📚 Confidence Intervals and Their Applications | 置信区间及其应用
Estimating a population parameter from sample data is at the heart of statistical inference. A confidence interval provides a range of plausible values for a parameter, together with a level of confidence. In A-Level Further Mathematics, you are expected to construct and interpret confidence intervals for population means, using both the normal distribution and the t-distribution.
从样本数据估计总体参数是统计推断的核心。置信区间给出了参数的一个合理取值区间，并附带一个置信水平。在 A-Level 进阶数学中，你需要掌握用正态分布和 t 分布构造并解读总体均值置信区间。
1. What is a Confidence Interval? | 什么是置信区间？
imate ± margin of error. The point estimate for the population mean μ is the sample mean x̄.
置信区间是对未知参数的区间估计，由样本数据计算，一般形式为点估计 ± 误差界限。总体均值 μ 的点估计是样本均值 x̄。
The margin of error reflects the uncertainty in the estimate. It depends on the variability of the data, the sample size, and the chosen confidence level. A wider interval gives more confidence but less precision, while a narrower interval is more precise but less certain.
误差界限反映了估计的不确定性，它取决于数据的离散程度、样本量以及所选置信水平。区间越宽，置信程度越高但精度越低；区间越窄，精度越高但确定性越弱。
2. The Sampling Distribution of the Sample Mean | 样本均值的抽样分布
If X ~ N(μ, σ²), then the sample mean from a random sample of size n is distributed as x̄ ~ N(μ, σ²/n). Standardizing gives
若 X ~ N(μ, σ²)，则来自容量为 n 的随机样本的样本均值满足 x̄ ~ N(μ, σ²/n)。标准化后得到
Z = (x̄ - μ)/(σ/√n) ~ N(0, 1)
This result is the foundation for constructing confidence intervals for μ when the population variance σ² is known. The standard deviation of x̄, written σ/√n, is called the standard error.
这一结论是 σ² 已知时构造 μ 的置信区间的基础。x̄ 的标准差 σ/√n 称为标准误差。
3. Interval for a Normal Mean with Known Variance | 已知方差时正态总体均值的区间
he critical value such that P(-z* < Z < z*) = C, where C is the confidence level. For a 95% level, z* = 1.96. Rearranging the inequality inside the probability statement gives the C% confidence interval for μ.
设 z* 为满足 P(-z* < Z < z*) = C 的临界值，其中 C 为置信水平。对于 95% 置信水平，z* = 1.96。对概率式中的不等式重新整理，即可得到 μ 的 C% 置信区间。
x̄ ± z* × σ/√n
For example, if n = 25, x̄ = 50, σ = 8 and C = 95%, then the interval is 50 ± 1.96 × 8/√25, which gives (46.864, 53.136).
例如，若 n = 25，x̄ = 50，σ = 8，C = 95%，则区间为 50 ± 1.96 × 8/√25，即 (46.864, 53.136)。
4. Using the Central Limit Theorem | 利用中心极限定理
When the population is not normal, the exact distribution of x̄ may be unknown. However, for a sufficiently large sample size, the Central Limit Theorem states that x̄ is approximately normally distributed regardless of the population shape.
当总体不服从正态分布时，x̄ 的精确分布可能未知。但根据中心极限定理，当样本量足够大时，无论总体形态如何，x̄ 近似服从正态分布。
e, n ≥ 30 is often used as a rough guide. If σ is known and n is large, the same formula x̄ ± z* × σ/√n can be applied. If σ is unknown, replace σ by s, the sample standard deviation.
实际中常用 n ≥ 30 作为粗略标准。若 σ 已知且 n 较大，仍可使用公式 x̄ ± z* × σ/√n。若 σ 未知，则以样本标准差 s 代替 σ。
5. Interval for a Normal Mean with Unknown Variance | 未知方差时正态总体均值的区间
When σ is unknown and the population is normal, we use the sample standard deviation s and the t distribution. The confidence interval becomes
当 σ 未知且总体为正态分布时，我们使用样本标准差 s 和 t 分布。置信区间变为
x̄ ± t* × s/√n
Here t* is the critical value from a t distribution with n − 1 degrees of freedom. The t distribution has heavier tails than the normal distribution, so t* is larger than z* for small n. As n increases, t* approaches z*.
其中 t* 是自由度为 n − 1 的 t 分布临界值。t 分布的尾部比正态分布更厚，因此在小样本时 t* 大于 z*。随着 n 增大，t* 逐渐接近 z*。
For example, with n = 16, x̄ = 20, s = 4 and a 95% e degrees of freedom are 15. From the t table, t* ≈ 2.131. The interval is 20 ± 2.131 × 4/√16, giving (17.869, 22.131).
例如，n = 16，x̄ = 20，s = 4，95% 置信水平，自由度为 15。查 t 表得 t* ≈ 2.131。区间为 20 ± 2.131 × 4/√16，即 (17.869, 22.131)。
6. Interpreting a Confidence Interval Correctly | 正确解读置信区间
A 95% confidence interval means that if we repeated the sampling process many times, about 95% of the intervals constructed in that way would contain the true population mean μ. The probability applies to the procedure, not to a single interval.
95% 置信区间的含义是：如果重复抽样多次，用同样方式构造的区间中约有 95% 会包含总体均值 μ。这个概率针对的是构造过程，而不是某一个具体区间。
Once the interval is observed, it either contains μ or it does not. It is therefore incorrect to say that the true mean has a 95% chance of lying inside a particular observed interval, such as (46.864, 53.136).
一旦区间已经观测到，它要么包含 μ，要么不包含 μ。因此，说“总体均值有 95% 概率落在某个已观测区间（如 (46.864, 53.136)）内”是错误的。
7. Margin of Error and Confidence Level | 误差界限与置信水平
The margin of error is E = z* × σ/√n for a known variance, or E = t* × s/√n when the variance is unknown. It measures half the width of the confidence interval.
方差已知时误差界限为 E = z* × σ/√n；方差未知时为 E = t* × s/√n。它等于置信区间宽度的一半。
Increasing the sample size n decreases E, because √n appears in the denominator. Increasing the confidence level, for example from 95% to 99%, increases z* or t* and therefore widens the interval. This trade-off between precision and confidence is fundamental.
增大样本量 n 会使 E 变小，因为 √n 位于分母。提高置信水平，例如从 95% 提高到 99%，会使 z* 或 t* 变大，从而加宽区间。这种精度与置信度之间的权衡是基本思想。
8. Determining the Sample Size | 确定样本量
specified margin of error E. For a known σ and a chosen confidence level, set E = z* × σ/√n and solve for n.
有时需要达到指定的误差界限 E。在 σ 已知并选定置信水平时，令 E = z* × σ/√n，解出 n。
n ≥ (z* × σ / E)²
Because n must be an integer, always round up to the nearest whole number. If σ is unknown, use a previous estimate of σ, such as an earlier sample standard deviation, or carry out a small pilot study.
由于 n 必须是整数，最终结果要向上取整。若 σ 未知，可使用 σ 的先前估计值，例如以往样本的标准差，或先进行一次小规模试点研究。
9. Applications and Worked Example | 应用与例题
is known to be normally distributed with σ = 15 g. A random sample of 100 bags has mean x̄ = 502 g. Find the 95% confidence interval for the true mean fill weight μ.
某工厂生产袋装大米。已知装袋重量服从正态分布，σ = 15 g。随机抽取 100 袋，样本均值 x̄ = 502 g。求总体均值 μ 的 95% 置信区间。
Since σ is known, use x̄ ± z* × σ/√n. Here z* = 1.96 and √n 1.96 × 15/10 = 2.94. The interval is (502 − 2.94, 502 + 2.94), giving (499.06, 504.94) g.
σ 已知，使用 x̄ ± z* × σ/√n。此处 z* = 1.96，√n = 10，所以误差界限为 1.96 × 15/10 = 2.94。区间为 (502 − 2.94, 502 + 2.94)，即 (499.06, 504.94) g。
othesis testing. If a claim states that μ = 500 g, we would not reject this claim at the 5% significance level, because 500 lies inside the 95% confidence interval.
该区间与假设检验自然关联。若声称 μ = 500 g，在 5% 显著性水平下我们不能拒绝这一说法，因为 500 位于 95% 置信区间之内。
10. Common Misconceptions and Exam Advice | 常见误解与考试建议
A common mistake is to interpret a 95% confidence interval as containing 95% of the sample data. A confidence interval is for the population parameter, not for individual observations. Another error is claiming that μ has a 95% probability of being in the interval; the correct interpretation is about long-run frequency.
常见错误之一是把 95% 置信区间理解为包含 95% 的样本数据。置信区间针对的是总体参数，而不是单个观测值。另一个错误是声称 μ 有 95% 概率位于该区间内；正确的解读是关于长期频率的。
In the exam, always state which distribution you are using. If σ is known, use z*; if σ is unknown and n is small, use t* with n − 1 degrees of freedom. Write the formula clearly, substitute values accurately, and quote the final interval with correct units.
考试中，务必说明你使用的是哪种分布。σ 已知时用 z*；σ 未知且 n 较小时用自由度为 n − 1 的 t*。清晰写出公式，准确代入数值，并用正确单位写出最终区间。
Finally, check whether the condition for using the normal distribution is satisfied. For a normal population, any n works. For non-normal populations, rely on the Central Limit Theorem and use a large enough sample size.
最后，检查是否满足使用正态分布的条件。若总体正态，任意 n 均可；若总体非正态，则依赖中心极限定理并保证样本量足够大。
Published by TutorHao | Further 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