📚 A-Level OCR Statistics: Quick Reference Handbook of Formulas and Theorems | A-Level OCR 统计:公式定理速查手册
Need a one-stop reference for every key formula and theorem in OCR A-Level Statistics? This handbook is designed for rapid revision. It consolidates probability rules, distribution properties, sampling theory, confidence intervals, hypothesis tests, correlation, regression, and chi‑squared tests into a clear, bilingual format. Each section pairs a concise English explanation with its Chinese counterpart so you can cross‑reference instantly.
想要一份 OCR A-Level 统计所有核心公式定理的一站式速查手册吗?本文专为快速复习打造。它将概率法则、分布性质、抽样理论、置信区间、假设检验、相关、回归以及卡方检验浓缩为清晰的双语格式。每个小节都用精炼的英文解释配以中文对应,方便即时对照。
1. Basic Probability Rules | 基本概率法则
Addition Rule: P(A ∪ B) = P(A) + P(B) − P(A ∩ B). For mutually exclusive events, P(A ∩ B) = 0, hence P(A ∪ B) = P(A) + P(B).
加法法则: P(A ∪ B) = P(A) + P(B) − P(A ∩ B)。对于互斥事件,P(A ∩ B) = 0,因此 P(A ∪ B) = P(A) + P(B)。
Multiplication Rule: P(A ∩ B) = P(A) × P(B|A) = P(B) × P(A|B). If A and B are independent, P(A|B) = P(A) and P(B|A) = P(B); then P(A ∩ B) = P(A) × P(B).
乘法法则: P(A ∩ B) = P(A) × P(B|A) = P(B) × P(A|B)。若 A 与 B 独立,P(A|B) = P(A) 且 P(B|A) = P(B),则 P(A ∩ B) = P(A) × P(B)。
Complement Rule: P(A’) = 1 − P(A).
补集法则: P(A’) = 1 − P(A)。
Conditional Probability: P(A|B) = P(A ∩ B) / P(B), provided P(B) > 0.
条件概率: P(A|B) = P(A ∩ B) / P(B),前提 P(B) > 0。
2. Discrete Random Variables | 离散随机变量
Notation & Sum of Probabilities: For a discrete random variable X with values xi, Σ P(X = xi) = 1.
记法与概率之和: 对于离散随机变量 X,其取值 xi 满足 Σ P(X = xi) = 1。
Expected Value (Mean): E(X) = μ = Σ xi · P(X = xi).
期望值(均值): E(X) = μ = Σ xi · P(X = xi)。
Variance: Var(X) = E(X²) − [E(X)]² = Σ xi² P(X = xi) − μ². Standard deviation σ = √Var(X).
方差: Var(X) = E(X²) − [E(X)]² = Σ xi² P(X = xi) − μ²。标准差 σ = √Var(X)。
Linear Transformations: E(aX + b) = a E(X) + b, Var(aX + b) = a² Var(X).
线性变换: E(aX + b) = a E(X) + b, Var(aX + b) = a² Var(X)。
3. Binomial Distribution | 二项分布
Conditions: Fixed number of trials n, each trial independent, two possible outcomes (success/failure), constant probability of success p.
适用条件: 固定试验次数 n,各次试验独立,每次只有成功或失败两种结果,成功概率 p 保持不变。
X ~ B(n, p)
Probability Mass Function: P(X = k) = nCk pk (1 − p)n−k, k = 0, 1, 2, …, n.
(or nCk = n! / [k!(n−k)!])
概率质量函数: P(X = k) = nCk pk (1 − p)n−k, k = 0, 1, 2, …, n。
(或 nCk = n! / [k!(n−k)!])
Mean & Variance: E(X) = np, Var(X) = np(1 − p).
均值与方差: E(X) = np, Var(X) = np(1 − p)。
4. Poisson Distribution | 泊松分布
Conditions: Events occur singly, randomly, independently, and at a constant average rate λ over a fixed interval (time, length, area, etc.).
适用条件: 事件在固定区间内单个、随机、独立地发生,且平均发生率 λ 恒定。
X ~ Po(λ)
Probability Mass Function: P(X = k) = e−λ λk / k!, k = 0, 1, 2, …
概率质量函数: P(X = k) = e−λ λk / k!, k = 0, 1, 2, …
Mean & Variance: E(X) = λ, Var(X) = λ.
均值与方差: E(X) = λ, Var(X) = λ。
Binomial to Poisson Approximation: If n is large and p is small (n > 50, np < 10), then B(n, p) ≈ Po(np).
二项逼近泊松: 当 n 很大且 p 很小(n > 50,np < 10)时,B(n, p) 可近似为 Po(np)。
5. Normal Distribution | 正态分布
X ~ N(μ, σ²)
Standardisation: Z = (X − μ) / σ → Z ~ N(0, 1). Use standard normal tables Φ(z) for cumulative probabilities.
标准化: Z = (X − μ) / σ → Z ~ N(0, 1)。使用标准正态分布表 Φ(z) 求累积概率。
Symmetry Property: Φ(−z) = 1 − Φ(z). P(Z > z) = 1 − Φ(z).
对称性: Φ(−z) = 1 − Φ(z),P(Z > z) = 1 − Φ(z)。
Finding x-values from probabilities: If P(X < x) = p, then x = μ + σ × zp, where zp is the z‑value such that Φ(zp) = p (inverse normal).
由概率求 x 值: 若 P(X < x) = p,则 x = μ + σ × zp,其中 zp 满足 Φ(zp) = p(逆正态)。
Approximations to Normal:
• Binomial: when np > 5 and n(1−p) > 5, B(n, p) ≈ N(np, np(1−p)) with continuity correction ±½.
• Poisson: when λ > 10, Po(λ) ≈ N(λ, λ) with continuity correction ±½.
正态近似:
• 二项:当 np > 5 且 n(1−p) > 5 时,B(n, p) ≈ N(np, np(1−p)),需连续性校正 ±½。
• 泊松:当 λ > 10 时,Po(λ) ≈ N(λ, λ),需连续性校正 ±½。
6. Geometric Distribution | 几何分布
Conditions: Independent trials, each with constant probability of success p. The random variable X is the number of trials needed to obtain the first success.
适用条件: 独立试验,每次成功概率 p 恒定。随机变量 X 表示获得首次成功所需的试验次数。
X ~ Geo(p)
Probability Mass Function: P(X = k) = (1 − p)k−1 p, k = 1, 2, 3, …
概率质量函数: P(X = k) = (1 − p)k−1 p, k = 1, 2, 3, …
Cumulative Distribution: P(X ≤ k) = 1 − (1 − p)k.
累积分布: P(X ≤ k) = 1 − (1 − p)k。
Mean & Variance: E(X) = 1/p, Var(X) = (1 − p) / p².
均值与方差: E(X) = 1/p, Var(X) = (1 − p) / p²。
7. Sampling & Central Limit Theorem | 抽样与中心极限定理
Sample Mean: x̄ = (Σ xi) / n. It is an unbiased estimator of the population mean μ.
样本均值: x̄ = (Σ xi) / n。它是总体均值 μ 的无偏估计量。
Sampling Distribution of the Mean: If X ~ N(μ, σ²) or n is large, then X̄ ~ N(μ, σ²/n). Standard error = σ/√n.
均值的抽样分布: 若 X ~ N(μ, σ²) 或 n 很大,则 X̄ ~ N(μ, σ²/n)。标准误差 = σ/√n。
Central Limit Theorem (CLT): For any population with mean μ and variance σ² (σ² < ∞), the distribution of X̄ approaches N(μ, σ²/n) as n → ∞ (typically n ≥ 30).
中心极限定理 (CLT): 对于任何均值为 μ、方差为 σ²(有限)的总体,当 n → ∞(通常 n ≥ 30)时,X̄ 的分布趋近于 N(μ, σ²/n)。
Unbiased Estimate of Population Variance: s² = Σ (xi − x̄)² / (n − 1).
总体方差的无偏估计: s² = Σ (xi − x̄)² / (n − 1)。
8. Confidence Intervals | 置信区间
Confidence Interval for Mean (σ known): x̄ ± z* × σ/√n. Typical z* values: 90% → 1.645, 95% → 1.960, 99% → 2.576.
均值置信区间(σ 已知): x̄ ± z* × σ/√n。常用 z* 值:90% → 1.645,95% → 1.960,99% → 2.576。
Confidence Interval for a Proportion: p̂ ± z* × √[p̂(1 − p̂)/n], where p̂ = number of successes / n. Conditions: np̂ > 5 and n(1−p̂) > 5.
比例置信区间: p̂ ± z* × √[p̂(1 − p̂)/n],其中 p̂ = 成功次数 / n。条件:np̂ > 5 且 n(1−p̂) > 5。
Interpretation: If many samples are taken, the stated percentage (e.g., 95%) of such confidence intervals will contain the true population parameter.
解释: 若多次抽样,则所声称的百分比(如 95%)的置信区间将包含真实的总体参数。
9. Hypothesis Testing | 假设检验
Structure: State null H0 and alternative H1 hypotheses, significance level α, test statistic, critical region, and conclusion (reject H0 or not).
结构: 陈述原假设 H0 与备择假设 H1,设定显著性水平 α,计算检验统计量,确定临界域,得出结论(拒绝或不拒绝 H0)。
p‑value approach: Probability, under H0, of obtaining a result at least as extreme as the observed statistic. Reject H0 if p‑value < α.
p 值法: 在 H0 为真的前提下,获得至少与观测统计量一样极端的结果的概率。若 p 值 < α,则拒绝 H0。
| Test | Null H0 | Test Statistic | Critical Region |
|---|---|---|---|
| Binomial single proportion | p = p0 | X ~ B(n, p0) | Use binomial CDF |
| Poisson mean | λ = λ0 | X ~ Po(λ0) | Use Poisson CDF |
| z‑test (mean, σ known) | μ = μ0 | z = (x̄ − μ0) / (σ/√n) | z > zα or z < −zα |
| 检验 | 原假设 H0 | 检验统计量 | 临界域 |
|---|---|---|---|
| 二项单比例检验 | p = p0 | X ~ B(n, p0) | 使用二项累积分布 |
| 泊松均值检验 | λ = λ0 | X ~ Po(λ0) | 使用泊松累积分布 |
| z 检验(均值,σ 已知) | μ = μ0 | z = (x̄ − μ0) / (σ/√n) | z > zα 或 z < −zα |
Type I & Type II Errors: Type I: rejecting H0 when it is true (P = α). Type II: failing to reject H0 when it is false (P = β).
第一类与第二类错误: 第一类错误:H0 为真却被拒绝(概率 α)。第二类错误:H0 为假却未被拒绝(概率 β)。
10. Correlation and Regression | 相关与回归
Pearson’s Product–Moment Correlation Coefficient:
r = Sxy / √(Sxx · Syy)
where Sxx = Σ(xi − x̄)², Syy = Σ(yi − ȳ)², Sxy = Σ(xi − x̄)(yi − ȳ). −1 ≤ r ≤ 1. r near ±1 indicates strong linear correlation.
皮尔逊积矩相关系数:
r = Sxy / √(Sxx · Syy)
其中 Sxx = Σ(xi − x̄)², Syy = Σ(yi − ȳ)², Sxy = Σ(xi − x̄)(yi − ȳ)。−1 ≤ r ≤ 1。r 接近 ±1 表示强线性相关。
Spearman’s Rank Correlation: For ranked data, rs = 1 − (6 Σ di²) / [n(n²−1)], where di = difference in ranks. Used when data is non‑linear or ordinal.
斯皮尔曼秩相关系数: 对排序数据,rs = 1 − (6 Σ di²) / [n(n²−1)],其中 di 为秩差。适用于非线性或有序数据。
Least Squares Regression Line: y = a + b x, where b = Sxy / Sxx, a = ȳ − b x̄.
最小二乘回归直线: y = a + b x,其中 b = Sxy / Sxx, a = ȳ − b x̄。
Residuals: ei = observed yi − predicted yi. The sum of residuals is zero. Residual plots help check linearity and constant variance.
残差: ei = 观测值 yi − 预测值 yi。残差和为零。残差图可检查线性与等方差性。
11. Chi‑Squared Tests | 卡方检验
Chi‑Squared Statistic: χ² = Σ (O − E)² / E, where O = observed frequency, E = expected frequency. The statistic follows approximately χ² distribution provided all E ≥ 5.
卡方统计量: χ² = Σ (O − E)² / E,其中 O 为观测频数,E 为期望频数。该统计量近似服从卡方分布,前提是所有 E ≥ 5。
1. Goodness‑of‑Fit Test: Tests whether observed frequencies differ from a hypothesised distribution. Degrees of freedom (df) = k − 1 (k = number of categories). If parameters are estimated from data, subtract one df for each parameter estimated.
1. 拟合优度检验: 检验观测频数与假设分布是否吻合。自由度 df = k − 1(k 为类别数)。若从数据估计参数,每估计一个参数则再减 1 个自由度。
2. Test for Association (Contingency Table): H0: variables are independent. Expected frequency for cell (i, j) = (row totali × column totalj) / grand total. df = (r − 1)(c − 1).
2. 关联性检验(列联表):
Published by TutorHao | A-Level 统计 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