Cumulative Distribution Function: Interpretation and Application | 累积分布函数：解读与应用

📚 Cumulative Distribution Function: Interpretation and Application | 累积分布函数：解读与应用

The cumulative distribution function (CDF) is one of the most powerful tools in statistics, bridging the gap between probability theory and real-world data analysis. In Edexcel A-Level Further Mathematics, mastering the CDF is essential for tackling questions on continuous and discrete random variables, hypothesis testing, and confidence intervals.

累积分布函数（CDF）是统计学中最强大的工具之一，它连接了概率理论与现实世界的数据分析。在 Edexcel A-Level 进阶数学中，掌握累积分布函数对于解决连续型和离散型随机变量、假设检验以及置信区间等问题至关重要。

1. Definition of the Cumulative Distribution Function | 累积分布函数的定义

For a random variable \(X\), the cumulative distribution function is defined as \(F(x) = P(X \leq x)\). It gives the probability that the random variable takes a value less than or equal to a specific number \(x\). This function accumulates all probabilities from the left tail of the distribution up to the point \(x\).

对于随机变量 \(X\)，累积分布函数定义为 \(F(x) = P(X \leq x)\)，它表示随机变量取值小于或等于某个特定数 \(x\) 的概率。该函数从分布的左尾开始累积所有概率，一直到点 \(x\) 为止。

F(x) = P(X ≤ x) = ∫₋∞ˣ f(t) dt (continuous)

F(x) = P(X ≤ x) = Σₜ₌₋∞ˣ P(X = t) (discrete)

2. Key Properties of the CDF | 累积分布函数的关键性质

Every cumulative distribution function must satisfy four essential properties. First, \(F(x)\) is non-decreasing, meaning as \(x\) increases, the cumulative probability never decreases. Second, the limits hold: \(F(-\infty) = 0\) and \(F(+\infty) = 1\). Third, \(F(x)\) is right-continuous for discrete distributions. Fourth, for any two values \(a < b\), we have \(P(a < X \leq b) = F(b) - F(a)\).

每一个累积分布函数都必须满足四个基本性质。第一，\(F(x)\) 是单调不减的，也就是说随着 \(x\) 增大，累积概率绝不会减小。第二，极限条件成立：\(F(-\infty) = 0\) 且 \(F(+\infty) = 1\)。第三，对于离散分布，\(F(x)\) 是右连续的。第四，对于任意两个值 \(a < b\)，有 \(P(a < X \leq b) = F(b) - F(a)\)。

Non-decreasing: If \(x₁ < x₂\), then \(F(x₁) \leq F(x₂)\).
Monotonicity: 若 \(x₁ < x₂\)，则 \(F(x₁) \leq F(x₂)\)，概率不会倒退。
Bounded range: \(0 \leq F(x) \leq 1\) for all real \(x\).
有界范围： 对所有实数 \(x\)，均有 \(0 \leq F(x) \leq 1\)。

3. CDF for Discrete Random Variables | 离散型随机变量的累积分布函数

For a discrete random variable such as the Binomial or Poisson distribution, the CDF is a step function. Each step occurs at a possible value of the random variable, and the height of the step equals the probability mass at that point. For example, if \(X \sim \text{Binomial}(n, p)\), the CDF is \(F(x) = \sum_{k=0}^{x} \binom{n}{k} p^k (1-p)^{n-k}\).

对于离散型随机变量，如二项分布或泊松分布，累积分布函数是一个阶梯函数。每一步出现在随机变量的一个可能取值处，步长的高度等于该点的概率质量。例如，若 \(X \sim \text{Binomial}(n, p)\)，则累积分布函数为 \(F(x) = \sum_{k=0}^{x} \binom{n}{k} p^k (1-p)^{n-k}\)。

F(x) = Σₖ₌₀ˣ C(n,k) pᵏ(1-p)ⁿ⁻ᵏ

When using statistical tables for discrete distributions, the tabulated values are precisely these cumulative probabilities. This is why the binomial table in your formula booklet lists \(P(X \leq x)\) rather than individual probabilities.

在使用离散分布的统计表时，表中列出的数值正是这些累积概率。这就是为什么公式手册中的二项分布表列出 \(P(X \leq x)\) 而不是单个概率值的原因。

4. CDF for Continuous Random Variables | 连续型随机变量的累积分布函数

For a continuous random variable, the CDF is obtained by integrating the probability density function (PDF). If \(X\) has PDF \(f(x)\), then \(F(x) = \int_{-\infty}^{x} f(t) \, dt\). The CDF is a smooth, continuous function, and by the Fundamental Theorem of Calculus, the PDF is the derivative of the CDF: \(f(x) = F'(x)\).

对于连续型随机变量，累积分布函数通过对概率密度函数（PDF）积分得到。若 \(X\) 的概率密度函数为 \(f(x)\)，则 \(F(x) = \int_{-\infty}^{x} f(t) \, dt\)。累积分布函数是光滑的连续函数，根据微积分基本定理，概率密度函数是累积分布函数的导数：\(f(x) = F'(x)\)。

f(x) = d/dx F(x) ⇔ F(x) = ∫ f(x) dx

This reciprocal relationship is fundamental: knowing one function allows you to derive the other. In the Edexcel syllabus, you are often given a piecewise PDF and asked to find the CDF by integration, or given a CDF and asked to differentiate to obtain the PDF.

这种互逆关系是基础性的：知道其中一个函数就可以推导出另一个。在 Edexcel 考纲中，你常会被给出一段分段概率密度函数，要求通过积分求累积分布函数；或者给出累积分布函数，要求通过微分求得概率密度函数。

5. Calculating Probabilities Using the CDF | 使用累积分布函数计算概率

The CDF simplifies probability calculations enormously. For any values \(a < b\), the probability that \(X\) lies in the interval \((a, b]\) is simply \(F(b) - F(a)\). This avoids multiple integrations or complex summations, especially when dealing with piecewise-defined distributions.

累积分布函数极大地简化了概率计算。对于任意 \(a < b\)，随机变量 \(X\) 落在区间 \((a, b]\) 内的概率就是 \(F(b) - F(a)\)。这避免了多次积分或复杂的求和，尤其在处理分段定义的分布时尤为便利。

Interval probability: \(P(a < X \leq b) = F(b) - F(a)\)
区间概率： \(P(a < X \leq b) = F(b) - F(a)\)，注意开闭区间在连续情形下无差异。
Right-tail probability: \(P(X \geq a) = 1 – F(a)\)
右尾概率： \(P(X \geq a) = 1 – F(a)\)，这在假设检验中常用。
Point probability (discrete): \(P(X = a) = F(a) – F(a^-)\)
单点概率（离散）： \(P(X = a) = F(a) – F(a⁻)\)，即左右极限之差。

P(a < X ≤ b) = F(b) − F(a)

6. Finding the Median and Quartiles | 求中位数和四分位数

The CDF provides a direct route to quantiles. The median \(m\) is the value such that \(F(m) = 0.5\), meaning half of the distribution lies below \(m\). The lower quartile \(Q₁\) satisfies \(F(Q₁) = 0.25\), and the upper quartile \(Q₃\) satisfies \(F(Q₃) = 0.75\). These quantiles are found by solving the equation \(F(x) = p\) for \(x\).

累积分布函数为求分位数提供了直接途径。中位数 \(m\) 是满足 \(F(m) = 0.5\) 的值，意味着分布的一半位于 \(m\) 之下。下四分位数 \(Q₁\) 满足 \(F(Q₁) = 0.25\)，上四分位数 \(Q₃\) 满足 \(F(Q₃) = 0.75\)。这些分位数通过解方程 \(F(x) = p\) 来求得。

Median: F(m) = 0.5 | Q₁: F(Q₁) = 0.25 | Q₃: F(Q₃) = 0.75

For continuous distributions, this often involves inverting the CDF function. For example, if \(X\) has PDF \(f(x) = 2x\) on \([0, 1]\), then \(F(x) = x²\). Setting \(F(m) = 0.5\) gives \(m² = 0.5\), hence \(m = \sqrt{0.5} \approx 0.707\).

对于连续分布，这通常涉及对累积分布函数求逆。例如，若 \(X\) 在 \([0, 1]\) 上的概率密度函数为 \(f(x) = 2x\)，则 \(F(x) = x²\)。令 \(F(m) = 0.5\)，得 \(m² = 0.5\)，因此 \(m = \sqrt{0.5} \approx 0.707\)。

7. Expectation Using the CDF | 利用累积分布函数求期望

Although the standard formula for expectation uses the PDF directly, there is an elegant alternative: for a non-negative continuous random variable, \(E[X] = \int_0^\infty (1 – F(x)) \, dx\). This formula is particularly useful when the CDF is simpler to integrate than the PDF, or when the PDF is not explicitly known.

尽管期望的标准公式直接使用概率密度函数，但有一个优雅的替代方法：对于非负连续随机变量，\(E[X] = \int_0^\infty (1 – F(x)) \, dx\)。这个公式在累积分布函数比概率密度函数更易积分时，或者当概率密度函数未知时特别有用。

E[X] = ∫₀^∞ [1 − F(x)] dx (for X ≥ 0)

Similarly, for discrete random variables, the expectation can be computed as \(E[X] = \sum_{x} (1 – F(x))\) over non-negative integers. This alternative formulation appears in more advanced questions and can simplify calculations significantly.

类似地，对于离散随机变量，期望可以通过 \(E[X] = \sum_{x} (1 – F(x))\) 在非负整数上求和来计算。这种替代公式出现在更高级的题目中，可以显著简化计算。

8. The Standard Normal CDF and Φ(z) | 标准正态累积分布函数与 Φ(z)

The most commonly encountered CDF in A-Level Further Mathematics is the standard normal cumulative distribution function, denoted \(\Phi(z)\). This is the CDF of the standard normal distribution \(Z \sim N(0, 1)\). Unfortunately, \(\Phi(z)\) has no closed-form expression in elementary functions, which is why statistical tables are essential. The values in the normal distribution table in your formula booklet are precisely \(\Phi(z)\) values for \(z \geq 0\).

在 A-Level 进阶数学中最常遇到的累积分布函数是标准正态累积分布函数，记作 \(\Phi(z)\)。它是标准正态分布 \(Z \sim N(0, 1)\) 的累积分布函数。遗憾的是，\(\Phi(z)\) 没有初等函数的闭式表达式，因此统计表至关重要。公式手册中正态分布表中的值正是 \(z \geq 0\) 时 \(\Phi(z)\) 的值。

Φ(z) = P(Z ≤ z) = ∫₋∞ᶻ (1/√(2π)) e^(−t²/2) dt

Symmetry property: \(\Phi(-z) = 1 – \Phi(z)\) for all \(z\).
对称性： 对任意 \(z\)，有 \(\Phi(-z) = 1 – \Phi(z)\)。这是使用正态表的关键。
Standardisation: If \(X \sim N(\mu, \sigma²)\), then \(Z = (X – \mu)/\sigma \sim N(0, 1)\).
标准化： 若 \(X \sim N(\mu, \sigma²)\)，则 \(Z = (X – \mu)/\sigma \sim N(0, 1)\)。所有正态概率都通过标准化转换求解。

9. Inverse Cumulative Distribution Function | 逆累积分布函数

The inverse CDF, also called the quantile function, answers the question “what value \(x\) corresponds to a given cumulative probability \(p\)?” In Edexcel questions, you will encounter this when finding critical values for hypothesis tests or confidence intervals. For the standard normal distribution, this is denoted \(\Phi^{-1}(p)\) or sometimes \(z_p\).

逆累积分布函数，也称分位数函数，回答的问题是”哪个值 \(x\) 对应给定的累积概率 \(p\)？”在 Edexcel 的题目中，你会在求假设检验的临界值或置信区间时遇到它。对于标准正态分布，这记作 \(\Phi^{-1}(p)\) 或有时记作 \(z_p\)。

For example, the critical value for a two-tailed 5% significance test is \(\Phi^{-1}(0.975) = 1.96\), since \(P(Z \leq 1.96) = 0.975\). Similarly, \(\Phi^{-1}(0.95) = 1.645\) for a one-tailed test at the 5% level. In inverse normal problems, you may be asked to find a value \(a\) such that \(P(X \leq a) = p\), which requires solving \(F(a) = p\) where \(F\) is the CDF of \(X\).

例如，双尾 5% 显著性检验的临界值是 \(\Phi^{-1}(0.975) = 1.96\)，因为 \(P(Z \leq 1.96) = 0.975\)。类似地，\(\Phi^{-1}(0.95) = 1.645\) 对应单尾 5% 水平检验。在逆正态问题中，你可能会被要求找到满足 \(P(X \leq a) = p\) 的值 \(a\)，这需要解方程 \(F(a) = p\)，其中 \(F\) 是 \(X\) 的累积分布函数。

10. Working with Piecewise CDFs | 处理分段累积分布函数

Many examination questions present a CDF defined piecewise over different intervals. The key skills are: verifying it is a valid CDF (checking that it is non-decreasing and tends to 1), extracting probabilistic statements, and finding the PDF by differentiation. Care must be taken at the boundary points where the pieces meet.

许多考试题目给出在不同区间上分段定义的累积分布函数。关键技能包括：验证它是有效的累积分布函数（检查单调不减且趋于 1）、提取概率陈述、以及通过微分求概率密度函数。在分段交界点处需要格外小心。

Consider the CDF defined by:

考虑如下分段累积分布函数：

F(x) = 0 for x < 0; F(x) = x²/4 for 0 ≤ x < 1; F(x) = x/4 for 1 ≤ x < 2; F(x) = 1 for x ≥ 2

To find the PDF, differentiate each piece: \(f(x) = x/2\) on \([0, 1)\), \(f(x) = 1/4\) on \([1, 2)\), and \(f(x) = 0\) elsewhere. Always verify that \(F\) is continuous at the boundaries; in this example, \(F(1) = 1/4\) from both pieces, confirming continuity.

要求概率密度函数，只需对每一段求导：在 \([0, 1)\) 上 \(f(x) = x/2\)，在 \([1, 2)\) 上 \(f(x) = 1/4\)，其余区间为 0。务必验证 \(F\) 在边界点是否连续；在本例中，两段在 \(F(1) = 1/4\) 处一致，确认连续性成立。

11. Common Examination Pitfalls | 考试中常见的陷阱

Students often make predictable errors when working with CDFs. The first is confusing the CDF with the PDF: remembering that \(F(x)\) accumulates probability up to \(x\), while \(f(x)\) gives the density at a single point. The second error is misapplying the formula \(P(a < X \leq b) = F(b) - F(a)\) for discrete distributions, where endpoints matter. The third is forgetting that \(F(x)\) can never be negative or exceed 1.

学生在处理累积分布函数时常犯一些可预测的错误。第一个错误是混淆累积分布函数和概率密度函数：记住 \(F(x)\) 累积到 \(x\) 为止的概率，而 \(f(x)\) 给出单个点的密度。第二个错误是在离散分布中误用公式 \(P(a < X \leq b) = F(b) - F(a)\)，此时端点取值至关重要。第三个错误是忘记 \(F(x)\) 永远不能为负或超过 1。

Check the limits: Always verify \(F(-\infty) = 0\) and \(F(\infty) = 1\).
检查极限： 始终验证 \(F(-\infty) = 0\) 和 \(F(\infty) = 1\)。
Watch discrete endpoints: For \(P(X < a)\) use \(F(a⁻)\), not \(F(a)\).
注意离散端点： 求 \(P(X < a)\) 时用 \(F(a⁻)\)，而不是 \(F(a)\)。
Verify continuity: For continuous distributions, \(P(X = a) = 0\), so \(F(a⁻) = F(a)\).
验证连续性： 对于连续分布，\(P(X = a) = 0\)，所以 \(F(a⁻) = F(a)\)。

12. Worked Example: From PDF to CDF to Probability | 实例演练：从概率密度函数到累积分布函数再到概率

Let us work through a complete example. Suppose a continuous random variable \(X\) has PDF \(f(x) = 3x²\) for \(0 \leq x \leq 1\), and \(f(x) = 0\) otherwise. First, verify this is a valid PDF: \(\int_0^1 3x² \, dx = [x³]_0^1 = 1\). Now find the CDF: \(F(x) = \int_0^x 3t² \, dt = x³\) for \(0 \leq x \leq 1\).

让我们完整地解一道例题。假设连续随机变量 \(X\) 在 \(0 \leq x \leq 1\) 上的概率密度函数为 \(f(x) = 3x²\)，否则为 0。首先验证它是有效的概率密度函数：\(\int_0^1 3x² \, dx = [x³]_0^1 = 1\)。现在求累积分布函数：\(F(x) = \int_0^x 3t² \, dt = x³\)，其中 \(0 \leq x \leq 1\)。

F(x) = 0 (x < 0); F(x) = x³ (0 ≤ x ≤ 1); F(x) = 1 (x > 1)

Now suppose we want \(P(0.2 < X < 0.6)\). Using the CDF: \(F(0.6) - F(0.2) = 0.6³ - 0.2³ = 0.216 - 0.008 = 0.208\). We can also find the median: set \(F(m) = 0.5\), giving \(m³ = 0.5\), so \(m = \sqrt[3]{0.5} \approx 0.794\). The interquartile range is \(Q₃ - Q₁ = \sqrt[3]{0.75} - \sqrt[3]{0.25} \approx 0.909 - 0.630 = 0.279\).

现在假设我们要求 \(P(0.2 < X < 0.6)\)。利用累积分布函数：\(F(0.6) - F(0.2) = 0.6³ - 0.2³ = 0.216 - 0.008 = 0.208\)。我们还可以求中位数：令 \(F(m) = 0.5\)，得 \(m³ = 0.5\)，所以 \(m = \sqrt[3]{0.5} \approx 0.794\)。四分位距为 \(Q₃ - Q₁ = \sqrt[3]{0.75} - \sqrt[3]{0.25} \approx 0.909 - 0.630 = 0.279\)。

This example demonstrates how the CDF consolidates all probability calculations into simple arithmetic once it has been determined. It is the single most efficient tool in your statistical arsenal.

这个例子展示了累积分布函数一旦确定，就能将所有概率计算简化为简单算术。它是你统计工具库中最高效的单一工具。

In conclusion, the cumulative distribution function is not merely a theoretical construct but a practical computational device. It unifies discrete and continuous probability, enables efficient probability extraction, and underpins quantile-based inference. For Edexcel A-Level Further Mathematics, proficiency in interpreting, constructing, and applying the CDF is non-negotiable. Regular practice with piecewise functions, standard normal tables, and inverse problems will build the fluency required for top marks.

总而言之，累积分布函数不仅仅是一个理论概念，更是一个实用的计算工具。它统一了离散与连续概率，实现了高效的概率提取，并支撑了基于分位数的推断。对于 Edexcel A-Level 进阶数学，熟练地解读、构造和应用累积分布函数是必不可少的。通过定期练习分段函数、标准正态分布表和逆问题，你将建立起获得高分所需的熟练度。

Published by TutorHao | Further Mathematics Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

Cumulative Distribution Function: Interpretation and Application | 累积分布函数：解读与应用