📚 Essay Writing Framework and Model Answers for Year 12 WJEC Computer Science | Year 12 WJEC 计算机：论文写作框架与范文
Extended writing questions in WJEC AS Computer Science require more than just recalling facts; they demand structured arguments, precise technical vocabulary, and the ability to evaluate competing ideas. Mastering a clear essay framework not only boosts your confidence but also helps you secure the highest marks in Component 1. This guide breaks down a proven writing method and provides annotated model answers so you can see exactly how to apply it.
在 WJEC AS 计算机科学的考试中，拓展写作题不仅需要你回忆知识点，还要求你能组织有逻辑的论证、使用准确的技术词汇，并能评估不同的观点。掌握一个清晰的论文写作框架不仅能提升你的信心，还能帮助你在 Component 1 中拿到最高分。本指南将拆解一个行之有效的写作方法，并提供带注释的范文，让你清楚地看到如何实际运用。
1. Understanding the WJEC Essay Question | 理解 WJEC 论文题目
Before you start writing, identify the command word: 'discuss', 'compare', 'evaluate', or 'justify'. 'Discuss' means you must explore different sides of an issue. 'Compare' requires you to highlight similarities and differences, often with a concluding judgement. 'Evaluate' asks you to weigh up strengths and weaknesses to form a reasoned conclusion. Highlighting the key technical terms in the question ensures you stay focused on the specific area of the specification.
在你开始写作之前，先找出题目中的指令词：'discuss'（讨论）、'compare'（比较）、'evaluate'（评价）或 'justify'（论证）。'Discuss' 意味着你必须探讨问题的不同方面。'Compare' 要求你突出相似点和不同点，通常还需要给出有依据的结论。'Evaluate' 则要求你权衡优缺点，从而形成一个合理的结论。把问题中的关键科技术语标亮，能确保你始终紧扣考纲中特定的领域。
2. The PEEL Paragraph Structure | PEEL 段落结构
Every analytical paragraph should follow PEEL: Point – state the main idea of the paragraph in one clear sentence. Evidence – support it with a technical example, an algorithm snippet or a reference to a case study. Explanation – elaborate on how your evidence proves the point, using correct terminology. Link – connect the paragraph back to the question or forward to the next argument. This structure prevents rambling and keeps your writing examiner-friendly.
每一个分析性段落都应该遵循 PEEL 结构：P（观点）—— 用一句清晰的话陈述段落的主要观点。E（证据）—— 用一个技术实例、一段算法片段或一个案例研究来支撑你的观点。E（解释）—— 详细阐述你的证据如何证明了观点，并使用正确的术语。L（衔接）—— 把该段落与问题联系起来，或者自然地过渡到下一个论点。这个结构能有效防止跑题，并使你的文章更易于考官阅读。
3. Introduction: Setting the Scene | 引言：设定背景
Begin with two or three sentences that define the scope of the essay. Restate the question in your own words and briefly signpost the key areas you will cover. For a question on the environmental impact of data centres, your introduction could acknowledge the recent surge in cloud computing and then state that you will examine energy consumption, e-waste and potential green solutions. Avoid lengthy background; jump straight into the technical focus the examiner expects.
用两到三个句子开头，界定论文的范围。用自己的话复述问题，并简要地预告你将讨论的关键领域。对于一个关于数据中心环境影响的题目，你可以先承认云计算近期的快速发展，然后表明你将探讨能源消耗、电子废弃物以及潜在的绿色解决方案。避免冗长的背景介绍；直接切入考官期待的技术焦点。
4. Main Body: Building Arguments with Evidence | 主体：用证据构建论点
Each main body paragraph should tackle one distinct aspect. If you are comparing insertion sort and quicksort, use one paragraph to explain time complexity with concrete numbers, another to discuss space complexity, and a third to explore stability and adaptability. Use phrases like 'In contrast', 'Furthermore', and 'On the other hand' to make the logical flow explicit. Technical accuracy matters more than elegant prose, so spell out acronyms on first use and apply Big O notation precisely.
每一个主体段落都应该处理一个不同的方面。如果你在比较插入排序和快速排序，可以用一个段落结合具体数字解释时间复杂度，另一个段落讨论空间复杂度，第三个段落则探讨稳定性和自适应性。使用 'In contrast'（相比之下）、'Furthermore'（此外）和 'On the other hand'（另一方面）这样的短语来使逻辑脉络更加清晰。技术准确比辞藻华丽更重要，因此首次出现的缩略词要写出全称，并准确地使用大 O 符号。
5. Using Technical Terminology Correctly | 正确使用技术术语
WJEC examiners look for precise language: say 'linear search has an average-case time complexity of O(n)' instead of 'it is slow'. Use terms such as 'abstraction', 'decomposition', 'caching', 'volatile memory', and 'protocol stack' where they genuinely fit. If you introduce a term like 'lossy compression', immediately illustrate it with an example such as JPEG discarding high-frequency colour data. Misusing terminology signals uncertainty, so only use words you can define confidently.
WJEC 考官看重用词精确：要写 'linear search has an average-case time complexity of O(n)'（线性搜索的平均时间复杂度为 O(n)），而不是简单地说 'it is slow'。在合适的地方使用诸如 'abstraction'（抽象）、'decomposition'（分解）、'caching'（缓存）、'volatile memory'（易失性存储器）和 'protocol stack'（协议栈）等术语。如果你引入一个像 'lossy compression'（有损压缩）这样的术语，要立刻用一个例子来阐释，比如 JPEG 会丢弃高频颜色数据。用错术语会表明你掌握不扎实，因此只使用你能自信定义的词。
6. Addressing Ethical, Legal, and Environmental Issues | 论述伦理、法律与环境问题
Many 8-12 mark essays include a directive to discuss wider issues. Create a simple mental checklist: ethical (bias, fairness, digital divide), legal (GDPR, Computer Misuse Act, Copyright, Designs and Patents Act) and environmental (rare-earth mining, energy usage, recycling). When writing about facial recognition, you might contrast its crime-fighting benefits with the legal requirement for explicit consent under GDPR and the ethical concern of algorithmic racial bias.
许多 8-12 分的论文题都包含讨论更广泛影响的要求。你可以列一个简单的思维清单：伦理方面（偏见、公平性、数字鸿沟）、法律方面（GDPR、《计算机滥用法》、《版权、外观设计和专利法》）以及环境方面（稀土开采、能源消耗、回收）。当写到人脸识别时，你可以对比它在打击犯罪方面的益处与 GDPR 要求获取明确同意的法律规定，以及算法导致种族偏见的伦理问题。
7. Comparing Algorithms: Time and Space Complexity | 比较算法：时间复杂度与空间复杂度
A typical compare question might ask you to contrast two sorting algorithms. State the complexities using standard notation and then explain what those values mean for real data sets. For instance:
一个典型的比较题可能会让你对比两种排序算法。使用标准符号写出它们的复杂度，然后解释这些数值对于真实数据集意味着什么。例如：
Bubble sort: O(n²) time, O(1) space
Merge sort: O(n log n) time, O(n) space
Bubble sort's quadratic growth means that doubling the input size roughly quadruples the number of comparisons, making it impractical for more than a few thousand items. Merge sort's linearithmic time keeps it efficient on large arrays, but the additional O(n) auxiliary space can be a constraint on memory-limited embedded systems. Always tie the theoretical complexity back to a practical scenario.
冒泡排序的二次增长意味着输入规模翻倍时，比较次数大约会增加到四倍，这使得它对多于几千条数据的情况不实用。归并排序的线性对数时间使其能高效处理大型数组，但是 O(n) 的额外辅助空间可能会在内存受限的嵌入式系统中成为瓶颈。始终要把理论上的复杂度与一个实际场景联系起来。
8. Evaluating Impacts of Digital Technology | 评估数字技术的影响
Evaluation requires balanced judgement. Use a 'benefit, drawback, mitigation' pattern. For example, when discussing cloud storage: 'Cloud storage enables seamless collaboration and automatic backups (benefit). However, it creates a dependency on stable internet connectivity and raises data sovereignty concerns because servers may reside in different legal jurisdictions (drawback). This can be partially mitigated by hybrid cloud models that keep sensitive data on-premises (mitigation).' Such a structure demonstrates critical thinking.
评价需要有平衡的判断。可以使用'益处、弊端、缓解措施'的模式。例如，在讨论云存储时：'云存储实现了无缝协作和自动备份（益处）。然而，它造成了对稳定互联网连接的依赖，并引发了数据主权问题，因为服务器可能位于不同的司法管辖区（弊端）。这可以通过将敏感数据保留在本地的混合云模型来部分缓解（缓解措施）。'这种结构体现了批判性思维。
9. Model Essay: The Ethics of AI in Decision-Making | 范文：人工智能在决策中的伦理问题
Question: 'Discuss the ethical implications of using artificial intelligence in automated decision-making systems.'
题目：'讨论在自动化决策系统中使用人工智能的伦理影响。'
Artificial intelligence is increasingly used to make decisions in recruitment, finance and law enforcement. While AI offers speed and consistency, it also introduces significant ethical risks related to transparency, bias and accountability that must be carefully examined.
人工智能正越来越多地被用于招聘、金融和执法领域的决策。虽然人工智能提供了速度和一致性，但它也引发了与透明度、偏见和问责相关的重大伦理风险，这些都必须仔细审视。
One major ethical concern is algorithmic bias. Machine learning models trained on historical data can perpetuate existing inequalities. For instance, a recruitment tool developed by a large technology firm was found to penalise CVs containing the word 'women's', because it had learned from past hiring patterns that favoured male candidates. Such bias is often unintentional, yet its impact can systematically exclude protected groups, violating principles of fairness and equality.
一个主要的伦理问题是算法偏见。基于历史数据训练的机器学习模型可能会延续现有的不平等。例如，一家大型科技公司开发的招聘工具，被发现会惩罚包含 'women's' 一词的简历，因为它从过去倾向男性候选人的招聘模式中进行了学习。这种偏见往往是无意的，但其影响可能会系统性地排斥受保护群体，违反公平和平等的原则。
A second issue is the lack of transparency, often called the 'black box' problem. Deep neural networks can contain millions of parameters, making it nearly impossible to explain how a specific decision was reached. When a loan application is rejected or a defendant is given a high-risk score, the individual has a right to an explanation under regulations like GDPR. If the logic remains opaque, it undermines trust and makes legal redress extremely difficult.
第二个问题在于缺乏透明度，通常被称为'黑箱'问题。深度神经网络可能包含数百万个参数，几乎不可能解释它是如何做出一个具体决定的。当一笔贷款申请被拒绝，或者一名被告被给予高风险评分时，根据 GDPR 等法规，个人有权获得解释。如果决策逻辑始终不透明，就会破坏人们的信任，并使法律救济变得极其困难。
Nevertheless, it is important to recognise that AI can also reduce human bias when properly audited. Unlike human recruiters, an algorithm does not suffer from fatigue or mood swings. With rigorous testing for fairness, diverse training data and ongoing human oversight, AI systems can support more equitable outcomes. The ethical path lies not in abandoning AI, but in designing governance frameworks that mandate regular audits, transparency reports and clear lines of accountability.
然而，也需认识到，在适当审计下，人工智能同样可以减少人类偏见。与人类招聘官不同，算法不会疲倦也不受情绪影响。通过对公平性的严格测试、多样化的训练数据以及持续的人类监督，人工智能系统能够支持更公平的结果。伦理之路不在于放弃人工智能，而在于设计强制要求定期审计、透明度报告和明确问责线的治理框架。
10. Model Essay: Choosing Between Sorting Algorithms | 范文：排序算法的选择
Question: 'Compare the efficiency of bubble sort and merge sort, and justify when each might be preferred.'
题目：'比较冒泡排序和归并排序的效率，并说明在什么情况下应优先选用哪一种。'
Sorting is a fundamental operation in computer science, and the choice of algorithm can dramatically affect performance. Bubble sort and merge sort represent two ends of the efficiency spectrum, and their suitability depends on data size, memory constraints and implementation context.
排序是计算机科学中的一项基本操作，算法的选择会显著影响性能。冒泡排序和归并排序分别代表了效率光谱的两端，它们是否适用取决于数据规模、内存限制和实现环境。
Bubble sort repeatedly steps through a list, comparing adjacent elements and swapping them if they are in the wrong order. Its average and worst-case time complexity is O(n²). For an array of 10 000 elements, this could entail approximately 50 million comparisons, making it extremely slow. However, bubble sort operates in-place and requires only O(1) additional memory. In very small data sets, such as sorting a handful of sensor readings on a microcontroller with 2 KB of RAM, its simple code and minimal memory footprint can make it a pragmatic choice.
冒泡排序会重复遍历一个列表，比较相邻元素并在顺序错误时交换它们。其平均和最坏情况的时间复杂度都是 O(n²)。对于一个包含一万个元素的数组，这可能需要进行大约五千万次比较，因而非常缓慢。然而，冒泡排序是原地排序，只需要 O(1) 的额外内存。在非常小的数据集里，比如在一个只有 2 KB 内存的微控制器上为少量传感器读数排序时，它简单的代码和极小的内存占用，使它成为一个实际的选择。
Merge sort adopts a divide-and-conquer approach, recursively splitting the list into halves until single-element sublists remain, then merging them back together in sorted order. Its time complexity is O(n log n) in all cases, which scales far better than bubble sort. For modern desktop applications processing millions of records, merge sort's efficiency is essential. The trade-off is its O(n) auxiliary space requirement, because it needs temporary arrays during the merge phase. On devices with abundant memory, this is acceptable; on embedded systems, it may be a constraint.
归并排序采用分治策略，递归地将列表分成两半，直到只剩下单元素子列表，然后再将它们合并成有序的序列。它在所有情况下的时间复杂度均为 O(n log n)，扩展性远胜于冒泡排序。对于需要处理数百万条记录的现代桌面应用来说，归并排序的效率是必不可少的。代价在于它需要 O(n) 的辅助空间，因为合并阶段需要临时数组。在内存充裕的设备上这是可以接受的，但在嵌入式系统中这可能是一个制约因素。
In conclusion, bubble sort is justified only for educational purposes or severely memory-constrained tiny data sets. Merge sort, or its optimised variants, is the default choice for general-purpose sorting where n is large and memory is available. The decision ultimately balances time complexity against space complexity within the specific system limits.
总之，冒泡排序仅在教学场景或内存极度受限且数据极小的情形下才有存在的理由。归并排序或其优化变体，才是当 n 较大且内存可用时通用排序的默认选择。最终的决策，其实就是在特定系统限制下对时间复杂度和空间复杂度做出的权衡。
11. Common Pitfalls and How to Avoid Them | 常见错误及避免方法
Even strong candidates lose marks through avoidable mistakes. Describing instead of analysing: if the question says 'evaluate', do not simply list features; give a supported judgement. Ignoring the counter-argument: a one-sided essay rarely reaches the top band. Always dedicate at least one paragraph to an opposing view. Vague technical statements: writing 'it is faster' earns no credit; instead, compare the growth rates using Big O with sample input sizes. Finally, poor time management: allocate roughly 1.5 minutes per mark and reserve the last five minutes for a quick proofread to catch missing notation or spelling errors in keywords.
即使是实力强的考生也会因一些可以避免的错误而丢分。描述代替分析：如果题目要求 'evaluate'，不要只是罗列特性，而要给出有依据的判断。忽视反方论点：片面论述的论文几乎不可能拿到最高档次。要确保至少用一个段落来讨论对立的观点。含糊的技术表述：写 'it is faster' 是拿不到分的；相反，要用大 O 符号结合示例输入规模来比较增长速率。最后，时间管理不佳：大致为每 1 分分配 1.5 分钟，并留出最后五分钟快速通读，以检查遗漏的符号或关键词拼写错误。
12. Revision Checklist for Essay Writing | 论文写作复习检查清单
Use this checklist when practising past-paper questions: (1) Have I underlined the command word and key terms? (2) Is my introduction focused and does it signpost the essay? (3) Does every body paragraph contain a distinct technical point supported by specific evidence? (4) Have I included at least one counter-argument where required? (5) Am I using correct notation such as O(log₂ n) and stating units where needed? (6) Have I linked each paragraph back to the question? (7) Is my conclusion a justified summary, not a repetition? Practising with this framework will make your writing more precise, confident and exam-ready.
练习历年真题时，请使用这份检查清单：（1）我划出指令词和关键术语了吗？（2）我的引言是否聚焦，并简要预告了全文内容？（3）每个主体段落是否都包含一个明确的技术论点，并有具体证据支持？（4）在需要时，我是否包含了至少一个反方论点？（5）我是否使用了正确的符号，如 O(log₂ n)，并在必要时注明了单位？（6）我是否把每个段落都联系回了问题？（7）我的结论是一个有理有据的总结，而不是对前文的简单重复吗？用这个框架多加练习，你的写作会更加精确、自信，并做好充分的考试准备。
Published by TutorHao | WJEC 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