📚 Year 13 Edexcel Statistics: Key Knowledge Points Review | Year 13 Edexcel 统计：核心知识点梳理
In Year 13, Edexcel A Level Mathematics builds upon the statistics foundations laid in Year 12, introducing more advanced inferential techniques and continuous probability models. The core themes involve the normal distribution as a model for real-world data, hypothesis testing for population means and correlation coefficients, and the normal approximation to the binomial. A thorough understanding of these concepts is essential for tackling exam questions that require both calculator fluency and the ability to interpret statistical evidence in context.
在 Year 13，Edexcel A Level 数学在 Year 12 打下的基础之上，引入了更深入的推断方法和连续概率模型。核心主题包括用正态分布为实际数据建模、对总体均值和相关系数进行假设检验，以及用正态分布近似二项分布。透彻理解这些概念对于应对考试至关重要，这类题目既要求熟练使用计算器，又需要准确地结合情境解释统计证据。
1. Conditional Probability and Bayes’ Theorem | 条件概率与贝叶斯定理
Conditional probability is the probability of an event occurring given that another event has already occurred. The formula P(A|B) = P(A ∩ B) / P(B) captures this relationship, and it underpins many Year 13 probability problems involving tree diagrams, two-way tables, and Venn diagrams. When events are not independent, calculating P(A|B) often requires careful multiplication along branches.
条件概率是指在已知另一事件已发生的条件下，某事件发生的概率。公式 P(A|B) = P(A ∩ B) / P(B) 概括了这一关系，它也是许多 Year 13 概率问题的基础，这些问题常涉及树状图、双向表格和韦恩图。当事件不独立时，计算 P(A|B) 常常需要沿树状图分支细心地进行乘法运算。
Bayes’ theorem provides a systematic way to reverse conditional probabilities, linking P(A|B) to P(B|A). In an exam, you may be asked to calculate the probability that a defective item came from a specific machine, or that a positive test result truly indicates a disease. The theorem is expressed as P(A|B) = [P(B|A) × P(A)] / P(B), where P(B) is the total probability of B across all partitions.
贝叶斯定理提供了一种将条件概率“反转”的系统方法，从而将 P(A|B) 与 P(B|A) 联系起来。考试中可能会要求你计算有缺陷的零件来自某台特定机器的概率，或者某项阳性检测结果真实地预示疾病的概率。定理的表达式为 P(A|B) = [P(B|A) × P(A)] / P(B)，其中 P(B) 是 B 在所有划分下的全概率。
| Example: A box has two types of coins: 70% fair, 30% biased (with P(Heads)=0.8). If a randomly chosen coin shows heads, what is the probability it is biased? P(Biased|H) = P(H|Biased)×P(Biased) / [P(H|Fair)P(Fair)+P(H|Biased)P(Biased)] = (0.8×0.3)/(0.5×0.7+0.8×0.3) = 0.24/0.59 ≈ 0.407. |
| 示例：一个盒子里有两种硬币：70% 是公平硬币，30% 是有偏硬币（P(正面)=0.8）。随机取一枚硬币投出正面，问它是偏币的概率是多少？ P(偏|H) = P(H|偏)×P(偏) / [P(H|公平)P(公平)+P(H|偏)P(偏)] = (0.8×0.3)/(0.5×0.7+0.8×0.3) = 0.24/0.59 ≈ 0.407。 |
2. The Normal Distribution | 正态分布
At the heart of Year 13 statistics is the normal distribution, defined by its bell-shaped probability density curve. It is fully specified by two parameters: the population mean μ and the population variance σ². We write X ~ N(μ, σ²). Continuous variables such as heights, weights, and measurement errors are often modelled using a normal distribution, and the total area under the curve equals 1.
Year 13 统计的核心是正态分布，其标志性的钟形概率密度曲线由两个参数完全决定：总体均值 μ 和总体方差 σ²，记作 X ~ N(μ, σ²)。诸如身高、体重和测量误差这类连续变量，往往可以用正态分布建模，而且曲线下的总面积等于 1。
The probability that X falls within an interval is found by calculating the area under the curve between two x-values. In practice, we use a calculator’s normal cumulative distribution function (normal CDF) with inputs: lower bound, upper bound, μ, σ. For example, P(X < a) uses a very low lower bound, e.g. -1099. It is crucial to draw a sketch and shade the required region before using technology, as this helps prevent sign errors.
X 落在某个区间内的概率是通过计算曲线上两个 x 值之间的面积来得到的。实际计算中，我们使用计算器的正态累积分布函数，输入：下界、上界、μ、σ。例如，计算 P(X < a) 时，会以一个非常小的值作为下界，如 -1099。在动用计算器前，画出草图并涂出所需区域非常重要，这有助于避免符号错误。
- About 68% of values lie within μ ± σ; about 95% within μ ± 2σ; about 99.7% within μ ± 3σ.
- 大约 68% 的值落在 μ ± σ 之间；约 95% 落在 μ ± 2σ 之间；约 99.7% 落在 μ ± 3σ 之间。
3. Standard Normal and Inverse Normal | 标准正态与逆正态
Any normal variable X ~ N(μ, σ²) can be transformed into the standard normal variable Z by the transformation Z = (X − μ) / σ. Then Z ~ N(0, 1) has mean 0 and standard deviation 1. This transformation is useful for comparing values from different normal distributions and for finding probabilities using standard normal tables, although calculators have made direct use of tables less common.
任何正态变量 X ~ N(μ, σ²) 都可以通过变换 Z = (X − μ) / σ 转化为标准正态变量 Z，其分布为 Z ~ N(0, 1)，均值为 0，标准差为 1。这种变换有助于比较来自不同正态分布的值，并可使用标准正态表求概率，不过计算器的普及使得直接查表不那么常见了。
The inverse normal operation answers the question: “Given a left-hand cumulative probability p, what value a gives P(X < a) = p?” On your calculator, this is the ‘Inverse Normal’ function; you enter the area (probability), μ, and σ. You can also find the bounds of a symmetric central interval containing, say, 95% of the data, by using the fact that each tail contains 2.5%.
逆正态运算回答的问题是：“给定一个左尾累积概率 p，满足 P(X < a) = p 的 a 值是多少？”在计算器上，使用“逆正态”功能并输入面积（概率）、μ 和 σ。你也可以求出包含 95% 数据的对称中心区间的界限，利用两侧尾部各含 2.5% 这一事实。
For a 95% symmetric interval: a = invNorm(0.025, μ, σ), b = invNorm(0.975, μ, σ).
95% 对称区间：a = invNorm(0.025, μ, σ)，b = invNorm(0.975, μ, σ)。
4. Normal Approximation to the Binomial Distribution | 二项分布的正态近似
When a binomial distribution B(n, p) has a large n and p is not too close to 0 or 1, its shape approximates a normal curve. The usual rule of thumb is that both np > 5 and nq > 5, where q = 1 − p. Then the approximating normal distribution has μ = np and σ² = npq. This allows us to use normal probability methods to approximate binomial probabilities, which is particularly useful when calculating large cumulative sums.
当二项分布 B(n, p) 的 n 足够大且 p 不太接近 0 或 1 时，其形状近似于一条正态曲线。通常的经验法则是要求 np > 5 且 nq > 5（其中 q = 1 − p）。此时，用以近似的正态分布的参数为 μ = np，σ² = npq。这使得我们能够用正态概率方法来近似二项概率，尤其在计算大量累积概率时非常有用。
Because the binomial is discrete and the normal is continuous, a continuity correction must be applied. For example, P(X ≤ 10) for the binomial becomes P(X < 10.5) under the normal model, while P(X ≥ 10) becomes P(X > 9.5). In exam questions, always state the continuity correction explicitly and show the binomial statement converting to a normal statement.
由于二项分布是离散的而正态分布是连续的，必须进行连续性校正。例如，二项分布的 P(X ≤ 10) 在正态模型下变为 P(X < 10.5)，而 P(X ≥ 10) 则变为 P(X > 9.5)。在考试题中，务必清楚地写明连续性校正，并展示从二项表达式到正态表达式的转化过程。
5. Correlation and the Product Moment Correlation Coefficient | 相关性与积矩相关系数
The product moment correlation coefficient (PMCC), denoted by r, measures the linear strength and direction of the relationship between two variables. Its value ranges from −1 (perfect negative correlation) to +1 (perfect positive correlation), with 0 indicating no linear correlation. The formula is r = Sxy / √(Sxx Syy), where Sxy = Σ(x−x̄)(y−ȳ) and similarly for Sxx, Syy.
积矩相关系数（PMCC），记作 r，衡量的是两个变量之间线性关系的强度和方向。它的取值范围从 −1（完全负相关）到 +1（完全正相关），0 表示无线性相关。计算公式为 r = Sxy / √(Sxx Syy)，其中 Sxy = Σ(x−x̄)(y−ȳ)，Sxx 和 Syy 同理。
In practice, use your calculator’s statistics mode to enter bivariate data and obtain r directly. The PMCC is unaffected by linear changes of scale (coding), such as converting units or shifting by a constant. You may need to interpret r in context: a high positive r suggests that as one variable increases, so does the other, but does not necessarily imply causation.
在实际操作中，可使用计算器的统计模式输入双变量数据，直接得到 r 值。PMCC 不受线性尺度变换（编码）的影响，比如单位转换或加上常数。你可能还需要结合具体情境解释 r：较高的正 r 意味着随着一个变量增加，另一个也同样增加，但这不一定意味着因果关系。
6. Hypothesis Testing for Zero Correlation | 零相关假设检验
We often test whether the population correlation coefficient ρ is zero against a one- or two-tailed alternative. The null hypothesis is H₀: ρ = 0, meaning no linear relationship. The test statistic uses the sample r and the critical value from a t-distribution with n − 2 degrees of freedom, or more directly, the PMCC table of critical values provided in the formula booklet.
我们经常检验总体相关系数 ρ 是否为零，备择假设可以是单尾或双尾。零假设为 H₀: ρ = 0，表示没有线性关系。检验统计量基于样本 r 以及来自自由度为 n − 2 的 t 分布的临界值，或者更直接地，使用公式表中提供的 PMCC 临界值表。
The test statistic is t = r √(n − 2) / √(1 − r²), but in Edexcel exams you compare r with the table critical value directly. If |r| exceeds the critical value, reject H₀ and conclude there is evidence of correlation. Always interpret the conclusion in the context of the question, and state whether the correlation is positive or negative.
检验统计量为 t = r √(n − 2) / √(1 − r²)，但在 Edexcel 考试中，通常直接将 r 与表格中的临界值进行比较。如果 |r| 超过临界值，则拒绝 H₀，认为有证据表明存在相关性。始终要将结论放在题目情境中进行解释，并说明相关的正负方向。
| Steps | 步骤 |
|---|---|
| 1. State H₀: ρ = 0, H₁: ρ ≠ 0 (or >0, <0) 2. Choose significance level, e.g. 5% 3. Find critical value from table for n−2 df 4. Calculate r from data 5. Compare |r| with critical value and conclude |
1. 写出 H₀: ρ = 0，H₁: ρ ≠ 0（或 >0, <0） 2. 选择显著性水平，例如 5% 3. 从表中查得 n−2 自由度对应的临界值 4. 根据数据计算 r 5. 将 |r| 与临界值比较并给出结论 |
7. Hypothesis Testing for the Mean (Known Variance) | 均值的假设检验（已知方差）
When a population is known to be normally distributed with a known variance σ², we can test a claimed value for the population mean μ. The null hypothesis is H₀: μ = μ₀. The test statistic is Z = (x̄ − μ₀) / (σ/√n), which follows a standard normal distribution. You compare this z-value with the critical z-value corresponding to the significance level, using ±1.96 for a 5% two-tailed test.
当已知总体服从正态分布且方差 σ² 已知时，我们可以检验对总体均值 μ 的声明。零假设为 H₀: μ = μ₀。检验统计量为 Z = (x̄ − μ₀) / (σ/√n)，它服从标准正态分布。将该 z 值与对应显著性水平的临界 z 值进行比较，例如对于 5% 的双尾检验，临界值是 ±1.96。
Alternatively, calculate the p-value — the probability of obtaining a sample mean as extreme as observed, given H₀ is true — and compare it to the significance level. If the p-value < significance level, reject H₀. In an exam, you must state your conclusion clearly in words, referring back to the original claim.
另一种方法是计算 p 值——在 H₀ 为真的条件下，得到像观测值那样极端样本均值的概率——并与显著性水平比较。如果 p 值小于显著性水平，则拒绝 H₀。在考试中，你必须用清晰的语言陈述结论，并回扣原声明。
Z = (x̄ − μ₀) / (σ/√n)
8. Hypothesis Testing for Proportions (Normal Approximation) | 比例假设检验（正态近似）
To test a claim about a population proportion p, the sample proportion p̂ is used. If np₀ > 5 and n(1 − p₀) > 5, the binomial distribution of the number of successes is approximately normal, and the test statistic is Z = (p̂ − p₀) / √(p₀(1 − p₀)/n). A continuity correction is generally applied to the observed number of successes, e.g. using X ± 0.5, before calculating p̂.
要检验关于总体比例 p 的声明，需使用样本比例 p̂。若 np₀ > 5 且 n(1 − p₀) > 5，成功次数所服从的二项分布近似正态，检验统计量为 Z = (p̂ − p₀) / √(p₀(1 − p₀)/n)。通常在计算 p̂ 之前，先对观测到的成功次数进行连续性校正，即使用 X ± 0.5。
For example, if testing H₁: p > p₀ and 63 successes are observed in 200 trials, the corrected number is 62.5. Then p̂ = 62.5/200 = 0.3125, and the z-statistic is compared with the critical value. Remember to check that the conditions for the normal approximation hold; otherwise, a binomial exact test may be required.
例如，如果检验 H₁: p > p₀ 并在 200 次试验中观察到 63 次成功，校正后的成功次数为 62.5，于是 p̂ = 62.5/200 = 0.3125，再将 z 统计量与临界值比较。务必先检查正态近似的条件是否满足；若不满足，则可能需要使用精确的二项检验。
9. Data Coding and Its Effects on Mean and Variance | 数据编码及对均值和方差的影响
Coding is the process of transforming original data by a linear change, y = (x − a)/b or y = (x − c). It simplifies the arithmetic for calculating the mean and standard deviation by hand. If y = (x − a)/b, then the mean transforms as x̄ = a + bȳ, and the standard deviation as σx = bσy. Knowing these relationships allows you to recover the original summary statistics from coded values.
编码是指通过线性变换 y = (x − a)/b 或 y = (x − c) 来转换原始数据的过程。它简化了手工计算均值和标准差的运算。若 y = (x − a)/b，则均值的转换为 x̄ = a + bȳ，标准差的转换为 σx = bσy。了解这些关系后，你就能从编码后的值反推出原始数据的汇总统计。
When working with bivariate data, the PMCC r and the equation of the regression line are affected by coding. Specifically, r remains unchanged under any linear transformation of x or y, but the regression coefficients adjust. If y on x is coded as v = (x − a)/c and w = (y − b)/d, the regression coefficient in the original units is recovered by reversing the coding.
在处理双变量数据时，编码会影响 PMCC 和回归直线方程。具体来说，r 在对 x 或 y 进行任何线性变换时都保持不变，但回归系数会随之调整。如果 y 对 x 的回归中，v = (x − a)/c, w = (y − b)/d，则需要通过逆向编码来还原原始单位下的回归系数。
By mastering these Year 13 topics and regularly practising past paper questions, you will build the confidence to apply statistical reasoning under exam conditions. Always remember to annotate your working, sketch distributions, and interpret your findings in the given context.
通过掌握这些 Year 13 主题并定期练习历年真题，你将建立起在考试条件下运用统计推理的信心。切记要标注解题步骤、画出分布草图，并结合给定情境解释你的发现。
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