📚 Edexcel Further Statistics 2: Core Topics and Exam Techniques | 爱德思进阶统计2：核心主题与考试技巧
Edexcel AS and A Level Further Mathematics Further Statistics 2 builds on Further Statistics 1 by introducing continuous probability models, advanced hypothesis tests and non-parametric methods. This article summarises the key concepts from the textbook by Greg Attwood, Ian Bettison, Alan Clegg and colleagues, helping you consolidate definitions, formulas and exam strategies.
爱德思 AS 与 A Level 进阶数学 进阶统计2 在进阶统计1 的基础上引入连续概率模型、高级假设检验和非参数方法。本文总结了 Greg Attwood、Ian Bettison、Alan Clegg 等编写的教材中的核心概念，帮助你巩固定义、公式和考试策略。
1. Overview of Further Statistics 2 | 进阶统计2 概览
FS2 covers continuous random variables, continuous distributions such as the uniform and the exponential, Poisson hypothesis testing, chi-squared tests, t-tests, confidence intervals, correlation and non-parametric tests. The emphasis is on applying statistical models to real-world data and on interpreting results in context rather than on manipulating algebra alone.
进阶统计2 涵盖连续随机变量、均匀分布与指数分布等连续分布、泊松假设检验、卡方检验、t 检验、置信区间、相关分析以及非参数检验。重点在于将统计模型应用于真实数据，并在具体情境中解释结果，而不只是做代数运算。
- Continuous models: probability density functions, cumulative distribution functions, expectation and variance.
- Named continuous distributions: the continuous uniform distribution and the exponential distribution.
- Hypothesis testing: Poisson means, chi-squared goodness of fit and contingency tables, one-sample and two-sample t-tests.
- Estimation: confidence intervals for a mean using the normal and t-distributions.
- Association: the product moment correlation coefficient and Spearman's rank correlation coefficient.
- Non-parametric methods: the Wilcoxon signed-rank test and the Mann-Whitney U test.
In some routes the chi-squared test and Poisson hypothesis testing are met first in Further Statistics 1 and then revised and extended in FS2, so always check your own specification and the order in which your school teaches the topics. Whatever the route, exam papers usually mix pure calculation with written interpretation, so memorising formulas alone is not enough. You need to state hypotheses, identify the correct distribution and its parameter, compare the test statistic with the critical value, and write a conclusion that refers back to the original problem.
在部分考试路径中，卡方检验与泊松假设检验会先在进阶统计1 中学习，再在进阶统计2 中复习与拓展，因此请务必核对你自己的考试大纲以及学校的教学顺序。无论如何，考试通常混合纯计算与文字解释，因此仅记住公式是不够的。你需要写出假设、确定正确的分布及其参数、将检验统计量与临界值比较，并写出回到原问题情境的结论。
2. Continuous Random Variables and the PDF | 连续随机变量与概率密度函数
A continuous random variable X takes values over an interval and is described by a probability density function f(x). The density itself is not a probability: it only becomes a probability when it is integrated over an interval. Two conditions must hold for a valid density.
连续随机变量 X 在一个区间上取值，由概率密度函数 f(x) 描述。密度函数本身不是概率，只有在一个区间上积分后才成为概率。一个有效的密度函数必须满足两个条件。
f(x) ≥ 0 for all x, and ∫₋∞⁺∞ f(x) dx = 1
(X ≤ x) and P(X < x) are equal, which is a common source of confusion when a question says 'at most' or 'fewer than'.
因此概率通过积分求得；由于单点处的面积为 0，对任意 x 都有 P(X = x) = 0。这意味着 P(X ≤ x) 与 P(X < x) 相等，这也是题目出现 '至多' 或 '少于' 时常见的困惑点。
P(a < X < b) = ∫ₐᵇ f(x) dx
range of x, including the range where f(x) = 0, because the limits of integration come directly from that statement.
当题目给出的 f(x) 含有未知常数 k 时，标准做法是令总面积为 1 并解出 k。务必写出 x 的完整范围，包括 f(x) = 0 的部分，因为积分上下限直接来自这个范围。
3. The CDF, Median and Percentiles | 累积分布函数、中位数与百分位数
tained by integrating the density from the lower end of the range up to x. It is a non-decreasing function that starts at 0 and finishes at 1.
累积分布函数定义为 F(x) = P(X ≤ x)，通过把密度函数从取值范围的下端积分到 x 得到。它是一个不减函数，起点为 0，终点为 1。
F(x) = P(X ≤ x) = ∫₋∞ˣ f(t) dt
: the p-th percentile satisfies F(p) = p/100. The lower quartile is the 25th percentile and the upper quartile is the 75th percentile.
中位数 m 是把分布分成两半相等的取值，因此满足 F(m) = 0.5。百分位数的求法完全相同：第 p 百分位数满足 F(p) = p/100。下四分位数即第 25 百分位数，上四分位数即第 75 百分位数。
iecewise density you must check each piece separately, and for a smooth density you differentiate and solve f' ere in the range, the mode may be at the boundary rather than at a stationary point, so always sketch the density if you are unsure.
众数是使密度 f(x) 取得最大值的 x 值。对于分段密度，必须分别检查每一段；对于光滑密度，则求导并解 f'(x) = 0，再确认它是极大值而非极小值。如果 f(x) 在范围内某处为 0，众数可能出现在边界而不是驻点，因此不确定时应先画草图。
4. Expectation and Variance of a Continuous Variable | 连续随机变量的期望与方差
Expectation and variance for a continuous random variable follow exactly the same logic as the discrete case, with a sum replaced by an integral and a probability replaced by a density.
连续随机变量的期望与方差在逻辑上与离散情形完全一致，只是把求和换成积分、把概率换成密度。
E(X) = ∫ x f(x) dx E(X²) = ∫ x² f(x) dx
Var(X) = E(X²) − [E(X)]² SD(X) = √Var(X)
line x = c, then E(X) = c without any integration at all, and the median is also c. This shortcut is worth looking for before starting a long integral.
X 的函数的期望是自然的推广；而利用对称性求 E(X) 往往更快。如果密度关于直线 x = c 对称，则无需积分即可得到 E(X) = c，中位数也等于 c。在开始冗长积分之前，值得先寻找这个捷径。
E(g(X)) = ∫ g(x) f(x) dx
When a question asks for E(1/X), E(X² + 3) or E(eX are false. For the variance, always subtract the square of the mean at the end, and keep enough decimal places in E(X²) so that the final answer is accurate to three significant figures.
当题目要求 E(1/X)、E(X² + 3) 或 E(eX) 时，应使用上式，而不是试图对 E(X) 做变换。考试中非常常见的错误是写成 E(1/X) = 1/E(X) 或 E(X³) = [E(X)]³，这两个式子都是错误的。求方差时最后一定要减去均值的平方，并在 E(X²) 中保留足够的小数位，使最终答案精确到三位有效数字。
5. The Continuous Uniform Distribution | 连续均匀分布
The continuous uniform distribution models a variable whose density is constant over an interval [a, b]. Because every value in the interval is 'equally likely', probabilities reduce to simple fractions of interval length, which makes this the easiest continuous model in the module.
连续均匀分布用于描述在区间 [a, b] 上密度为常数的变量。由于区间内每个取值 '同样可能'，概率可简化为区间长度的简单比例，这使它成为本模块中最容易的连续模型。
| Quantity | Uniform X ~ U(a, b) |
| Density | f(x) = 1/(b − a) for a ≤ x ≤ b, 0 otherwise |
| CDF | F(x) = (x − a)/(b − a) for a ≤ x ≤ b |
| Mean | E(X) = (a + b)/2 |
| Variance | Var(X) = (b − a)²/12 |
| Median | (a + b)/2 |
In application questions, modelling a random variable as uniform often means that the original data were rounded or that a value is equally likely anywhere in a range. Typical contexts include the time a student arrives within a ten-minute window, or the position of a point chosen at random along a rod. Read carefully for words such as 'rounded to the nearest', because a value rounded to the nearest minute is uniform over an interval of length 1, not over the whole hour.
在应用题中，把随机变量建模为均匀分布，通常意味着原始数据经过取整，或某个取值在某个范围内等可能。典型情境包括学生在这十分钟窗口内到达的时间，或在一根杆上随机选取一点的位置。要特别留意 '四舍五入到最近的' 之类的措辞，因为四舍五入到最近分钟的值在长度为 1 的区间上均匀分布，而不是在整个小时上均匀分布。
P(c < X < d) = (d − c)/(b − a) for a ≤ c < d ≤ b
6. The Exponential Distribution | 指数分布
The exponential distribution models the waiting time between events in a Poisson process. If events occur at random at a constant average rate λ, then the time X between successive events follows an exponential distribution with parameter λ.
指数分布用于描述泊松过程中事件之间的等待时间。如果事件以恒定平均速率 λ 随机发生，那么相邻两次事件之间的时间 X 服从参数为 λ 的指数分布。
f(x) = λe−λx, x ≥ 0 F(x) = 1 − e−λx
E(X) = 1/λ Var(X) = 1/λ²
a high rate λ means a short average waiting time. The survival probability P(X > x) = e−λx is often quicker to use than the CDF, especially for 'greater than' questions. The median is found from e−λx 2 / λ, which is noticeably smaller than the mean because the distribution is right-skewed.
注意参数方向：速率 λ 越大，平均等待时间越短。生存概率 P(X > x) = e−λx 通常比使用累积分布函数更快，尤其是 '大于' 类问题。中位数由 e−λx = 0.5 解得 x = ln 2 / λ，由于分布右偏，它明显小于均值。
oryless, which means that knowing you have already waited does not change the distribution of the remaining wait. Formally, P(X > s + t | s examined both as a calculation and as an interpretation question, and it is the key difference between the exponential model and, for example, a uniform waiting-time model.
指数分布具有无记忆性，也就是说，知道你已等待了一段时间，并不会改变剩余等待时间的分布。形式上，P(X > s + t | X > s) = P(X > t)。这一性质既可能以计算题出现，也可能以解释题出现，它是指数模型与均匀等待时间模型等之间的关键区别。
7. Hypothesis Testing for a Poisson Mean | 泊松均值的假设检验
A Poisson hypothesis test asks whether the observed data are consistent with a claimed mean rate. The four-step structure is fixed and marks are awarded for each step, so follow it every time.
泊松假设检验用于判断观测数据是否与某个声称的平均速率一致。四步结构是固定的，每一步都有分数，因此每次都要按顺序执行。
- Step 1 - Hypotheses:广告 / Sponsored官网正版 国际课程数学核心词汇 唯寻国际教育 全英 学科 GCSE A Level IB MYP IB IBDP 美高…数学 备考教辅 · 正版现货 · 与本文内容相关查看详情 / View on Taobao参考价 ¥19.7
