📚 Inverse Normal Distribution Function: Principles and Usage | 逆正态分布函数的原理与使用
The normal distribution is one of the most fundamental probability distributions in statistics. Its cumulative distribution function (CDF) gives the probability that a random variable falls below a given value. The inverse normal distribution function, often denoted as Φ⁻¹(p), reverses this process: given a probability p, it returns the corresponding value on the normal curve. This function is essential for constructing confidence intervals, conducting hypothesis tests, and determining critical values in statistical analysis.
正态分布是统计学中最基本的概率分布之一。其累积分布函数（CDF）给出了随机变量小于某个数值的概率。逆正态分布函数通常记为 Φ⁻¹(p)，它逆转了这一过程：给定概率 p，返回正态曲线上对应的数值。这一函数在构造置信区间、进行假设检验以及确定统计分析的临界值中至关重要。
1. The Normal Distribution and Its Cumulative Function | 正态分布及其累积函数
A continuous random variable X follows a normal distribution with mean μ and standard deviation σ. Its probability density function is symmetric and bell-shaped. The cumulative distribution function, F(x), is the area under the density curve to the left of x, representing P(X ≤ x). For the standard normal distribution, where μ = 0 and σ = 1, the CDF is denoted by Φ(z).
连续随机变量 X 服从均值为 μ、标准差为 σ 的正态分布。其概率密度函数对称且呈钟形。累积分布函数 F(x) 是密度曲线下 x 左侧的面积，表示 P(X ≤ x)。对于标准正态分布，即 μ = 0、σ = 1 时，其累积分布函数记为 Φ(z)。
The CDF is a strictly increasing function, which means it has a unique inverse. If Φ(z) = p, then z = Φ⁻¹(p). This inverse function is also known as the quantile function or the percent point function.
累积分布函数是严格递增的，因此它具有唯一的反函数。若 Φ(z) = p，则 z = Φ⁻¹(p)。这一反函数也被称为分位数函数或百分点函数。
2. Definition of the Inverse Normal Function | 逆正态分布函数的定义
For a given probability p, where 0 < p < 1, the inverse standard normal function returns the z-value such that the area under the standard normal curve to the left of z equals p. In symbols:
对于给定的概率 p（0 < p < 1），逆标准正态函数返回这样的 z 值，使得标准正态曲线在 z 左侧的面积等于 p。用符号表示：
z = Φ⁻¹(p) ⇔ Φ(z) = p
This function is widely used in statistics. For example, the 95th percentile of the standard normal distribution is approximately 1.645, since the area to the left of 1.645 is 0.95.
这一函数在统计学中应用广泛。例如，标准正态分布的第 95 百分位数约为 1.645，因为 1.645 左侧的面积为 0.95。
3. Standard Normal Distribution and Z-Scores | 标准正态分布与 Z 分数
In practice, any normal distribution can be transformed to the standard normal distribution using the z-score formula:
在实际应用中，任何正态分布都可以通过 z 分数公式转化为标准正态分布：
z = (x – μ) / σ
Here, x is the original value, μ is the mean, and σ is the standard deviation. The z-score measures how many standard deviations x is from the mean. When using the inverse normal function, we often first find the z-value, then convert back to the original scale using:
其中 x 是原始值，μ 是均值，σ 是标准差。z 分数衡量 x 距离均值多少个标准差。使用逆正态函数时，我们通常先求出 z 值，再通过下式转换回原始尺度：
x = μ + z × σ
4. Computing the Inverse Normal Value | 计算逆正态值
There is no simple closed-form expression for Φ⁻¹(p), so numerical methods are required. Modern calculators and statistical software use iterative algorithms, such as the Beasley-Springer-Moro algorithm or the Acklam algorithm, to compute highly accurate approximations. For educational purposes, standard normal tables provide the forward mapping from z to Φ(z); to use them inversely, one locates the probability in the table and reads the corresponding z-value.
Φ⁻¹(p) 没有简单的封闭表达式，因此需要使用数值方法。现代计算器和统计软件采用迭代算法，如 Beasley-Springer-Moro 算法或 Acklam 算法，以计算高精度的近似值。在教学场景中，标准正态分布表提供了从 z 到 Φ(z) 的正向查询；若要逆用，则在表中找到概率，再读取对应的 z 值。
For example, to find z such that Φ(z) = 0.975, one looks for 0.9750 in the body of the table. The corresponding row and column give z ≈ 1.96. This value is famously used for 95% confidence intervals.
例如，要求 Φ(z) = 0.975 时的 z 值，在表体中找到 0.9750，其对应的行和列给出 z ≈ 1.96。这个值常用于 95% 置信区间。
5. Example: Finding Critical Values | 示例：求临界值
Suppose we need the critical value for a two-tailed test with significance level α = 0.05. The total probability in the tails is 0.05, so each tail contains 0.025. The critical z-values are Φ⁻¹(0.025) and Φ⁻¹(0.975).
假设我们需要显著性水平 α = 0.05 的双侧检验临界值。尾部总概率为 0.05，因此每个尾部包含 0.025。临界 z 值为 Φ⁻¹(0.025) 和 Φ⁻¹(0.975)。
From standard tables or a calculator, Φ⁻¹(0.025) ≈ -1.96 and Φ⁻¹(0.975) ≈ 1.96. Thus, if a test statistic falls outside the interval [-1.96, 1.96], we reject the null hypothesis at the 5% level.
通过标准表或计算器，Φ⁻¹(0.025) ≈ -1.96，Φ⁻¹(0.975) ≈ 1.96。因此，如果检验统计量落在区间 [-1.96, 1.96] 之外，我们在 5% 水平上拒绝原假设。
6. Using Statistical Tables vs. Calculators | 统计表与计算器的使用
Statistical tables were historically the primary tool for finding normal probabilities and quantiles. They are limited by finite precision and typically cover only the standard normal distribution. Calculators and software such as Python, R, or Excel provide direct functions like NORM.INV or qnorm, which are faster and more accurate for arbitrary probabilities.
统计表在历史上是查找正态概率和分位数的主要工具。它们受限于有限的精度，通常只覆盖标准正态分布。计算器和软件（如 Python、R 或 Excel）提供直接函数，如 NORM.INV 或 qnorm，对于任意概率更快且更准确。
For example, in Excel, =NORM.INV(0.95, 100, 15) directly returns the 95th percentile of a normal distribution with mean 100 and standard deviation 15, which is approximately 124.67. This illustrates how the inverse normal function can work directly on the original scale without converting to z-scores separately.
例如，在 Excel 中，=NORM.INV(0.95, 100, 15) 直接返回均值为 100、标准差为 15 的正态分布的第 95 百分位数，约为 124.67。这展示了逆正态函数如何直接在原始尺度上工作，而无需单独转换为 z 分数。
7. Confidence Intervals and the Inverse Normal | 置信区间与逆正态分布
One of the most frequent uses of the inverse normal function is constructing confidence intervals for a population mean. For a known population standard deviation σ, a 100(1-α)% confidence interval is given by:
逆正态函数最常见的用途之一是构造总体均值的置信区间。当总体标准差 σ 已知时，100(1-α)% 置信区间为：
x̄ ± z_{α/2} × (σ / √n)
Here, x̄ is the sample mean, n is the sample size, and z_{α/2} = Φ⁻¹(1 – α/2) is the critical value. For a 95% confidence level, α = 0.05, so z_{α/2} = 1.96. This critical value ensures that exactly 95% of the standard normal distribution lies between -1.96 and 1.96.
其中 x̄ 是样本均值，n 是样本量，z_{α/2} = Φ⁻¹(1 – α/2) 是临界值。对于 95% 置信水平，α = 0.05，因此 z_{α/2} = 1.96。这个临界值保证了标准正态分布恰好有 95% 落在 -1.96 与 1.96 之间。
8. Sample Size Determination | 样本量确定
The inverse normal function also plays a crucial role in planning studies. To estimate a population mean with a desired margin of error E and a confidence level 1 – α, the required sample size is:
逆正态函数在研究设计中也起着关键作用。为了以期望的误差范围 E 和置信水平 1 – α 估计总体均值，所需样本量为：
n = (z_{α/2} × σ / E)²
For example, suppose σ = 10 and we want a 95% confidence interval with a margin of error of 2. Here, z_{α/2} = 1.96, so n = (1.96 × 10 / 2)² ≈ 96.04, which we round up to 97. This calculation relies directly on the inverse normal quantile.
例如，假设 σ = 10，我们想要 95% 置信区间且误差范围为 2。此时 z_{α/2} = 1.96，因此 n = (1.96 × 10 / 2)² ≈ 96.04，取整为 97。这一计算直接依赖于逆正态分位数。
9. Common Pitfalls and Tips | 常见误区与提示
One common error is confusing one-tailed and two-tailed probabilities. When constructing a two-sided confidence interval, the probability associated with z_{α/2} is 1 – α/2, not 1 – α. Another pitfall is using the inverse normal function when the sample size is small and the population variance is unknown; in such cases, the t-distribution is more appropriate.
一个常见错误是混淆单尾与双尾概率。在构造双侧置信区间时，与 z_{α/2} 相关的概率是 1 – α/2，而不是 1 – α。另一个误区是当样本量较小且总体方差未知时使用逆正态函数；此时使用 t 分布更为合适。
Always check whether the input probability is in the left tail. If you need the value that leaves a probability q in the right tail, use Φ⁻¹(1 – q). Also, remember that the inverse normal function is symmetric: Φ⁻¹(p) = -Φ⁻¹(1 – p).
始终检查输入概率是否为左尾概率。如果需要的值在右尾留下概率 q，则应使用 Φ⁻¹(1 – q)。此外，记住逆正态函数具有对称性：Φ⁻¹(p) = -Φ⁻¹(1 – p)。
10. Applications Across Fields | 跨领域的应用
The inverse normal function appears in finance (Value at Risk), engineering (tolerance design), psychology (norm-referenced scoring), and quality control (control charts). For instance, in statistical process control, upper and lower control limits are often set at mean ± 3 standard deviations, corresponding to Φ⁻¹(0.99865) ≈ 2.998, which is approximated as 3.
逆正态函数出现在金融（风险价值）、工程（公差设计）、心理学（常模参照评分）以及质量控制（控制图）等领域。例如，在统计过程控制中，上下控制限通常设为均值 ± 3 个标准差，对应于 Φ⁻¹(0.99865) ≈ 2.998，通常近似为 3。
In educational testing, percentile ranks are computed by applying the inverse normal function to raw scores. If a student’s score is at the 90th percentile, the corresponding z-score is 1.2816, which can then be translated to the student’s score using the test mean and standard deviation.
在教育测试中，百分位排名通过对原始分数应用逆正态函数计算。如果学生的分数位于第 90 百分位，则对应的 z 分数为 1.2816，可根据考试的均值和标准差将其转换为学生的原始分数。
11. Working with Non-Standard Normal Distributions | 处理非标准正态分布
When the mean and standard deviation are not 0 and 1, the inverse normal function can be applied directly if the software supports parameters. For example, Python’s scipy.stats.norm.ppf(p, loc, scale) returns the quantile for a normal distribution with given location and scale. This saves the step of converting to z-scores and back.
当均值和标准差不是 0 和 1 时，如果软件支持参数，可以直接应用逆正态函数。例如，Python 的 scipy.stats.norm.ppf(p, loc, scale) 返回具有给定位置和尺度的正态分布的分位数。这就省去了先转换为 z 分数再转换回来的步骤。
Mathematically, the relationship between the general inverse normal and the standard inverse normal is:
数学上，一般逆正态分布与标准逆正态分布之间的关系为：
Q(p) = μ + σ × Φ⁻¹(p)
This formula is used widely in simulation and risk analysis. For example, to generate a random observation from a normal distribution, one can generate a uniform random probability p and then apply this transformation.
该公式广泛用于模拟和风险分析。例如，要生成一个来自正态分布的随机观察值，可以先生成一个均匀随机概率 p，然后应用这一变换。
12. Summary and Final Thoughts | 总结与要点回顾
The inverse normal distribution function is a powerful tool that connects probabilities to values. Understanding its principle helps you interpret statistical outputs, create confidence intervals, determine critical values, and plan experiments. The key idea is that for any probability p, Φ⁻¹(p) gives the point on the standard normal curve where the cumulative probability equals p.
逆正态分布函数是一个强大的工具，它将概率与数值联系起来。理解其原理有助于解读统计输出、创建置信区间、确定临界值以及规划实验。核心思想是：对于任意概率 p，Φ⁻¹(p) 给出标准正态曲线上累积概率等于 p 的那个点。
When using this function, always clarify whether you need a one-tailed or two-tailed probability, check the tails, and choose the appropriate distribution. With practice, the inverse normal function becomes as intuitive as the normal distribution itself.
使用该函数时，务必明确需要单尾还是双尾概率，检查尾部方向，并选择恰当的分布。通过练习，逆正态函数将变得和正态分布本身一样直观。
Published by TutorHao | Mathematics Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
屏轩国际教育cambridge primary/secondary checkpoint, cat4, ukiset,ukcat,igcse,alevel,PAT,STEP,MAT, ibdp,ap,ssat,sat,sat2课程辅导，国外大学本科硕士研究生博士课程论文辅导Cancel reply