📚 Year 13 WJEC Statistics: Experimental/Practical Assessment Key Points | WJEC Year 13 统计：实验/实践考核要点
Experimental and practical assessments in Year 13 WJEC Statistics require you to demonstrate a deep understanding of the entire statistical enquiry cycle. This component is not only about performing calculations but also about planning, collecting quality data, applying appropriate techniques, interpreting results in context, and critically evaluating your findings. The following guide breaks down the key areas you need to master to excel in your practical assessment.
WJEC Year 13 统计的实验与实践考核要求你全面掌握统计调查循环的每一个环节。这一部分不仅关乎计算能力，更强调如何规划项目、收集优质数据、选用恰当的分析方法、结合背景解释结果，并对整个研究进行批判性评价。以下指南将逐一梳理你必须掌握的核心考核要点，帮助你在实践中脱颖而出。
1. The Statistical Enquiry Cycle | 统计调查循环
Every practical investigation follows the statistical enquiry cycle. You should be able to identify and execute each stage: formulating a problem or hypothesis, designing an experiment or survey, collecting data, processing and presenting data, performing statistical analysis, interpreting results, and evaluating the whole process. Examiners will look for evidence that you have consciously moved through these stages rather than jumping straight to calculations.
每一项实践研究都遵循统计调查循环。你必须清晰地识别并执行以下阶段：提出问题或假设、设计实验或调查、收集数据、处理与展示数据、进行统计分析、解释结果，以及对整个过程进行评价。考官希望看到你有意识地按照这些阶段推进，而不是直接跳到计算部分。
Understanding this cycle helps you structure your written report logically. Your introduction should set out the plan, the middle sections should present your data and analysis, and the final sections should interpret and evaluate. Keeping the cycle in mind also reduces the risk of omitting crucial steps like checking assumptions or considering ethical issues.
理解这一循环能帮助你逻辑清晰地组织书面报告。引言部分应阐述计划，中间部分呈现数据和分析，结尾部分则进行解释与评价。牢记此循环还能降低遗漏关键步骤（如检查统计假设或考虑伦理问题）的风险。
2. Formulating a Clear Hypothesis or Research Question | 明确假设或研究问题的提出
Your project must begin with a focused, measurable research question or a pair of hypotheses. For an experiment, this is often a null hypothesis (H₀) and an alternative hypothesis (H₁). For example, H₀: There is no difference in mean reaction time between treatment A and treatment B. The hypotheses must be stated precisely using parameters (μ₁ = μ₂) and in the context of your study. Avoid vague aims like “I want to see if something has an effect.”
你的项目必须以一个聚焦、可量化的研究问题或一对假设开篇。对于实验而言，通常是原假设（H₀）与备择假设（H₁）。例如，H₀：处理A与处理B的平均反应时间没有差异。假设必须用参数形式精确表述（如 μ₁ = μ₂），并紧密结合研究背景。避免模糊不清的目标，如“我想看看某事物是否有影响”。
If you are conducting a survey-based investigation, your research question might involve association between categorical variables, for instance: “Is there an association between daily exercise frequency and stress levels among Year 13 students?” This then leads to a chi-squared test for independence. The quality of your hypothesis directly influences the clarity of your design and choice of test.
如果你进行的是基于问卷的调查，研究问题可能涉及分类变量之间的关联，例如：“Year 13 学生每日锻炼频率与压力水平之间是否存在关联？”这将导向独立性卡方检验。假设的质量直接决定了后续设计和检验选择的清晰度。
3. Designing the Experiment or Survey | 实验或调查设计
Good design is the foundation of reliable conclusions. In an experiment, you must describe how treatments are allocated, whether you use a completely randomised design, matched pairs, or a double-blind procedure. You must identify the response variable, factors, levels, and controls. For instance, “30 participants were randomly divided into two groups of 15. Group C consumed a caffeine drink, while Group P received a placebo. Reaction time was measured using a computer-based test.”
良好的设计是可靠结论的基石。在实验中，你必须说明处理如何分配，是采用完全随机化设计、配对设计还是双盲程序。你需要明确响应变量、因子、水平及控制措施。例如：“30 名参与者被随机分为两组，每组 15 人。C 组饮用含咖啡因饮料，P 组服用安慰剂。反应时间通过计算机测试进行测量。”
For surveys, design includes defining your target population, creating an unbiased questionnaire, and planning how to pilot the questions. You need to explain why a particular data collection method (online, face-to-face, postal) was chosen. Always mention measures taken to reduce non-sampling errors like unclear wording or leading questions.
对于调查而言，设计包括界定目标总体、编制无偏问卷，以及规划如何先导测试问题。你需要解释为何选择特定的数据收集方式（在线、面对面、邮寄）。务必提及为减少非抽样误差（如问题表述不清或带有导向性）所采取的措施。
4. Sampling Methods and Data Collection | 抽样方法与数据收集
You must describe your sampling method and justify its appropriateness. Common techniques include simple random sampling, stratified sampling, systematic sampling, cluster sampling, and quota sampling. For example, “A stratified sample by gender and year group was used to ensure proportional representation.” Be honest about practical constraints and whether your sample is truly random or a convenience sample.
你必须描述抽样方法并论证其合理性。常用技术包括简单随机抽样、分层抽样、系统抽样、整群抽样和配额抽样。例如：“按性别和年级组进行分层抽样，以确保成比例的代表性。”要诚实地说明实际操作中的限制，以及你的样本究竟是真正的随机样本还是便利样本。
Data collection must be systematic and replicable. Keep raw data neatly organised in tables with clear headings and units. If you use secondary data, cite the source and comment on its reliability. Record exactly how measurements were taken, any equipment used, and steps to standardise the process. The examiner expects a level of detail that would allow another student to reproduce your investigation.
数据收集必须系统化且可重复。将原始数据整齐地组织在表格中，标明清晰的表头和单位。若使用二手数据，需要注明来源并评价其可靠性。准确记录测量方法、所用设备以及标准化流程的步骤。考官期待你提供足够细节，使另一名学生能够复制你的研究。
5. Ensuring Data Quality and Reliability | 确保数据质量与可靠性
Data quality is judged by accuracy, precision, and absence of bias. You should discuss potential sources of measurement error and how you minimised them. For example, “Reaction times were recorded to the nearest millisecond, and each participant completed three practice trials to reduce learning effects.” If using a questionnaire, you might test for internal consistency using Cronbach’s alpha or conduct a pilot to refine ambiguous items.
数据质量的评判标准包括准确性、精确性和无偏性。你应讨论潜在的测量误差来源以及如何将其最小化。例如：“反应时间精确记录至毫秒，每位参与者先完成三次练习试验以减少学习效应。”若使用问卷，可通过克隆巴赫系数检验内部一致性，或通过先导测试来修正模糊题项。
Reliability refers to the consistency of your measuring instrument – would you get similar results on repeated trials? You can improve reliability by using established instruments, standardising conditions, and training observers. Always distinguish between random errors (which can be reduced by averaging) and systematic errors (which require calibration or design changes).
信度指的是测量工具的一致性——重复试验能否得到相似结果？你可以通过使用成熟的工具、统一实验条件和培训观察者来提高信度。务必区分随机误差（可通过取平均值减小）和系统误差（需通过校准或修改设计来解决）。
6. Data Processing and Organisation | 数据处理与整理
Raw data is rarely ready for analysis. You need to show how you cleaned the data: handling missing values, removing obvious outliers with justification, and coding categorical variables for software. Present clean data in a well-structured table. For paired data, show both before and after values. If using transformations (e.g., log transform to stabilise variance), explain your reasoning.
原始数据很少能直接用于分析。你需要展示如何清理数据：处理缺失值、在合理理由下剔除明显异常值，以及为软件分析编码分类变量。用清晰结构的表格呈现清理后的数据。对于配对数据，应同时显示处理前与处理后的值。若使用了数据变换（如为稳定方差取对数），解释你的理由。
Summary statistics should be calculated as a first step: means, medians, standard deviations, quartiles, and range. Present these in a summary table. This step not only gives you a feel for the data but also helps you choose the right inferential method. For instance, large differences between mean and median may indicate skewness, pointing towards non-parametric tests.
首先计算概括统计量：均值、中位数、标准差、四分位数和极差。将这些呈现在汇总表中。这一步不仅让你初步把握数据特征，还能帮助你选择正确的推断方法。例如，均值与中位数差距很大可能表明数据偏斜，提示应使用非参数检验。
7. Choosing Appropriate Statistical Techniques | 选择适当的统计方法
Your choice of test must be justified with reference to the type of data, the study design, and which assumptions are met. For comparing two means, you might use a two-sample t-test (independent or paired) after checking normality and equal variances. For associations between two categorical variables, a chi-squared test of independence is appropriate. For correlation, use Pearson’s r if data are normally distributed; otherwise, consider Spearman’s rank.
你选择的检验方法必须基于数据类型、研究设计和满足的假设条件进行论证。比较两个均值时，可检查正态性和方差齐性后采用双样本 t 检验（独立或配对）。对于两个分类变量的关联，适用独立性卡方检验。若数据呈正态分布，用皮尔逊相关系数 r；否则考虑斯皮尔曼秩相关系数。
Year 13 practicals may also involve regression analysis, where you test the significance of a slope using a t-test or construct a prediction interval. Always state the test statistic distribution (e.g., t-distribution with df = n₁ + n₂ − 2), calculate the statistic manually or via technology, and find the p-value. Never simply report output without explaining what it means.
Year 13 的实践项目还可能涉及回归分析，运用 t 检验来检验斜率的显著性，或构建预测区间。始终写明检验统计量的分布（例如，自由度为 n₁ + n₂ − 2 的 t 分布），手动或用技术计算统计量，并求出 p 值。切勿只报告软件输出而不解释其含义。
8. Performing Calculations and Using Technology | 进行计算与使用技术
You are expected to carry out calculations accurately either by hand (showing steps for key values) or with the aid of statistical software/spreadsheets. When using technology, you must still provide key intermediate values such as sums of squares, standard errors, and test statistics. Include screenshots or printouts labelled clearly. For hypothesis tests, always compare the p-value to the significance level (commonly α = 0.05) and state your decision rule.
你应当准确进行计算，既可以手动（展示关键值的计算步骤），也可以借助统计软件或电子表格。即使使用技术，仍需提供关键中间值，如平方和、标准误和检验统计量。包含清晰标注的截图或打印输出。进行假设检验时，始终将 p 值与显著性水平（通常取 α = 0.05）进行比较，并陈述你的决策规则。
Confidence intervals add great value to your analysis. For example, “The 95% confidence interval for the difference in mean reaction time is (12.4 ms, 28.6 ms).” Interpret the interval correctly: “We are 95% confident that the true difference lies between 12.4 and 28.6 ms.” This shows a mature understanding beyond simple ‘reject/fail to reject’ conclusions.
置信区间能为你的分析增添巨大价值。例如，“平均反应时间差值的 95% 置信区间为（12.4 ms, 28.6 ms）。”正确解读该区间：“我们有 95% 的把握认为真实差值在 12.4 至 28.6 ms 之间。”这体现了超越单纯“拒绝/不拒绝”论断的成熟理解。
9. Interpreting Results and Drawing Conclusions | 结果解释与得出结论
Interpretation must be in the context of your original problem. Do not just say “p < 0.05, so we reject H₀." Instead, explain: "There is sufficient evidence at the 5% level to suggest that the caffeine group has a lower mean reaction time. This result is consistent with the theory that caffeine increases alertness." Where results are not significant, avoid concluding 'no effect'—discuss the possibility of insufficient power.
解释必须结合原始问题背景。不要仅仅说“p < 0.05，因此拒绝 H₀”。而要这样阐述：“在 5% 显著性水平下，有充分证据表明咖啡因组的平均反应时间更低。这一结果与咖啡因提高警觉性的理论一致。”当结果不显著时，避免得出“无效应”的结论——应讨论检验效能不足的可能性。
Link your findings to real-world implications. If your study was on student stress and exercise, suggest practical recommendations like “Since a moderate association was found, schools might consider promoting physical activity as part of wellbeing programmes.” Always remain cautious; association does not imply causation, especially in observational studies.
将发现与现实生活意义联系起来。如果你的研究关注学生压力与锻炼，可给出实际建议，例如：“由于发现二者存在中等关联，学校不妨考虑将体育活动纳入健康促进计划。”始终保持谨慎；关联并不意味着因果，尤其在观察性研究中更是如此。
10. Evaluating Limitations and Suggesting Improvements | 评估局限性并提出改进建议
No practical investigation is perfect. High marks are awarded for honest, insightful evaluation. Identify specific limitations, such as small sample size, convenience sampling, measurement error, or confounding variables you could not control. For each limitation, propose a realistic improvement. For example, “The sample size of 30 limited statistical power. A future study could involve at least 50 participants per group to detect smaller effects.”
任何实践研究都不完美。高分属于那些诚实且富有洞察力的评价。找出具体的局限性，如样本量小、便利抽样、测量误差或未能控制的混杂变量。针对每项局限性，提出切实可行的改进建议。例如，“30 人的样本量限制了统计效能，未来的研究可以每组至少招募 50 人，以检测更小的效应。”
Also comment on the generalisability of your findings. If your participants were all Year 13 students from one school, state that the results may not apply to all age groups or regions. Discuss any ethical constraints that affected your design, and how you might address them differently if resources allowed. This critical reflection demonstrates the highest level of statistical thinking.
此外，还要评价你研究结论的推广性。若参与者均来自同一所学校的 Year 13 学生，应说明结果可能不适用于所有年龄段或地区。讨论影响你设计方案的伦理限制，并设想在资源允许的情况下你会如何不同地处理。这种批判性反思体现了最高层次的统计思维。
11. Presenting Findings Effectively | 有效展示发现
Visual presentation is part of the assessment. Use appropriate graphs: box plots for comparing distributions, scatter plots with a line of best fit for correlation and regression, and bar charts or segmented bar charts for categorical data. Every graph must have a clear title, labelled axes with units, and a brief caption. Avoid unnecessary 3D effects or clutter that obscures the data.
视觉呈现是考核的一部分。选用恰当的图表：箱线图用于比较分布，散点图配最佳拟合线用于相关与回归，条形图或堆积条形图用于分类数据。每幅图都应有清晰的标题、带单位的坐标轴标签，以及简要图注。避免不必要的 3D 效果或杂乱元素干扰数据呈现。
Your final report should read like a scientific paper: abstract, introduction, methodology, results, discussion, and references. Writing clearly and concisely is essential. Use the correct statistical terminology and format numbers and symbols consistently. Demonstrate that you can communicate complex statistical ideas to a non-specialist reader, which is a key skill assessed in the practical.
你的最终报告应像一篇科学论文那样结构完整：摘要、引言、方法、结果、讨论和参考文献。清晰简练的写作至关重要。使用正确的统计术语，数字和符号格式保持一致。展现你能将复杂的统计思想传达给非专业读者，这是实践考核中评估的一项关键能力。
12. Adhering to Ethical Guidelines | 遵守伦理准则
Ethical considerations are mandatory. You must obtain informed consent from participants, explain that they can withdraw at any time, and keep all data anonymous and confidential. If your project involves sensitive topics, you need to describe how you minimised distress. Even if your school does not have a formal ethics committee, you should still document your ethical steps in the report.
伦理考量不可或缺。你必须获得参与者的知情同意，告知他们可随时退出，并确保所有数据匿名且保密。若项目涉及敏感话题，你需说明如何降低不适感。即使学校没有正式伦理委员会，你也应在报告中记录所采取的伦理措施。
When using secondary data, respect copyright and data protection laws. Always cite sources properly. Never fabricate or manipulate data to achieve a desired outcome – this is a serious academic offence. Integrity throughout the project not only meets WJEC requirements but also builds your credibility as a young statistician.
使用二手数据时，要尊重版权和数据保护法规。始终规范引用来源。切勿为得到理想结果而伪造或篡改数据，这是严重的学术违规。全程保持诚信不仅满足 WJEC 要求，更能为你作为青年统计学者建立信誉。
Published by TutorHao | Statistics Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
屏轩国际教育cambridge primary/secondary checkpoint, cat4, ukiset,ukcat,igcse,alevel,PAT,STEP,MAT, ibdp,ap,ssat,sat,sat2课程辅导，国外大学本科硕士研究生博士课程论文辅导Cancel reply