📚 IGCSE CAIE Statistics: Past Paper Deep Dive | IGCSE CAIE 统计：历年真题深度解析
Past papers are the most reliable resource for IGCSE CAIE Statistics. They reveal recurring question types, common traps and the exact wording examiners expect.
历年真题是 IGCSE CAIE 统计最可靠的备考资料。它们能揭示反复出现的题型、常见陷阱以及考官期望的精确表达。
This article breaks down the most tested topics and shows how to turn past-paper practice into higher marks, based on examiner reports and official mark schemes.
本文基于考官报告和官方评分标准，拆解最高频考点，并说明如何通过真题练习有效提分。
1. Exam Structure and Mark Allocation | 考试结构与分值分配
The CAIE IGCSE Statistics paper usually lasts 2 hours 15 minutes and carries 100 marks. Questions are a mix of short-answer items and longer structured tasks based on real datasets.
CAIE IGCSE 统计考试通常为 2 小时 15 分钟，满分 100 分。试题包括简答题和基于真实数据集的较长结构化任务。
Past papers show that marks are spread across data handling, probability, summary statistics and interpretation. Do not spend too long on drawing one chart; often the interpretation carries more marks than the drawing itself.
历年真题显示，分值分布在数据处理、概率、汇总统计和结果解释中。不要在绘制某一张图上花费过多时间，解释往往比绘图本身占更多分数。
In many sessions, roughly 40% of marks come from statistical calculations and 30% from diagrams, while the remaining 30% test interpretation and communication in context.
许多考季中，约 40% 的分值来自统计计算，30% 来自图表，其余 30% 考查结合背景的解释与表达。
| Question type 题型 | Typical marks 典型分值 | Examiner focus 考官关注点 |
|---|---|---|
| Data representation 数据表示 | 10-15 | Accurate plotting and reading 准确绘图与读数 |
| Summary statistics 汇总统计 | 12-18 | Correct formulas and comparison 正确公式与比较 |
| Probability 概率 | 15-20 | Tree diagrams and rules 树状图与规则 |
2. Data Representation: Charts and Diagrams | 数据表示：图表
A very frequent past-paper task is completing a cumulative frequency table and drawing a smooth cumulative frequency curve. You must plot points at upper class boundaries, not midpoints.
真题中非常常见的任务是补全累积频数表并绘制平滑的累积频率曲线。你必须在组上限处描点，而不是组中点。
Histograms require frequency density. If a bar is missing, calculate frequency density = frequency ÷ class width, then read the vertical scale carefully.
直方图需要用到频率密度。如果缺少某个条形，先计算频率密度 = 频数 ÷ 组距，然后仔细读取纵轴刻度。
In box-and-whisker plots, examiners often ask for median, quartiles and interquartile range. Use the cumulative frequency graph to locate Q₁, Q₂ and Q₃ accurately.
在箱线图中，考官经常要求求中位数、四分位数和四分位距。要利用累积频率图准确定位 Q₁、Q₂ 和 Q₃。
When comparing two distributions, always refer to both the median and the interquartile range. For example, ‘Dataset B has a higher median and a smaller IQR, so it is generally higher and more consistent.’
比较两个分布时，一定要同时提到中位数和四分位距。例如 ‘数据集 B 的中位数更高且 IQR 更小，因此整体更高且更稳定。’
3. Measures of Central Tendency and Spread | 集中趋势与离散程度
Past papers regularly include grouped frequency tables where you must estimate the mean using midpoints. The formula is Σfx ÷ Σf, where x is the class midpoint.
真题经常给出分组频数表，要求利用组中点估计均值。公式为 Σfx ÷ Σf，其中 x 为组中点。
Be ready to explain why an estimate is used: you no longer have the original raw data, only class intervals. The true mean can differ from the estimated mean.
要准备好解释为什么使用估计值：原始数据已经不在，只有组区间，因此真实均值可能与估计均值不同。
The standard deviation is a measure of spread. A lower standard deviation means the data are more consistent, while a higher value means greater variability.
标准差是一种离散程度的度量。标准差越小，数据越稳定；标准差越大，数据波动越大。
When comparing two datasets, always quote both a measure of centre and a measure of spread, then make a comparative statement in context.
比较两个数据集时，一定要同时引用集中趋势和离散程度的度量，并结合背景作出比较性结论。
4. Probability Rules and Tree Diagrams | 概率规则与树状图
Tree diagrams are tested almost every session. Label each branch with its probability, and remember that probabilities on branches from the same point must sum to 1.
树状图几乎每一考季都会涉及。要标注每条分支的概率，并记住从同一点出发的分支概率之和必须为 1。
For sequential events, multiply along branches to find combined probabilities. If there are alternative paths, add those probabilities together.
对于连续事件，沿分支相乘得到组合概率。如果存在多种可能的路径，则将这些概率相加。
Conditional probability questions often use the wording ‘given that’. Use the formula P(A|B) = P(A and B) ÷ P(B), or use the restricted sample space method.
条件概率题常使用 ‘given that’ 的表述。使用公式 P(A|B) = P(A 且 B) ÷ P(B)，或者使用缩小样本空间的方法。
Without replacement means the probabilities change after the first selection. Many candidates forget to update the denominators on the second set of branches.
‘不放回’ 意味着第一次选择后概率发生改变。很多考生忘记在第二层分支更新分母。
5. Permutations and Combinations | 排列与组合
The key distinction is order. Use permutations nPr when order matters, and combinations nCr when order does not matter.
关键区别在于顺序。顺序重要时使用排列 nPr，顺序不重要时使用组合 nCr。
Past papers often include code or committee problems. For example, choosing 3 people from 10 is 10C3, but arranging 3 people in a line is 10P3.
真题常包含密码或委员会问题。例如，从 10 人中选 3 人用 10C3，而将 3 人排成一列用 10P3。
Check whether repetition is allowed. For codes, repetition may be allowed, so the number of arrangements is nr rather than nPr.
检查是否允许重复。对于密码，可能允许重复，因此排列数为 nr 而非 nPr。
Always simplify factorial expressions step by step. Calculators can evaluate nCr and nPr, but you must show the correct substitution to gain method marks.
始终逐步化简阶乘表达式。计算器可以计算 nCr 和 nPr，但你必须写出正确代入才能获得方法分。
6. Binomial Distribution | 二项分布
The binomial distribution applies to a fixed number of independent trials with only two outcomes and a constant probability of success p.
二项分布适用于固定次数的独立试验，每次只有两种结果，且成功概率 p 不变。
The probability formula is P(X = r) = nCr × pr × (1 − p)n−r. Use it when the question asks for exactly r successes.
概率公式为 P(X = r) = nCr × pr × (1 − p)n−r。当题目要求恰好 r 次成功时使用。
For ‘at least’ or ‘fewer than’ questions, add several individual probabilities or use cumulative tables if available.
对于 ‘至少’ 或 ‘少于’ 的题目，可将若干个单独概率相加，或使用累积概率表（如提供）。
Mean of a binomial distribution is np and variance is np(1−p). Past papers sometimes ask for these as a quick check before probability calculations.
二项分布的均值为 np，方差为 np(1−p)。真题有时会在概率计算前要求这些值作为快速检验。
7. Scatter Graphs and Correlation | 散点图与相关性
In scatter graph questions, describe correlation as positive, negative or none, and state its strength as strong, moderate or weak.
在散点图题中，要描述相关性为正相关、负相关或无相关，并说明强度为强、中等或弱。
A line of best fit must follow the trend, pass through the mean point (x̄, ȳ) and be drawn with roughly equal numbers of points above and below it.
最佳拟合线必须沿趋势方向，经过均值点 (x̄, ȳ)，并且线上方和下方的点数大致相等。
Use the line of best fit only within the range of the data. Predicting outside the range is extrapolation and is considered unreliable.
最佳拟合线仅应在数据范围内使用。超出该范围的预测属于外推，被认为不可靠。
The product-moment correlation coefficient r ranges from −1 to 1. Values closer to ±1 indicate stronger linear correlation; r = 0 indicates no linear correlation.
积矩相关系数 r 的范围是 −1 到 1。越接近 ±1 表示线性相关越强；r = 0 表示不存在线性相关。
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