📚 AQA GCSE Statistics: Key Points for Experimental and Practical Assessments | AQA GCSE 统计：实验与实践考核要点
The experimental and practical components of AQA GCSE Statistics go beyond textbook theory, asking you to plan, collect, process and evaluate data in realistic contexts. Mastery of the statistical enquiry cycle, sound sampling, experimental design and critical reflection are all assessed. This guide distils the key points you need for success in practical assessments and exam questions that test these skills.
AQA GCSE 统计的实验与实践部分不仅要求掌握课本理论，更需要在真实情境中规划、收集、处理和评价数据。统计调查周期、合理的抽样、实验设计以及批判性反思都是考核重点。本指南提炼了您在实践评估以及考查这些技能的试题中取得成功所需的关键要点。
1. The Statistical Enquiry Cycle | 统计调查周期
al task expects you to move through these stages methodically and document your reasoning.
AQA 实践评估建立在统计调查周期 (SEC) 之上：问题 → 计划 → 数据 → 分析 → 结论。每项实践任务都要求你有条理地经过这些阶段并记录自己的推理过程。
State the hypothesis clearly: a null hypothesis (H₀) and an alternative hypothesis (H₁) should be defined before data collection.
清晰陈述假设：在收集数据之前应定义原假设 (H₀) 和备择假设 (H₁)。
Plan for variability: acknowledge that data will vary and decide how you will handle outliers, missing values and bias.
考虑变异性：承认数据会有变化，并决定如何处理异常值、缺失值和偏差。
After analysis, the conclusion must refer back to the original problem and hypothesis, stating whether the evidence supports H₁ at a given significance level.
分析之后，结论必须回扣最初的问题和假设，说明在给定的显著性水平下证据是否支持 H₁。
2. Planning Investigations | 调查规划
Before any data is gathered, a robust plan is essential. The plan must identify the population, the variables, data types and the methods of collection.
在收集任何数据之前，稳健的计划必不可少。计划必须明确总体、变量、数据类型和收集方法。
Define the population and sampling frame: e.g., 'all Year 11 students in my school'. The sampling frame is the list of individuals from which the sample is drawn.
定义总体和抽样框：例如“我所在学校所有 Year 11 学生”。抽样框是用于抽取样本的个体名单。
Identify variable types: qualitative (categorical) or quantitative (discrete/continuous). This affects which charts and summary statistics are appropriate.
识别变量类型：定性（分类）或定量（离散/连续）。这会影响哪些图表和汇总统计量是合适的。
Decide on data collection tools: questionnaire, experiment, observation or use of secondary data. Each tool must be justified for validity and reliability.
确定数据收集工具：问卷、实验、观察或使用二手数据。每种工具都必须说明其有效性和可靠性的理由。
Pilot studies should be mentioned to test the procedure and refine questions.
应提及预研究，以测试流程并优化问题。
3. Sampling Techniques | 抽样技术
Choosing the right sampling method is critical for representativeness and minimising bias. AQA expects you to know the advantages and disadvantages of each.
选择合适的抽样方法对于代表性和减少偏差至关重要。AQA 要求你了解每种方法的优缺点。
| Simple random sampling – every member has an equal chance; requires a sampling frame. | 简单随机抽样 – 每个成员有均等机会；需要抽样框。 |
| Stratified sampling – population divided into strata; sample proportional to stratum size; improves representativeness. | 分层抽样 – 总体划分为层；样本与层大小成比例；提高代表性。 |
| Systematic sampling – choose every kth individual; easy to implement but periodicity can introduce bias. | 系统抽样 – 选择每隔 k 个的个体；易实施，但周期性可能引入偏差。 |
| Cluster sampling – population naturally divided into clusters; randomly select whole clusters; useful when a full list is unavailable. | 整群抽样 – 总体自然分为群；随机选择整群；在无法获得完整名单时有用。 |
| Quota sampling – non-probability; interviewer selects fixed numbers in categories; quick but can be highly biased. | 配额抽样 – 非概率；访问者按类别选择固定数量；快速但可能高度偏差。 |
In practical write‑ups, always justify why a particular sampling method was chosen and discuss its limitations for your specific context.
在实践报告中，务必说明选择特定抽样方法的理由，并针对具体情境讨论其局限性。
4. Designing Experiments | 实验设计
When an investigation involves changing one variable to see its effect on another, it becomes an experiment. Control, randomisation and replication are the three pillars.
当调查涉及改变一个变量以观察其对另一个变量的影响时，就构成了实验。控制、随机化和重复是三大支柱。
Independent variable (IV) – what you deliberately change. Dependent variable (DV) – what you measure as a result. All other variables that could affect the DV must be controlled as much as possible.
自变量 (IV) – 你有意改变的变量。因变量 (DV) – 你测量的结果。所有可能影响因变量的其他变量必须尽可能控制。
Random allocation: assign subjects to treatment groups using a random mechanism to avoid selection bias. Use random number tables or software.
随机分配：使用随机机制将受试者分配到处理组，以避免选择偏差。使用随机数表或软件。
Replication: repeat the experiment several times or use a large enough sample to reduce the effect of random variation and increase precision.
重复：多次重复实验或使用足够大的样本，以减少随机变异的影响并提高精确度。
Common designs: completely randomised design, matched pairs (where subjects are paired on a relevant characteristic) and block design to reduce variability.
常见设计：完全随机设计、配对设计（根据相关特征将受试者配对）以及区组设计以减少变异性。
In practical assessments, describe your control group or baseline measurement explicitly.
在实践评估中，要明确描述控制组或基线测量。
5. Data Collection Tools | 数据收集工具
Questionnaires, interviews and observation sheets must be designed to produce valid, reliable and unbiased data. Question design is a subtle art.
问卷、访谈和观察表的设计必须能产生有效、可靠且无偏的数据。问题设计是一门微妙的艺术。
Closed questions provide quantitative data that is easy to analyse (e.g., multiple choice, rating scales 1–5). Open questions give rich qualitative data but are harder to summarise.
封闭式问题 提供易于分析的定量数据（如多选、1–5 评分量表）。开放式问题 提供丰富的定性数据，但难以总结。
Avoid leading questions (e.g., 'Don't you agree that…'), double-barrelled questions and ambiguous wording. Use a pilot study to spot these.
避免引导性问题（如“难道您不赞同……”）、双重问题和模糊措辞。通过预研究来发现这些问题。
For interviews, have a structured protocol; for observations, use clearly defined categories and record sheets to ensure inter‑observer reliability.
对于访谈，应有结构化提纲；对于观察，应使用明确定义的类别和记录表，以确保观察者间信度。
When using secondary data, always evaluate its source: is it trustworthy, recent and relevant?
使用二手数据时，始终评估其来源：是否可信、近期且相关？
6. Processing and Presenting Data | 数据处理与展示
Raw data must be cleaned, organised and summarised before meaningful analysis can begin. Choose graphical representations appropriate to data type.
原始数据必须经过清理、组织和汇总，才能开始有意义的分析。选择适合数据类型的图形表示。
For categorical data: bar charts, pie charts (but be careful with angles), pictograms and comparative component bar charts.
对于分类数据：条形图、饼图（注意角度计算）、象形图和对比复合条形图。
cy ÷ class width), cumulative frequency curves and box plots. Always label axes, include units and provide a title.
对于离散/连续数据：直方图（频率密度 = 频率 ÷ 组距）、累积频率曲线和箱线图。务必标注坐标轴、包含单位并提供标题。
Frequency density = Frequency / Class width
频率密度 = 频率 / 组距
median, mode, range, interquartile range (IQR) and standard deviation. Know that IQR = Q₃ – Q₁ and that outliers can be flagged using Q₁ – 1.5×IQR and Q₃ + 1.5×IQR.
汇总统计：平均数、中位数、众数、极差、四分位距 (IQR) 和标准差。掌握 IQR = Q₃ – Q₁，以及利用 Q₁ – 1.5×IQR 和 Q₃ + 1.5×IQR 标记异常值。
In experimental practicals, calculate effect sizes or percentage change to quantify differences between groups.
在实验实践中，计算效应量或百分比变化以量化组间差异。
7. Probability in Practice | 实践中的概率
Probability underpins risk assessment and experimental outcomes. You may need to estimate probabilities from data or apply probability rules.
概率是风险评估和实验结果的基础。您可能需要从数据中估计概率或应用概率规则。
P(Event) = Number of favourable outcomes / Total number of outcomes
P(事件) = 有利结果数 / 总结果数
When comparing experimental design, use the idea of expected frequencies under a null model. For example, in a chi‑squared test of independence (covered more in Year 11), you calculate expected values and compare with observed.
在比较实验设计时，利用零模型下的期望频率概念。例如，在独立性卡方检验（Year 11 深入涉及）中，您计算期望值并与观测值比较。
Relative frequency from a large number of trials provides an estimate of probability. Understand that the more trials, the closer the relative frequency tends to settle around the true probability (Law of Large Numbers).
大量试验得到的相对频率提供了概率的估计。要理解试验次数越多，相对频率倾向于稳定在真实概率附近（大数定律）。
Use tree diagrams and two‑way tables to organise outcomes and calculate combined probabilities in practical contexts.
使用树形图和双向表来组织结果，并在实践中计算组合概率。
8. Analysing and Interpreting Results | 分析与解释结果
Analysis goes beyond describing data; it seeks to answer the original question and test the hypothesis. Interpret measures of central tendency and spread in context.
分析不仅仅是描述数据；它旨在回答最初的问题并检验假设。要结合实际解释集中趋势和离散程度的度量。
When comparing two groups, use comparative box plots or histograms. Key comparisons: medians, IQRs, ranges and overall shape (skewness). Write sentences such as 'The median for group A was higher, suggesting a general tendency towards…'
比较两组时，使用对比箱线图或直方图。关键比较：中位数、IQR、极差和总体形状（偏度）。写出类似“A 组的中位数更高，表明总体上倾向于……”的句子。
Statistical significance: in practical work, you might informally assess overlap of box plots or differences in means relative to variation. Formal tests (e.g., t‑test, Mann–Whitney) may be referenced but the focus is on reasoning.
统计显著性：在实践工作中，您可以非正式地评估箱线图的重叠情况或均值差异相对于变异的大小。可能提及正式的检验（如 t 检验、Mann–Whitney），但重点是推理。
Always state whether the findings support the alternative hypothesis or if the null hypothesis cannot be rejected. Avoid 'prove'.
始终陈述结果是支持备择假设还是不能拒绝原假设。避免使用“证明”。
Comment on unexpected results and possible explanatory factors rather than ignoring them.
对意外结果和可能的解释性因素加以评论，而不是忽略它们。
9. Evaluating the Process | 评价过程
No practical investigation is perfect. AQA rewards honest, detailed evaluation that identifies limitations and suggests realistic improvements.
没有实践调查是完美的。AQA 奖励诚实、详细的评价，识别局限性并提出切实可行的改进建议。
Reliability: would repeating the procedure yield similar results? Discuss sample size, measurement consistency and control of extraneous variables.
信度：重复该流程会得到类似的结果吗？讨论样本大小、测量一致性以及无关变量的控制。
Validity: did you actually measure what you intended to? For questionnaires, check content validity; for experiments, check internal validity (was it really the IV causing the change?).
效度：您是否确实测量了想要测量的东西？对于问卷，检查内容效度；对于实验，检查内部效度（确实是 IV 导致了变化吗？）。
Identify sources of bias: selection bias, measurement bias, non‑response bias, observer effect. Suggest how each could be reduced next time.
识别偏差来源：选择偏差、测量偏差、无回应偏差、观察者效应。建议下次如何减少每种偏差。
Generalisation: to what population can the conclusions be reasonably extended? Discuss the representativeness of your sample.
推广：结论可以合理地推广到哪个总体？讨论样本的代表性。
Use specific evidence from your data to support evaluative points, e.g., 'Three outliers above the upper fence inflated the mean, so the median is a better summary.'
使用数据中的具体证据支持评价观点，例如“三个超过上须线的异常值夸大了平均数，因此中位数是更好的汇总指标”。
10. Ethical and Practical Considerations | 伦理与实际考量
Ethics matter in any data collection involving people. AQA expects awareness of GDPR principles, informed consent and confidentiality.
任何涉及人的数据收集中，伦理都很重要。AQA 期望你了解 GDPR 原则、知情同意和保密性。
Participants must give informed consent (or parental consent for under‑16s). They must know the purpose, what data is collected, and their right to withdraw without penalty.
参与者必须给予知情同意（16 岁以下需家长同意）。他们必须知道目的、收集什么数据，以及有权随时退出而不受惩罚。
Data must be anonymised and stored securely. When presenting results, do not identify individuals.
数据必须匿名并安全存储。在展示结果时，不要透露个人身份。
For experiments, avoid causing distress or harm. For sensitive topics, provide support information.
实验中避免造成困扰或伤害。对于敏感话题，提供支持信息。
Practical constraints are equally important: time, cost, access to subjects and equipment. Acknowledge these in your evaluation and explain how they affected your design choices.
实际限制同样重要：时间、成本、受试者和设备的获取。在评价中承认这些限制，并说明它们如何影响你的设计选择。
11. Common Mistakes in Practical Work | 实践工作中的常见错误
Mistakes can cost marks in both controlled assessments and exams. Recognising them is the first step to avoiding them.
错误在受控评估和考试中都会失分。认识错误是避免错误的第一步。
Mixing up correlation and causation: an observed correlation (e.g., ice‑cream sales and drowning) does not imply one causes the other. Always look for confounding variables.
混淆相关与因果：观察到的相关（如冰淇淋销量和溺水）并不意味着因果关系。始终寻找混杂变量。
Poorly drawn diagrams: axes not scaled properly, no labels, bars not touching in histograms where appropriate, lines of best fit that don't reflect the trend.
绘图不当：坐标轴标度不合适、没有标签、直方图中该接触的条形未接触、最佳拟合线不能反映趋势。
Using mean for skewed data: when the distribution is skewed, the median and IQR are more robust. Explain your choice.
对偏态数据使用平均数：当分布偏斜时，中位数和 IQR 更稳健。解释你的选择。
Ignoring sample size: small samples lead to high uncertainty. State that results are tentative and further research is needed.
忽略样本大小：小样本导致高度不确定性。声明结果是初步的，需要进一步研究。
Failing to refer back to the hypothesis in the conclusion, or making sweeping generalisations beyond the data.
在结论中没有回扣假设，或做出超出数据范围的过度推广。
12. Exam-Style Practical Questions | 考试中的实践题型
AQA exam papers include questions that simulate practical scenarios. They may provide a description of a flawed study and ask you to critique or redesign it.
AQA 试卷包含模拟实践情境的题目。可能给出一项有缺陷的研究描述，要求你进行评论或重新设计。
Typical command words: 'Evaluate the sampling method', 'Describe how you would improve the experiment', 'Identify sources of bias', 'Explain why the sample is not representative'.
典型指令词：“评价该抽样方法”、“描述你将如何改进实验”、“识别偏差来源”、“解释为什么样本不具有代表性”。
When answering, structure your response using the statistical enquiry cycle: identify the flaw, explain its impact, suggest a specific improvement, and note this improvement's effect on reliability/validity.
回答时，使用统计调查周期结构化您的答案：识别缺陷，解释其影响，提出具体的改进，并说明该改进对信度/效度的影响。
For questions that ask you to 'Design an investigation', provide a clear aim, hypothesis, detailed method including sampling, variables, controls, and a brief plan for analysis.
对于要求“设计一项调查”的题目，提供明确的目的、假设、详细方法（包括抽样、变量、控制）以及简要的分析计划。
Practise past paper questions that require you to draw conclusions from given output (e.g., box plots, summary statistics) and evaluate the design. Marks are awarded for statistical language (e.g., median, IQR, spread, skew) and contextual interpretation.
练习真题中要求从给定输出（如箱线图、汇总统计量）得出结论并评价设计的题目。分值会根据统计术语（如中位数、IQR、离散度、偏度）和情境解释给予。
Remember, even in an exam-only specification, these practical reasoning skills are tested and carry significant weight. Treat every data‑handling question like a mini‑investigation.
请记住，即使仅通过笔试考核，这些实践推理技能也会被测试，并占有相当大的权重。把每一道数据处理题都视为一次小型调查。
Published by TutorHao | Statistics Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
或截图后用「扫一扫 → 相册」选择本图