📚 A-Level Further Mathematics: Student’s t-Distribution Percentile Points | A-Level进阶数学：学生t分布百分位点
The Student’s t-distribution is one of the most important continuous probability distributions in A-Level Further Mathematics. It arises naturally when estimating the mean of a normally distributed population when the sample size is small and the population variance is unknown. In this article, we focus on the percentile points of the t-distribution — often denoted as tₙ(α) or t(α; ν) — which are essential for hypothesis testing and confidence interval construction.
学生t分布是A-Level进阶数学中最重要的连续概率分布之一。当样本量较小且总体方差未知时，估计正态总体均值自然会产生t分布。本文聚焦t分布的百分位点，通常记为tₙ(α)或t(α; ν)，它们在假设检验和置信区间构造中至关重要。
1. Definition and Origin of the t-Distribution | t分布的定义与起源
The t-distribution was developed by William Sealy Gosset, who published under the pseudonym “Student.” It is defined as the ratio of a standard normal random variable Z to the square root of an independent chi-squared random variable divided by its degrees of freedom. If Z ~ N(0, 1) and U ~ χ²(ν) independently, then the random variable
t分布由威廉·西利·戈塞特（笔名”Student”）提出。它定义为标准正态随机变量Z与独立卡方随机变量除以其自由度后开方的比值。若Z ~ N(0, 1)且U ~ χ²(ν)相互独立，则随机变量
T = Z / √(U/ν)
follows a t-distribution with ν degrees of freedom, denoted by t(ν). The probability density function is symmetric about zero and has heavier tails than the standard normal distribution.
服从自由度为ν的t分布，记为t(ν)。其概率密度函数关于零点对称，且尾部比标准正态分布更厚。
2. Degrees of Freedom | 自由度
The parameter ν (Greek letter nu) represents the degrees of freedom. In the context of a single-sample t-test, ν = n − 1, where n is the sample size. The degrees of freedom control the shape of the distribution: smaller ν gives a flatter, more spread-out distribution with thicker tails, while larger ν makes the t-distribution approach the standard normal distribution.
参数ν（希腊字母nu）表示自由度。在单样本t检验中，ν = n − 1，其中n是样本量。自由度决定分布形状：较小的ν使分布更平坦、更分散、尾部更厚；较大的ν使t分布趋近于标准正态分布。
ν → ∞ ⇒ t(ν) → N(0, 1)
This limiting behaviour is a key principle: when the sample size is large, the t-distribution and the standard normal distribution become nearly indistinguishable.
这一极限性质是核心原理：当样本量很大时，t分布与标准正态分布几乎无法区分。
3. Percentile Points: Notation and Meaning | 百分位点：记号与含义
For a given probability α and degrees of freedom ν, the percentile point t_ν(α) is defined as the value such that the area under the t-density curve to its left equals α. In other words, P(T ≤ t_ν(α)) = α. However, in A-Level tables, it is common to present upper-tail percentage points, where t_ν(p) indicates the value such that P(T > t_ν(p)) = p. Always check the convention used in your exam board’s statistical tables.
对于给定的概率α和自由度ν，百分位点t_ν(α)定义为t分布密度曲线下其左侧面积等于α的数值。换言之，P(T ≤ t_ν(α)) = α。然而，在A-Level统计表中，通常给出上尾百分位点，其中t_ν(p)表示满足P(T > t_ν(p)) = p的数值。务必确认你所在考试局的统计表采用何种约定。
Two-sided test: P(T > t_ν(α/2)) = α/2
For two-tailed tests, the critical region is split equally: each tail contains α/2. Thus the critical values are ±t_ν(α/2).
对于双尾检验，拒绝域均分到两侧：每个尾部含α/2。因此临界值为±t_ν(α/2)。
4. Reading the t-Distribution Table | 查t分布表
A typical t-table is arranged with rows for degrees of freedom and columns for tail probabilities such as 0.10, 0.05, 0.025, 0.010, and 0.005. The entry at the intersection gives the t-value with that tail area. For example, for ν = 10 and a one-tailed probability of 0.05, the table gives t₁₀(0.05) ≈ 1.812.
典型的t表以行表示自由度，以列表示尾部概率如0.10、0.05、0.025、0.010和0.005。行列交叉处的数值即为对应尾部面积的t值。例如，ν = 10且单尾概率为0.05时，表给出t₁₀(0.05) ≈ 1.812。
| ν \ p | 0.10 | 0.05 | 0.025 | 0.01 | 0.005 |
| 1 | 3.078 | 6.314 | 12.706 | 31.821 | 63.657 |
| 5 | 1.476 | 2.015 | 2.571 | 3.365 | 4.032 |
| 10 | 1.372 | 1.812 | 2.228 | 2.764 | 3.169 |
| 20 | 1.325 | 1.725 | 2.086 | 2.528 | 2.845 |
| 30 | 1.310 | 1.697 | 2.042 | 2.457 | 2.750 |
| ∞ | 1.282 | 1.645 | 1.960 | 2.326 | 2.576 |
Notice that as ν increases, the entries approach the corresponding standard normal z-values. This confirms the limiting result stated earlier.
注意随着ν增大，表中数值逐步接近对应的标准正态z值。这印证了前面提到的极限结论。
5. Symmetry Property | 对称性
Because the t-distribution is symmetric about zero, the lower-tail percentile points satisfy
由于t分布关于零对称，下尾百分位点满足
t_ν(1 − α) = − t_ν(α)
For example, if t₁₀(0.05) = 1.812, then t₁₀(0.95) = −1.812. This property halves the number of values you need to memorise or look up.
例如，若t₁₀(0.05) = 1.812，则t₁₀(0.95) = −1.812。这一性质使你需要记忆或查找的数值数量减半。
6. Using t-Table in Hypothesis Testing | 假设检验中的t表应用
Consider a two-tailed test at the 5% significance level with 8 degrees of freedom. The critical value is t₈(0.025) ≈ 2.306. If the calculated test statistic satisfies |T| > 2.306, you reject the null hypothesis at the 5% level. For a one-tailed test with the same significance level, the critical value is t₈(0.05) ≈ 1.860, and you reject H₀ only if T > 1.860 (for an upper-tailed test).
考虑自由度为8、显著性水平5%的双尾检验。临界值为t₈(0.025) ≈ 2.306。若计算出的检验统计量满足|T| > 2.306，则在5%水平下拒绝原假设。对于相同显著性水平的单尾检验，临界值为t₈(0.05) ≈ 1.860，且仅在T > 1.860时拒绝H₀（对于上尾检验而言）。
Two-tailed: reject H₀ if |T| > t_ν(α/2)
One-tailed upper: reject H₀ if T > t_ν(α)
7. Confidence Intervals Using t Percentile Points | 利用t百分位点构造置信区间
For a random sample x₁, x₂, …, xₙ from a normal population with unknown mean μ and unknown variance σ², a 100(1 − α)% confidence interval for μ is
对于来自正态总体、均值μ和方差σ²均未知的随机样本x₁, x₂, …, xₙ，μ的100(1 − α)%置信区间为
x̄ ± t_{n−1}(α/2) × (s/√n)
where x̄ is the sample mean, s is the sample standard deviation, n is the sample size, and t_{n−1}(α/2) is the two-tailed critical value. For example, if n = 12, α = 0.05, then t₁₁(0.025) ≈ 2.201. Multiplying this by the standard error s/√12 gives the margin of error.
其中x̄是样本均值，s是样本标准差，n是样本量，t_{n−1}(α/2)是双尾临界值。例如，若n = 12，α = 0.05，则t₁₁(0.025) ≈ 2.201。将其乘以标准误s/√12即得误差范围。
8. Interpolation for Unlisted Degrees of Freedom | 未列出自由度的插值
Most t-tables only give values for selected degrees of freedom, such as 1, 2, …, 30, 40, 60, 120, ∞. If your degrees of freedom lie between two tabulated values, use linear interpolation. However, when ν is large (say > 30), you may safely use the next lower tabulated ν as a conservative approximation.
大多数t表只列出选定的自由度，如1, 2, …, 30, 40, 60, 120, ∞。若你的自由度介于两个表列值之间，可采用线性插值。然而，当ν较大（如大于30）时，可安全地使用下一个较小的表列ν作为保守近似。
Interpolated t ≈ t₁ + (t₂ − t₁) × (ν − ν₁)/(ν₂ − ν₁)
This approximation is sufficiently accurate for exam purposes, provided the interval between ν₁ and ν₂ is not too wide.
只要ν₁与ν₂之间的间隔不太宽，这种近似在考试中已经足够准确。
9. Relationship Between t and Standard Normal | t分布与标准正态分布的关系
For large ν, t_ν(α) ≈ z_α, the standard normal percentile point. More precisely, t_∞(α) = z_α. The Edexcel formula booklet includes this in the tables, and exam questions sometimes ask you to state which distribution would be used if the variance were known.
当ν较大时，t_ν(α) ≈ z_α，即标准正态百分位点。更准确地说，t_∞(α) = z_α。Edexcel公式册的统计表中包含这一点，考试题有时要求你说明当方差已知时应使用哪种分布。
Known σ² ⇒ use N(0, 1) or z-test; Unknown σ² ⇒ use t(n − 1)
10. Worked Example: One-Sample t-Test | 例题：单样本t检验
A sample of 9 measurements gives a mean of 50.2 and a standard deviation of 2.4. Test at the 1% significance level whether the population mean is significantly different from 48.
某9次测量得样本均值为50.2，标准差为2.4。在1%显著性水平下检验总体均值是否显著不同于48。
T = (50.2 − 48) / (2.4/√9) = 2.2 / 0.8 = 2.75
With n − 1 = 8 degrees of freedom, the two-tailed critical value at α = 0.01 is t₈(0.005) ≈ 3.355. Since |T| = 2.75 < 3.355, we do not reject H₀. There is insufficient evidence to claim μ ≠ 48.
自由度n − 1 = 8，双尾α = 0.01的临界值为t₈(0.005) ≈ 3.355。由于|T| = 2.75 < 3.355，我们不拒绝H₀。没有充分证据表明μ ≠ 48。
11. Common Mistakes and Exam Tips | 常见错误与考试提示
Students often confuse tail probabilities with cumulative probabilities. When asked for t₈(0.025), some mistakenly use the cumulative entry 0.975. Always check the column heading: if the table lists “tail probability”, the given p is the area in one tail. If the table lists “cumulative probability”, use 1 − p to find the corresponding percentile point.
学生常把尾部概率与累积概率混淆。当要求t₈(0.025)时，有人误用累积值0.975。务必检查列标题：若表列表示”尾部概率”，则所给p是单尾面积；若表列表示”累积概率”，则用1 − p来查找对应百分位点。
- Always record degrees of freedom as n − 1 for a single sample.
- Draw a diagram of the t-curve and shade the critical region.
- Use symmetry: t_ν(1 − α) = −t_ν(α).
- Do not round intermediate values; round only the final answer.
- 单样本的自由度始终记为n − 1。
- 画出t曲线示意图并标注拒绝域。
- 利用对称性：t_ν(1 − α) = −t_ν(α)。
- 中间值不要四舍五入，只对最终答案舍入。
12. Summary and Final Advice | 总结与最后建议
Student’s t-distribution percentile points are a bridge between theory and application. By mastering the table, remembering the symmetry property, and understanding the role of degrees of freedom, you can handle any t-distribution question in the Edexcel A-Level Further Mathematics exam. Practise past papers and always verify whether the required probability is one-tailed or two-tailed before looking up the critical value.
学生t分布的百分位点是连接理论与应用的桥梁。通过掌握查表方法、牢记对称性、理解自由度的作用，你能应对Edexcel A-Level进阶数学考试中任何t分布问题。多做真题，并在查临界值前务必确认所需概率是单尾还是双尾。
Published by TutorHao | Further 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