². The sample mean is X̄ = (ΣXᵢ)/n. Applying the rules of expectation and variance:
设X₁, X₂, …, Xₙ是来自均值为μ、方差为σ²的总体的独立观测值。样本均值为X̄ = (ΣXᵢ)/n。利用期望与方差规则可得：
E(X̄) = μ
Var(X̄) = σ²/n
This shows that the sample mean is an unbiased estimator of μ and becomes more precise as n increases – its standard error is σ/√n.
这表明样本均值是μ的无偏估计，且随着n增大变得更加精确——其标准误为σ/√n。
If the original population is normally distributed, then X̄ is exactly normal: X̄ ∼ N(μ, σ²/n). This exact result is a direct consequence of the linear combination property of normals.
如果原始总体服从正态分布，则X̄恰好服从正态分布：X̄ ∼ N(μ, σ²/n)。这一精确结果是正态变量线性组合性质的直接推论。
8. Applying the Central Limit Theorem | 中心极限定理的应用
When the original population is not normal but has finite mean μ and variance σ², the Central Limit Theorem (CLT) states that for sufficiently large n, the distribution of the sample mean X̄ is approximately normal:
当原始总体并非正态但具有有限均值μ和方差σ²时，中心极限定理（CLT）指出，对于充分大的n，样本均值X̄的分布近似正态：
X̄ ≈ N(μ, σ²/n) provided n is large
a rule of thumb for the CLT to hold, unless the original distribution is extremely skewed.
同样，总和ΣXᵢ近似服从N(nμ, nσ²)。在Edexcel考试中，通常以n ≥ 30作为中心极限定理成立的参考标准，除非原始分布非常偏斜。
This theorem allows you to answer probability questions about total times, total weights or average scores even when the underlying individual distribution is unknown, as long as you know the mean and variance.
该定理使你能够回答关于总时间、总重量或平均分数的概率问题，即使单个观测值的分布未知，只要你已知均值和方差。
9. Common Mistakes and Exam Tips | 常见错误与考试技巧
Mistake 1: Forgetting to square constants when finding variance. Always write a² and b² explicitly. This is the most frequently tested detail.
错误1：求方差时忘记将常数平方。始终明确写出a²与b²。这是考查频率最高的细节。
Mistake 2: Subtracting variances for a difference. Var(X – Y) = Var(X) + Var(Y) if X and Y are independent. Check your signs carefully.
错误2：计算差时减去方差。若X和Y独立，Var(X – Y) = Var(X) + Var(Y)。仔细检查符号。
Mistake 3: Assuming sums maintain the same distribution without verifying independence and parameter conditions. Binomial sums require equal p; Poisson sums require independence.
错误3：在未验证独立性与参数条件的情况下，就假设和保持相同分布。二项之和要求相等的p；泊松之和要求独立。
Exam Tip: Always define your random variables clearly at the start of a solution. State distributions, parameters, and the assumption of independence. This secures method marks.
考试技巧：解题开始时始终清晰定义随机变量，写明分布、参数及独立性假设。这可以确保拿到方法分数。
Another tip: When working with the CLT, apply a continuity correction if you approximate a discrete distribution (e.g., binomial or Poisson) with a normal distribution, unless the question instructs otherwise.
另一个技巧：在应用CLT时，如果用正态分布逼近离散分布（如二项或泊松），除非题目另有要求，应进行连续性校正。
Published by TutorHao | 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