📚 Essay Writing Framework and Model Essays for Year 12 AQA Computer Science | AS阶段AQA计算机科学论文写作框架与范文
For many Year 12 AQA Computer Science students, extended-response questions in Paper 1 and Paper 2 can feel daunting. These essay-style tasks test not only your subject knowledge but also your ability to structure arguments, use precise terminology, and evaluate competing ideas. This article provides a practical framework for planning and writing high-quality essays, together with two annotated model answers that illustrate how to meet AQA’s assessment objectives.
对许多 Year 12 AQA 计算机科学的学生来说，Paper 1 和 Paper 2 中的长答题往往令人望而生畏。这类论文式题目不仅考查学科知识，还检验你组织论证、使用精确术语以及评价不同观点的能力。本文将提供一个实用的规划与写作框架，并附上两篇带注的范文，展示如何达成 AQA 的评估目标。
1. Understanding the Essay Question | 理解论文题目
Before you write a single word, read the question at least twice and underline the key terms. Identify the topic scope, the command words (e.g. ‘discuss’, ‘compare’, ‘evaluate’), and any constraints such as ‘with reference to a named example’. Misinterpreting one of these can cause your entire essay to miss the mark.
在动笔之前，请至少把题目读两遍并划出关键词。找好主题范围、指令词（如“讨论”“比较”“评价”）以及任何限制条件，例如“结合一个命名的例子”。只要误读了其中一项，整篇论文就可能偏离要求。
- Topic scope: e.g. ethical issues, algorithms, system architecture
主题范围：如伦理问题、算法、系统架构 - Command word: ‘discuss’ requires balanced arguments; ‘compare’ needs similarities and differences
指令词：“讨论”要求正反论证；“比较”须列出相似与不同之处 - Constraints: may specify a scenario, number of points, or a named example
限制条件：可能指定场景、要点数量或具体实例
2. Deconstructing the Command Words | 解析指令词
AQA essays are driven by command words, and each demands a slightly different approach. ‘Explain’ means giving reasons for how or why something works. ‘Discuss’ expects exploration of both sides of an issue. ‘Evaluate’ goes further by requiring a judgement backed by evidence. Misjudging the command word is one of the most common reasons for lost marks.
AQA 的论文题由指令词驱动，不同指令词要求略有差异。“解释”要求说明某事物如何或者为何运行。“讨论”须探究问题的两面。“评价”则需要更进一步，提出有证据支撑的判断。误判指令词是失分最常见的原因之一。
| Command Word | What It Asks For | 指令词 | 要求 |
|---|---|---|---|
| Explain | Set out purposes or reasons | 解释 | 陈述目的或原因 |
| Discuss | Present different viewpoints and their implications | 讨论 | 呈现不同观点及其影响 |
| Compare | Identify similarities and differences | 比较 | 找出相同点和不同点 |
| Evaluate | Make a reasoned judgement, weighing up evidence | 评价 | 权衡证据，做出理性判断 |
3. Planning Your Essay Structure | 规划论文结构
Spend the first 5–8 minutes of your writing time planning a skeleton. A clear structure signals to the examiner that your answer is coherent. For a 12-mark ‘discuss’ question, aim for an introduction, three or four body paragraphs, and a conclusion. Each body paragraph should make one distinct point, supported by a concrete example.
在写作的最初 5–8 分钟先搭好骨架。清晰的结构向考官表明你的答案是有条理的。对于 12 分的“讨论”题，目标是引言、三到四个主体段和一个结论。每个主体段只论述一个独立观点，并用具体实例支撑。
A recommended plan for a typical essay:
以下是典型论文的推荐规划：
- Introduction paragraph: define key terms and state your line of argument
引言段：定义关键术语，表明你的论证主线 - Body paragraph 1: first point + technical reasoning + example
主体段 1：第一个论点 + 技术解释 + 例子 - Body paragraph 2: second point + technical reasoning + example
主体段 2：第二个论点 + 技术解释 + 例子 - Body paragraph 3 (counter-argument): alternative view + rebuttal or limitation
主体段 3（反面论点）：替代观点 + 反驳或局限性 - Conclusion: summarise and deliver a focused judgement
结论：总结并给出明确的判断
4. Introduction: Setting the Scene | 引言：设定背景
Your introduction should be brief but purposeful. Start by unpacking the question’s key term or concept, then write a sentence that outlines the scope of your answer. Avoid vague statements like ‘This is a very important topic in computer science.’ Instead, define technical vocabulary and give the reader a roadmap.
引言应当简短而有目的。先厘清题目中的关键术语或概念，再用一句话概括你的回答范围。避免“这是计算机科学中非常重要的话题”这类模糊表述。相反，你要定义技术词汇，并给读者一张路线图。
For a question on the ethical issues of facial recognition, a strong introduction might state: ‘Facial recognition refers to the use of algorithms to identify individuals from digital images. This essay will discuss the ethical, legal and cultural implications of its deployment, considering both the benefits for security and the risks to privacy.’
对于面部识别伦理问题的题目，一段有力的引言可以这样写：“面部识别指使用算法从数字图像中辨认个体。本文将讨论其部署的伦理、法律和文化影响，同时考量对安全的益处和对隐私的风险。”
5. Main Body: Developing Your Arguments | 主体：展开论点
Each body paragraph should follow the PEEL structure: Point, Evidence, Explanation, Link. State your point clearly in the first sentence, back it with a technical detail or case study, explain how that evidence supports the point, and link back to the question. This keeps your writing analytical rather than descriptive.
每个主体段都应遵循 PEEL 结构：论点、证据、解释、回扣。在首句清晰提出论点，接着用技术细节或案例支撑，解释该证据如何支持论点，最后回扣题目。这能让你的写作保持分析性而非描述性。
For example, when arguing that facial recognition can reduce crime, you might present a statistic from a pilot scheme (evidence), explain how real-time identification leads to quicker arrests (explanation), and then link that back to the ‘discuss’ command by noting a counter-risk of mass surveillance (link).
例如，在论证面部识别能减少犯罪时，你可以提出一个试点项目的统计数据（证据），解释实时识别如何加快逮捕（解释），然后通过指出大规模监控的风险来回扣“讨论”指令词。
6. Using Technical Terminology Accurately | 准确使用技术术语
AQA examiners explicitly credit precise technical language. Terms like ‘machine learning model’, ‘bias in training data’, ‘time complexity O(n²)’, or ‘virtual memory thrashing’ must be used correctly and in context. A paragraph sprinkled with vague words such as ‘thing’ or ‘stuff’ will not score highly, even if the underlying idea is correct.
AQA 考官明确奖励精确的技术语言。“机器学习模型”“训练数据偏差”“时间复杂度 O(n²)”“虚拟内存抖动”等术语必须在正确的语境中使用。如果段落中充斥着“东西”之类模糊的词汇，即便核心观点正确也很难得高分。
Creating a glossary of subject-specific terms for each topic can help. Before the exam, list words like abstraction, decomposition, protocol, cache, phishing, Turing completeness and practise writing sentences that embed them naturally.
为每个主题建立一个学科术语表会很有帮助。考前列出诸如 抽象、分解、协议、缓存、钓鱼攻击、图灵完备 等词汇，并练习将它们自然嵌入句子。
7. Incorporating Examples and Case Studies | 结合实例与案例研究
Concrete examples demonstrate depth of understanding and are explicitly rewarded in the mark scheme. Whenever you make a claim, follow it with a real-world or textbook example. For instance, when discussing the Data Protection Act, reference a breach scenario; when explaining TCP/IP layers, name a protocol that operates at each layer.
具体实例能展示理解的深度，评分方案中明确给予奖励。每提出一个主张，就应紧跟一个真实的或教科书上的例子。例如，讨论《数据保护法》时引用一个数据泄露场景；解释 TCP/IP 分层时，说明各层所运行的协议。
For AQA Year 12 essays, examples drawn from classroom coding projects, well-known security incidents (e.g. WannaCry ransomware), or comparisons of sorting behaviour on different data sets all work well.
对 AQA Year 12 的论文而言，从课堂编程项目、知名的安全事件（如 WannaCry 勒索软件）或不同数据集上的排序行为比较中提取例子都很合适。
8. Drawing Balanced Conclusions | 得出均衡的结论
Your conclusion must not introduce new material. Instead, it should synthesise the arguments presented and deliver a clear, evaluative judgement. Phrases like ‘On balance, the benefits of encryption outweigh the risks because …’ or ‘While bubble sort is easier to implement, merge sort is preferable for large datasets due to …’ show higher-order thinking.
结论绝不能引入新材料。它应当综合已呈现的论点，并给出清晰、评价性的判断。使用诸如“总体而言，加密的好处超过风险，因为……”或“尽管冒泡排序更易实现，但对于大数据集，归并排序更优，因为……”之类的表述，可以展现高阶思维。
For a ‘discuss’ question, briefly acknowledge the counter-argument before restating your overall position. This demonstrates that you have weighed evidence rather than simply listed facts.
对于“讨论”题，在重申整体立场之前，简要承认反方论点。这表明你权衡了证据，而非简单罗列事实。
9. Common Pitfalls to Avoid | 常见误区
Many students lose marks by narrating instead of analysing. Describing a long history of computing or reiterating the question without adding value wastes time. Similarly, writing an essay without a plan often leads to repetition or incomplete coverage. Finally, ignoring the command word — for example, writing a purely descriptive answer when asked to evaluate — is a frequent and costly mistake.
许多学生因为叙述而失分，而不是进行分析。描述计算机的漫长历史或者复述题目而不增加任何价值，都是在浪费时间。类似地，未经规划直接写论文常常导致重复或覆盖不完整。最后，忽视指令词——例如要求“评价”却写出纯描述性答案——是一个常见且代价高昂的错误。
To avoid these, always ask yourself: ‘Am I answering the question?’, ‘Have I included technical detail?’, and ‘Does this paragraph move my argument forward?’
为避免这些，时刻问自己：“我回答题目了吗？”“我包含了技术细节吗？”“这一段是否推进了我的论证？”
10. Sample Essay: Ethical Issues of AI | 范文：人工智能的伦理问题
Below is a model answer for the question: "Discuss the ethical, legal and cultural issues arising from the widespread use of AI-powered facial recognition systems." The essay has been written to reflect AQA’s expectations for a Year 12 standard response.
以下是针对题目 “讨论由人工智能面部识别系统广泛使用引发的伦理、法律和文化问题。” 的范文。该论文旨在反映 AQA 对 Year 12 水平作答的期望。
Paragraph 1 – Introduction
Artificial-intelligence-powered facial recognition uses deep learning models to map and match facial features from images or video streams. While it offers significant benefits in law enforcement and access control, its deployment raises serious ethical concerns around consent, bias, and surveillance. This essay will discuss these issues, examining legal frameworks such as GDPR and cultural variations in the acceptance of public monitoring.
基于人工智能的面部识别使用深度学习模型来映射并匹配图像或视频流中的面部特征。虽然它在执法和出入控制方面有显著优势，但其部署在同意、偏见和监控方面引发了严重的伦理担忧。本文将讨论这些问题，考察 GDPR 等法律框架，以及公众监控接受度的文化差异。
Paragraph 2 – Ethical issues of consent and bias
A primary ethical concern is the absence of informed consent. In many public spaces, individuals are scanned without their knowledge. Furthermore, studies have shown that some facial recognition models exhibit racial and gender bias because they are trained on unrepresentative datasets. This can lead to higher false-positive rates for minority groups, resulting in unjust accusations.
一个主要伦理问题在于缺乏知情同意。在许多公共场所，个人在不知情的情况下被扫描。此外，研究表明，由于训练数据不具代表性，一些面部识别模型表现出种族和性别偏见。这可能导致少数群体较高的假阳性率，进而造成不公正的指控。
Paragraph 3 – Legal frameworks and data protection
Legally, the General Data Protection Regulation (GDPR) classifies biometric data as ‘special category data’. Organisations must have a lawful basis for processing, such as substantial public interest, and conduct a Data Protection Impact Assessment. However, enforcement varies between member states, and critics argue that exemptions for law enforcement create loopholes that weaken individual rights.
在法律上，《通用数据保护条例》（GDPR）将生物数据归类为“特殊类别数据”。组织必须有处理数据的合法依据，例如重大公共利益，并进行数据保护影响评估。然而，各成员国执法力度不一，批评者认为执法豁免条款造成了漏洞，削弱了个人权利。
Paragraph 4 – Cultural dimensions and social acceptance
Cultural attitudes towards facial recognition differ globally. In some East Asian countries, public acceptance of surveillance is higher, often linked to perceived improvements in public safety. In contrast, many European societies emphasise individual privacy and view pervasive facial recognition as a threat to civil liberties. These cultural norms influence both regulation and uptake.
全球各地对面部识别的文化态度不尽相同。在一些东亚国家，公众对监控的接受度较高，常与公共安全的提升挂钩。相比之下，许多欧洲社会强调个人隐私，将无处不在的面部识别视作对公民自由的威胁。这些文化规范同时影响着监管和技术的采纳。
Paragraph 5 – Conclusion
In conclusion, AI facial recognition embodies a classic tension between security and liberty. The ethical issues of consent and bias, coupled with uneven legal protections and divergent cultural norms, make its universal deployment deeply problematic. On balance, a carefully regulated approach, with mandatory bias audits and transparent consent mechanisms, offers the most ethically defensible path forward.
总之，人工智能面部识别体现了安全与自由之间的经典张力。同意与偏见的伦理问题，加上不均衡的法律保护和文化规范的差异，使其普及部署极富争议。总体而言，采取审慎监管的方法，包含强制偏见审计和透明的同意机制，才是伦理上最站得住脚的出路。
11. Sample Essay: Comparing Sorting Algorithms | 范文：排序算法比较
The following model answer addresses: "Compare the bubble sort and merge sort algorithms. In your answer you should consider their time complexities, memory usage, and suitability for different data sets."
以下范文回应：“比较冒泡排序和归并排序算法。在答案中请考虑它们的时间复杂度、内存使用，以及在不同数据集上的适用性。”
Paragraph 1 – Introduction
Sorting is a fundamental operation in computer science that organises data into a specified order. Bubble sort and merge sort represent two contrasting paradigms: a simple iterative comparison-based method and an efficient divide-and-conquer approach. This essay compares their algorithmic properties and explains when each might be preferred.
排序是计算机科学中将数据按指定顺序组织的基本操作。冒泡排序和归并排序代表了两种截然不同的范式：一种简单的迭代比较方法，与一种高效的分治方法。本文将比较它们的算法特性，并说明各自在何种情况下更受青睐。
Paragraph 2 – Bubble sort characteristics
Bubble sort repeatedly steps through a list, compares adjacent elements, and swaps them if they are in the wrong order. Its worst-case and average time complexity is O(n²). The algorithm is in-place, requiring only a constant amount of additional memory, O(1). Bubble sort is stable, meaning equal elements retain their relative order. For very small or already nearly sorted data sets, its simplicity can be an advantage, but performance degrades sharply as n grows.
冒泡排序反复遍历列表，比较相邻元素，若顺序错误则交换。其最坏和平均时间复杂度为 O(n²)。该算法是原地排序，仅需常数额外空间 O(1)。冒泡排序是稳定的，即相等元素保持原有相对顺序。对于极小的或已近乎有序的数据集，其简单性可能是一种优势，但随着 n 增长，性能急剧下降。
Paragraph 3 – Merge sort characteristics
Merge sort divides the unsorted list into n sublists, each containing one element, and then repeatedly merges sublists to produce new sorted sublists until only one remains. Its worst-case time complexity is O(n log n). However, merge sort requires additional memory proportional to the size of the input, O(n), because it creates temporary arrays during merging. It is also stable and performs consistently regardless of the initial order of data.
归并排序将无序列表划分为 n 个子列表，每个子列表含一个元素，然后反复合并子列表，直至最终剩下一个有序列表。其最坏时间复杂度为 O(n log n)。然而，归并排序需要与输入规模成正比的额外内存 O(n)，因为在合并过程中要创建临时数组。它同样是稳定的，并且无论数据初始顺序如何，性能都很稳定。
Paragraph 4 – Direct comparison and suitability
Comparing these two, merge sort scales far better to large datasets due to its logarithmic factor. In an exam context, if you are asked to sort 10,000 records, bubble sort would perform approximately 100 million comparisons in the worst case, whereas merge sort performs around 132,000 — a dramatic difference. However, in embedded systems with severely limited memory, bubble sort’s O(1) space usage may be a deciding factor. Additionally, for teaching purposes or small n, bubble sort’s transparency makes it valuable.
比较二者，归并排序因对数因子在大数据集上扩展性远胜前者。在考试情境下，如果要排序 10,000 条记录，冒泡排序在最坏情况下大约执行 1 亿次比较，而归并排序约执行 13.2 万次——差距显著。但在内存极度受限的嵌入式系统中，冒泡排序的 O(1) 空间占用可能成为决定性因素。此外，在教学目的或 n 很小的情况下，冒泡排序的直观透明让它颇具价值。
Paragraph 5 – Conclusion
In summary, while bubble sort is conceptually simple and memory-efficient, its O(n²) time complexity renders it unsuitable for production-scale applications. Merge sort’s O(n log n) efficiency makes it the default choice for reliable sorting, provided sufficient memory is available. A thoughtful developer selects the algorithm based on data size, initial ordering, memory constraints, and stability requirements — not on a single metric alone.
总之，冒泡排序虽然概念简单且内存效率高，但其 O(n²) 的时间复杂度使其不适合生产环境的大规模应用。只要内存充足，归并排序 O(n log n) 的效率使它成为可靠排序的默认选择。深思熟虑的开发者会根据数据规模、初始顺序、内存限制和稳定性要求来选择算法，而非仅凭单一指标。
Published by TutorHao | Computer Science 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