📚 Year 12 OCR Computer Science: Exam Techniques and Mark Schemes | Year 12 OCR 计算机科学：答题技巧与评分标准
Mastering the content of the OCR AS Computer Science specification is only half the battle. To convert your knowledge into marks, you need to understand exactly what examiners are looking for and how to structure your responses. This guide breaks down the key command words, the way marks are allocated across the two components (Computer Systems and Algorithms & Programming), and the proven techniques that top students use to avoid losing easy marks. We will explore everything from writing precise definitions to tracing algorithms and drawing clean diagrams, ensuring you can approach every question with confidence.
掌握 OCR AS 计算机科学大纲的内容只是成功的一半。要把你的知识转化为分数，你必须准确理解考官想要什么，以及如何组织你的答案。本指南拆解了关键的命令词、两个组成部分（计算机系统与算法及编程）中分数的分配方式，以及优秀学生用来避免丢失简单分的久经考验的技巧。我们将探讨从写出精准的定义到追踪算法和绘制清晰的图表等方方面面，确保你能自信地应对每一道题目。
1. Understanding OCR Command Words | 理解 OCR 的命令词
OCR exam papers use specific command words that tell you the depth and structure of the answer required. Misreading a ‘describe’ question as an ‘explain’ question is one of the most common ways to waste marks. The three main command words at AS Level are Describe, Explain, and Compare. There are also shorter-answer commands like State, Give, Identify, and Define. Each one demands a different type of thinking and writing.
OCR 的试卷使用特定的命令词，它们告诉你答案所需的深度和结构。把 “描述” 题误读成 “解释” 题是浪费分数的最常见方式之一。在 AS Level 中，三个主要的命令词是 Describe、Explain 和 Compare。还有一些简答题命令词，如 State、Give、Identify 和 Define。每一个都要求不同的思维和写作方式。
| Command Word | Meaning | 命令词 | 含义 |
|---|---|---|---|
| State / Give / Identify | Recall a fact, term, or short phrase. No explanation needed. | 说出 / 给出 / 识别 | 回忆一个事实、术语或短语。不需要解释。 |
| Define | Give a precise meaning. Often requires a one-sentence formal definition. | 定义 | 给出精确含义。通常需要一句话的正式定义。 |
| Describe | Write what something is or what happens. Focus on characteristics, features, or steps. No reasons or causes. | 描述 | 写出某事是什么或发生了什么。关注特征、步骤。不涉及原因或缘由。 |
| Explain | Describe and then give reasons or causes. Use ‘because’, ‘so that’, ‘this means’. Show how or why something works. | 解释 | 先描述，再给出理由或原因。使用 “因为”、”以便”、”这意味着”。展示某物如何或为什么运作。 |
| Compare | Identify similarities and differences. You must compare from the same aspect, not just write two separate descriptions. | 比较 | 识别相同点和不同点。必须从同一方面进行比较，而不是分开描述两个事物。 |
Every time you read a question, circle the command word and stick to it. If the question says ‘Identify’, do not waste time explaining — you will gain no extra marks and may run out of time later.
每次读题时，圈出命令词并严格遵循。如果题目说 “识别”，不要浪费时间解释 —— 你不会得到额外分数，还可能导致后面时间不够。
2. How Marks Are Allocated in OCR AS Papers | OCR AS 试卷的分数分配方式
Understanding mark allocation helps you decide how much detail to write and what speed to work at. Both papers (Computer Systems, 70 marks, 1 hour 15 minutes; Algorithms & Programming, 70 marks, 1 hour 15 minutes) include short questions, structured questions, and some extended writing. In the Computer Systems paper, marks often come from accurate recall of definitions, diagrams, and comparison tables. In the Algorithms paper, marks heavily reward correct tracing, algorithm design, and logical reasoning. Each paper has Assessment Objectives: AO1 (Recall), AO2 (Apply), AO3 (Analyse/Evaluate). Roughly, AO2 and AO3 account for more than half the marks.
理解分数分配可以帮助你决定答案的详细程度和答题速度。两份试卷（计算机系统，70分，1小时15分钟；算法与编程，70分，1小时15分钟）都包含简答题、结构化试题和一些扩展写作题。在计算机系统试卷中，分数通常来自对定义、图表和比较表格的准确回忆。在算法试卷中，分数大量奖励正确的追踪、算法设计和逻辑推理。每份试卷都有评估目标：AO1（回忆）、AO2（应用）、AO3（分析与评价）。大致上，AO2 和 AO3 合计超过总分的一半。
For a 4-mark ‘Describe’ question, giving four separate, accurate points will usually secure full marks, even if they are not connected. For a 4-mark ‘Explain’ question, you often need two points each with a reason (2 marks for the explanation chain). Look at the number of marks as a clue: one mark per distinct point, definition, or cause-and-effect link.
对于一道4分的 “描述” 题，给出四个独立的、准确的要点通常就能拿满分，即使它们之间没有关联。对于一道4分的 “解释” 题，你通常需要两个要点，每个都带有原因（每个解释链条值2分）。把题目分值当作线索：每1分对应一个明确的要点、定义或因果关联。
3. Writing High-Scoring Describe Answers | 写出高分的描述类答案
Describe questions test your ability to recall the characteristics of a concept or the steps of a process. Do not add reasons unless the question also asks ‘and explain’. For example, ‘Describe the role of the CPU in the fetch-decode-execute cycle’ expects you to outline what happens at each stage, not why the cycle exists. Write in clear, sequential statements. Use bullet points if the question suggests structure, but always use full sentences in prose form if the paper expects it.
描述题考查的是你回忆一个概念的特征或一个过程的步骤的能力。除非题目也要求 “并解释”，否则不要添加原因。例如，”描述 CPU 在取指-译码-执行周期中的角色” 这道题，期望你概述每个阶段发生了什么，而不是为什么这个周期存在。用清晰、有顺序的陈述来写。如果题目暗示可以用结构，可以使用项目符号，但如果试卷要求，总是用完整的句子以散文形式书写。
Many students lose marks by being vague. Instead of ‘The ALU does calculations’, write ‘The Arithmetic Logic Unit (ALU) performs arithmetic operations such as addition and logical operations such as AND, OR, and NOT’. Specific terminology is rewarded. In topics like memory (RAM, ROM, virtual memory) or network topologies, name the exact components and their functions.
许多学生因表述含糊而丢分。与其写 “ALU 做计算”，不如写 “算术逻辑单元（ALU）执行加法等算术运算和 AND、OR、NOT 等逻辑运算”。具体的术语会得到奖励。在内存（RAM、ROM、虚拟内存）或网络拓扑等主题中，要说出确切的组件及其功能。
4. Crafting Effective Explain Answers | 精心构思有效的解释类答案
Explain questions require a cause-and-effect relationship. The examiner expects a statement of the fact and then the reasoning that links it to the outcome. Use connective words: ‘therefore’, ‘as a result’, ‘because’, ‘this means that’. A common pattern is: ‘X does … because … which leads to …’. For example, ‘Explain why increasing the clock speed of a processor improves performance’ should be answered as: ‘A higher clock speed means more cycles per second are completed (fact). This causes the processor to fetch and execute more instructions per second (reason). As a result, the overall execution time for a program decreases (outcome).’
解释题要求有因果关系。考官期待你先陈述事实，然后给出将其与结果联系起来的推理。使用连接词：”因此”、”结果是”、”因为”、”这意味着”。一个常见的模式是：”X 做了……，因为……，这导致了……”。例如，”解释为什么提高处理器的时钟速度会改善性能” 应该这样回答：”更高的时钟速度意味着每秒完成更多的周期（事实）。这使得处理器每秒能获取和执行更多的指令（原因）。因此，一个程序的整体执行时间减少了（结果）。”
Avoid circular explanations. Saying ‘It improves performance because it makes the computer faster’ adds nothing. Always link back to the underlying mechanism. For logic circuits, explain how the gates combine to produce the output. For virtual memory, explain how the use of secondary storage as extension of RAM prevents the computer from crashing when RAM is full, albeit at slower speeds.
避免循环解释。说 “它提高了性能，因为它让电脑更快了” 没有任何价值。始终要回到根本机制上。对于逻辑电路，解释门电路如何组合产生输出。对于虚拟内存，解释将辅助存储器用作 RAM 的扩展如何防止 RAM 满时电脑崩溃，尽管速度较慢。
5. Mastering Compare Questions with Parallel Points | 用平行要点掌握比较题
Compare questions are frequently answered poorly because students write a paragraph about the first item, then a paragraph about the second — this is a description, not a comparison. A comparison must discuss the same aspect for both items side by side. Use phrases like ‘Unlike X, Y …’, ‘Both X and Y …’, ‘X is … whereas Y is …’. Aim for at least two similarities and two differences for a typical 6‑mark compare question.
比较题常常答得很差，因为学生先写一段关于第一个事物的文字，再写一段关于第二个事物的文字 —— 这是描述，不是比较。比较必须并排讨论两个事物在同一个方面的异同。使用诸如 “与 X 不同，Y……”、”X 和 Y 都……”、”X 是……而 Y 是……” 的短语。对于一道典型的6分比较题，要至少找到两个相同点和两个不同点。
For instance, comparing RAM and ROM: ‘Both RAM and ROM are primary memory located close to the CPU. However, RAM is volatile, meaning it loses data when power is off, whereas ROM is non-volatile and retains data. RAM is used to store currently running programs and data, while ROM typically stores the boot program (BIOS).’ This structure directly addresses the mark scheme and shows analytical skill.
例如，比较 RAM 和 ROM：”RAM 和 ROM 都是位于 CPU 附近的主存储器。然而，RAM 是易失性的，意味着断电时数据会丢失，而 ROM 是非易失性的，能保留数据。RAM 用于存储当前正在运行的程序和数据，而 ROM 通常存储启动程序（BIOS）。” 这种结构直接切合评分方案，并展示了分析技巧。
6. Algorithm Tracing and Dry Runs | 算法追踪和手工执行
In Paper 2, tracing an algorithm is one of the most reliable ways to gain marks. You are often given an algorithm in pseudocode and asked to show the output or the contents of variables at each step. Always set up a trace table with a column for each variable. Work line by line, updating values carefully. Common errors include missing the order of operations in a expression like total ← total + count inside a loop, or forgetting that a WHILE loop checks the condition at the start.
在试卷二中，追踪算法是拿分最可靠的方式之一。题目通常会给出伪代码形式的算法，并要求你展示每一步的输出或变量的内容。始终建立一个追踪表，为每个变量设一列。逐行执行，仔细更新数值。常见错误包括在循环中的像 total ← total + count 这样的表达式里弄错运算顺序，或者忘记了 WHILE 循环是在开始时检查条件。
When writing your own algorithms, use the exact pseudocode style OCR expects. Use INDENTATION for loops and IF statements, and use the assignment operator ← rather than =. Clearly separate your solution into logical steps: input, processing, output. Marks are given for the correct idea, not for syntax like a programming language. However, the logic must be unambiguous. For example, if you need to find the maximum value in a list, initialise a variable maxVal ← list[0] and then compare each element. Write out the exact condition: IF list[i] > maxVal THEN maxVal ← list[i].
在书写自己的算法时，使用 OCR 期望的精确伪代码风格。用缩进表示循环和 IF 语句，并使用赋值运算符 ← 而不是 =。将你的解决方案清晰地分成逻辑步骤：输入、处理、输出。分数是给正确的思路，而不是像编程语言那样的语法。然而，逻辑必须明确。例如，如果你需要找列表中最大值，初始化一个变量 maxVal ← list[0]，然后比较每一个元素。写出精确的条件：IF list[i] > maxVal THEN maxVal ← list[i]。
7. Diagram Questions: CPU, Logic Gates, and Networks | 图表题：CPU、逻辑门和网络
Diagram questions reward clarity and labelling. When drawing a CPU architecture, include the ALU, Control Unit, registers (PC, MAR, MDR, ACC), and the buses (data, address, control). Always draw directional arrows on buses and label them. In logic gate circuits, draw the standard symbols for AND, OR, NOT, NAND, etc., and clearly show how inputs lead to outputs. For a truth table with n inputs, ensure you have 2ⁿ rows and that the binary input sequence is systematic.
图表题奖励清晰和标注。在画 CPU 架构时，要包括 ALU、控制单元、寄存器（PC、MAR、MDR、ACC）以及总线（数据总线、地址总线、控制总线）。始终在总线上画方向箭头并加以标注。在逻辑门电路中，画出 AND、OR、NOT、NAND 等的标准符号，并清楚地展示输入如何到达输出。对于有 n 个输入的真值表，确保你有 2ⁿ 行，并且二进制输入序列是系统的。
Network diagrams require you to draw devices like switches, routers, servers, and clients in a topology such as star or mesh. Label the transmission medium (e.g., fibre optic cable) and indicate data flow. A common trap is drawing a bus topology when the question explicitly asks for a star topology. Read the question carefully. Marks are often half for drawing and half for labelling.
网络图要求你画出像交换机、路由器、服务器和客户端等设备，并以某种拓扑（如星型或网状）连接。标注传输介质（例如光纤电缆），并指出数据流向。一个常见的陷阱是，题目明确要求星型拓扑时却画了总线拓扑。仔细读题。分数通常一半给画图，一半给标注。
8. Tackling Boolean Algebra and Logic Simplification | 攻克布尔代数与逻辑化简
Questions on Boolean algebra test your ability to manipulate expressions using the laws of Boolean logic. You must know the key axioms: A ∧ 0 = 0, A ∨ 1 = 1, A ∧ ¬A = 0, A ∨ (B ∧ C) = (A ∨ B) ∧ (A ∨ C), and De Morgan’s laws: ¬(A ∧ B) = ¬A ∨ ¬B. Marks are awarded for each correct substitution step. Show your working — if you make a slip, you can still pick up method marks.
布尔代数题考查你使用逻辑定律操作表达式的能力。你必须知道关键公理：A ∧ 0 = 0、A ∨ 1 = 1、A ∧ ¬A = 0、A ∨ (B ∧ C) = (A ∨ B) ∧ (A ∨ C)，以及德摩根定律：¬(A ∧ B) = ¬A ∨ ¬B。每个正确的替换步骤都有分数。展示你的推导过程 —— 如果你犯了个小错，仍然可以得到步骤分。
When simplifying, do one step at a time and write the law you used next to it, e.g., ‘A ∧ (A ∨ B) = A (absorption law)’. This proves your understanding and helps the examiner award marks. For circuit expressions, you might be asked to write an expression from a diagram or vice versa. Break the circuit into layers and label intermediate outputs.
化简时，一次只做一步，并在旁边写出你使用的定律，例如 “A ∧ (A ∨ B) = A（吸收律）”。这证明了你的理解，也帮助考官给分。对于电路表达式，你可能会被要求从图表写出表达式或将表达式画成电路。将电路分层，并标注中间输出。
9. Handling Questions on Data Representation and Binary | 应对数据表示与二进制问题
Data representation topics, including binary, hexadecimal, floating point, and character sets, appear in both papers. When asked to convert a denary number to binary or hex, show the column headings or division steps. In a subtraction using two’s complement, first find the complement of the subtrahend and add. For floating point, the mark scheme looks for correct signs, mantissa, and exponent, often in a specified number of bits. Practise normalisation: moving the binary point and adjusting the exponent accordingly.
数据表示的主题，包括二进制、十六进制、浮点数和字符集，在两份试卷中都会出现。当被要求将十进制数转换为二进制或十六进制时，展示列标题或除法步骤。在利用二进制补码做减法时，先求出减数的补码，然后相加。对于浮点数，评分方案检查正确的符号、尾数和阶码，通常要求指定位数。练习规范化：移动二进制小数点并相应调整阶码。
For character sets, do not just write ‘ASCII uses 7 bits’. Link it to the number of characters: ‘ASCII originally used 7 bits, allowing 2⁷ = 128 characters, whereas Unicode uses up to 32 bits, supporting over a million code points to represent global scripts and emojis.’ Such precise detail scores full marks.
对于字符集，不要只写 “ASCII 使用7比特”。要把它与字符数量联系起来：”ASCII 最初使用7比特，允许 2⁷ = 128 个字符，而 Unicode 最多使用32比特，支持超过一百万个码位，来表示全球文字和表情符号。” 这种精确的细节能拿满分。
10. Answering Cybersecurity and Network Questions | 回答网络安全和网络问题
Network and security questions often combine technical terms with real-world scenarios. When asked to ‘Identify a network security threat’, name the threat (e.g., DDoS attack, SQL injection, malware) and describe its mechanism. For ‘Explain how a firewall protects a network’, you must state that it examines incoming and outgoing packets against a set of rules, and can block or allow traffic. Use specific terminology like ‘packet filtering’, ‘stateful inspection’, ‘proxy server’. Do not confuse a firewall with encryption — they are entirely different.
网络和安全问题通常将技术术语与现实场景结合。当被要求 “识别一种网络安全威胁” 时，要说出威胁的名称（例如，DDoS 攻击、SQL 注入、恶意软件）并描述其机制。对于 “解释防火墙如何保护网络”，你必须说明它根据一组规则检查进出的数据包，并能阻止或允许流量。使用具体术语，如 “包过滤”、”状态检测”、”代理服务器”。不要将防火墙与加密混淆 —— 它们完全不同。
In the compare questions on networks, you might be asked to discuss client-server vs peer-to-peer networks. Use these pairs: ‘In a client-server network, files are stored centrally on a server, whereas in a peer-to-peer network, files are distributed across peers. Client-server is easier to manage and back up, but has a single point of failure. Peer-to-peer is cheaper to set up but can be less secure.’ Such structured answers hit the marking points directly.
在关于网络的比较题中，你可能会被要求讨论客户端-服务器网络与对等网络。使用这些对比对：”在客户端-服务器网络中，文件集中存储在服务器上，而在对等网络中，文件分布在各个对等点上。客户端-服务器更容易管理和备份，但存在单点故障。对等网络搭建成本低，但可能安全性较差。” 这种结构化的答案直接命中了得分点。
11. Time Management and Question Selection | 时间管理与选题策略
With only 75 minutes per paper, time pressure is real. As a rule of thumb, spend one minute per mark, but allocate a little less time on quick recall and more on extended answers. In the Algorithms paper, don’t get stuck on a tricky trace question — skip and return to it after completing the algorithm design question, which often has more marks that are easier to secure if you keep a clear head.
每份试卷只有75分钟，时间压力是真实存在的。根据经验，每1分花1分钟，但可以在快速回忆题上少花点时间，而在扩展题上多花点时间。在算法试卷中，不要卡在一道棘手的追踪题上 —— 先跳过去，完成算法设计题后再回来做，算法设计题通常分值更高，而且只要头脑清醒，更容易得分。
Before starting any question, read the whole paper for 2 minutes. Note which topics appear and where you feel strongest. Always answer the questions you find easiest first to secure the marks and build confidence. In extended writing questions (e.g., 8‑mark ‘Explain’ or ‘Discuss’), plan your answer for 1 minute using a quick spider diagram. This ensures a logical flow and prevents rambling.
在开始答任何题之前，花2分钟浏览全卷。记下出现了哪些主题，以及你感觉最有把握的是哪些。总是先回答你觉得最简单的题目，以确保拿到分数并建立信心。在扩展写作题（例如，8分的 “解释” 或 “讨论”）中，花1分钟用快速蛛网图进行规划。这能确保逻辑流畅，防止东拉西扯。
12. Avoiding Common Pitfalls and Final Checks | 避免常见陷阱与最终检查
The most common exam mistakes include writing too little for the mark count, giving a description when an explanation is required, missing out units (e.g., writing ‘3’ instead of ‘3 GHz’), and leaving out the final output statement in an algorithm. Also, many students misinterpret pseudocode because they read it too quickly. Pay special attention to indentation and loop boundaries. For example, a WHILE loop with condition count ≤ 10 will iterate 10 times, not 11, if it starts at 1 and increments.
最常见的考试失误包括：写的字数与分值不匹配；需要解释时却写了描述；漏掉了单位（例如写 “3” 而不是 “3 GHz”）；以及在算法中遗漏最终的输出语句。还有，许多学生误读伪代码，因为他们读得太快。要特别注意缩进和循环边界。例如，条件为 count ≤ 10 的 WHILE 循环，如果从1开始并递增，那么它会迭代10次，而不是11次。
If you have time at the end, review your answers and check for these four things: (1) Have you used correct technical terminology? (2) Did you respond to the command word? (3) Are your trace tables complete and consistent? (4) Have you drawn and labelled all diagrams? Even two minutes of checking can recover 5% – 10% of marks.
如果最后还有时间，回顾你的答案并检查这四点：（1）你是否使用了正确的技术术语？（2）你是否回应了命令词的要求？（3）你的追踪表是否完整且一致？（4）你是否画图并标注了所有图表？即使只有两分钟的检查，也能挽回5% – 10%的分数。
Finally, use past papers from the OCR website or your teacher, and always mark them against the official mark scheme. Note the exact phrasing that examiners reward and incorporate it into your own answers. This active revision method is the fastest way to improve your grade.
最后，使用来自 OCR 网站或你的老师的历年真题，并始终对照官方评分方案进行批改。记下考官奖励的确切措辞，并将其融入到你自己的答案中。这种主动的复习方法是提高成绩的最快途径。
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