📚 AS OCR Statistics: Essay Writing Framework and Model Answers | AS OCR 统计：论文写作框架与范文
In AS OCR Statistics, extended response questions often require you to produce a structured statistical report or essay. This article provides a practical framework for constructing such responses, together with a full model answer annotated with revision tips. Mastering this format boosts both clarity and marks.
在 AS OCR 统计考试中，拓展型题目常要求你撰写结构化的统计报告或论文。本文提供了一个实用的写作框架，并附有带注释的完整范文。掌握这一格式能同时提升表达的清晰度与得分。
1. Understanding the Assessment Objectives | 理解评分目标
Statistical essays in OCR AS are assessed on knowledge of techniques (AO1), application to data and context (AO2), and interpretation of results (AO3). A top-mark answer must demonstrate accurate calculations, logical flow, and critical evaluation of findings within the given scenario.
OCR AS 统计论文的评分依据是对技术的掌握（AO1）、对数据与情境的应用（AO2）以及对结果的解读（AO3）。高分答案必须在给定情境中展示准确的计算、合乎逻辑的论述流程以及对结论的批判性评价。
Examiners look for your ability to select the right statistical tools, justify them, carry them out correctly, and finally link the numerical evidence back to the original research question. Every paragraph should serve a clear purpose.
考官看重你选择恰当的统计工具、说明理由、正确执行、并最终将数值证据与原始研究问题关联起来的能力。每一段文字都应该有明确的目的。
2. The Three-Part Essay Structure | 三段式论文结构
An effective statistical essay follows a simple three-part skeleton: Plan – Do – Evaluate. In the 'Plan' stage you set up the question, hypotheses, and data design. In 'Do' you present calculations, diagrams, and test results. In 'Evaluate' you interpret findings and comment on limitations.
一篇有效的统计论文遵循简单的三部骨架：计划 – 执行 – 评估。在“计划”阶段，你设定问题、假设和数据设计；在“执行”阶段，你展示计算、图表和检验结果；在“评估”阶段，你解读发现并评论局限性。
This framework aligns with the statistical enquiry cycle taught in AS: specify the problem, collect/process data, analyse, and conclude. Use it to organise your paragraphs so the examiner can follow your reasoning easily.
这一框架与 AS 所教的统计探究周期一致：明确问题、收集/处理数据、分析、得出结论。用它来组织段落，考官就能轻松地跟随你的推理过程。
| Structure | What to include |
|---|---|
| Plan | Context, hypotheses, variables, data source |
| Do | Summary statistics, diagrams, probability/model calculations, test statistic, p-value |
| Evaluate | Conclusion in context, reliability, limitations, suggestions |
3. Introduction: Framing the Problem | 引言：构建问题框架
and clearly phrases a null and alternative hypothesis using correct notation. For example: 'H₀: μ = 3.0 hours' vs 'H₁: μ > 3.0 hours'.
以简短的段落开篇，说明调查目的，指出关注的总和变量，并使用正确的符号清晰地表述零假设与备择假设。例如：“H₀: μ = 3.0 小时” 与 “H₁: μ > 3.0 小时”。
Always link the hypotheses to the real-world context given in the question. This shows the examiner you are not performing a mechanical calculation but engaging with the scenario.
始终将假设与题目提供的现实情境挂钩。这能向考官表明你不是在机械地计算，而是真正融入了情景。
A typical introduction might also mention the type of data (continuous, discrete, categorical) and the sampling method, if known. Keep it concise – four to five lines are sufficient.
典型的引言还可能提及数据类型（连续、离散、分类）以及抽样方法（若已知）。保持简明——四到五行便足够。
4. Describing the Data and Summary Statistics | 描述数据与概要统计量
After introducing the context, present the data or a summary table. Give appropriate measures of central tendency and spread – mean and standard deviation for symmetric data, median and IQR for skewed data. Justify your choice of measure based on the shape of the distribution.
介绍完情境后，呈现数据或汇总表。给出恰当的集中趋势和离散指标——对称数据用均值和标准差，偏斜数据用中位数和四分位距。根据分布形状说明选择该指标的理由。
Include a diagram such as a box plot or histogram, referring to it in the text. Narrate the key features visible in the diagram: 'The box plot shows a slight positive skew, with the median closer to the lower quartile.'
纳入箱线图或直方图等图示，并在正文中提及。描述图中可见的关键特征：“箱线图显示出轻微的正偏斜，中位数较靠近下四分位数。”
Use clear sentences rather than a list of numbers. For example: 'The sample mean screen time was 3.4 hours with a standard deviation of 0.8 hours, suggesting moderate variation among students.'
使用清晰的句子而非数字堆砌。例如：“样本平均屏幕使用时间为 3.4 小时，标准差为 0.8 小时，表明学生之间存在中等程度的差异。”
5. Choosing and Justifying the Statistical Test | 选择并论证统计检验
st) and for the binomial proportion. State clearly which test you are using and why: e.g., ' e is unknown and the sample size n=30 is fairly large, a one-sample t-test is appropriate.'
AS OCR 涵盖关于总体均值的假设检验（z 检验或 t 检验）以及关于二项比率的检验。明确陈述你使用哪种检验及其理由，例如：“由于总体方差未知且样本量 n=30 较大，适用单样本 t 检验。”
heorem for n≥30). For a binomial test, confirm that the observations are independent and that np and nq satisfy the normality approximation if used.
核查检验所需的假设条件。对均值的 t 检验，要提到抽样分布的近似正态性（n≥30 时依据中心极限定理）。对二项检验，确认观测值相互独立，并且若使用正态近似，需满足 np 与 nq 条件。
Examiners reward candidates who do not blindly run a test but pause to reflect on its suitability.
考官会奖励那些不盲目执行检验，而是停下来思考其适用性的考生。
6. Performing Calculations and Displaying Work | 执行计算与展示步骤
Show all steps of your hypothesis test clearly. Write the test statistic formula, substitute values, and calculate the result. For a t-test: t = (x̄ - μ₀) / (s/√n). Arrange the work so the reader can trace every substitution.
清晰地展示假设检验的所有步骤。写出检验统计量公式，代入数值，计算其结果。对于 t 检验：t = (x̄ - μ₀) / (s/√n)。将步骤排版得让读者能追踪每一次代入。
If you are using a critical value approach, give the critical value from tables. If using p-value, compare it to the significance level (usually 5%). State clearly whether the observed statistic lies in the rejection region.
若使用临界值法，给出查表所得的临界值。若使用 p 值，将其与显著性水平（通常 5%）比较。明确说明观测值统计量是否落在拒绝域内。
Round numbers appropriately but avoid premature rounding. It is good practice to keep at least three significant figures during calculations.
适当四舍五入，但避免过早舍入。好的做法是在计算过程中保持至少三位有效数字。
7. Interpreting the Results in Context | 结合情境解读结果
State the decision ('Reject H₀' or 'Do not reject H₀') and immediately translate it into plain English tied to the original problem. Do not just write 'Reject H₀'; say: 'There is sufficient evidence at the 5% level to suggest that the mean daily screen time of students exceeds 3 hours.'
陈述判定（“拒绝 H₀”或“不拒绝 H₀”），并立刻将其转译为与原始问题相关的朴素英语。不要只写“拒绝 H₀”；要写：“在 5% 显著性水平下有充分证据表明学生日均屏幕使用时间超过 3 小时。”
This is where many candidates lose marks – they stop at the mathematical conclusion and neglect the contextual interpretation required by AO3. Always finish the test paragraph with a sentence that directly answers the research question.
这是许多考生失分之处——他们停在数学结论上，忽视了 AO3 所要求的结合情境的解读。始终在检验段落的末尾加上一句直接回答研究问题的话。
8. Evaluation: Limitations and Reliability | 评估：局限性与可靠性
To reach the highest bands, you must discuss the reliability of your conclusion. Mention possible sources of bias, sample size, sampling method, and any assumptions that might have been violated. For example: 'The sample was a convenience sample from one school, so results may not generalise to all students in the region.'
要想达到最高档次，你必须讨论结论的可靠性。提及可能存在的偏误来源、样本量、抽样方法以及任何可能被违反的假设。例如：“样本是来自一所学校的便利样本，因此结果可能无法推广至该地区所有学生。”
Suggest a practical improvement: 'A stratified sample across year groups could improve representativeness.' This shows the examiner you are thinking like a statistician, not just a calculator.
提出一个切实可行的改进建议：“按年级分层抽样可提高代表性。”这向考官表明你以统计学家的方式思考，而不仅仅是一个计算器。
Keep the evaluation balanced – acknowledge what the analysis does well, then critique. Use phrases like 'A limitation of this test is...' or 'The assumption of independence is reasonable because...'
保持评估平衡——先肯定分析做得好的地方，再给予批评。使用像“该检验的一个局限是……”或“独立性的假设是合理的，因为……”这样的短语。
9. Model Answer – Part 1: Introduction to Analysis | 范文 – 第一部分：引言至分析
Below is an example response to the question: 'Investigate whether the mean weekly pocket money for Year 12 students at a school is less than £15. A random sample of 25 students gave a mean of £13.20 and a standard deviation of £3.50.'
以下是一个示例回答，问题为：“调查某校 12 年级学生每周零花钱的均值是否小于 15 英镑。一个由 25 名学生组成的随机样本给出的均值为 13.20 英镑，标准差为 3.50 英镑。”
Model text: The aim is to test if the population mean weekly pocket money, μ, is below £ sumed independent. Because n<30, normality of the population is assumed. The sample mean x̄=13.20 and s=3.50. As σ is unknown, a one-sample t-test at 5% significance is used.
范文： 目的是检验总体平均每周零花钱 μ 是否低于 15 英镑。假设：H₀: μ = 15，H₁: μ < 15 （单侧检验）。变量为连续型；数据来自 n=25 的随机样本，假定观测独立。由于 n<30，需假定总体服从正态分布。样本均值 x̄=13.20，s=3.50。由于 σ 未知，在 5% 显著性水平下使用单样本 t 检验。
Test statistic: t = (13.20 - 15) / (3.50/√25 / 0.70 = -2.571. Degrees of freedom = 24. The critical value from t-tables (one-tailed, 5% nce -2.571 < -1.711, the test statistic falls in the rejection region. Hence we reject H₀.
检验统计量：t = (13.20 - 15) / (3.50/√25) = -1.80 / 0.70 = -2.571。自由度 = 24。从 t 分布表查得临界值（单侧，5%）为 -1.711。由于 -2.571 < -1.711，检验统计量落入拒绝域。因此我们拒绝 H₀。
The analysis is clearly laid out step by step, with all substitutions shown. A diagram of the t-distribution with the rejection region shaded could be added for extra clarity.
分析步骤清晰，所有代入均已展示。可附加一张 t 分布图并标明拒绝域以增加清晰度。
10. Model Answer – Part 2: Conclusion and Evaluation | 范文 – 第二部分：结论与评估
Model text continued: There is sufficient evidence at the 5% level to conclude that the mean weekly pocket money for Year 12 students is less than £15. In context, this suggests students may have limited disposable income compared to a benchmark value, which could affect decisions about school trip costs.
范文续： 在 5% 显著性水平下有充分证据得出，12 年级学生每周平均零花钱低于 15 英镑。结合情境，这表明与基准值相比学生的可支配收入可能有限，这可能影响有关学校旅行费用的决策。
5). The assumption of normality cannot be verified without the raw data; a box plot of the sample would help. A larger sample or a different year group could be investigated to generalise the findings.
关于可靠性，随机抽样方法是一个优点，但样本量相对较小（n=25）。没有原始数据便无法验证正态性假设；样本的箱线图会有所帮助。可以扩大样本或调查其他年级以推广结论。
A suggestion for further work: record individual amounts to check for outliers and to construct a confidence interval, which would give a range of plausible values for the population mean. Overall, the t-test provides a valid initial conclusion provided the normality assumption is not seriously violated.
后续研究建议：记录个体金额以检查异常值并构建置信区间，这会给出总体均值的一个合理范围。总体而言，只要正态性假设未被严重违背，该 t 检验提供了一个有效的初步结论。
11. Common Pitfalls and How to Avoid Them | 常见陷阱与规避方法
Many students lose marks by writing 'accept H₀' or stating the conclusion without linking to the context. Always use the phrase 'do not reject H₀' and then explain what this means in words. Never say 'there is proof' – statistical tests provide evidence, not proof.
许多学生因写“接受 H₀”或陈述结论时没有联系情境而失分。始终使用“不拒绝 H₀”这一措辞，然后用语言解释其含义。永远不要写“有证据证明”——统计检验提供的是证据，而非证明。
Other errors include: confusing one-tailed and two-tailed tests, failing to check assumptions, and rounding the test statistic too early. Set up hypotheses before looking at the sample data, as the direction must be based on the research question, not the data.
其他错误包括：混淆单侧与双侧检验，未检查假设，以及过早舍入检验统计量。在查看样本数据之前设定假设，因为方向必须基于研究问题，而非数据。
A checklist before finishing your answer can help: Have I stated H₀ and H₁ correctly? Did I justify my test? Are calculations clear? Did I interpret the p-value or critical value correctly? Did I conclude in context? Did I discuss limitations?
在完成答案前使用清单会有帮助：我是否正确地陈述了 H₀ 和 H₁？我是否解释了选择该检验的理由？计算是否清晰？我是否正确解读了 p 值或临界值？我是否结合情境作出了结论？我是否讨论了局限性？
12. Practice and Revision Tips | 练习与复习建议
To master the essay format, practise writing full responses under timed conditions using past paper questions. Annotate your own answers with the framework headings (Plan-Do-Evaluate) to check you have covered all elements.
为掌握论文格式，在限时条件下使用往年真题练习撰写完整答案。用框架标题（计划-执行-评估）为自己的答案加注，检查是否覆盖了所有要素。
Read examiner reports from OCR to understand what distinguishes a grade A response from a grade C response. Build a bank of useful phrases such as 'The distribution appears symmetric, therefore the mean is an appropriate measure of central tendency.'
阅读 OCR 的考官报告，以理解 A 等级与 C 等级答案的区别。建立一个有用短语库，如“分布看似对称，因此均值是恰当的集中趋势指标”。
Finally, always keep a calculator handy and become fluent with the statistical functions. Fast, accurate computation frees mental energy for interpretation and evaluation, which carry the highest marks in extended questions.
最后，随时备好计算器，并熟练使用统计功能。快速、准确的计算能为解读和评估（在拓展题中分值最高）腾出脑力。
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