📚 IGCSE OCR Computer Science: Common Mistake Analysis | IGCSE OCR 计算机科学：易错题精讲
Many IGCSE OCR Computer Science candidates lose marks not because they don’t understand the concepts, but because they fall into predictable traps set by examiners. This article walks you through the most frequently misunderstood topics, typical errors, and how to avoid them. Each section focuses on a specific exam-style question, explains the common mistake, and provides a clear, correct approach. Mastering these will boost your confidence and your grade.
许多 IGCSE OCR 计算机科学考生丢分并非因为不理解概念，而是落入了出题人设下的常见陷阱。本文带你逐一剖析最容易混淆的知识点、典型错误及避坑方法。每一节围绕一道考试风格的题目展开，解释常见错误，并给出清晰正确的解法。掌握这些内容将提升你的信心和分数。
1. Binary Addition and Overflow | 二进制加法与溢出
Students often forget that when adding two 8-bit binary numbers, the result might require 9 bits. If the question says “using 8-bit registers”, an overflow occurs when the carry into the most significant bit (MSB) differs from the carry out. A typical mistake is to just write the 9-bit answer without discussing overflow, or to claim overflow whenever there is a carry out of the MSB, ignoring the carry in.
学生常常忘记，当两个 8 位二进制数相加时，结果可能需要 9 位。如果题目说“使用 8 位寄存器”，当最高有效位（MSB）的进位与出位不同时，就会发生溢出。典型错误是直接写下 9 位答案而不讨论溢出，或者只要 MSB 有进位出就声称溢出，却忽略了进位入的情况。
Example: Add 10101010₂ and 01100110₂ using 8-bit registers. Identify if overflow occurs.
示例：使用 8 位寄存器将 10101010₂ 与 01100110₂ 相加。判断是否发生溢出。
10101010
+ 01100110
──────────
100010000 (9 bits)
Carry in to MSB = 1, carry out = 1. Since they are equal, no overflow. The extra bit is ignored, but the 8-bit result is 00010000₂. Many incorrectly say overflow because a 9th bit appeared. The correct explanation: overflow only occurs when the sign bit is corrupted due to the sum of two numbers with the same sign producing a result with a different sign.
MSB 的进位入 = 1，进位出 = 1。由于它们相等，无溢出。多余位被忽略，8 位结果是 00010000₂。许多人因为出现了第 9 位而错误地认为有溢出。正确解释：溢出仅当两个同符号数相加产生了不同符号的结果，导致符号位损坏时才发生。
2. Identifying Logic Gates from Truth Tables | 从真值表识别逻辑门
A common exam question gives a truth table and asks which logic gate it represents. Candidates often misread the input order or confuse AND with NAND, OR with NOR. The trap is that the output column might be inverted compared to the standard gate. Always check if the output is 1 only when both inputs are 1 (AND), or 1 when at least one input is 1 (OR), and then see if it’s the opposite (NAND/NOR).
常见的考题是给出真值表，问它代表哪个逻辑门。考生经常看错输入顺序，或混淆 AND 与 NAND、OR 与 NOR。陷阱在于输出列可能与标准门相反。务必检查输出是否仅当两个输入均为 1 时为 1（AND），还是至少一个输入为 1 时为 1（OR），然后看它是否取反（NAND/NOR）。
Typical error: For a table showing 0,0→1; 0,1→1; 1,0→1; 1,1→0, many rush to say AND because they focus on the last row. Actually it is NAND, since AND would give 0,0,0,1.
典型错误：对于一个显示 0,0→1; 0,1→1; 1,0→1; 1,1→0 的表，许多人因关注最后一行而匆忙说是 AND。实际上它是 NAND，因为 AND 会给出 0,0,0,1。
To avoid errors, write the expected output of candidate gates next to the given table and compare systematically. Draw a quick AND truth table, then NOT it for NAND. This double-check prevents silly marks lost.
为了避免错误，在给定表格旁边写下候选门的预期输出，并系统地进行比较。快速画一个 AND 真值表，然后对其取反得到 NAND。这种双重检查可以防止无谓失分。
3. Tracing Pseudocode Loops | 追踪伪代码循环
Loops with counters and conditions inside pseudocode are a frequent source of mistakes. Students tend to miscount the number of iterations, especially with REPEAT…UNTIL loops where the condition is checked at the end. Forgetting that the loop body executes at least once is a classic blunder. Another pitfall is updating a variable inside a WHILE loop incorrectly, leading to an infinite loop in theory – examiners often test this by asking for the final value.
带有计数器和条件的循环是伪代码中常见的错误来源。学生往往会数错迭代次数，尤其是 REPEAT…UNTIL 循环，其条件在末尾检查。忘记循环体至少执行一次是一个典型失误。另一个陷阱是在 WHILE 循环内部错误地更新变量，导致理论上无限循环——考官常通过询问最终值来测试这一点。
Example:
count ← 0 total ← 0 WHILE count < 5 total ← total + count count ← count + 2 ENDWHILE OUTPUT total
Many assume the loop runs 5 times (count = 0 to 4). But count increments by 2, so values: 0,2,4 => loop stops when count=6. Total = 0+2+4 = 6. The correct output is 6, not 10. Always simulate step by step.
许多人假设计数器从 0 到 4 循环 5 次。但 count 每次加 2，所以值为：0,2,4 => 当 count=6 时停止循环。total = 0+2+4 = 6。正确输出是 6，而不是 10。务必逐步模拟。
4. Arrays vs Records | 数组与记录
Candidates often mix up the use cases for arrays (lists of same data type) and records (fields of possibly different types). A typical exam question gives a scenario like storing student names and marks, and asks which data structure is appropriate. Choosing an array where a record is needed loses marks because an array of name strings and a separate array of integers cannot guarantee the link between a specific name and mark. A record with fields "name" and "mark" keeps related data together.
考生常混淆数组（同类型数据列表）和记录（可能不同类型的字段）的使用场景。典型的考题给出存储学生姓名和成绩的场景，问哪种数据结构合适。在需要记录时选择了数组会失分，因为字符串姓名数组和另一个整数成绩数组无法保证特定姓名与成绩之间的关联。带有“name”和“mark”字段的记录将相关数据保存在一起。
Another error: using a 2D array to represent a record, which is valid but less efficient and harder to read. The answer expects explicit field names. In OCR pseudocode, records are created with Name = "" etc. Emphasize that records are for entities with attributes.
另一个错误：用二维数组来表示记录，虽然有效但效率低且难读。答案期望明确的字段名。在 OCR 伪代码中，记录用 Name = "" 等创建。要强调记录用于具有属性的实体。
5. Network Topologies: Star vs Mesh | 网络拓扑：星型与网状
When asked to recommend a topology for a given scenario, students frequently confuse the advantages. A star topology has a central switch; if it fails, the network goes down – that's a disadvantage. But it's cheaper and easier to manage. A full mesh provides redundancy but is expensive due to cabling. A common mistake is saying "star is more reliable than mesh" or "mesh is always faster" without context. The correct answer must relate to the scenario: e.g., "A hospital network needs maximum uptime, so full mesh is justified despite cost."
当被要求针对某个场景推荐拓扑时，学生经常混淆优缺点。星型拓扑有一个中央交换机；如果它出故障，整个网络瘫痪——这是一个缺点。但它更便宜且易于管理。全互联网状拓扑提供冗余，但因布线昂贵。常见错误是脱离上下文说“星型比网状更可靠”或“网状总是更快”。正确答案必须与场景相关：例如，“医院网络需要最大正常运行时间，因此尽管成本高，全互联网状是合理的。”
Also, be precise about partial mesh vs full mesh. Partial mesh is more common: only critical nodes have multiple connections. Use correct terminology: "dedicated connections", "single point of failure".
此外，要准确区分部分网状和全互联网状。部分网状更常见：只有关键节点有多条连接。使用正确术语：“专用连接”、“单点故障”。
6. Cybersecurity: Phishing vs Pharming | 网络安全：网络钓鱼与域欺骗
Social engineering threats are tested regularly. A classic mix-up is phishing versus pharming. Phishing involves fraudulent emails or messages that trick users into revealing personal information. Pharming redirects website traffic to a fake site without the user’s knowledge, often via DNS poisoning. Students often write "phishing is when a website is fake" – that's actually pharming. The distinction lies in the method: phishing is "pull" (user clicks a link), pharming is "push" (traffic automatically redirected).
社会工程威胁经常被考。典型的混淆是网络钓鱼与域欺骗。网络钓鱼涉及欺诈性电子邮件或消息，诱骗用户泄露个人信息。域欺骗在用户不知情下将网站流量重定向到虚假网站，通常通过 DNS 投毒实现。学生常写“网络钓鱼就是虚假网站”——那实际上是域欺骗。区别在于方法：网络钓鱼是“拉”（用户点击链接），域欺骗是“推”（流量自动重定向）。
Similarly, distinguish between a virus (needs host file) and a worm (self-replicating, spreads independently). In an exam question, describing a worm as "attached to an email" misses the mark – that's a virus. Worms exploit network vulnerabilities.
类似地，区分病毒（需要宿主文件）和蠕虫（自我复制、独立传播）。在考题中，将蠕虫描述为“附在电子邮件上”不得分——那是病毒。蠕虫利用网络漏洞。
7. Ethical and Legal Issues: Data Protection | 伦理与法律问题：数据保护
Questions about the Data Protection Act (DPA) and Computer Misuse Act (CMA) trip up many students. A common error is citing the wrong act. For hacking, the CMA is the relevant law; for storing personal data, it's the DPA. Another trap: explaining "what" the law says instead of "how" it applies to the scenario. For example, "The company must keep data secure" is a principle, but the examiner wants "The company must encrypt the database and restrict access to authorised staff, otherwise they breach the DPA."
关于《数据保护法》（DPA）和《计算机滥用法》（CMA）的题目会让许多学生出错。常见错误是引用错误的法案。对于黑客攻击，适用的是 CMA；对于存储个人数据，适用的是 DPA。另一个陷阱：解释法律“说什么”而不是“如何”应用于场景。例如，“公司必须保证数据安全”是一条原则，但考官想要的是“公司必须加密数据库并限制仅有授权人员访问，否则就违反了 DPA。”
Also, remember the eight principles of DPA: data must be fairly processed, used for specified purposes, adequate, accurate, not kept longer than necessary, processed in line with rights, secure, and not transferred without adequate protection. Be specific.
此外，记住 DPA 的八项原则：数据必须公平处理、用于特定目的、充分、准确、保存不超过必要时间、按权利处理、安全、不转移到无充分保护的国家。要具体。
8. CPU Components and the FDE Cycle | 中央处理器组件与取指-译码-执行周期
A typical 4-mark question asks to describe the Fetch-Decode-Execute (FDE) cycle. Students often omit the role of the Program Counter (PC) or confuse it with the Memory Address Register (MAR). The PC holds the address of the next instruction; it increments after fetch. MAR holds the address of the data/instruction currently being accessed. Many write "PC holds the instruction"; that's wrong – the CIR (Current Instruction Register) holds the instruction. Mislabeling registers costs easy marks.
一个典型的 4 分题要求描述取指-译码-执行（FDE）周期。学生常忽略程序计数器（PC）的作用，或将其与内存地址寄存器（MAR）混淆。PC 存放下一条指令的地址；它在取指后递增。MAR 存放当前正在访问的数据/指令的地址。许多人写“PC 存放指令”；这是错的——当前指令寄存器（CIR）存放指令。标错寄存器会白白失分。
For fetch: PC contents copied to MAR, address sent via address bus, read signal on control bus, instruction from RAM placed on data bus into MDR, then copied to CIR, PC incremented. Being precise with bus names and register transfers is crucial. Always mention the buses involved.
对于取指：PC 内容复制到 MAR，地址通过地址总线发送，控制总线发读信号，RAM 中的指令通过数据总线放入 MDR，然后复制到 CIR，PC 递增。准确说出总线名称和寄存器传输至关重要。始终提及所涉及的总线。
9. Storage Units Conversion | 存储单位换算
Converting between bytes, kilobytes, megabytes, etc., seems trivial, yet errors abound due to two factors: binary vs decimal prefixes, and calculation slips. OCR expects you to know that 1 KiB = 1024 bytes, 1 MiB = 1024 KiB. However, some questions use traditional KB = 1000 bytes for simplified scenarios, but exam mark schemes usually accept both if justified. The bigger mistake is forgetting to multiply when finding file sizes of images or sound. E.g., image size = width × height × bit depth (in bits), then convert to bytes.
字节、千字节、兆字节之间的换算看似简单，但因为两个因素错误频发：二进制与十进制前缀，以及计算失误。OCR 期望你知道 1 KiB = 1024 字节，1 MiB = 1024 KiB。然而，有些题目在简化场景中会用传统 KB = 1000 字节，但考试评分方案通常若说明理由两者都接受。更大的错误是在计算图像或声音文件大小时忘记乘法。例如，图像大小 = 宽 × 高 × 位深（以位为单位），然后转换为字节。
Common pitfall: For a 5-minute audio sampled at 44.1 kHz, 16-bit, stereo. Students compute: 44.1 × 1000 × 16 × 2 × 60 × 5 bits. They often forget the "× 60" for seconds or confuse kHz with Hz. Always write units clearly and break calculation into steps: sample rate → per second bits → per minute → total. Avoid rounding early.
常见陷阱：对于一段 5 分钟、采样率 44.1 kHz、16 位、立体声的音频。学生计算：44.1 × 1000 × 16 × 2 × 60 × 5 位。他们常忘记“× 60”来换算秒，或混淆 kHz 与 Hz。务必清晰写出单位，分步计算：采样率 → 每秒位数 → 每分钟 → 总计。避免过早舍入。
10. Trace Tables for Algorithms | 算法追踪表
Trace tables are intended to be straightforward, but rushing causes misaligned columns and missing variable updates. When given a pseudocode with a FOR loop, candidates sometimes forget to record the final value of the loop variable after exit, or they don't track all variables. Also, conditional statements that modify multiple variables must be evaluated with current values from the previous row, not from memory. Always record every line of execution in order, even if no change.
追踪表本应简单，但匆忙会导致列不对齐和遗漏变量更新。当给出带有 FOR 循环的伪代码时，考生有时忘记记录循环变量退出后的最终值，或没追踪所有变量。此外，修改多个变量的条件语句必须使用前一行中的当前值来评估，而非凭记忆。始终按顺序记录每一行执行，即使没有变化。
Example: FOR i ← 1 TO 3: a ← a + i; IF a > 4 THEN b ← b + 1 ENDIF ENDFOR. A rushed trace might skip the IF check for i=1 (a=1→ a=2, a>4 false) and update b incorrectly. Construct a table with columns for i, a, b, and output. Fill row by row. This guarantees accuracy.
示例：FOR i ← 1 TO 3: a ← a + i; IF a > 4 THEN b ← b + 1 ENDIF ENDFOR。草率的追踪可能会跳过 i=1 时的 IF 检查（a=1→ a=2，a>4 假），并错误更新 b。建一个包含 i、a、b 和输出的表格，逐行填写。这保证准确性。
11. High-Level vs Low-Level Languages | 高级语言与低级语言
When comparing high-level languages (HLL) and low-level languages (assembly/machine code), students often provide vague statements like "HLL is easier". The mark scheme expects specific advantages: HLL is portable, easier to debug, has built-in libraries, one statement translates to many machine instructions, and is problem-oriented. Assembly language gives direct hardware control, produces efficient, compact code, and is machine-specific. A common error is claiming assembly is "faster to write" – that's false. Another: saying machine code is written in hexadecimal – machine code is binary, hexadecimal is a representation.
在比较高阶语言（HLL）与低阶语言（汇编/机器码）时，学生常给出模糊说法如“高阶语言更简单”。评分方案期望具体优势：高阶语言可移植、易于调试、有内置库、一条语句对应多条机器指令、面向问题。汇编语言提供直接硬件控制、生成高效紧凑代码、特定于机器。常见错误是声称汇编“编写更快”——那是错的。另一个：说机器码用十六进制编写——机器码是二进制，十六进制只是一种表示。
Also, understand that an assembler translates assembly to machine code, a compiler translates HLL to machine code (whole program at once), an interpreter translates and executes line by line. Mixing these up is frequent.
此外，要理解汇编器将汇编翻译成机器码，编译器将高阶语言翻译成机器码（整个程序一次性），解释器逐行翻译并执行。混淆这些很常见。
12. Binary Shifts and Arithmetic | 二进制移位与算术
Binary shifts are used for fast multiplication/division by powers of 2. Left shift multiplies, right shift divides (integer division). Errors happen when students forget that a right shift on a negative number in two's complement might need to preserve the sign bit (arithmetic shift). OCR typically focuses on logical shifts for unsigned numbers, but it's wise to check the context. Another mistake: performing a shift but then forgetting to pad with zeros, or mixing up the direction.
二进制移位用于快速乘以/除以 2 的幂。左移乘，右移除（整数除法）。当学生忘记对补码表示的负数右移时可能需要保留符号位（算术移位）时就会犯错。OCR 通常关注无符号数的逻辑移位，但检查上下文是明智的。另一个错误：执行移位但忘记用零填充，或搞混方向。
Example: Multiply 00001100₂ (12) by 4. Left shift twice: 00110000₂ (48). If a student shifts three times by mistake, they get 96, which is multiplication by 8. Always verify the number of shifts: 2ⁿ → n shifts.
示例：将 00001100₂（12）乘以 4。左移两次：00110000₂（48）。如果学生错误地移了三次，得到 96，那是乘以 8。始终验证移位次数：2ⁿ → n 次移位。
To avoid sign errors, note that left shift can cause overflow; if the MSB changes from 0 to 1 in unsigned representation, that's fine, but for signed two's complement, it may indicate overflow. Always interpret shifts in the representation asked for.
为了避免符号错误，注意左移可能导致溢出；如果在无符号表示中 MSB 从 0 变为 1，这没问题，但对有符号补码，可能表示溢出。始终按题目要求的表示法解释移位。
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