📚 Year 13 AQA Statistics: Core Knowledge Review | AQA 统计 Year 13 核心知识点梳理
Mastering AQA Statistics at Year 13 requires a solid grasp of probability models, inference techniques, and data analysis skills. This article reviews the essential topics – from discrete and continuous distributions to hypothesis testing, regression, and non‑parametric methods – providing you with a structured revision map in both English and Chinese.
要想在 Year 13 阶段拿下 AQA 统计学，必须牢固掌握概率模型、推断方法和数据分析技能。本文梳理了从离散与连续分布、假设检验、回归到非参数方法等核心专题，为你提供一份结构清晰的中英双语复习指南。
1. Probability and Discrete Distributions | 概率与离散分布
Discrete probability distributions assign probabilities to countable outcomes. For AQA Statistics, the two key models are the binomial distribution B(n, p) and the Poisson distribution Po(λ). The binomial counts successes in a fixed number n of independent trials with constant probability p. Its mean is np and variance np(1−p). The Poisson models the number of events occurring in a fixed interval of time or space, with mean and variance both equal to λ. Conditions for a Poisson distribution include events occurring independently at a constant average rate, with no simultaneous events.
离散型概率分布为可数结果赋予概率。在 AQA 统计中，两个关键模型是二项分布 B(n, p) 和泊松分布 Po(λ)。二项分布用于在固定 n 次独立试验中计数成功次数，每次成功概率 p 不变，均值为 np，方差为 np(1−p)。泊松分布则描述固定时间或空间区间内发生的事件数，其均值与方差都等于 λ。泊松分布的条件包括事件独立发生、平均发生率恒定且不会同时发生两个事件。
Both distributions can be used to calculate probabilities using the probability mass function (PMF) or cumulative distribution tables. The binomial PMF is P(X=k)=ⁿCₖ pᵏ (1−p)ⁿ⁻ᵏ, while the Poisson PMF is P(X=k)=e^{−λ} λᵏ / k!. Recognising the distribution from a scenario is a fundamental skill.
两种分布都可通过概率质量函数（PMF）或累积分布表计算概率。二项 PMF 为 P(X=k)=ⁿCₖ pᵏ (1−p)ⁿ⁻ᵏ，泊松 PMF 为 P(X=k)=e^{−λ} λᵏ / k!。能够从实际问题中识别出正确的分布是一项基本技能。
2. Continuous Distributions – Normal and Exponential | 连续分布 – 正态与指数分布
The normal distribution N(μ, σ²) is the cornerstone of continuous data analysis. Its bell‑shaped curve is symmetric about the mean μ. Standardisation transforms any normal variable X into Z = (X−μ)/σ, where Z ~ N(0,1). Tables of the standard normal distribution provide Φ(z), the cumulative probability P(Z ≤ z). Inverse normal calculations find the value corresponding to a given probability, which is essential for confidence intervals and critical values.
正态分布 N(μ, σ²) 是连续数据分析的基石。它的钟形曲线关于均值 μ 对称。标准化可将任意正态变量 X 转换为 Z = (X−μ)/σ，其中 Z ~ N(0,1)。标准正态分布表给出 Φ(z) 即累积概率 P(Z ≤ z)。逆向正态计算能根据给定概率找到对应数值，这对于置信区间和临界值至关重要。
The exponential distribution Exp(λ) is also examined in AQA Statistics. Its probability density function is f(x)=λe^{−λx} for x≥0, with mean 1/λ and variance 1/λ². The memoryless property states that P(X > s+t | X > s) = P(X > t). It is often used to model waiting times.
指数分布 Exp(λ) 同样是 AQA 统计的考点。其概率密度函数为 f(x)=λe^{−λx}（x≥0），均值为 1/λ，方差为 1/λ²。无记忆性质意味着 P(X > s+t | X > s) = P(X > t)。它常用于建模等待时间。
3. Sampling and the Central Limit Theorem | 抽样与中心极限定理
Understanding the distinction between a population and a sample is vital. A statistic calculated from a sample, such as the sample mean x̄, is a random variable. Its distribution is called the sampling distribution. The Central Limit Theorem (CLT) states that for a sufficiently large sample size (typically n ≥ 30), the distribution of the sample mean is approximately normal with mean μ and variance σ²/n, regardless of the population distribution. If the population is normal, x̄ is exactly normal for any n.
理解总体与样本的区别至关重要。由样本计算出的统计量（如样本均值 x̄）是一个随机变量，其分布称为抽样分布。中心极限定理（CLT）指出，当样本量足够大（通常 n ≥ 30）时，无论总体分布如何，样本均值的分布近似于均值为 μ、方差为 σ²/n 的正态分布。若总体本身服从正态分布，则 x̄ 对任意样本量都精确服从正态分布。
This theorem underpins many inference procedures. When σ is unknown, the sample variance s² is used, and the t‑distribution with n−1 degrees of freedom comes into play for small samples from a normal population.
这一定理是众多推断方法的基础。当 σ 未知时，会使用样本方差 s²，对于来自正态总体的小样本，则需要用到自由度为 n−1 的 t 分布。
4. Estimation and Confidence Intervals | 估计与置信区间
A point estimate gives a single value for a population parameter, while a confidence interval provides a range of plausible values. For a normal population with known σ, a 95% confidence interval for the mean μ is given by x̄ ± z₀.₀₅₍₂₎ × (σ/√n), where z₀.₀₅₍₂₎ = 1.96 is the critical value. If σ is unknown, replace σ/√n with the estimated standard error s/√n and use the t‑distribution critical value.
点估计给出总体参数的单一数值，而置信区间则提供一个可信值范围。对于已知 σ 的正态总体，均值 μ 的 95% 置信区间为 x̄ ± z₀.₀₅₍₂₎ × (σ/√n)，其中 z₀.₀₅₍₂₎ = 1.96 是临界值。若 σ 未知，则将 σ/√n 替换为估计标准误 s/√n，并使用 t 分布的临界值。
Confidence intervals for proportions are constructed using p̂ ± z × √(p̂(1−p̂)/n), provided the success/failure condition (np̂≥10, n(1−p̂)≥10) is met. The width of the interval depends on the sample size and confidence level: higher confidence produces a wider interval.
比例的置信区间使用 p̂ ± z × √(p̂(1−p̂)/n) 构建，前提是满足成功/失败条件 (np̂≥10, n(1−p̂)≥10)。区间的宽度取决于样本量和置信水平：置信水平越高，区间越宽。
5. Hypothesis Testing – Fundamentals and One‑Sample Tests | 假设检验 – 基础与单样本检验
A hypothesis test is a formal procedure for deciding whether a claim about a population parameter is supported by sample evidence. The null hypothesis H₀ is the status quo, while the alternative hypothesis H₁ represents the claim to be tested. Tests can be one‑tailed (directional) or two‑tailed (non‑directional). The test statistic is computed from the sample and compared to a critical region defined by the significance level α. Alternatively, the p‑value – the probability of obtaining a result at least as extreme as the observed one, assuming H₀ is true – is compared with α.
假设检验是一种利用样本证据判断总体参数声明是否成立的规范流程。原假设 H₀ 是现状，备择假设 H₁ 代表待检验的声明。检验可分为单尾（定向）或双尾（非定向）。检验统计量由样本计算得出，并与显著性水平 α 定义的临界域比较。也可以计算 p 值——在原假设为真的前提下，观察到至少如此极端结果的概率——并与 α 对比。
Common one‑sample tests include the z‑test for a mean (σ known or large sample via CLT), the t‑test for a mean (σ unknown, normal population), and tests for a binomial proportion or Poisson mean. Errors can occur: Type I error (rejecting a true H₀) has probability α, while Type II error (failing to reject a false H₀) has probability β. The power of a test is 1−β.
常见的单样本检验包括均值的 z 检验（σ 已知或利用 CLT 的大样本）、均值的 t 检验（σ 未知且来自正态总体），以及二项比例或泊松均值的检验。检验可能出错：第一类错误（拒真）的概率为 α，第二类错误（取伪）的概率为 β。检验的功效为 1−β。
6. Two‑Sample and Paired t‑Tests | 双样本与配对 t 检验
When comparing two population means, the independent two‑sample t‑test assesses whether the difference is statistically significant. It assumes independent random samples from normal populations with equal (or unequal) variances. The test statistic for the equal‑variance case uses a pooled variance estimator, while Welch’s t‑test for unequal variances uses a corrected formula and a reduced degrees of freedom calculated via the Welch–Satterthwaite equation.
比较两个总体均值时，独立双样本 t 检验可评估差异是否具有统计显著性。它假设独立随机样本来自方差相等（或不等）的正态总体。等方差情况下的检验统计量使用合并方差估计量，而针对不等方差的 Welch t 检验则使用修正公式，并通过 Welch–Satterthwaite 方程计算减半的自由度。
A paired t‑test is appropriate when two measurements are taken on the same individual or matched pairs. It reduces the data to differences dᵢ and performs a one‑sample t‑test on the mean difference, often increasing power by eliminating between‑subject variability.
当对同一个体或配对对象进行两次测量时，应使用配对 t 检验。该方法将数据简化为差值 dᵢ，并对平均差值进行单样本 t 检验，通过消除个体间变异通常可提高检验功效。
7. Chi‑Squared Tests: Goodness of Fit and Independence | 卡方检验：拟合优度与独立性
Chi‑squared (χ²) tests handle categorical data. The goodness‑of‑fit test compares observed frequencies with those expected under a hypothesised distribution. The test statistic is χ² = Σ (Oᵢ − Eᵢ)² / Eᵢ, where Oᵢ and Eᵢ are the observed and expected frequencies. The null hypothesis is that the data follow the specified distribution; the degrees of freedom are (number of categories − 1 − number of estimated parameters).
卡方（χ²）检验处理分类数据。拟合优度检验比较观测频数与假设分布下的期望频数。检验统计量为 χ² = Σ (Oᵢ − Eᵢ)² / Eᵢ，其中 Oᵢ 与 Eᵢ 分别是观测频数和期望频数。原假设为数据来自指定分布；自由度为（类别数 − 1 − 估计的参数个数）。
The chi‑squared test for independence examines whether two categorical variables are associated. Data are summarised in a contingency table; expected frequencies are calculated from row and column totals. The degrees of freedom are (r−1)(c−1) for an r×c table. A small expected frequency in any cell (<5) may require combining categories or using a continuity correction.
独立性卡方检验考察两个分类变量是否有关联。数据整理为列联表，期望频数由行列合计计算得出。r×c 表的自由度为 (r−1)(c−1)。若任一单元格的期望频数小于 5，可能需要合并类别或使用连续性校正。
8. Correlation and Linear Regression | 相关与线性回归
The product‑moment correlation coefficient (PMCC) r measures the strength and direction of a linear relationship between two variables. Its value lies between −1 and +1. A hypothesis test for zero correlation (H₀: ρ=0) uses the test statistic t = r√(n−2) / √(1−r²), which follows a t‑distribution with n−2 degrees of freedom. Spearman’s rank correlation coefficient is a non‑parametric alternative based on ranks, useful for monotonic relationships or non‑normal data.
积矩相关系数（PMCC）r 衡量两变量间线性关系的强度与方向，其值介于 −1 和 +1 之间。检验零相关（H₀: ρ=0）使用统计量 t = r√(n−2) / √(1−r²)，它服从自由度为 n−2 的 t 分布。Spearman 秩相关系数是基于排名的非参数替代方法，适用于单调关系或非正态数据。
The least‑squares regression line y = a + bx predicts y from x. The slope b = Sxy / Sxx and intercept a = ȳ − b x̄. Residual analysis checks model assumptions: residuals should be random, have constant variance, and be approximately normal. The coefficient of determination r² indicates the proportion of variation in y explained by x.
最小二乘回归直线 y = a + bx 用于根据 x 预测 y。斜率 b = Sxy / Sxx，截距 a = ȳ − b x̄。残差分析用于检验模型假设：残差应呈随机分布、具有恒定方差并近似正态。决定系数 r² 表示 y 的变异中可由 x 解释的比例。
9. Non‑Parametric Tests – Sign and Wilcoxon | 非参数检验 – 符号检验与 Wilcoxon 检验
Non‑parametric tests do not assume a specific population distribution. The sign test is used for paired data or a single sample to test the median. It counts the number of positive differences (or signs) and compares this to a binomial distribution with p=0.5. It makes no assumption about symmetry but can be less powerful than parametric alternatives.
非参数检验不假定特定的总体分布。符号检验用于配对数据或单样本中检验中位数，它计算正差值（或符号）的个数，并对照 p=0.5 的二项分布。该检验不要求对称性，但功效可能低于参数方法。
The Wilcoxon signed‑rank test assumes the population of differences is symmetric. It ranks the absolute differences in magnitude, attaches the original signs, and sums the positive (or negative) ranks to form the test statistic. It provides greater power than the sign test when the symmetry condition is met. For two independent samples, the Wilcoxon rank‑sum (Mann–Whitney U) test compares medians by ranking all observations together and summing the ranks in one group.
Wilcoxon 符号秩检验假定差值总体对称。它按绝对值大小排序，附上原始符号，再对正（或负）秩求和作为检验统计量。当对称条件满足时，该检验比符号检验功效更高。对于两个独立样本，Wilcoxon 秩和检验（Mann–Whitney U 检验）通过将所有观测值统一排序并对其中一组计算秩和，来比较两组的中位数。
10. The Statistical Enquiry Cycle and Data Analysis | 统计探究循环与数据分析
AQA Statistics emphasises the entire statistical process: from posing a question, planning data collection (including sampling methods such as simple random, stratified, cluster or systematic), to gathering and cleaning data, exploratory analysis (using measures of centre, spread, and graphical displays), formal inference, and finally interpretation and communication. Understanding bias, confounding variables, and the limitations of observational studies is crucial for valid conclusions.
AQA 统计学强调完整的统计流程：从提出问题、规划数据收集（包括简单随机抽样、分层抽样、整群抽样和系统抽样等）、收集与清理数据、探索性分析（利用中心度量、离散度量和图形展示）、进行正规推断，直至结果的解读与传达。理解偏倚、混杂变量以及观察性研究的局限性，是得出有效结论的关键。
Proficiency with large data sets is also assessed. You should be able to summarise, visualise, and reason statistically with real‑world data, using technology where appropriate. Recognising patterns, outliers, and trends, as well as making informed decisions based on statistical evidence, are the ultimate goals of the course.
大规模数据集的处理能力也是考查内容。你应该能够利用技术对真实数据进行总结、可视化并进行统计推理。识别模式、异常值和趋势，以及基于统计证据做出明智决策，是本课程的最终目标。
Published by TutorHao | Statistics 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