📚 IGCSE WJEC Statistics: Practical Investigation and Assessment Essentials | IGCSE WJEC 统计：实验与实践考核要点
In IGCSE WJEC Statistics, the practical investigation is not just about doing calculations. It tests your ability to plan a statistical enquiry, collect suitable data, process it accurately, present findings clearly, and evaluate the whole process critically. Examiners want to see evidence of statistical reasoning from the first hypothesis to the final conclusion.
在 IGCSE WJEC 统计中，实践调查不只是做计算。它考查你规划统计调查、收集合适数据、准确处理数据、清晰展示结果以及批判性评估整个过程的能力。考官希望看到从最初假设到最后结论的完整统计推理证据。
This article covers the key assessment points for practical and experimental tasks in WJEC Statistics. Use it as a checklist when planning, carrying out and writing up any investigation.
本文涵盖 WJEC 统计实践与实验任务的关键考核要点。在规划、实施和撰写任何调查报告时，都可以把它当作自查清单。
1. Understanding WJEC Practical Assessment Aims | 理解 WJEC 实践考核目标
The practical component of WJEC Statistics assesses how well you can apply statistical techniques to a real or realistic problem. You are expected to identify a question, form a hypothesis, collect and process data, and draw conclusions based on evidence.
WJEC 统计的实践部分考查你把统计技术应用到真实或现实问题中的能力。要求你明确问题、提出假设、收集并处理数据，并根据证据得出结论。
Assessment focuses on four main areas: planning, data collection, processing and presentation, and interpretation and evaluation. Each section of your report should show clear links back to the original hypothesis.
考核主要关注四个方面：规划、数据收集、处理与展示、解释与评估。报告的每一部分都应与原始假设清晰关联。
Marks are usually awarded for method as well as final answers. Showing correct working, labelling diagrams, and explaining decisions are just as important as getting a number right.
分数通常既给方法分也给最终答案分。展示正确过程、标注图表、解释决策与算对数字同样重要。
2. Planning a Statistical Investigation | 规划统计调查
Every practical investigation should begin with a clear aim and a testable hypothesis. A hypothesis is a statement that can be supported or refuted by data, such as 'Students who revise longer score higher marks'.
每项实践调查都应以明确的目标和可检验的假设开始。假设是一个可以被数据支持或反驳的陈述，例如 '复习时间较长的学生得分更高'。
Define the population and variables precisely. State whether the variable is categorical, discrete or continuous, because this affects which charts and calculations are appropriate.
要准确定义总体和变量。说明变量是分类变量、离散变量还是连续变量，因为这会影响适用的图表和计算方法。
A good plan includes sample size, data collection method, ethical considerations, and how you will reduce bias. Planning errors are often more damaging than calculation errors.
一个好的计划包括样本量、数据收集方法、伦理考虑以及如何减少偏差。规划错误往往比计算错误影响更大。
on plan: aim → hypothesis → variables → sample → data collection → processing → conclusion
良好调查计划：目标 → 假设 → 变量 → 样本 → 数据收集 → 处理 → 结论
3. Choosing Sampling Methods | 选择抽样方法
Choosing the right sampling method is essential for reliable conclusions. Random sampling gives every member of the population an equal chance of being selected, reducing selection bias.
选择合适的抽样方法对可靠结论至关重要。随机抽样使总体中每个成员都有相同被选中的机会，从而减少选择偏差。
Stratified sampling divides the population into groups and samples proportionally from each group. It is useful when the population has distinct subgroups, such as year groups or gender.
分层抽样将总体分组，并按比例从每组中抽样。当总体有明显的子群体（如年级或性别）时非常有用。
Systematic sampling selects every nth item from a list, while cluster sampling selects whole groups. Convenience sampling is easy but often unrepresentative, so it should be avoided or clearly justified.
系统抽样从列表中每隔 n 个选取一个，整群抽样则选择整个群体。便利抽样容易操作但往往不具代表性，因此应避免使用或给出明确理由。
| Method 方法 | Strength 优点 | Weakness 缺点 |
|---|---|---|
| Simple random 简单随机 | Unbiased 无偏 | Needs full population list 需要完整名单 |
| Stratified 分层 | Represents subgroups 代表子群体 | More complex to organise 组织更复杂 |
| Systematic 系统 | Simple to use 使用简单 | Can miss patterns 可能遗漏规律 |
| Convenience 便利 | Quick and cheap 快速便宜 | High bias risk 偏差风险高 |
4. Designing Data Collection Tools | 设计数据收集工具
Questionnaires, observation sheets and recording tables must be designed to produce accurate, usable data. Questions should be clear, specific and free from leading language.
问卷、观察表和记录表的设计必须能产生准确、可用的数据。问题应清晰、具体，且不含引导性语言。
Avoid overlapping response boxes, vague options such as 'often' without a time frame, and double-barrelled questions that ask two things at once.
避免响应框重叠、没有时间范围的模糊选项（如 '经常'），以及同时问两件事的双重问题。
Pilot your data collection tool with a small group before full use. Pilot studies help identify confusing wording, missing categories and practical problems.
在全面使用前，先用小群体试测数据收集工具。试测研究有助于发现措辞混淆、缺少类别和实际操作问题。
- Use closed questions for easy analysis: 'How many hours do you sleep? 0-5, 6-7, 8-9, 10+'
- 使用封闭式问题便于分析：'你睡多少小时？0-5, 6-7, 8-9, 10+'
- Use open questions sparingly when detailed opinions are needed.
- 当需要详细意见时，谨慎使用开放式问题。
5. Carrying Out Experiments and Simulations | 开展实验与模拟
In WJEC Statistics, experiments may involve physical trials such as rolling dice, measuring reaction times, or comparing two conditions. Simulations use random numbers or technology to model real processes.
在 WJEC 统计中，实验可能涉及物理试验，如掷骰子、测量反应时间或比较两种条件。模拟则使用随机数或技术来建模真实过程。
Control variables that are not being tested so that any change in the outcome can be attributed to the explanatory variable. Repeating trials increases reliability by reducing random variation.
控制未测试的变量，使结果变化能够归因于解释变量。重复试验可以减少随机变异，提高可靠性。
For simulation, describe the model clearly. State how random numbers are generated and how they map to outcomes, for example using 00-99 to represent a 30% success rate.
对于模拟，要清楚描述模型。说明如何生成随机数以及如何映射到结果，例如用 00-99 表示 30% 的成功率。
Number of trials n ↑ → variability of estimates ↓ → more reliable conclusions
试验次数 n 增加 → 估计的变异性减小 → 结论更可靠
6. Processing and Cleaning Raw Data | 处理与清洗原始数据
Raw data usually contains errors, missing values and outliers. Cleaning data means checking for impossible values, correcting entry mistakes, and deciding how to handle gaps.
原始数据通常包含错误、缺失值和异常值。清洗数据意味着检查不可能的值、纠正录入错误，并决定如何处理空缺。
Missing data can be omitted if random and small in number, but this should be stated. If a value is clearly wrong, such as a height of 2500 cm, it should be corrected or excluded with justification.
缺失数据如果随机且数量少，可以删除，但应说明。如果数值明显错误，如身高 2500 厘米，应更正或排除并说明理由。
× IQR is an outlier. Investigate outliers before deciding whether to keep or remove them.
异常值应使用四分位距规则识别：低于 Q1 − 1.5 × IQR 或高于 Q3 + 1.5 × IQR 的值为异常值。在决定保留还是删除之前，应调查异常值。
IQR = Q3 − Q1
四分位距 = Q3 − Q1
7. Presenting Data Graphically | 图形化展示数据
Choose the correct chart for each data type. Bar charts are for categorical data, histograms for continuous data, and pie charts for proportions of a whole.
为每种数据类型选择正确的图表。条形图用于分类数据，直方图用于连续数据，饼图用于整体的比例。
For comparing two distributions, use comparative box plots or back-to-back stem-and-leaf diagrams. For bivariate numerical data, use scatter graphs to show correlation.
要比较两个分布，可使用对比箱线图或背靠背茎叶图。对于双变量数值数据，使用散点图展示相关性。
Every graph must have a title, labelled axes with units, and a key if more than one series is shown. Scale should be linear and appropriate; do not start a frequency axis at a misleading value without good reason.
每张图必须有标题、带单位的坐标轴标签，如果有多组数据则需要图例。刻度应线性且合适；没有充分理由，不应从误导性数值开始频率轴。
In practical work, charts can be drawn by hand or produced using software. In either case, accuracy and clarity are more important than decoration.
在实践工作中，图表可以手绘或用软件生成。无论哪种方式，准确和清晰比装饰更重要。
8. Calculating Summary Statistics | 计算汇总统计量
Summary statistics condense a dataset into a few key numbers. Measures of location include mean, median and mode; measures of spread include range, interquartile range and standard deviation.
汇总统计量将数据集浓缩为几个关键数字。位置度量包括平均数、中位数和众数；离散度量包括极差、四分位距和标准差。
Use the median and IQR when data is skewed or contains outliers, because they are resistant. Use the mean and standard deviation when data is roughly symmetric and outliers are not a problem.
当数据偏斜或包含异常值时，使用中位数和四分位距，因为它们具有抗性。当数据大致对称且异常值不成问题时，使用平均数和标准差。
Sample mean: x̄ = Σx / n
样本平均数：x̄ = Σx / n
Sample standard deviation: s = √[Σ(x − x̄)² / (n − 1)]
样本标准差：s = √[Σ(x − x̄)² / (n − 1)]
Always show your working. For grouped data, use midpoints and clearly state that results are estimates.
始终展示计算过程。对于分组数据，使用组中值，并明确说明结果是估计值。
9. Interpreting Results in Context | 结合实际解释结果
A calculated value means nothing until it is interpreted in the context of the original problem. For example, a median of 6 hours sleep means something different for adults than for teenagers.
计算值只有在结合原始问题背景解释时才有意义。例如，中位数 6 小时睡眠对成年人和青少年的含义不同。
Compare your results with the hypothesis. Does the data support or refute it? Avoid saying 'prove' because statistical evidence is never absolute proof.
将结果与假设进行比较。数据是支持还是反驳假设？避免说 '证明'，因为统计证据从来不是绝对证明。
Use comparative language such as 'higher than', 'more spread out', or 'positively correlated'. Link every conclusion back to the variables and population being studied.
使用比较性语言，如 '高于'、'更分散' 或 '正相关'。将每个结论都联系回所研究的变量和总体。
Interpretation: statistic → context → comparison → conclusion
解释路径：统计量 → 背景 → 比较 → 结论
10. Evaluating Reliability and Validity | 评估可靠性与有效性
Reliability means whether results would be similar if the investigation were repeated. Larger samples and repeated trials usually increase reliability.
可靠性指如果重复调查，结果是否相似。更大的样本和重复试验通常会提高可靠性。
Validity means whether the investigation actually measures what it claims to measure. A biased sample, poorly worded question or uncontrolled variable can reduce validity.
有效性指调查是否真正测量了它声称要测量的内容。有偏样本、措辞不当的问题或未控制的变量都会降低有效性。
Discuss sources of bias, such as non-response, self-selection and measurement error. Suggest realistic improvements, not just 'use a bigger sample'.
讨论偏差来源，如无回应、自选择和测量误差。提出切合实际的改进意见，而不只是 '使用更大样本'。
Every practical investigation should end with an evaluation paragraph that honestly addresses limitations and how they might affect conclusions.
每项实践调查都应以评估段落结束，诚实地说明局限性及其对结论的可能影响。
11. Writing the Investigation Report | 撰写调查报告
The report should follow a logical structure: introduction, methodology, results, analysis, conclusion and evaluation. Use headings and subheadings to guide the reader.
报告应遵循逻辑结构：引言、方法、结果、分析、结论和评估。使用标题和小标题引导读者。
Write in clear, formal English and avoid informal phrases. Present raw data in an appendix if it is large, but include summary tables and key graphs in the main body.
使用清晰、正式的英语，避免口语化表达。如果原始数据较多，可放在附录中，但主要部分应包含汇总表和关键图表。
Number all tables and graphs, and refer to them in the text, such as 'Table 2 shows the time taken'. Every visual should have a purpose.
给所有表格和图表编号，并在正文中引用，如 '表 2 显示了所用时间'。每个图表都应有其目的。
- Introduction: aim, hypothesis, variables 引言：目标、假设、变量
- Method: sample, data collection, controls 方法：样本、数据收集、控制变量
- Results: tables, graphs, summary statistics 结果：表格、图表、汇总统计
- Analysis: interpretation in context 分析：结合实际解释
- Conclusion and evaluation: answer hypothesis, discuss limitations 结论与评估：回答假设、讨论局限性
12. Common Errors and Exam Tips | 常见错误与考试提示
One common error is choosing a graph that does not match the data type, such as using a pie chart for continuous data. Always check whether variables are categorical or numerical.
一个常见错误是选择了与数据类型不匹配的图表，例如对连续数据使用饼图。务必检查变量是分类变量还是数值变量。
Another error is ignoring units on axes and in calculations. Include units in every table heading and graph label, and state them in conclusions.
另一个错误是忽略坐标轴和计算中的单位。在每个表头和图表标签中都应包含单位，并在结论中说明单位。
Do not confuse correlation with causation. A scatter graph may show that two variables are linked, but that does not mean one causes the other.
不要混淆相关与因果。散点图可能显示两个变量有关联，但这并不意味着一个导致另一个。
Finally, manage time carefully in practical assessments. Spend enough time on planning and evaluation, not just on drawing graphs and calculating statistics.
最后，在实践考核中合理管理时间。需要在规划和评估上投入足够时间，而不只是绘图和计算统计量。
| Common error 常见错误 | How to avoid 如何避免 |
|---|---|
| Wrong chart type 图表类型错误 | Check data type first 先检查数据类型 |
| Missing units 缺少单位 | Label all axes and tables 标注所有轴和表格 |
| Causation confusion 因果混淆 | Say 'linked with' not 'causes' 说 '与...相关' 而非 '导致' |
| Weak evaluation 评估薄弱 | Address bias and limitations 讨论偏差和局限性 |
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