Year 11 CAIE Statistics: Key Points for Experimental/Practical Assessment | CAIE 11年级统计：实验/实践考核要点

📚 Year 11 CAIE Statistics: Key Points for Experimental/Practical Assessment | CAIE 11年级统计：实验/实践考核要点

In the CAIE IGCSE Statistics examination, questions on experimental design and practical data handling carry significant weight. Whether you are asked to critique a survey method, plan a simple experiment, or interpret the results of a simulation, you need to demonstrate a clear understanding of statistical principles applied to real-world contexts. This article outlines the essential points you must master to excel in experimental and practical assessment tasks.

在 CAIE IGCSE 统计考试中，实验设计和实践数据处理题目占有重要分值。无论题目要求你评论一项调查方法、设计一个简单实验，还是解读模拟结果，你都需要清晰展现将统计原理应用于真实情境的能力。本文梳理了你必须掌握的关键要点，帮助你在实验和实践考核任务中取得优异成绩。

1. Understanding the Purpose of Statistical Experiments | 理解统计实验的目的

A statistical experiment is a process of collecting data to answer a specific question or investigate a hypothesis, such as ‘Does a new fertiliser increase plant growth?’ The aim must be clearly defined before any data collection begins, because it shapes the choice of method, sample size, and variables to be measured.

统计实验是为回答特定问题或研究假设（如’新肥料能否促进植物生长？’）而收集数据的过程。实验目的必须在数据收集前明确界定，因为它决定了方法选择、样本容量和需要测量的变量。

For practical assessment, you may be asked to state a suitable hypothesis in a null form (e.g., ‘There is no difference in mean height between plants given fertiliser A and those given no fertiliser’) or to identify the objective of a given study from its description.

在实践考核中，你可能需要以零假设形式（例如’施用肥料A与不施肥的植株平均高度没有差异’）陈述合适的假设，或根据描述识别给定研究的目标。

2. Designing a Statistical Experiment | 设计统计实验

Design covers the overall plan: what to measure, how to control extraneous factors, and how to allocate treatments. A key principle is randomisation — assigning subjects or items to different groups by chance to reduce bias. For example, in a drug trial, patients are randomly allocated to the treatment group or the placebo group.

设计涵盖整体方案：测量什么、如何控制无关因素、如何分配处理。一个关键原则是随机化——通过随机方式将受试对象分配到不同组别，以减少偏差。例如，在药物试验中，患者被随机分配到治疗组或安慰剂组。

You should know how to use random number tables or a simple lottery method to achieve randomisation. Also understand the importance of replication — using enough subjects or repeated measurements to obtain reliable results. The design must be practical: if you propose measuring the height of every tree in a forest, it may be impossible; a sample of 50 trees in selected plots is more realistic.

你应该知道如何使用随机数表或简单的抽签法实现随机化，还要理解重复的重要性——使用足够的受试对象或重复测量以获得可靠结果。设计必须切实可行：若你提议测量森林中每棵树的高度，可能无法实现；而选取样方内的50棵树则更切实际。

3. Sampling Techniques | 抽样技术

In many practical scenarios, you cannot measure the whole population, so you take a sample. CAIE expects you to describe and evaluate sampling methods:

在许多实践场景中，你无法测量整个总体，因此需要抽样。CAIE 期望你能描述并评价以下抽样方法：

Simple random sampling: every member has an equal chance of being selected. Use a random number generator or draw names from a hat.
简单随机抽样：每个成员被选中的概率相等。使用随机数生成器或从帽子中抽签。
Systematic sampling: choose a starting point at random, then select every kth member. Quick but can be biased if there is a hidden pattern.
系统抽样：随机选定起点，然后每隔k个选取一个成员。速度快，但如果存在隐藏模式可能产生偏差。
Stratified sampling: split the population into distinct groups (strata) and take a random sample from each in proportion to group size. Ensures representation of key subgroups.
分层抽样：将总体划分为不同的组（层），按各组大小比例从每层中抽取随机样本。确保关键子群的代表性。
Cluster sampling: divide the population into clusters (e.g., schools, villages), randomly select some clusters, and sample all or randomly within them. Cost-effective but can increase sampling error.
整群抽样：将总体划分为群（如学校、村庄），随机选取若干群，并对群内全部或随机抽样。节约成本但可能增大抽样误差。

You must be prepared to recommend a method and justify it based on cost, time, accuracy, and the nature of the population.

你必须能够根据成本、时间、精度和总体性质，推荐一种方法并给出理由。

4. Data Collection Methods | 数据收集方法

Data can be obtained by observation, survey, interview, or experiment. Each has strengths and limitations. Questionnaires must be worded neutrally to avoid leading questions — e.g., ‘Don’t you agree that the new park is beautiful?’ introduces bias. Closed questions give fixed choices, making analysis easier; open questions allow richer but harder-to-analyse responses.

数据可以通过观察、调查、访谈或实验获取。每种方法各有优缺点。问卷措辞必须中立，避免引导性问题——例如’难道你不同意新公园很美吗？’会引入偏差。封闭式问题提供固定选项，便于分析；开放式问题能得到更丰富的回答，但分析难度较大。

When collecting data manually, you need clear recording sheets or tables designed in advance. Include columns for date, time, treatment group, and the response variable. For an experiment, consider using a control group to isolate the effect of the treatment. Always state the units of measurement.

人工收集数据时，需要事先设计清晰的记录表或表格，包含日期、时间、处理组和响应变量等列。对于实验，应考虑使用对照组以隔离处理效应。务必注明测量单位。

5. Organizing and Recording Data | 组织和记录数据

After collection, raw data must be organised into a frequency table or a tally chart. Tally marks (||||, with the fifth crossing through) help count occurrences efficiently. For grouped continuous data, choose class intervals that are of equal width where possible, and specify boundaries clearly. No gaps should exist between intervals (e.g., 0 ≤ x < 10, 10 ≤ x < 20, etc.).

收集之后，原始数据必须整理成频数表或划记表。划记符号（||||，第五个划穿前面四个）有助于高效计数。对于分组连续数据，尽可能选择等宽的组距，并清楚说明边界。组与组之间不应有间隙（例如0 ≤ x < 10，10 ≤ x < 20等）。

Check for data entry errors: values that are impossibly large or negative given the context. A clean, well-labelled table is a basic requirement in practical answers. Marks are often awarded for correct tally and frequency columns, clear headings, and appropriate units.

检查是否存在数据录入错误：在给定情境下不可能出现极大值或负值。整洁明了、标注清晰的表格是实践答题的基本要求。正确划记和频数列、清晰的标题以及适当的单位常常能获得分数。

6. Descriptive Statistics and Summary Measures | 描述性统计和汇总指标

Once data are tabulated, compute appropriate summary statistics. For central tendency: the mean, median, and mode each have their uses. The mean uses all values but is affected by outliers; the median is robust; the mode is the most frequent category and is especially useful for qualitative data.

数据列表后，计算合适的汇总统计量。关于集中趋势：平均数、中位数和众数各有用途。平均数使用了所有数值，但受异常值影响；中位数具有稳健性；众数是最频繁出现的类别，对定性数据尤其有用。

Measures of spread include the range, interquartile range (IQR), and standard deviation. The range is simple but sensitive to a single extreme value. IQR (Q3 – Q1) describes the middle 50% of data and is less affected by outliers. For IGCSE, you are expected to calculate quartiles from a list or cumulative frequency graph. Variance and standard deviation measure average deviation from the mean; you must know both the formula and its application:

离散度量包括极差、四分位距（IQR）和标准差。极差简单，但对单个极端值敏感。IQR（Q3 – Q1）描述中间50%的数据，受异常值影响较小。IGCSE 要求你能从列表或累积频数图计算四分位数。方差和标准差衡量与平均数的平均偏差；你需要掌握公式及其应用：

σ = √[ Σ(x – x̄)² / n ] for a population; s = √[ Σ(x – x̄)² / (n – 1) ] for a sample.

总体标准差 σ = √[ Σ(x – x̄)² / n ]；样本标准差 s = √[ Σ(x – x̄)² / (n – 1) ]。

7. Graphical Representation of Data | 数据的图形表示

Selecting the right graph is crucial. Bar charts compare discrete categories; height shows frequency or value, and bars are separated. Histograms represent grouped continuous data, with area proportional to frequency — if class widths are unequal, calculate frequency density (frequency ÷ class width) for the vertical axis. Pie charts show proportions of a whole; angle = (category frequency / total) × 360°.

选择合适的图形至关重要。条形图比较离散类别，高度表示频数或数值，条形之间有空隙。直方图表示分组连续数据，面积与频数成正比——若组距不等，需用频数密度（频数 ÷ 组距）作为纵轴。饼图显示整体中各部分的比例；角度 = （类别频数 / 总数）× 360°。

Line graphs and scatter diagrams serve different purposes: line graphs show trends over time (time series), while scatter diagrams reveal correlation between two continuous variables. A cumulative frequency curve (ogive) is used to estimate medians, quartiles, and percentiles, and to construct box-and-whisker plots. Always label axes, provide a title, and use a sensible scale.

折线图和散点图各有用途：折线图展示随时间变化的趋势（时间序列），散点图揭示两个连续变量间的相关性。累积频数曲线（ogive）用于估算中位数、四分位数和百分位数，并用于绘制箱形图。务必为坐标轴添加标签、提供标题，并使用合理的刻度。

8. Probability Experiments and Simulations | 概率实验与模拟

Probability can be estimated by conducting an experiment: tossing coins, rolling dice, spinning spinners, or using random number tables to simulate events. The experimental probability is the relative frequency after a large number of trials:

概率可以通过实验来估计：抛硬币、掷骰子、转盘，或使用随机数表模拟事件。实验概率是在大量试验后的相对频率：

P(event) ≈ (number of favourable outcomes) / (total trials)

P(事件) ≈ （有利结果出现次数） / （总试验次数）

The more trials, the closer the experimental probability should be to the theoretical probability — this is the Law of Large Numbers. In an exam, you might be given results of a simulation and asked to estimate a probability, or you may have to design a simulation using random digits to model a real-world process (like estimating the number of rainy days). When designing, clearly state how a random number maps to an outcome.

试验次数越多，实验概率越接近理论概率——这是大数定律。考试中可能给出模拟结果并要求估计概率，或者你需要设计一个使用随机数字的模拟来为现实过程建模（如估计雨天数量）。设计时，要明确说明随机数字如何对应到结果。

9. Analyzing Experimental Outcomes | 分析实验结果

Analysis goes beyond calculation: you compare groups, identify patterns, and check for anomalies. For example, an outlier in a data set might be a data entry mistake or a genuine extreme. You should be able to describe the shape of a distribution (symmetrical, positively/negatively skewed) and link it to the context. Comparing two groups might involve comparing medians and interquartile ranges, or means and standard deviations, depending on skewness.

分析不仅限于计算：需要比较组别、识别模式并检查异常。例如，数据集中的异常值可能是录入错误，也可能是真实的极端情况。你应能描述分布的形状（对称、正偏斜/负偏斜），并与情境联系起来。比较两组数据时，可根据偏斜程度比较中位数和四分位距，或比较平均数和标准差。

Always support your conclusions with numerical evidence. Instead of saying ‘Group A performed better,’ say ‘The median score for Group A was 78, compared to 65 for Group B, indicating a higher typical performance.’ Use summary statistics to back up any claim.

始终用数值证据支持你的结论。不要说’组A表现更好’，而要说’组A的中位数得分为78，组B为65，表明典型表现更高’。用汇总统计量支撑任何论断。

10. Interpreting Results and Drawing Conclusions | 解释结果并得出结论

Interpreting means explaining what the statistics tell you in the context of the original problem. If an experiment tested whether a new teaching method improves test scores, the conclusion should relate directly to that hypothesis. Avoid overgeneralisation: a result observed in one class of 25 students cannot be assumed true for all students worldwide.

解释意味着在原始问题背景下说明统计量告诉你什么。如果一项实验检测了新的教学方法是否提高考试成绩，结论应直接与假设相关。避免过度泛化：在一个25人班级中观察到的结果，不能假定对全世界的学生都成立。

Consider the reliability of results: larger samples and smaller variability increase confidence. Mention any limitations — e.g., ‘The sample was taken from only one school, so results may not be representative.’ A good conclusion acknowledges uncertainty and suggests improvements for future studies.

考虑结果的可靠性：样本越大、变异性越小，置信度越高。提及任何局限性——例如’样本仅来自一所学校，因此结果可能不具代表性’。好的结论会承认不确定性，并对未来的研究提出改进建议。

11. Common Pitfalls and How to Avoid Them | 常见陷阱及避免方法

Bias in sampling: A voluntary response sample (e.g., an online poll) often over-represents people with strong opinions. Use probability-based methods instead.
抽样偏差：自愿回应样本（如在线投票）往往过度代表了持有强烈意见的人群。应改用基于概率的方法。
Poor question wording: Leading or ambiguous questions distort responses. Pilot test your questionnaire on a small group first.
问题措辞不当：引导性或模棱两可的问题会扭曲回答。先在小范围内试点测试问卷。
Ignoring confounding variables: In an experiment measuring the effect of a study technique on exam scores, if the treatment group also received extra tuition, you cannot isolate the effect of the technique. Control as many variables as possible.
忽视混杂变量：在测量学习方法对考试成绩影响的实验中，如果处理组还接受了额外辅导，你就无法分离出学习方法的效应。尽可能控制更多变量。
Inappropriate choice of average: Using the mean for skewed data can mislead; the median is more appropriate.
平均数选用不当：对偏斜数据使用平均数可能产生误导；使用中位数更合适。
Misleading graphs: Truncating the y-axis, using uneven intervals without adjusting area, or 3D effects that distort proportions. Always check that your graph is honest and clearly labelled.
误导性图表：截断y轴、组距不匀而又不调整面积，或使用扭曲比例的3D效果。务必检查图表是否真实且标注清晰。

By anticipating these common errors, you can both avoid them in your own work and critique others’ designs effectively.

通过预知这些常见错误，你可以在自己的工作中避免它们，也能有效地评价他人的设计。

12. Presenting Your Findings | 展示你的发现

Whether for a written report or a structured exam answer, clarity and logical flow matter. Start by restating the aim, summarise the methodology, present key results in tables and graphs, then provide the analysis and conclusion. Use precise language: ‘The results suggest’ rather than ‘It proves’, because statistical evidence is rarely absolute.

无论是撰写报告还是组织考试答案，清晰和逻辑流畅都很重要。先重述目的，概述方法，用表格和图表展示关键结果，然后提供分析和结论。使用精确的语言：’结果表明’而非’这证明了’，因为统计证据很少是绝对的。

In CAIE assessments, marks are often reserved for a final comment on the validity or limitations of the investigation. Address the sample size, potential bias, and whether the findings can be generalised. This shows evaluative thinking and fully satisfies the practical assessment criteria.

在 CAIE 考核中，最后对调查有效性或局限性的评论往往可获分数。谈及样本大小、潜在偏差以及研究结果是否可推广。这展现了评估性思维，能完全满足实践考核的标准。

Published by TutorHao | Statistics Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

Year 11 CAIE Statistics: Key Points for Experimental/Practical Assessment | CAIE 11年级统计：实验/实践考核要点