📚 Year 10 Eduqas Computer Science: In-depth Past Paper Analysis | Eduqas 计算机科学10年级：历年真题深度解析
Eduqas GCSE Computer Science past papers are a goldmine for Year 10 students aiming to understand examiner expectations. This article dissects recurring themes, common pitfalls, and high-yield topics across Components 1 and 2. By examining real questions, you will learn how to apply theory to unfamiliar contexts, structure responses for maximum marks, and avoid the mistakes that cost grades year after year.
Eduqas GCSE 计算机科学的历年真题是10年级学生理解考官要求的绝佳资源。本文深度剖析试卷一和试卷二中反复出现的主题、常见失分点和高频考点。通过分析真实考题，你将学会如何将理论应用于陌生情境，如何组织答案以获得最高分数，并避免那些年复一年导致失分的错误。
Exam Structure at a Glance
- Component 1: Understanding Computer Science – 1 hour 45 minutes, 62.5% of total marks
- Component 2: Computational Thinking and Programming – 2 hours, 37.5% of total marks (on-screen exam)
考试结构速览
- 试卷一：计算机科学原理 – 1小时45分钟，占总分62.5%
- 试卷二：计算思维与编程 – 2小时，占总分37.5%（上机考试）
1. Decoding Assessment Objectives in Past Papers | 从真题中解读评估目标
Eduqas questions are mapped to three Assessment Objectives: AO1 (demonstrate knowledge and understanding), AO2 (apply knowledge to problems), and AO3 (analyse and evaluate). In Component 1, AO1 dominates short-answer questions on hardware and binary, while AO2 and AO3 appear in longer 6-8 mark questions on networks or ethics.
Eduqas 的试题对应三个评估目标：AO1（展示知识与理解）、AO2（应用知识解决问题）和 AO3（分析与评估）。在试卷一中，AO1 主导关于硬件和二进制的简答题，而 AO2 和 AO3 出现在有关网络或伦理的 6-8 分较长问题中。
Past papers show that students lose marks by writing purely descriptive answers when evaluation is required. For example, when asked to compare the environmental impact of cloud storage versus local servers, mere definition of cloud computing will only reach AO1. You must also weigh pros and cons with specific evidence to hit AO3.
历年真题表明，学生因在需要评估时只写出描述性答案而失分。例如，当被要求比较云存储与本地服务器对环境的影响时，仅仅定义云计算只能达到 AO1。你还必须用具体的证据权衡利弊，才能触及 AO3。
Common command words such as 'explain', 'compare', and 'evaluate' must be read carefully. An 'evaluate' question always expects a supported conclusion.
常见的指令词如“解释”、“比较”和“评估”必须仔细阅读。“评估”类问题总期待一个带有依据的结论。
2. Algorithm Tracing and Pseudocode Performance | 算法追踪与伪代码表现
Component 2 on-screen exams frequently ask you to trace pseudocode and produced outputs. A classic past-paper trap is failing to handle loops correctly when the counter increments at the start or end. Consider a while loop that checks a condition before executing; students often miss that the condition is re-evaluated after each full iteration.
试卷二的上机考试常要求追踪伪代码并给出输出结果。一个经典的真题陷阱是，当计数器在循环开始或结束时递增，学生无法正确处理循环。例如一个 while 循环在执行前检查条件；学生经常忽略每次完整迭代后条件会重新评估。
Examiners report that trace tables are often incomplete. Always include all variables mentioned, and update them line-by-line. A typical question might trace a linear search: you must show each comparison value and the index change.
考官报告显示追踪表常常不完整。务必包含所有涉及的变量，并逐行更新。一个典型问题可能追踪线性搜索：你必须展示每一次比较值和索引的变化。
Common algorithm pitfalls from past papers include misunderstanding the difference between count ← count + 1 and count ← count + num, and confusing selection (if) with iteration (for). In one exam, many candidates assumed that a repeat...until loop would execute the increment even if the condition was met, which is incorrect.
历年真题中常见的算法误区包括误解 count ← count + 1 与 count ← count + num 的区别，以及混淆选择结构（if）与迭代结构（for）。在一次考试中，许多考生认为 repeat...until 循环即使条件满足也会执行增量，这是不正确的。
3. Binary, Hexadecimal, and Data Representation | 二进制、十六进制与数据表示
Binary addition, two's complement, and hexadecimal conversions are virtually guaranteed on Component 1. Past papers reveal that marks are often lost on leading zeros and sign bits. For instance, representing -7 in an 8-bit two's complement requires finding positive 7 (00000111), flipping bits (11111000), and adding 1 to get 11111001.
二进制加法、二进制补码和十六进制转换几乎是试卷一的必考内容。真题显示，前导零和符号位经常导致失分。例如，用 8 位二进制补码表示 -7 需要找到正数 7（00000111），取反（11111000），再加 1 得到 11111001。
s 1+1=0 carry 1; 1+1+1=1 carry 1. Always label overflow if the result exceeds the bit width.
在进行二进制加法时，学生必须检查溢出和进位。一个常见错误是将 1+1+1 写成 10 而不是 11。正确的规则是 1+1=0 进位1；1+1+1=1 进位1。如果结果超出位宽，务必标注溢出。
Hexadecimal questions often ask you to convert a binary pattern into hex. Group bits in fours from the right. Past paper analysis shows that for an 8-bit number like 10110111, candidates split into 1011 (B) and 0111 (7) but forget to pad the second nibble to four bits. The correct answer is B7.
十六进制题常要求将二进制模式转换为十六进制。从右侧开始将位划分为四位一组。真题分析显示，对于像 10110111 这样的 8 位数，考生会分为 1011（B）和 0111（7），但忘记将第二个半字节补足四位。正确答案是 B7。
| Denary | Binary (8-bit) | Hexadecimal |
| 25 | 00011001 | 19 |
| -25 (two's comp.) | 11100111 | E7 |
十进/二进/十六进对比表
4. Data Structures and Arrays in Typical Questions | 数据结构与数组在典型题中的应用
Past paper questions on 1D and 2D arrays test indexing, assignment, and traversals. The Eduqas pseudocode uses square brackets with a zero-based index. Many students fail to distinguish between the array length and the highest index: for an array scores[5], the indices are 0 to 4.
历年真题中关于一维和二维数组的题目考查索引、赋值和遍历。Eduqas 的伪代码使用方括号和从零开始的索引。许多学生无法区分数组长度和最高索引：对于数组 scores[5]，索引是 0 到 4。
A common exam task is to find the maximum or minimum value. Past papers show that candidates often forget to initialise the max variable to the first element or to handle empty arrays. A robust algorithm sets max ← arr[0] and then iterates from index 1.
一个常见的考题任务是找出最大值或最小值。真题显示，考生经常忘记将 max 变量初始化为第一个元素，或者忽略了空数组的处理。健壮的算法设置 max ← arr[0]，然后从索引1开始迭代。
For 2D arrays, questions on grid-based games or pixel images are popular. A typical pitfall is mixing rows and columns. In grid[row, col], the first index is the row. Past paper examiners noted that students accessing grid[3][2] often reversed the coordinates.
对于二维数组，基于网格游戏或像素图像的问题颇受欢迎。一个典型陷阱是混淆行和列。在 grid[row, col] 中，第一个索引是行。历年考官的评语指出，学生访问 grid[3][2] 时常常把坐标颠倒了。
5. Programming Constructs and Error Spotting | 编程结构与错误识别
Component 2 includes a section where you must correct code or identify syntax and logic errors. Missing punctuation such as semicolons or using a single equals sign (assignment) instead of double equals (comparison) are the most common mistakes identified in examiners' reports.
试卷二包含一个部分，你需要纠正代码或找出语法和逻辑错误。丢失标点如分号，或使用单等号（赋值）而非双等号（比较）是考官报告中最常指出的错误。
Logic errors, such as infinite loops or off-by-one errors, also feature heavily. For example, a for loop with for i = 0 to 10 iterates 11 times, not 10. If you intend to process exactly 10 elements, use for i = 0 to 9.
逻辑错误，如无限循环或差一错误，也大量出现。例如，使用 for i = 0 to 10 的 for 循环迭代 11 次，而非 10 次。如果你打算处理恰好 10 个元素，应使用 for i = 0 to 9。
Past papers demand that you dry-run code with unexpected inputs. Try inputs like negative numbers, zero, or empty strings. This reveals hidden bugs and demonstrates AO2 application.
真题要求你用意外输入进行代码走查。尝试负数、零或空字符串输入。这会揭示隐藏的错误，并展示 AO2 的应用能力。
6. Computer Hardware, Storage, and System Performance | 计算机硬件、存储与系统性能
Questions on the fetch-decode-execute cycle are very frequent. You must be able to label the program counter, MAR, MDR, and accumulator, and describe each step sequentially. Past papers often ask "Explain how the fetch-decode-execute cycle processes an ADD instruction." Many students ignore the decode step and jump directly to execution.
关于取指-译码-执行周期的问题非常频繁。你必须能够标注程序计数器、MAR、MDR 和累加器，并按顺序描述每一步。真题常问“解释取指-译码-执行周期是如何处理一条 ADD 指令的。”许多学生忽略译码步骤，直接跳到执行。
Embedded systems appear in AO3 evaluation questions. A typical past question asks to evaluate the use of an embedded system in a washing machine. Strong answers discuss limited RAM, low power consumption, reduced cost, and difficulty of updates compared to a general-purpose computer.
嵌入式系统出现在 AO3 评估题中。一个典型的真题要求评估洗衣机中使用嵌入式系统的情况。出色的答案讨论了有限的 RAM、低功耗、低成本，以及与通用计算机相比难以更新。
Storage comparisons (magnetic vs. solid-state vs. optical) require precise technical language. Do not just say "SSD is faster"; state the absence of moving parts and lower latency. Examiner feedback highlights that vague comparisons cap marks at half.
存储设备比较（磁性、固态与光学）需要精确的技术语言。不要只说“SSD 更快”；要说明没有活动部件和更低的延迟。考官反馈强调，模糊的比较会使分数封顶在一半。
7. Networking, Topologies, and Protocols | 网络、拓扑与协议
Past papers test star, mesh, and bus topologies along with their resilience, cost, and scalability. A common error is stating that a single cable failure in a star network brings the whole network down; that is true for bus but not for star.
历年真题考查星形、网状和总线拓扑，以及它们的容错性、成本和可扩展性。一个常见错误是声称星形网络中单一电缆故障会导致整个网络瘫痪；这适用于总线拓扑而非星形。
| Topology | Resilience | Cost |
| Star | High (one link failure isolates one node) | High (cabling, switch) |
| Mesh | Very high (redundant paths) | Very high (many connections) |
| Bus | Low (backbone failure crashes network) | Low (single cable) |
拓扑特性对照表
Protocol questions often focus on TCP/IP and the layers model. You need to explain how packets are reassembled using sequence numbers and why packet switching is efficient. Exam reports show that students struggle to name layers: application, transport, internet, link.
协议题常聚焦于 TCP/IP 和分层模型。你需要解释数据包如何使用序号重新组装，以及为什么分组交换是高效的。考试报告显示学生很难说出各层名称：应用层、传输层、网际层、链路层。
8. Cybersecurity Threats and Prevention Methods | 网络安全威胁与防御方法
Malware types, social engineering, and network policies are examined almost every series. A classic past-paper question reads: "Describe how a phishing attack works and suggest two ways to protect against it." Many candidates describe 'phishing' as fishing in water, losing marks.
恶意软件类型、社会工程学和网络策略几乎每个系列都会考查。一个经典的真题是：“描述网络钓鱼攻击如何运作，并提出两种防护方法。”许多考生将“phishing”描述为在水中钓鱼，导致失分。
You must distinguish between a virus (self-replicating, attaches to files) and a worm (self-propagating without a host file). For prevention, always link methods to specific threats: firewalls filter packets based on rules, and anti-malware uses signature and heuristic detection.
你必须区分病毒（自我复制，附着到文件）和蠕虫（无需宿主文件自行传播）。对于防御，始终将方法与具体威胁关联：防火墙基于规则过滤数据包，反恶意软件使用特征码和启发式检测。
Brute-force attacks and denial-of-service are popular context-based questions. A strong answer explains how rate limiting and CAPTCHA mitigate brute force, while DoS defence involves traffic analysis and blackholing. Past papers reward cited definitions of acronyms like CAPTCHA.
暴力破解和拒绝服务是流行的情境类问题。一个出色的答案会解释速率限制和验证码如何缓解暴力破解，而 DoS 防御涉及流量分析和黑洞路由。历年真题会奖励引用首字母缩写 CAPTCHA 定义的做法。
9. Ethical, Legal, and Environmental Impact Questions | 伦理、法律与环境影响问题
Eduqas frequently includes an extended 8-mark debate question in Component 1 on topics such as AI surveillance or e-waste. Analysis of mark schemes shows that balanced arguments with at least two points for and two against, plus a justified conclusion, are essential for top marks.
Eduqas 常在试卷一中包含一道 8 分的扩展辩论题，主题如 AI 监控或电子垃圾。评分方案分析显示，至少两条正方论点和两条反方论点，外加有理有据的结论，是取得高分的关键。
Legal references must be precise: the Data Protection Act 2018, Computer Misuse Act 1990, and Copyright, Designs and Patents Act 1988. A common error is quoting outdated legislation or mixing up the acts. The Computer Misuse Act covers unauthorised access, not data theft.
法律引用必须精确：《2018年数据保护法》、《1990年计算机滥用法》和《1988年版权、设计与专利法》。常见错误是引用过时的法律或混淆各项法案。《计算机滥用法》涉及未经授权的访问，而非数据盗窃。
Environmental questions often require the carbon footprint of manufacturing vs. streaming. Past papers show that students fail to mention the embedded carbon in device manufacturing and only focus on operational energy. This limits AO3 marks.
环境问题常要求比较生产制造与流媒体播放的碳足迹。真题显示，学生未能提及设备制造中的隐含碳排放，只关注运行能耗。这限制了 AO3 得分。
10. Common Traps in Past Paper Mark Schemes | 历年评分方案中的常见陷阱
Examiners deduct marks for over-simplified answers. For instance, saying "RAM is volatile memory" without expanding that contents are lost when power is off often gains only one mark out of two. Always elaborate with 'so when the computer turns off...'.
考官会因答案过于简化而扣分。例如，只说“RAM 是易失性存储器”，却未说明断电时内容会丢失，通常只能从两分中得到一分。务必补充“因此当计算机关闭时……”。
gram code must follow Eduqas pseudocode conventions. Marks are lost for using = instead of ← using {} for comments instead of //. Review the official pseudocode guide before the exam.
程序代码必须遵循 Eduqas 的伪代码规范。用 = 而不是 ← 赋值，或使用 {} 而不是 // 进行注释，都会失分。考前务必复习官方伪代码指南。
For 6-mark compare questions, a table with clear headings (Feature, Solution A, Solution B) often scores higher than paragraphs. The mark scheme favours explicit comparative statements.
对于 6 分的比较题，带有清晰标题（特性、方案A、方案B）的表格通常比段落得分更高。评分方案偏好明确的对比陈述。
11. Practical Strategies from Top-Scoring Candidates | 高分考生的实战策略
Top performers in Component 2 use the on-screen exam environment to run and test pseudocode whenever allowed. They keep a rough programming area to sketch logic before committing to a final answer.
在试卷二中表现优异者会在允许的情况下利用上机考试环境运行和测试伪代码。他们会保留一个粗略的程序设计区，在确定最终答案前勾勒逻辑。
For Component 1, they allocate 1.5 minutes per mark. A 4-mark question gets about 6 minutes. This discipline prevents rushing the final 8-mark essay. They also underline command words to avoid misreading.
对于试卷一，他们会分配给每分 1.5 分钟。一道 4 分的题约花 6 分钟。这种自律防止了最后 8 分论述题的仓促作答。他们还会在指令词下划线以避免误读。
Common advice from examiner feedback: if a question says "state and explain," two distinct parts are needed. "State" gives a fact, "Explain" gives the reason or process. Many answers combine both and lose the state mark.
来自考官反馈的常见建议：如果一道题说“陈述并解释”，需要有两个不同的部分。“陈述”给出事实，“解释”给出原因或过程。许多答案混为一谈，丢了陈述分。
12. Reviewing Real Past Paper Segments | 回顾真实真题片段
Consider a Component 1 question: "The processor uses the fetch-decode-execute cycle. Describe what happens during the decode stage of the cycle when the processor is completing an instruction to add two values." Strong answer: "The Control Unit decodes the fetched instruction from the MDR. It interprets the opcode as ADD and identifies the operands (the addresses of the two values in memory)." Weak answer: "The instruction is broken down." The latter gains minimal marks.
考虑一道试卷一题目：“处理器使用取指-译码-执行周期。描述当处理器完成一条将两个值相加的指令时，在周期的译码阶段发生了什么。”优秀答案：“控制单元对从 MDR 中取出的指令进行译码。它将操作码解释为 ADD，并识别操作数（内存中两个值的地址）。”薄弱答案：“指令被分解。”后者只能获得极少的分数。
Another Component 2 segment asked to write pseudocode to find the average of marks above 50 in an array. Many candidates initialised total to 0 and increment count only inside the if block, but forgot to output total/count after the loop. Always check that your algorithm processes all required data.
另一道试卷二题目要求编写伪代码，找出一数组中高于 50 分的成绩平均值。许多考生将 total 初始化为 0，并只在 if 块内递增 count，但忘记在循环后输出 total/count。务必检查算法是否处理了所有必要数据。
These examples underscore that reading mark schemes and examiner comments is as valuable as attempting past papers. Patterns of penalised mistakes repeat, and you can pre-empt them.
这些例子强调，阅读评分方案和考官评语与练习真题同样有价值。被扣分的错误模式不断重复，你可以提前预防。
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