Year 11 Eduqas Statistics: Key Points for Practical Investigation | Year 11 Eduqas 统计：实验/实践考核要点

📚 Year 11 Eduqas Statistics: Key Points for Practical Investigation | Year 11 Eduqas 统计：实验/实践考核要点

Mastering the practical investigation component in Year 11 Eduqas Statistics means being able to design, carry out and critique a statistical enquiry from start to finish. This article gathers the essential assessment points—from planning and sampling to analysis and evaluation—so you can approach your controlled assessment or exam tasks with confidence.

掌握 Year 11 Eduqas 统计中的实验/实践考核部分，意味着你要能够从头到尾设计、实施并评判一项统计调查。本文汇总了从方案设计、抽样方法到数据分析与评估的所有重要考核要点，帮助你在受控评估或考试任务中自信应对。

1. The Statistical Enquiry Cycle (PPDAC) | 统计调查周期 (PPDAC)

The Eduqas practical investigation is structured around a cycle often referred to as PPDAC: Problem, Plan, Data, Analysis, Conclusion. You must show evidence of moving logically through each stage rather than jumping straight to calculations.

Eduqas 的实践调查围绕一个通常称为 PPDAC 的周期展开：问题、计划、数据、分析、结论。你必须展示出有逻辑地经历每一个阶段，而不是直接跳到计算环节。

A clear problem statement defines the population, the variables of interest and the purpose of the investigation. Without this, the rest of the enquiry lacks direction.

清晰的问题陈述会界定总体、感兴趣的变量以及调查目的。没有这一环节，后续的整个调查就缺乏方向。

The plan stage turns the problem into a practical strategy: choosing data sources, sampling methods and recording instruments. The data stage involves actually collecting and cleaning the data. Analysis is where you apply statistical techniques, and the conclusion interprets findings in context and reflects on limitations.

计划阶段将问题转化为可行的策略：选择数据来源、抽样方法和记录工具。数据阶段涉及实际收集和清理数据。分析阶段运用统计方法，结论则在具体情境中解读发现并反思局限性。

2. Formulating a Hypothesis and Research Question | 提出假设与研究问题

A well-written hypothesis states a prediction about the relationship between two variables, for example: ‘Students who spend more time on homework score higher in tests.’ The null hypothesis would be that there is no relationship, and the alternative hypothesis is what you aim to investigate.

一个撰写得当的假设会陈述对两个变量之间关系的预测，如：“花更多时间做作业的学生考试得分更高。” 原假设是两者没有关系，备择假设则是你要探究的目标。

Eduqas examiners expect you to distinguish between a research question (Is there a link between revision hours and exam results?) and a statistical hypothesis that can be tested with data. Always express hypotheses in terms of population parameters where possible, using symbols like μ for mean or ρ for correlation.

Eduqas 考官期望你区分研究问题（如：复习时间与考试成绩之间是否存在关联？）和可以用数据检验的统计假设。尽量用总体参数来表达假设，如用 μ 表示均值、ρ 表示相关系数。

Your hypothesis must be directional or non-directional as appropriate. A one-tailed hypothesis (e.g. ‘higher caffeine intake raises heart rate’) is acceptable when prior research supports the direction; otherwise use a two-tailed version.

你的假设必须根据情况设为定向或非定向。当先前研究支持某个方向时，可以使用单尾假设（如“更高的咖啡因摄入会使心率升高”），否则应使用双尾形式。

3. Planning Data Collection: Primary vs Secondary | 规划数据收集：一手数据与二手数据

Primary data is gathered by the investigator for the specific purpose of the enquiry, such as conducting a traffic survey or measuring plant growth. Secondary data relies on existing sources like government statistics, published tables or online datasets.

一手数据由调查者为了特定的调查目的亲自收集，如进行交通流量调查或测量植物生长。二手数据则依赖现有资料，如政府统计、已发布的表格或在线数据集。

In the plan you must justify your choice between primary and secondary data. Primary data gives control over accuracy and relevance, but takes time; secondary data is quicker but may come with biases or missing values you have to handle.

在计划中你必须为一手数据或二手数据的选择提供理由。一手数据能让你控制准确性和相关性，但耗费时间；二手数据更快，但可能包含偏差或缺失值，需要你进行处理。

Make clear how each variable will be measured, the units of measurement and the instrument (stopwatch, questionnaire, ruler). For secondary data, cite the source and comment on its reliability and validity.

要明确说明每个变量将如何测量、使用的测量单位以及工具（秒表、问卷、尺子）。如果是二手数据，要注明来源，并评述其可靠性和效度。

4. Sampling Methods: Random, Stratified, Systematic and More | 抽样方法：随机、分层、系统及其他

Selecting an adequate sample is a critical assessment objective. You need to know simple random sampling, stratified sampling, systematic sampling, cluster sampling and quota sampling, along with their strengths and weaknesses for different tasks.

选择恰当的样本是一项关键的考核目标。你需要了解简单随机抽样、分层抽样、系统抽样、整群抽样和配额抽样，以及在面对不同任务时各自的优缺点。

Simple random sampling: every member of the population has an equal chance of being selected. It avoids bias but needs a complete sampling frame.

简单随机抽样：总体中每个成员都有同等的被选中机会。它避免偏差，但需要一个完整的抽样框。
Stratified sampling: the population is divided into groups (strata), and a random sample is taken from each stratum. It ensures representation of key subgroups.

分层抽样：将总体划分为若干层，然后从每一层中随机抽样。这能确保关键子群体的代表性。
Systematic sampling: choose every nth item from a list. It is easy to implement but may introduce periodicity bias if a hidden pattern exists.

系统抽样：从列表中每隔 n 个选取一个。操作简便，但如果存在隐藏的模式，可能会引入周期性偏差。

Cluster sampling is useful when the population is naturally divided into clusters, such as schools or postcode areas; you randomly select whole clusters. Quota sampling is a non-random method where interviewers fill set numbers for each category—quick but prone to interviewer bias.

整群抽样适用于总体天然地分为一些群组（如学校或邮政编码区域）的情况；你随机选择整个群组。配额抽样是一种非随机方法，由调查员按每个类别填满定额——很快，但容易受到调查员偏差的影响。

In your practical write-up, always state the sampling method, the sampling frame used, sample size and how you dealt with potential non-response.

在实践报告中，务必陈述所用抽样方法、抽样框、样本量以及你如何处理可能出现的无应答。

5. Designing Experiments: Control, Randomisation, Replication | 设计实验：控制、随机化、重复

When the investigation is an experiment, you must show understanding of the three principles of experimental design: control, randomisation and replication. These are the bedrock of collecting valid data and making fair comparisons.

当调查本身是一项实验时，你必须展现对实验设计三大原则的理解：控制、随机化和重复。这是收集有效数据并进行公平比较的基石。

Control means keeping extraneous variables constant so that any observed effect can be attributed to the explanatory variable. Identify potential confounding variables (e.g. temperature, time of day) and explain how you will control them.

控制意味着保持外来变量恒定，这样观察到的任何效应都可以归因于解释变量。要识别潜在的混杂变量（如温度、时间段），并说明你将如何控制它们。

Randomisation is the process of allocating subjects to treatment groups by chance. This minimises selection bias and helps balance out unknown confounding factors. Use a random number generator or drawing lots, and document the method.

随机化是指通过随机方式将受试者分配到不同处理组。这能最大程度减少选择偏差，并帮助平衡未知的混杂因素。使用随机数生成器或抽签，并记录下方法。

Replication means having a sufficient number of repeated measurements or experimental units. A single plant or one trial is not enough; replication allows you to estimate experimental error and increases the reliability of conclusions.

重复意味着有足够数量的重复测量或实验单元。单单一株植物或一次试验是不够的；重复让你能够估计实验误差，并提高结论的可靠性。

For a practical exam scenario, you might be asked to sketch an experimental layout, such as a completely randomised design or a randomised block design. Be prepared to explain why blocking is important when there is a known source of variation, like different soil types in a field experiment.

在实践考试情境中，可能会要求你草拟一个实验布局，如完全随机设计或随机区组设计。要准备好解释为什么当存在已知变异来源时（如田间试验中的不同土壤类型）区组化是重要的。

6. Questionnaire Design and Avoiding Bias | 问卷设计与避免偏差

A badly designed questionnaire can ruin a practical investigation. Assessment points focus on wording clarity, avoiding leading questions, using a suitable response format (open vs closed) and piloting the instrument.

设计糟糕的问卷会毁掉一次实践调查。考核要点聚焦于措辞清晰、避免诱导性问题、使用合适的回答格式（开放式 vs 封闭式）以及预测试工具。

Closed questions provide tick boxes or scales (Likert: 1–5), making data easy to analyse. Open questions give richer detail but require more time to categorise.

封闭式问题提供复选框或量表（李克特量表：1–5），使数据易于分析。开放式问题能获取更丰富的细节，但需要更多时间进行分类。
Leading questions such as ‘Don’t you agree that homework is stressful?’ should be replaced by neutral wording: ‘How stressful do you find homework on a scale of 1 to 5?’

像“难道你不认为作业很有压力吗？”这样的诱导性问题应该改为中性措辞：“在1到5的量表上，你觉得作业有多大压力？”
Avoid double-barrelled questions that ask about two things at once: ‘How satisfied are you with the price and quality?’ should be split into two separate items.

避免同时问两件事的双管问题：“你对价格和质量的满意度如何？” 应拆分为两个独立题目。

Piloting means testing the questionnaire on a small group to spot unclear instructions or ambiguous wording. Always mention piloting in your evaluation to show good practice.

预测试即在小组中测试问卷，以发现不清晰的说明或模糊的措辞。在你的评估中务必提到预测试，以展示良好实践。

7. Types of Data: Quantitative, Qualitative, Discrete, Continuous | 数据类型：定量、定性、离散、连续

Correctly classifying data is often tested in the practical context. Quantitative data are numerical and can be discrete (counts, like number of siblings) or continuous (measurements, like height in cm). Qualitative data are categorical and non-numerical, such as eye colour or mode of transport.

正确分类数据类型经常在实践情境中受到考查。定量数据是数值型的，可以是离散的（如兄弟姐妹数量的计数）或连续的（如身高的测量值，单位厘米）。定性数据是类别性的且非数值型，如眼睛颜色或交通方式。

Your choice of data type influences the statistical tools you can use later: you would not calculate a mean for ordinal qualitative data (e.g. satisfaction ranks). Knowing the level of measurement—nominal, ordinal, interval or ratio—helps you select the right graphs and summary statistics.

你选择的数据类型会影响之后可用的统计工具：你不会为有序定性数据（如满意度排名）计算平均数。了解测量层级——名义、顺序、等距或等比——有助于你选择合适的图表和汇总统计量。

In the plan, clearly state the data type for each variable and justify why you will treat it as discrete or continuous. For grouped data, explain the reason for grouping, such as easier handling of large datasets or confidentiality.

在计划中，要清楚说明每个变量的数据类型，并解释为何将其视为离散或连续。对于分组数据，要解释分组的原因，比如为了更方便处理大型数据集或为了保密。

8. Collecting Data Reliably and Ethically | 可靠且合乎伦理地收集数据

Reliability means the data should be consistent if the investigation is repeated under similar conditions. You can improve reliability by standardising procedures, using calibrated instruments and taking repeated measurements to identify outliers.

可靠性是指如果在相似条件下重复调查，数据应当一致。你可以通过标准化操作、使用校准仪器以及进行重复测量以识别异常值来提高可靠性。

Validity refers to whether you are measuring what you intend to measure. A heart-rate monitor is a valid instrument for measuring pulse; asking ‘How fit do you feel?’ is not. Include checks such as recording the same observation twice and comparing.

效度指的是你测量的是否正是你想要测量的东西。心率监测器对于测量脉搏是有效度的工具；询问“你感觉自己多健康？”则不是。纳入一些检查，例如把同一个观察值记录两次并作比较。

Ethical data collection is a specific Eduqas requirement. For any human participants, you must obtain informed consent, ensure anonymity, allow the right to withdraw and avoid causing distress. Mention how you met these in the method section.

合乎伦理地收集数据是 Eduqas 的明确要求。对于任何人类参与者，你必须获得知情同意、确保匿名、允许退出权利，并避免造成困扰。在方法部分要说明你是如何满足这些要求的。

When using secondary data from websites or official sources, ethical practice involves acknowledging the source clearly and not misrepresenting the data.

当使用来自网站或官方来源的二手数据时，合乎伦理的做法包括清楚注明出处，并且不歪曲数据。

9. Processing and Representing Data | 处理与表示数据

Once data are collected, the processing stage involves cleaning, sorting and, where appropriate, recoding. Remove obvious errors, flag missing values and decide whether to exclude or impute them. All decisions must be transparent.

数据收集之后，处理阶段包括清理、排序，并在适当时候进行重新编码。移除明显的错误，标记缺失值，并决定是排除还是插补它们。所有决定都必须透明。

Choose graphical representations to match the data type: bar charts for categorical data, histograms for continuous grouped data, cumulative frequency curves for medians and quartiles, scatter diagrams for bivariate relationships and box plots for comparing distributions.

选择与数据类型相匹配的图形表示：条形图用于类别数据，直方图用于连续的分组数据，累积频率曲线用于中位数和四分位数，散点图用于双变量关系，箱线图用于比较分布。

For practical assessments, you must label axes clearly, use a sensible scale and give each chart a title. A common fault is plotting class intervals as discrete bars in a histogram without touching bars—remember histogram bars must touch for continuous data.

在实践评估中，你必须清晰地标注坐标轴，使用合宜的刻度，并为每张图表加上标题。一个常见错误是在直方图中将组距画成分离的条形——记住，对于连续数据，直方图的条形必须相连。

When summarising data in a table, include clear headings with units. For grouped frequency tables, you will need columns for class boundaries, midpoints, frequency and sometimes cumulative frequency.

当在表格中汇总数据时，要包含带单位的清晰表头。对于分组频数表，你需要有组界、组中值、频数等列，有时还需要累积频数列。

10. Analysing Data: Averages, Spread, Correlation | 分析数据：平均值、离散度、相关性

Analysis is not just running numbers—it must be linked back to the hypothesis. Begin with measures of central tendency: mean, median and mode. Know when each is most appropriate; the median is often better for skewed data or when outliers are present.

分析不仅仅是运算数字——它必须与假设联系起来。先从集中趋势的测量开始：平均数、中位数和众数。要知道每种测量在何时最为合适；对于偏态数据或存在异常值时，中位数通常更佳。

Measures of spread include range, interquartile range (IQR) and standard deviation. The IQR is found as Q₃ − Q₁ and pairs naturally with the median. Standard deviation measures the average distance from the mean and is used with normally distributed data.

离散度的测量包括极差、四分位距（IQR）和标准差。四分位距由 Q₃ − Q₁ 算出，与中位数自然配对。标准差测量的是各数据点与平均数的平均距离，常用于正态分布数据。

For bivariate data, calculate a correlation coefficient such as Spearman’s rank or Pearson’s r, but only after drawing a scatter diagram to check the relationship is linear and to spot outliers. Remember: correlation does not imply causation.

对于双变量数据，要计算相关系数，如斯皮尔曼等级相关系数或皮尔逊 r，但前提是先绘制散点图以检查关系是否为线性并发现异常值。记住：相关性并不意味着因果性。

When comparing two data sets, side-by-side box plots or back-to-back stem-and-leaf diagrams are effective. Always comment on both the average and the spread—not just ‘Group A is higher’. Say, for example, ‘The median of Group A is 12 cm higher than Group B, and its IQR is smaller, indicating more consistency.’

当比较两个数据集时，并列箱线图或背靠背茎叶图非常有效。务必同时评论平均水平和离散度——而不仅仅是“A组更高”。例如可以这样说：“A组中位数比B组高12厘米，且其四分位距更小，表明更为一致。”

11. Drawing Conclusions and Evaluating the Investigation | 得出结论与评估调查

A solid conclusion states whether the evidence supports the hypothesis, refers back to the original context and avoids overclaiming. Use phrases like ‘The data suggest…’ rather than ‘This proves…’

稳健的结论要说明证据是否支持假设，回扣到原始情境，并且避免过度断言。使用“数据表明……”这样的表述，而不是“这证明了……”。

Evaluate the whole investigation: discuss limitations in the sampling method, possible measurement errors, any uncontrolled variables and the impact of sample size. Distinguish between random errors (which can be averaged out) and systematic errors (which bias all results in one direction).

对整个调查进行评估：讨论抽样方法的局限性、可能的测量误差、任何未控制的变量以及样本量的影响。要区分随机误差（可以通过取平均值消除）和系统误差（使所有结果向一个方向偏倚）。

Suggest specific improvements rather than vague statements like ‘I would do it better next time.’ For example: ‘Using a stratified sample based on year group would increase representativeness,’ or ‘Measuring reaction time with computer software would reduce human reaction bias.’

要提出具体的改进建议，而不是笼统的“下次我会做得更好”。例如：“采用按年级分层的抽样方法可以提高代表性”，或者“使用计算机软件测量反应时间可减少人为反应偏差”。

Acknowledge any ethical issues that arose during data collection and how you resolved them. This demonstrates reflective practice, which is highly valued in the Eduqas marking scheme.

诚实地提及数据收集过程中出现的任何伦理问题以及你是如何解决的。这体现了反思性实践，在 Eduqas 评分方案中极受重视。

12. Common Pitfalls in Practical Statistics | 实验统计中的常见陷阱

Many students lose marks by confusing the hypothesis with a method description, mislabelling graphs, or calculating statistics without explaining their meaning. Reflection and commentary are as important as numerical accuracy.

许多学生由于把假设与方法描述混为一谈、图表标注错误，或只计算统计量而不解释其含义而丢分。反思和评论与数值准确性同等重要。

Other frequent errors include: using a sample that is too small, failing to identify outliers and their effect, presenting percentages without the raw frequencies, and over-reliance on pie charts when several categories make them unreadable.

其他常见错误包括：样本量太小，未能识别异常值及其影响，只呈现百分比而省略原始频数，以及在类别较多时过度依赖饼图导致其不易阅读。

When interpreting a scatter diagram with correlation, students sometimes forget to comment on the strength, direction and linearity of the association, or they fail to mention possible underlying variables. Always connect statistical findings back to the real-world problem you set out to explore.

在解读带有相关性的散点图时，学生有时会忘记评论关联的强度、方向和线性程度，或者未提及可能存在的潜在变量。务必把统计发现与你一开始要探究的现实世界问题联系起来。

Finally, in timed practical assessments, manage your time so that every stage of the PPDAC cycle is visible. An excellent analysis that lacks a plan or an evaluation cannot access the highest mark bands.

最后，在限时的实践评估中，要管理好时间，使 PPDAC 周期的每个阶段都有所体现。一份缺少计划或评估的出色分析无法触及最高分数段。

Published by TutorHao | Statistics Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

Year 11 Eduqas Statistics: Key Points for Practical Investigation | Year 11 Eduqas 统计：实验/实践考核要点