📚 Pre-U Cambridge Computer Science: Exam Techniques and Marking Criteria | Pre-U Cambridge 计算机：答题技巧与评分标准
The Cambridge Pre-U Computer Science examination is a rigorous assessment that tests not only your theoretical knowledge but also your ability to apply and evaluate concepts. Success hinges on mastering exam techniques and understanding precisely how marks are allocated. This article provides a comprehensive guide to the marking criteria and effective strategies for each question type.
剑桥 Pre-U 计算机科学考试是一项严格的评估，它不仅考查你的理论知识，还考查你应用和评估概念的能力。成功的关键在于掌握考试技巧并准确理解评分的分配方式。本文全面介绍了评分标准以及应对每种题型的有效策略。
1. Understanding Assessment Objectives | 理解评估目标
Pre-U Computer Science papers assess three main assessment objectives (AOs): AO1 (Knowledge and Understanding) accounts for roughly 30%, AO2 (Application) for 40%, and AO3 (Analysis and Evaluation) for 30%. Knowing this distribution helps you tailor your answers—AO1 questions require precise recall, AO2 demands deploying concepts in given scenarios, and AO3 expects critical judgement.
Pre-U 计算机科学试卷评估三个主要的评估目标（AO）：AO1（知识与理解）约占 30%，AO2（应用）占 40%，AO3（分析与评估）占 30%。了解这一分布有助于你调整答题方式——AO1 题目要求准确回忆，AO2 要求在给定场景中运用概念，AO3 则期待作出批判性判断。
For AO1, you must define terms like ‘virtual memory’ or list stages of the fetch–decode–execute cycle with flawless accuracy. Use the exact wording from the specification; paraphrasing may cause ambiguity. For AO2, practice applying algorithms (e.g., Dijkstra or binary search) to unfamiliar data, and for AO3, learn to justify choices, such as why one data structure outperforms another in a specific context.
对于 AO1，你必须准确无误地定义诸如 “虚拟内存” 等术语或列出取指–译码–执行周期的各个阶段。使用考纲中的精确措辞；转述可能造成歧义。对于 AO2，练习将算法（如 Dijkstra 或二分搜索）应用于不熟悉的数据，而对于 AO3，要学会论证选择的原因，例如为何某种数据结构在特定情境下优于另一种。
2. Deconstructing Command Words | 解构指令词
Command words such as ‘State’, ‘Describe’, ‘Explain’, ‘Compare’, and ‘Evaluate’ signal the depth required. ‘State’ needs a short, factual answer (1‑2 marks); ‘Describe’ requires a sequence of characteristics or steps (3‑4 marks); ‘Explain’ demands reasons or causes; ‘Compare’ looks for similarities and differences; ‘Evaluate’ requires a supported judgement. Match your sentence starters to the command: e.g., ‘One similarity is… another is…’ for Compare.
指令词如 “State”、”Describe”、”Explain”、”Compare” 和 “Evaluate” 提示所需的深度。”State” 需要简短的事实性回答（1‑2 分）；”Describe” 需要一系列特征或步骤（3‑4 分）；”Explain” 要求给出原因或起因；”Compare” 寻找相似点和不同点；”Evaluate” 要求作出有依据的判断。将你的开头句式与指令对应：例如，对于 “Compare” 可用 “一个相似点是……另一个是……”。
Many candidates lose marks by explaining when only description is needed, or by listing without giving reasons for ‘Explain’. Underline the command word in every question. The table below summarises common Pre-U command words and the expected response structure.
许多考生因题目只要求描述而进行解释，或对 “Explain” 题仅罗列而不给理由而丢分。务必划出每条题目中的指令词。下表总结了 Pre-U 常见指令词及其期待的回答结构。
| Command Word | Expected Response |
|---|---|
| State / Identify | Short, specific fact or name |
| Describe | Ordered characteristics or process steps |
| Explain | Reasons, causes, or how/why something works |
| Compare | Similarities and differences, using comparative language |
| Evaluate / Discuss | Advantages, disadvantages, and a justified conclusion |
3. Time Allocation Strategies | 时间分配策略
Paper 1 (Computer Systems) and Paper 2 (Algorithms and Programming) each last 2 hours and carry 75 marks. A practical rule is to spend roughly 1.5 minutes per mark. For a 7‑mark question, allocate about 10-11 minutes. Reserve 5‑10 minutes at the end to check your answers, especially numerical questions and OMR grids if any.
试卷一（计算机系统）和试卷二（算法与编程）各持续 2 小时，满分 75 分。一个实用的法则是大约每分 1.5 分钟。对于 7 分的题目，分配约 10‑11 分钟。最后留出 5‑10 分钟检查答案，特别是数字题和答题卡（若有）。
Begin with the questions you find easiest to secure early marks and build confidence. If you are stuck on a low‑mark question, mark it with a star and move on; return if time permits. Never leave an answer blank—even a partial definition or a brief pseudocode snippet can earn marks if it addresses part of the requirement.
从你认为最简单的题目入手，以锁定早期分数并建立信心。如果被某个低分题卡住，打个星号并继续前进；若有时间再回来。绝不要留空——即使是一个不完整的定义或一小段伪代码，只要触及部分要求，都可能得分。
4. Answering Structured Theory Questions | 回答结构化理论题
Theory questions on topics like operating systems, databases, or networking often ask you to describe a concept and then apply it. Write in clear paragraphs, using bullet points only if the question explicitly invites them. For a ‘Describe the role of the scheduler’ question, start by stating what the scheduler does, then mention key functions (e.g., process dispatching, context switching), and finally link to a real‑world analogy if it helps.
关于操作系统、数据库或网络等主题的理论题，常要求你先描述一个概念再应用之。用清晰的段落书写，除非题目明确允许，否则不要使用项目符号。对于 “Describe the role of the scheduler” 一题，先说明调度程序做什么，然后提及关键功能（如进程分派、上下文切换），最后若有助于理解可联系现实世界的类比。
Always incorporate technical vocabulary: ‘interrupt service routine’, ‘transaction consistency’, ‘normalisation to 3NF’ etc. Examiners reward precision. If a question is worth 6 marks, aim to make at least 6 distinct, relevant points, each linked directly to the stimulus material or specification context.
始终融入技术词汇：”中断服务程序”、”事务一致性”、”规范化至第三范式” 等。考官会奖励精确性。若一道题值 6 分，则争取给出至少 6 个清晰、相关的要点，每个要点直接关联题干材料或考纲语境。
5. Tackling Programming Questions | 应对编程题
Programming questions in Paper 2 often present an algorithm or incomplete code snippet. You may be asked to trace the code, identify errors, or write a new function in a language‑independent pseudocode (the Pre-U uses a structured English style). Always read the whole code segment before attempting any part. For trace tables, be systematic: label columns with variable names, and update row by row.
试卷二中的编程题常给出一段算法或不完整的代码片段。你可能需要追踪代码、找出错误或用一种不依赖特定语言的伪代码（Pre-U 采用结构化英语风格）编写新函数。务必先通读整段代码再着手任何部分。对于追踪表，要有条理：用变量名标注列，并逐行更新。
When writing new pseudocode, ensure you use consistent indentation, explicit loops (FOR, WHILE), and clear conditional structures (IF…THEN…ELSE…ENDIF). Comments are not mandatory but can clarify your logic if time permits. Even if you cannot solve the whole task, partial solutions that demonstrate correct input/output or loop structure can gain significant marks under AO2 and AO3.
书写新伪代码时，确保使用一致的缩进、明确的循环（FOR、WHILE）和清晰的条件结构（IF…THEN…ELSE…ENDIF）。注释不是强制要求，但时间允许时能说明你的逻辑。即便不能解出整个任务，展示正确输入/输出或循环结构的部分方案，也能在 AO2 和 AO3 下获得可观分数。
6. Handling Calculations and Binary Manipulations | 处理计算与二进制操作
Questions on binary/hexadecimal conversions, two’s complement, floating‑point representation, or Boolean algebra appear frequently. Show all your working steps clearly; even if the final answer is wrong, method marks are available. For example, when converting a negative number to two’s complement, write down the binary of the positive magnitude, invert bits, and add one—label each step.
有关二进制/十六进制转换、补码、浮点数表示或布尔代数的题目频繁出现。清晰地展示所有工作步骤；即便最终答案错误，方法分仍可获得。例如，将负数转换为补码时，写下正数幅值的二进制，按位取反，再加一——每一步都标注出来。
For Boolean expression simplification, use laws (De Morgan, distributive, etc.) systematically and state which law you are applying. In logic circuit diagrams, ensure your gates are neatly drawn and connections labelled. Practice with past papers to gain speed in floating‑point normalisation (mantissa and exponent adjustments), a common time‑consuming task.
在简化布尔表达式时，系统地使用定律（德摩根、分配律等），并说明应用了哪条定律。在逻辑电路图中，确保门画得整洁，连线标注清楚。通过历年真题练习提升浮点数规格化（尾数和指数调整）的速度，这是一个常见且耗时的任务。
7. The Mark Scheme and Common Pitfalls | 评分方案与常见陷阱
Official mark schemes reveal that marks are often awarded for specific keywords. For example, describing a firewall might require ‘packet filtering’ and ‘stateful inspection’. Missing these terms can cost marks even if your description is otherwise correct. Compare your practice answers with mark schemes to identify the exact phrasing examiners expect.
官方评分方案显示，分数往往给予特定的关键词。例如，描述防火墙可能需要 “包过滤” 和 “状态检测”。漏写这些术语可能丢分，即便你的描述在其他方面是正确的。将你的练习答案与评分方案对照，识别考官期待的确切措辞。
A pitfall in AO3 questions is making unsupported claims. Saying ‘a stack is faster than a queue’ without rationale earns no credit. Always justify with properties: ‘A stack provides LIFO access which suits depth‑first search because it backtracks naturally, whereas a queue’s FIFO structure fits breadth‑first traversal.’
在 AO3 题中的一个陷阱是作出无依据的断言。仅仅说 “堆栈比队列快” 而不给理由，是得不到分数的。始终用属性进行论证：”堆栈提供 LIFO 访问，适合深度优先搜索，因为它自然地回溯；而队列的 FIFO 结构适合广度优先遍历。”
8. Using Precise Technical Terminology | 使用精确的技术术语
Accuracy in terminology separates high achievers. Write ‘multiplexer’ not ‘device that selects signals’; use ‘relational database’ and ‘foreign key’ correctly. In the topic of object‑oriented programming, distinguish between ‘class’ and ‘object’, ‘inheritance’ and ‘composition’. A table of commonly confused pairs is useful for revision.
术语的准确性能区分优等生。写 “多路复用器” 而非 “选择信号的设备”；正确使用 “关系数据库” 和 “外键”。在面向对象编程主题中，区分 “类” 与 “对象”、”继承” 与 “组合”。复习时整理一份常见易混淆术语对表格会很有帮助。
When referring to protocols, specify the full name (e.g., ‘Transmission Control Protocol (TCP)’) at first mention, then abbreviate. In network layers, mention ‘OSI model’ or ‘TCP/IP stack’ consistently. Examiners are strict about spelling of key terms; ‘von Neumann’ must be capitalised correctly.
当提及协议时，首次出现使用全称（如 “传输控制协议 (TCP)”），然后可缩写。在网络分层中，一致地使用 “OSI 模型” 或 “TCP/IP 协议栈”。考官对关键术语的拼写非常严格；”冯·诺依曼” 的大小写必须正确。
9. Approaching Extended Response Questions | 应对扩展回答题
Extended responses (8‑12 marks) require structured essays. Plan on a rough note page: outline 3‑4 main paragraphs, each making a distinct point. For an ‘Evaluate the use of virtualisation in a data centre’ question, you might cover: definition, benefits (resource utilisation, isolation), drawbacks (overhead, single point of failure), and a reasoned conclusion. Link paragraphs with cohesive devices.
扩展回答题（8‑12 分）需要结构化的论述。在草稿纸上粗略规划：列出 3‑4 个主要段落，每段阐述一个独立观点。对于 “评估虚拟化在数据中心的应用” 一题，可涵盖：定义、优点（资源利用率、隔离性）、缺点（开销、单点故障）以及一个有理由的结论。使用衔接手段连接段落。
Depth is more important than breadth. A detailed exploration of two strong arguments with a nuanced conclusion often scores higher than a surface‑level list of many points. Use real‑world examples, such as ‘Amazon EC2 uses virtualisation to provision scalable instances’, to show applied knowledge (AO2). Always conclude by directly answering the question.
深度比广度更重要。深入探讨两个有力论据并给出细致入微的结论，通常比浮于表面地罗列许多要点得分更高。运用现实案例，如 “Amazon EC2 利用虚拟化技术提供可扩展的实例”，以展示应用知识（AO2）。务必以直接回答问题的方式作结。
10. Final Revision and Exam-Day Tips | 考前复习与考试日提示
In the final weeks, focus on practising full past papers under timed conditions, then self‑mark using the mark schemes. Maintain an ‘error log’ noting repeated mistakes—like forgetting to normalise floating‑point results or mixing up B‑trees and B+ trees—and review it regularly. Prioritise topics that appear frequently: data structures, sorting algorithms, CPU architecture, and network protocols are perennial favourites.
在最后几周，重点是在计时条件下完整地练习历年真题，然后用评分方案自行批改。准备一本 “错题日志”，记录反复出现的错误——如忘记规格化浮点结果或混淆 B 树与 B+ 树——并定期复习。优先处理高频主题：数据结构、排序算法、CPU 架构和网络协议是常年热门内容。
On exam day, read the entire paper during the first 5 minutes. Check for optional questions or confusing phrasing. Bring a clear pencil for diagrams, a ruler for logic circuits, and an approved calculator (where permitted). Keep an eye on the clock, but do not panic—calm, methodical answers consistently outperform rushed brilliance.
考试当天，用前 5 分钟通读全卷。检查有无选做题或含混的表述。带上清晰的铅笔用于图表、一把尺子用于逻辑电路，以及允许使用的计算器（若许可）。留意时间，但不要慌乱——冷静、有条不紊的作答始终胜过匆忙的灵光一现。
Published by TutorHao | Computer Science 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