📚 Year 13 WJEC Statistics: Experimental & Practical Assessment Essentials | WJEC 十三年级统计：实验与实践考核要点
In WJEC A Level Statistics, the practical assessment components — particularly Unit 4 Statistical Enquiry and Modelling — require you to demonstrate the ability to design, carry out, interpret and communicate a complete statistical investigation. Success depends not just on mathematical accuracy but on your grasp of the entire enquiry cycle, from formulating a hypothesis to presenting a coherent final report. This guide distils the essential experimental and practical assessment skills you need to master.
在 WJEC A Level 统计学中，实践考核部分——尤其是第四单元“统计探究与建模”——要求你展示设计、实施、解读并沟通完整统计调查的能力。成功不仅取决于数学的准确性，更取决于你对整个探究循环的把握，从提出假设到呈现条理清晰的最终报告。本指南提炼了你必须掌握的核心实验与实践考核技能。
1. Understanding the Statistical Enquiry Cycle | 理解统计探究循环
The practical assessment is built around the statistical enquiry cycle (also known as the PPDAC model: Problem, Plan, Data, Analysis, Conclusion). Every step must be logically connected and justified. Examiners look for evidence that you can move fluently from a real-world problem to a statistical solution and back again, reflecting critically on the process.
实践考核围绕着统计探究循环（也称 PPDAC 模型：问题、计划、数据、分析、结论）构建。每一步都必须有逻辑关联并加以论证。考官寻找的证据是你能否从现实世界的问题流畅地过渡到统计解决方案，再回到问题本身，并对整个过程进行批判性反思。
2. Formulating Clear Hypotheses and Research Questions | 拟定清晰的假设与研究问题
earch question. Then translate it into a null hypothesis (H₀) and an alternative hypothesis (H₁). For example, 'Does a new teaching method improve test scores?' re your hypotheses are testable with the data you can realistically collect and specify whether the test is one-tailed or two-tailed.
从一个精确、可衡量的研究问题开始。然后将其转化为原假设 (H₀) 和备择假设 (H₁)。例如，“新的教学方法是否提高了考试成绩？”会引出 H₀：μ₁ = μ₂ 和 H₁：μ₁ > μ₂。确保你的假设能用实际可收集的数据进行检验，并明确检验是单尾还是双尾。
3. Planning Data Collection: Observational vs Experimental Studies | 规划数据收集：观察研究与实验研究的对比
Distinguish between observational studies (where you simply record existing conditions) and designed experiments (where you actively impose treatments). In WJEC practical tasks, designed experiments often earn higher marks because they demonstrate understanding of control, randomisation and replication. State clearly which type of study you are using and justify your choice in terms of ethics, practicality and the ability to infer causation.
区分观察性研究（仅仅记录现有状况）和设计实验（主动施加处理）。在 WJEC 实践任务中，设计实验往往获得更高分数，因为它展示了对控制、随机化和重复的理解。明确说明你使用的研究类型，并从伦理、可行性以及推断因果关系的能力等角度论证你的选择。
4. Sampling Methods: Ensuring Representativeness | 抽样方法：确保代表性
Choosing the right sampling technique is crucial. Simple random sampling avoids bias, but stratified sampling ensures key subgroups are represented. In practical work, you might use systematic or cluster sampling when a full sampling frame is unavailable. Always discuss the potential for sampling bias, non-response error and the limitations of your chosen method. Define your target population and sampling frame explicitly.
选择合适的抽样技术至关重要。简单随机抽样可避免偏差，但分层抽样能确保关键子群得到代表。在实践中，当没有完整抽样框时，你可能使用系统抽样或整群抽样。务必讨论潜在的抽样偏差、无回答误差以及所选方法的局限性。明确定义你的目标总体和抽样框。
5. Designing Experiments: Control, Randomisation and Replication | 实验设计：控制、随机化与重复
A sound experiment rests on three pillars. Control: keep other variables constant or use a control group. Randomisation: assign subjects to treatments randomly to avoid confounding. Replication: use enough subjects to detect a meaningful effect. If you design a paired comparison or a block design, explain how it increases precision. Remember to mention blinding where relevant to reduce observer or subject bias.
一个扎实的实验建立在三个支柱之上。控制：保持其他变量恒定或设置对照组。随机化：将受试者随机分配到不同处理组，以避免混杂。重复：使用足够多的受试者以检测出有意义的效果。如果你设计的是配对比较或区组设计，要解释它如何提高精度。记得在适当处提及盲法，以减少观察者或受试者偏差。
6. Data Collection Tools and Questionnaire Design | 数据收集工具与问卷设计
d questions. Pilot your instrument on a small group to identify ambiguities. For physical experiments, specify measurement tools and note their precision (e.g. a digital scale reading to ±0.1 g). All data collection sheets should be well organised and ready before the investigation begins.
使用问卷时，设计必须清晰、无偏见且能产生可分析数据的问题。避免引导性问题或一题多问。在少数人中试测你的工具以找出模糊之处。对于物理实验，指定测量工具并记录其精度（如电子天平读数精确至±0.1 g）。所有数据收集表应在调查开始前整理齐全、准备就绪。
7. Data Cleaning and Preliminary Analysis | 数据清理与初步分析
Raw data almost always contains errors or outliers. Check for impossible values, missing entries and inconsistencies. Use graphical summaries — box plots, histograms and scatter plots — to explore shape, central tendency and spread. Calculate key summary statistics: mean (x̄), median, standard deviation (s), interquartile range. These insights guide which formal statistical tests are appropriate later.
原始数据几乎总是包含错误或异常值。检查不可能出现的值、缺失条目和不一致之处。使用箱线图、直方图和散点图等图形摘要来探索分布形状、集中趋势和离散程度。计算关键的摘要统计量：平均数 (x̄)、中位数、标准差 (s)、四分位距。这些见解可以指导后续哪些正式统计检验是合适的。
8. Selecting and Applying the Right Statistical Test | 选择合适的统计检验并应用
The practical exam expects you to match the test to the data type and design. A common decision table is shown below. For each test, state the assumptions (normality, equal variance, independence) and check them with appropriate diagnostics (e.g. Shapiro-Wilk test for normality, Levene’s test for homogeneity). Report the test statistic, degrees of freedom and p-value, interpreting the result in context.
实践考试要求你将检验方法与数据类型和研究设计相匹配。下表给出了常见的选择依据。对于每种检验，说明其假设条件（正态性、方差齐性、独立性）并用适当的诊断方法加以检验（如用 Shapiro-Wilk 检验正态性，Levene 检验方差齐性）。报告检验统计量、自由度和 p 值，并结合背景解释结果。
| Data & Design | Parametric Test | Non-parametric Alternative |
|---|---|---|
| Compare two means (independent groups) | Two-sample t-test | Mann-Whitney U test |
| Compare two means (paired) | Paired t-test | Wilcoxon signed-rank test |
| Association between two categorical variables | Chi-squared test (χ²) for independence | Fisher's exact test (small expected frequencies) |
| Strength of linear relationship | Pearson's correlation (r) | Spearman's rank correlation (ρ) |
9. Modelling and Regression Analysis | 建模与回归分析
For A2 practical tasks, you will often build a regression model to describe a relationship. Use least squares regression to find the line of best fit, then interpret the slope and intercept. Always produce a residual plot to check for patterns — random scatter supports the model, while curvature or a funnel shape indicates problems. Report the coefficient of determination, R², to measure how well the model explains the variability.
在 A2 实践任务中，你经常需要建立回归模型来描述一个关系。用最小二乘法找到最佳拟合线，然后解释斜率和截距。始终绘制残差图检查模式——随机散布支持模型，而弯曲或漏斗形状则表明存在问题。报告决定系数 R²，以衡量模型解释变异的程度。
Regression equation: y = a + bx, where b = Σ((x - x̄)(y - ȳ)) / Σ(x - x̄)² and a = ȳ - b x̄
10. Interpreting p-values and Drawing Valid Conclusions | 解释 p 值并得出有效结论
A small p-value (typically p < 0.05) leads you to reject H₀ in favour of H₁, but it does not prove the alternative is true; it merely measures the strength of evidence against H₀. Discuss the possibility of Type I error (false positive) and Type II error (false negative). When you fail to reject H₀, do not say 'accept H₀' — instead, conclude there is insufficient evidence to reject it. Always relate the statistical conclusion back to the original research question.
小的 p 值（通常 p < 0.05）引导你拒绝 H₀ 而支持 H₁，但这并不能证明备择假设为真；它仅仅度量了反对 H₀ 的证据强度。讨论发生第一类错误（假阳性）和第二类错误（假阴性）的可能性。当你未能拒绝 H₀ 时，不要说“接受 H₀”——而应得出结论：没有足够证据拒绝 H₀。始终将统计结论与原研究问题联系起来。
11. Communicating Findings: Report Structure and Visualisation | 交流发现：报告结构与可视化
Your practical write-up should follow a logical sequence: introduction (aim and hypotheses), methodology (design, sampling, tools), results (tables, graphs, test outputs), discussion (interpretation, limitations, relation to real world) and conclusion. Use clear, labelled diagrams — box plots for comparisons, scatter plots for relationships, and bar charts for categorical summaries. Every graph must have an informative title and labelled axes. Avoid clutter and unnecessary 3D effects.
你的实践报告应遵循逻辑顺序：引言（目的与假设）、方法（设计、抽样、工具）、结果（表格、图表、检验输出）、讨论（解释、局限性、与现实世界的联系）和结论。使用清晰的、带有标注的图形——比较用箱线图、关系用散点图、分类汇总用条形图。每张图都必须有信息丰富的标题和坐标轴标签。避免杂乱和不必要的三维效果。
12. Common Pitfalls and How to Avoid Them | 常见陷阱及避免方法
Many students lose marks by ignoring the enquiry cycle — performing tests without stating hypotheses, or failing to check test assumptions. Other frequent mistakes are using the wrong graph for the data type, confusing correlation with causation, and omitting units or labels. To avoid these, create a checklist: hypotheses stated, assumptions verified, appropriate test chosen, conclusions contextualised. Pause regularly to ask, 'Does this step follow from my original problem?'
许多学生因忽视探究循环而失分——未说明假设就进行检验，或未能检查检验前提。其他常见错误包括：对数据类型使用了错误的图形，混淆相关与因果关系，以及遗漏单位或标签。为避免这些问题，建立一份检查清单：假设已陈述、前提已验证、选择了恰当的检验、结论已结合背景。时常停下来问自己：“这一步是否源自我的原始问题？”
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