📚 Confidence Interval for the Mean with Known Variance (From Data) | 已知方差时均值的置信区间（基于数据）
In inferential statistics, a confidence interval provides a range of plausible values for a population parameter, such as the population mean μ. When we have sample data and the population variance σ² is known, we can construct a confidence interval for μ using the standard normal (Z) distribution. This is one of the most classical and fundamental techniques in statistics, often appearing early in IB Mathematics courses. Understanding how to calculate and interpret such intervals is crucial for data-based decision-making.
在推断统计学中，置信区间为总体参数（例如总体均值 μ）提供了一个合理的取值范围。当我们拥有样本数据且总体方差 σ² 已知时，可以利用标准正态（Z）分布构建总体均值 μ 的置信区间。这是统计学中最经典且基础的方法之一，通常在 IB 数学课程早期就会出现。理解如何计算并解释这类区间对于基于数据的决策至关重要。
1. What is a Confidence Interval? | 什么是置信区间？
A confidence interval (CI) is an interval estimate for a population parameter computed from sample data. Unlike a single point estimate like the sample mean x̄, a confidence interval accounts for sampling variability and gives a range within which the true parameter is likely to lie. The confidence level, typically 90%, 95%, or 99%, indicates how confident we are that the interval contains the parameter in repeated sampling.
置信区间（CI）是基于样本数据计算出的总体参数的区间估计。与样本均值 x̄ 这样的单个点估计不同，置信区间考虑了抽样变异性，并给出了真实参数可能落入的范围。置信水平通常为 90%、95% 或 99%，它表示在重复抽样中，我们有多次的把握认为该区间包含了参数真值。
For example, a 95% confidence interval for μ means that if we were to take many random samples and construct a confidence interval from each one, approximately 95% of those intervals would capture the true population mean μ. It does not mean that there is a 95% probability that μ lies within a single computed interval; the parameter is fixed, and the interval is random.
例如，一个 95% 置信区间对 μ 的解释是：如果我们重复抽取许多随机样本，并为每个样本构造一个置信区间，那么大约有 95% 的区间会捕获真实的总体均值 μ。这并不意味着 μ 落入某个具体计算出的区间的概率为 95%；参数是固定的，区间是随机的。
2. When Variance is Known: The Z-interval | 方差已知时：Z 区间
When the population variance σ² (or equivalently the standard deviation σ) is known, the sampling distribution of the sample mean x̄ follows a normal distribution with mean μ and standard deviation σ/√n, provided the population is normally distributed or the sample size is large enough (Central Limit Theorem). In this scenario, we standardize x̄ using Z = (x̄ – μ) / (σ/√n), which follows a standard normal distribution.
当总体方差 σ²（或等价地，标准差 σ）已知时，样本均值 x̄ 的抽样分布服从均值为 μ、标准差为 σ/√n 的正态分布，前提是总体服从正态分布或样本量足够大（中心极限定理）。在这种情况下，我们使用 Z = (x̄ – μ) / (σ/√n) 进行标准化，它服从标准正态分布。
This allows us to use Z critical values to build the confidence interval. We refer to this method as the Z-interval or one-sample Z-interval for the mean. It is distinct from the T-interval used when the population variance is unknown and must be estimated from the sample.
这使得我们可以使用 Z 临界值来构建置信区间。我们将这一方法称为均值 Z 区间或单样本 Z 区间。它区别于总体方差未知且必须由样本估计时使用的 T 区间。
3. The Formula and Its Components | 公式及其组成部分
The confidence interval for the population mean μ when σ is known is given by:
当 σ 已知时，总体均值 μ 的置信区间公式如下：
x̄ ± z* × (σ / √n)
where x̄ is the sample mean, z* is the critical value from the standard normal distribution corresponding to the desired confidence level, σ is the known population standard deviation, and n is the sample size. The quantity σ/√n is called the standard error of the mean, and z* × (σ/√n) is the margin of error.
其中 x̄ 是样本均值，z* 是对应于所需置信水平的标准正态分布临界值，σ 是已知的总体标准差，n 是样本量。量 σ/√n 称为均值的标准误，而 z* × (σ/√n) 是误差幅度。
The interval is symmetric around x̄ and extends a margin of error in both directions. A larger sample size, smaller σ, or lower confidence level reduces the margin of error, giving a narrower interval.
该区间以 x̄ 为中心对称，向两边延伸出一个误差幅度。较大的样本量、较小的 σ 或较低的置信水平会减小误差幅度，从而使区间变窄。
4. Finding the Critical Z-value | 查找临界 Z 值
The critical value z* is the number such that the area under the standard normal curve between -z* and z* equals the confidence level. For a given confidence level C, the significance level α = 1 – C, and z* is the upper (1 – α/2) quantile of the Z distribution, often denoted Zₐ/₂ or Z*. Common values can be memorized or found in statistical tables.
临界值 z* 是这样一个数：标准正态曲线下 -z* 与 z* 之间的面积等于置信水平。对于给定的置信水平 C，显著性水平 α = 1 – C，z* 是 Z 分布的上 (1 – α/2) 分位数，常记为 Zₐ/₂ 或 Z*。常见值可以记住或从统计表中查找。
Below is a quick reference for typical confidence levels:
下面是典型置信水平的快速参考：
| Confidence Level (C) | α | α/2 | Critical value z* |
|---|---|---|---|
| 90% | 0.10 | 0.05 | 1.645 |
| 95% | 0.05 | 0.025 | 1.960 |
| 99% | 0.01 | 0.005 | 2.576 |
For a two-sided interval, we always split α equally into the two tails. Using a Z-table, we find the z-score that leaves an area of α/2 in the right tail. For 95% confidence, α/2 = 0.025, and the z-score with a cumulative probability of 0.975 is 1.96.
对于双侧区间，我们总是将 α 平分为两个尾部。使用 Z 表，我们找出使右尾面积为 α/2 的 z 分数。对于 95% 置信水平，α/2 = 0.025，累积概率为 0.975 的 z 分数是 1.96。
5. Confidence Level and α | 置信水平与 α
The complementary probability α = 1 – C is the probability that the confidence interval does not contain the true population mean in repeated sampling. Choosing a higher confidence level increases z*, thereby widening the interval. Conversely, a lower confidence level narrows the interval but reduces our certainty.
互补概率 α = 1 – C 是指在重复抽样中置信区间不包含真实总体均值的概率。选择更高的置信水平会增大 z*，从而使区间变宽。相反，较低的置信水平会使区间变窄，但会降低我们的把握程度。
In practice, 95% is the most commonly used confidence level, striking a balance between precision and reliability. However, fields like medicine or engineering might demand 99% confidence for critical decisions, accepting a wider interval for greater assurance.
在实践中，95% 是最常用的置信水平，在精度和可靠性之间取得了平衡。然而，医学或工程等领域可能要求 99% 的置信水平来做出关键决策，接受更宽的区间以获得更大的保证。
6. From Data to Interval: A Worked Example | 从数据到区间：一个计算实例
Suppose a manufacturer claims that the lifetime of a light bulb is normally distributed with a known standard deviation of σ = 100 hours. A random sample of n = 36 bulbs is tested, and the sample mean lifetime is x̄ = 1250 hours. Construct a 95% confidence interval for the true mean lifetime μ.
假设一家制造商声称某种灯泡的寿命服从正态分布，已知标准差为 σ = 100 小时。随机抽取 n = 36 个灯泡进行测试，样本平均寿命为 x̄ = 1250 小时。构建真实平均寿命 μ 的 95% 置信区间。
We have all required components: x̄ = 1250, σ = 100, n = 36, so √n = 6. The standard error is σ/√n = 100/6 ≈ 16.667. For 95% confidence, z* = 1.96. The margin of error is 1.96 × 16.667 ≈ 32.667.
我们具备所有必要的组成部分：x̄ = 1250，σ = 100，n = 36，因此 √n = 6。标准误为 σ/√n = 100/6 ≈ 16.667。对于 95% 置信水平，z* = 1.96。误差幅度为 1.96 × 16.667 ≈ 32.667。
Confidence Interval = 1250 ± 32.67 → (1217.33, 1282.67)
Interpretation: We are 95% confident that the true mean lifetime of all such bulbs lies between 1217.33 hours and 1282.67 hours. Note that this does not mean 95% of individual bulbs fall within this range; it refers to the mean.
解释：我们有 95% 的把握认为所有这类灯泡的真实平均寿命介于 1217.33 小时和 1282.67 小时之间。请注意，这并不意味着 95% 的个别灯泡落在此范围内；它针对的是均值。
7. Interpreting the Confidence Interval | 解释置信区间
Correct interpretation is essential. The confidence level refers to the long-run proportion of similarly constructed intervals that would capture μ. It is incorrect to say “the probability that μ is in this interval is 95%”. Once the interval is computed, μ is either inside it or not; there is no probability associated with a realized interval.
正确的解释至关重要。置信水平指的是在重复构造类似区间的情况下，捕捉到 μ 的长期比例。说”μ 在这个区间内的概率是 95%”是不正确的。一旦计算出区间，μ 要么在里面，要么不在；实现的区间没有概率可言。
A more accurate phrasing: “If we were to repeat this sampling procedure many times, about 95% of the resulting confidence intervals would contain the true population mean μ.” This subtle distinction is frequently tested in IB examinations.
更准确的表述是：”如果我们重复这个抽样过程许多次，大约 95% 的置信区间结果会包含真实的总体均值 μ。”这种微妙的区别在 IB 考试中频繁考查。
8. Effect of Sample Size on Width | 样本量对区间宽度的影响
The width of the confidence interval is 2 × z* × (σ/√n). As the sample size n increases, the standard error σ/√n decreases, so the interval becomes narrower. This reflects the increased precision from using more data. The relationship is inversely proportional to the square root of n; thus, to halve the margin of error, the sample size must be quadrupled.
置信区间的宽度为 2 × z* × (σ/√n)。随着样本量 n 增加，标准误 σ/√n 减小，因此区间变窄。这反映了使用更多数据带来的精度提升。这种关系与 n 的平方根成反比；因此，要将误差幅度减半，样本量必须增加四倍。
For a fixed confidence level and known σ, the sample size required to achieve a desired margin of error E can be found by solving E = z* × (σ/√n) for n: n = (z* × σ / E)². This is a useful planning tool before data collection.
对于固定的置信水平和已知的 σ，达到期望误差幅度 E 所需的样本量可通过求解 E = z* × (σ/√n) 中的 n 得到：n = (z* × σ / E)²。这是数据收集前一个很有用的规划工具。
9. Assumptions and Conditions | 假设与条件
For the Z-interval to be valid, certain conditions must be satisfied:
要使 Z 区间有效，必须满足某些条件：
- Known population variance: σ must be known a priori, which is rare in practice but common in textbook problems.
- 已知总体方差：σ 必须是先验已知的，这在实践中很少见，但在教科书问题中很常见。
- Independence: Observations in the sample should be independent (random sampling, and sample size less than 10% of population if sampling without replacement).
- 独立性：样本中的观测值应是独立的（随机抽样，若为不放回抽样，样本量应小于总体的 10%）。
- Normality: The population distribution should be normal, or the sample size should be sufficiently large (n ≥ 30 is a common guideline) so that the Central Limit Theorem ensures approximate normality of x̄.
- 正态性：总体分布应为正态，或样本量足够大（通常指南为 n ≥ 30），使得中心极限定理能保证 x̄ 的近似正态性。
If the normality condition is violated with a small sample, the Z-interval may not be reliable, and non-parametric methods might be considered.
如果样本量小且违反正态性条件，Z 区间可能不可靠，应考虑非参数方法。
10. Common Misconceptions | 常见误解
Several pitfalls often trip up students:
有几个陷阱常使学生困惑：
- Confusing confidence level with probability of containing μ: As explained, the confidence level describes the method’s long-run accuracy, not a probability for a single interval.
- 将置信水平与包含 μ 的概率混淆：如前所述，置信水平描述的是此方法的长期准确性，而不是单个区间的概率。
- Using z* when σ is unknown: In real data situations, σ is rarely known. When it must be estimated by the sample standard deviation s, the appropriate distribution is the t-distribution, not Z. Using the z-interval in that case results in an interval that is too narrow and overly optimistic.
- 当 σ 未知时使用 z*：在真实数据情境中，σ 很少已知。当必须用样本标准差 s 估计时，正确的分布是 t 分布，而非 Z。此时若用 z 区间，会导致区间过窄而过于乐观。
- Misinterpreting the interval as a prediction interval for individual values: The interval is for the population mean, not for individual data points. A prediction interval for a single observation would be much wider.
- 将区间误解为对个体值的预测区间：此区间针对的是总体均值，而非个体数据点。单个观测值的预测区间会宽得多。
Awareness of these traps is vital for both problem-solving and conceptual understanding.
注意这些陷阱对解题和概念理解都至关重要。
11. Summary and Key Takeaways | 总结与重点
The confidence interval for a population mean with known variance is a foundational tool in statistical inference. Its formula x̄ ± z* × (σ/√n) directly links sample statistics, critical values, and precision. Mastery of this topic requires not only plugging numbers into the formula but also a deep understanding of what the interval represents, the assumptions behind it, and the correct interpretation of confidence.
已知方差情况下总体均值的置信区间是统计推断的基础工具。其公式 x̄ ± z* × (σ/√n) 将样本统计量、临界值和精度直接联系起来。掌握这一主题不仅需要将数字代入公式，还需要深刻理解区间所代表的含义、其背后的假设以及对置信度的正确解释。
Remember: when σ is known, use Z; check normality or large n; interpret the confidence level as a long-run capture rate; and never attach probability to a specific interval after it is calculated. These principles will serve you well in IB Mathematics and beyond.
记住：当 σ 已知时，使用 Z；检查正态性或大样本量；将置信水平解释为长期捕获率；切勿在计算出具体区间后附加概率。这些原则将在 IB 数学及更高层次的学习中为你提供帮助。
Published by TutorHao | IB 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