📚 AS CIE Computer Science: Exam Techniques and Marking Criteria | AS CIE 计算机：答题技巧与评分标准
Success in AS CIE Computer Science (9618) relies not only on knowing the theory and programming constructs, but also on understanding how examiners award marks. This article breaks down the key assessment objectives, command words, and practical strategies for both Paper 1 (Theory Fundamentals) and Paper 2 (Problem-solving and Programming). By mastering these techniques, you can avoid common pitfalls and present your answers exactly as the mark scheme expects.
要在 AS CIE 计算机科学 (9618) 中取得成功，不仅取决于掌握理论和编程结构，还取决于理解考官如何评分。本文详细分析了主要的评估目标、指令词，以及 Paper 1（理论基础）和 Paper 2（问题解决与编程）的实用策略。通过掌握这些技巧，你可以避开常见陷阱，完全按照评分标准的要求呈现答案。
1. Understanding the Exam Structure | 理解考试结构
AS Computer Science consists of two papers. Paper 1 tests knowledge of information representation, communication, hardware, processor fundamentals, security, and ethics. Paper 2 focuses on algorithm design, programming, databases, and Boolean logic. Knowing the weight of each topic helps you allocate revision time efficiently.
AS 计算机科学包含两套试卷。Paper 1 考查信息表示、通信、硬件、处理器基础、安全和伦理等知识。Paper 2 则侧重算法设计、编程、数据库和布尔逻辑。了解每个主题的分值权重有助于你高效分配复习时间。
Paper 1 has a duration of 1 hour 30 minutes and is worth 75 marks, with a mix of short-answer and structured questions. Paper 2 lasts 1 hour 45 minutes and carries 75 marks, requiring you to write pseudocode, draw flowcharts, and correct or complete program code. Familiarity with the question styles prevents surprises on exam day.
Paper 1 考试时间为 1 小时 30 分钟，满分 75 分，题型包括简答题和结构化问题。Paper 2 时长为 1 小时 45 分钟，满分 75 分，要求你撰写伪代码、绘制流程图，以及修正或补全程序代码。熟悉题型风格可以避免考试当天措手不及。
2. Command Words and What They Mean | 指令词及其含义
Every question uses command words such as ‘state’, ‘describe’, ‘explain’, ‘discuss’, ‘identify’, or ‘compare’. These words tell you the depth and style of response needed. Misinterpreting them is one of the quickest ways to lose marks.
每道题目都使用诸如“state”、“describe”、“explain”、“discuss”、“identify”或“compare”等指令词。这些词指示了你需要给出什么样的回答深度和风格。误解这些指令词是导致失分最快的途径之一。
‘State’ requires a brief, factual answer, often a single term or value. ‘Identify’ is similar but may ask you to pick out items from a given context. For ‘describe’, you must give a step-by-step account or characteristics without necessarily giving reasons. ‘Explain’ demands a clear cause-and-effect relationship: you must say why something happens. ‘Compare’ needs both similarities and differences, ideally supported with examples.
“State”要求给出简短、事实性的答案，通常是一个术语或数值。“Identify” 与之类似，但可能要求你从给定语境中选出项目。对于“describe”，你必须给出逐步的描述或特征，而不一定要给出原因。“Explain”则要求明确的因果关系：你必须说明某事为何发生。“Compare”需要同时指出相似点和不同点，最好辅以例子。
| Command Word | Expected Action | 指令词 | 期望回应 |
| State / Identify | Short, precise fact / name | 陈述 / 识别 | 简短、准确的事实 / 名称 |
| Describe | Account of steps or features | 描述 | 步骤或特征的叙述 |
| Explain | Reasoning and cause-and-effect | 解释 | 推理与因果关系 |
| Compare | Similarities and differences | 比较 | 相似点与不同点 |
| Discuss | Balanced review with pros/cons | 讨论 | 包含优缺点的平衡论述 |
Print the command word table and keep it beside you when attempting past papers. Before writing, underline the command word and think, ‘What exactly does the examiner want from me here?’ This small habit alone can improve your score significantly.
把指令词表格打印出来，做历年试卷时放在身边。提笔之前，在指令词下划线并思考：“考官究竟希望我回答什么？” 仅凭这个小习惯就能显著提高你的分数。
3. Answering Short-Answer Questions | 简答题答题技巧
Short-answer questions in Paper 1 often carry 1-3 marks. The number of marks is a direct guide to how many distinct points or steps to provide. A 2-mark question typically expects two separate pieces of information, not one point elaborated twice.
Paper 1 中的简答题通常分值为 1-3 分。分数直接提示你应该提供几个独立的要点或步骤。一个 2 分的题目通常期望两个独立的信息，而不是将一个要点详细阐述两遍。
If a question asks you to ‘State three reasons for …’, list exactly three, and label them clearly. Bullet points are acceptable and help the examiner identify each mark-bearing item. Avoid writing a dense paragraph where your distinct points could be missed.
如果题目要求“State three reasons for …”，就准确列出三点，并清晰地标明它们。可以使用项目符号，这有助于考官识别每个给分点。避免写一大段密密麻麻的文字，导致你的独立要点被遗漏。
When giving definitions, use precise technical terms as they appear in the syllabus. For example, define ‘protocol’ as ‘a set of rules governing communication between devices’, not ‘the way computers talk to each other’. The mark scheme is strict about key phrases.
下定义时，务必使用教学大纲中的精确技术术语。例如，将“protocol”定义为“a set of rules governing communication between devices”，而不是“the way computers talk to each other”。评分标准对关键短语非常严格。
4. Tackling Pseudocode and Flowchart Questions | 处理伪代码和流程图题
Paper 2 heavily tests algorithmic thinking through pseudocode and flowcharts. You must be comfortable with the CIE-preferred pseudocode syntax, which closely resembles Pascal. This includes keywords like WHILE…DO…ENDWHILE, IF…THEN…ELSE…ENDIF, FOR…TO…NEXT, OUTPUT, and INPUT.
Paper 2 通过伪代码和流程图大幅考查算法思维。你必须熟悉 CIE 偏好的伪代码语法，它非常接近 Pascal。这包括诸如 WHILE…DO…ENDWHILE、IF…THEN…ELSE…ENDIF、FOR…TO…NEXT、OUTPUT 和 INPUT 等关键字。
Consistency in indentation matters. Always indent the body of loops and selection constructs. The examiner does not explicitly deduct marks for ragged layout, but a clean structure reduces logical errors and makes it easier for you to trace variables. Use meaningful variable names like StudentMark, not x.
保持缩进一致很重要。循环和选择结构的函数体始终要缩进。考官不会明确扣分，但整洁的结构能减少逻辑错误，并让你更容易追踪变量。使用有意义的变量名，如 StudentMark，而非 x。
For flowchart questions, use the correct symbols: oval for Start/End, parallelogram for Input/Output, rectangle for Process, diamond for Decision. Arrows must show direction. When a question asks ‘Draw a flowchart’, never substitute it with pseudocode, as marks are specifically allocated to the visual representation.
对于流程图题，要使用正确的符号：椭圆形代表开始/结束，平行四边形代表输入/输出，矩形代表处理过程，菱形代表判断。箭头必须标出方向。当题目要求“Draw a flowchart”时，切勿用伪代码替代，因为分数是专门分配给可视化表示的。
5. Programming Questions in Paper 2 | Paper 2 编程题技巧
Programming questions may require you to write, complete, or debug code. Even though you are writing pseudocode, you must demonstrate understanding of core programming concepts: sequence, selection, iteration, data types, arrays, and string manipulation. Marks go to correct logic, not minor syntax flaws, but avoid careless mistakes like uninitialized variables.
编程题可能要求你撰写、补全或调试代码。尽管你写的是伪代码，但必须展示对核心编程概念的理解：顺序、选择、迭代、数据类型、数组和字符串操作。分数看重的是正确的逻辑，而不是微小的语法缺陷，但要避免变量未初始化等粗心错误。
When asked to correct a given code segment, identify the exact lines that contain errors, state what is wrong, and provide the corrected version. For instance, if a loop condition is WHILE Count < 10 DO but should include equality, write both the problem and the fix. Never simply rewrite the whole segment without pinpointing the error.
当被要求修正一个给定的代码段时，要找出包含错误的确切行，说明错在哪里，并提供修正后的代码。例如，如果一个循环条件是 WHILE Count < 10 DO 但实际应包含等号，那么既要写出问题也要写出修正方案。绝不要只是重写整个代码段而不指出错误所在。
Practice trace tables for loops and arrays. The exam often asks you to complete a trace table showing how variable values change at each iteration. Fill in values row by row; a single misplaced number can cascade errors. Double-check the final iteration to ensure the loop exits correctly.
针对循环和数组练习追踪表。考试常常要求你完成一张追踪表，显示每次迭代中变量值如何变化。逐行填入数值；一个错位的数字可能引发一系列错误。仔细检查最后一次迭代，确保循环正确退出。
6. Database and SQL Queries | 数据库与 SQL 查询
SQL questions typically ask you to write a query to retrieve specified data. CIE uses a limited set of SQL commands: SELECT…FROM…WHERE, ORDER BY, GROUP BY with HAVING, and aggregate functions like COUNT, SUM, AVG. Before writing, underline the key requirements in the question: which fields to display, which table(s) to use, what filter conditions apply, and how to sort or group results.
SQL 题通常要求你编写查询语句以检索指定数据。CIE 使用的 SQL 命令有限：SELECT…FROM…WHERE、ORDER BY、GROUP BY 与 HAVING，以及诸如 COUNT、SUM、AVG 等聚合函数。在编写之前，在题目要求中划出关键点：需要显示哪些字段，使用哪些表，应用什么筛选条件，以及如何排序或分组结果。
Always specify the exact field names and table names as they appear in the question schema. If the table is called STUDENT, do not write Students. For string matching, use LIKE with wildcards % (multiple characters) and _ (single character). When joining tables, ensure you match primary and foreign keys correctly.
务必使用题目模式中出现的准确字段名和表名。如果表名为 STUDENT，就不要写成 Students。对于字符串匹配，使用 LIKE 配合通配符 % (多个字符) 和 _ (单个字符)。在连接表时，确保正确匹配主键和外键。
Example query from a typical mark scheme:
SELECT StudentID, Name FROM STUDENT WHERE Grade > 75 ORDER BY Name ASC;
Notice the use of logical operators and precise syntax. Marks are awarded for each clause written correctly, so even if you cannot complete the entire query, write the parts you know; partial marks can be earned.
一份典型评分方案中的查询示例：
SELECT StudentID, Name FROM STUDENT WHERE Grade > 75 ORDER BY Name ASC;
请注意逻辑运算符和精确语法的使用。分数会按每个正确编写的子句给分，因此即使你无法完成整个查询，也要写下你知道的部分；部分得分是可能的。
7. Logic Gates and Truth Tables | 逻辑门与真值表
Questions on Boolean logic require you to draw logic circuits, complete truth tables, or derive expressions from a given scenario. Use the standard gate symbols: AND (D-shaped), OR (curved), NOT (triangle with circle). Clear labeling of inputs and outputs prevents confusion.
布尔逻辑题要求你绘制逻辑电路、完成真值表，或根据给定情景推导表达式。请使用标准门符号：与门 (D 形)、或门 (弧形)、非门 (三角形加圆圈)。清晰标记输入和输出可以避免混淆。
When constructing a truth table, list all possible combinations of n inputs, often in binary counting order (000, 001, 010…). Add columns for intermediate expressions if the circuit is complex, then the final output. Show your working: examiners often award marks for intermediate columns even if the final column has a mistake.
构建真值表时，列出 n 个输入的所有可能组合，通常按二进制计数顺序 (000, 001, 010…)。如果电路复杂，为中间表达式添加列，然后才是最终输出。展示你的推导过程：即使最终列有误，考官也常常会给中间列计分。
For writing Boolean expressions, use the standard notation: ¬A for NOT A (you may also use A’), A · B (or A AND B) for AND, A + B (or A OR B) for OR. Example:
X = (¬A · B) + (A · ¬B)
. This represents an XOR gate. Simplify expressions where indicated, but only if asked; over-simplifying where not required can introduce errors.
书写布尔表达式时，使用标准记法：¬A 表示 NOT A (也可用 A’)，A · B (或 A AND B) 表示与，A + B (或 A OR B) 表示或。例如：
X = (¬A · B) + (A · ¬B)
。这代表异或门。只有在题目要求时才化简表达式；在没有要求时过度化简可能引入错误。
8. Time Management and Mark Allocation | 时间管理与分数分配
Allocate time based on marks: spend roughly 1.2 minutes per mark in Paper 1 (75 marks in 90 minutes) and about 1.4 minutes per mark in Paper 2 (75 marks in 105 minutes). Use the first few minutes to skim through the paper and identify easy wins.
根据分值分配时间：Paper 1 大约每分钟配 1.2 分 (90 分钟, 75 分)，Paper 2 大约每分钟配 1.4 分 (105 分钟, 75 分)。利用开头几分钟浏览全卷，找出容易得分的题目。
Do not spend 15 minutes on a 2-mark question. If you are stuck, circle the question, leave space, and move on. Come back later with a fresher perspective. Avoid leaving any question completely unanswered; blank space earns zero, while a partial attempt could pick up a mark or two.
不要在一道 2 分的题目上花 15 分钟。如果被卡住，圈出题目，留出空白，继续往下做。稍后再以更清晰的思路回头解答。避免在任意一道题目上完全留白；空白得零分，而部分尝试可能拿下一两分。
In Paper 2, programming questions often demand more time. Check how many marks are assigned to each sub-part and plan accordingly. If a question is worth 7 marks and has 4 components, estimate the mark distribution: maybe 2, 2, 1, 2, and adjust your response length to match.
在 Paper 2 中，编程题常常需要更多时间。检查每个子部分的分值并相应规划。如果某道题 7 分，有 4 个组成部分，估计分值分布可能为：2, 2, 1, 2，并据此调整你的回答篇幅。
9. Common Mistakes to Avoid | 避免常见错误
One of the biggest mistakes is ignoring the context of the question. For example, a question about network security in a small-business setting wants practical, cost-effective measures, not a theoretical list of every possible encryption protocol. Tailor your answer to the scenario provided.
最大的错误之一是忽略题目的语境。例如，一道关于小型企业环境中的网络安全的题，希望得到的是实用、经济的措施，而不是所有可能加密协议的理论清单。根据所提供的情景来定制答案。
Another pitfall is vague language: ‘It is faster’ or ‘It is better’ without explanation. Always qualify with reasons: ‘It is faster because cache memory has a shorter access time than RAM, reducing the fetch-decode-execute cycle.’ This not only answers ‘what’ but also ‘why’.
另一个陷阱是语言模糊：“它更快”或“它更好”而没有解释。务必用理由来限定：“它更快，因为缓存的访问时间比 RAM 短，从而缩短了取指-译码-执行周期。” 这不仅回答了“是什么”，还回答了“为什么”。
Forgetting to include necessary units also costs marks. If you state a data transfer rate as ’10’, the examiner cannot tell if you mean 10 Mbps, 10 MB/s, or 10 bps. Similarly, when calculating storage size, always show working with units:
Size = 1000 × 64 bits = 64 000 bits = 8 000 bytes
.
忘记包含必要的单位也会导致失分。如果你把数据传输率写为“10”，考官无法分辨你指的是 10 Mbps、10 MB/s 还是 10 bps。同样，在计算存储大小时，始终展示带单位的推导过程：
Size = 1000 × 64 bits = 64 000 bits = 8 000 bytes
。
In pseudocode, mismatched IF…ENDIF or infinite loops are frequent logic errors. Always step through your algorithm mentally with a simple test case to verify that it terminates and produces the correct output.
在伪代码中，IF…ENDIF 不匹配或无限循环是常见的逻辑错误。务必在脑中用一个简单的测试用例逐步执行算法，以验证其终止并产生正确输出。
10. Using Technical Terminology Accurately | 准确使用技术术语
Examiners expect precise vocabulary. Referring to an ‘operating system’ when you mean ‘utility software’, or ‘interpretation’ when you mean ‘compilation’, can lose marks, especially in definition or distinction questions. Keep a glossary of the key terms from the syllabus and test yourself regularly.
考官期望精确的术语。如果你本意是“utility software”却说成“operating system”，或者指“compilation”却说成“interpretation”，都可能丢分，尤其在定义或辨析题中。制作一份教学大纲关键术语的词汇表，并定期自测。
When discussing ‘cloud computing’, distinguish between ‘public cloud’, ‘private cloud’, and ‘hybrid cloud’. When explaining ‘encryption’, specify ‘symmetric’ vs ‘asymmetric’ where relevant. Precision signals a higher level of understanding and often pushes your answer into the top marking band.
在讨论“cloud computing”时，区分“public cloud”、“private cloud”和“hybrid cloud”。在解释“encryption”时，如相关应指明“symmetric”与“asymmetric”。精确性标志着更深层的理解，常常能将你的答案推入最高评分区间。
11. How Marks Are Awarded: Levels of Response | 评分标准：分级应答
Extended written questions (4-8 marks) use ‘levels of response’ marking, especially in ethics or system evaluation. Examiners look for: a logical structure, coherent arguments, use of relevant examples, and a balanced conclusion. You are rewarded for quality, not just quantity.
扩展性写作题 (4-8 分) 使用“分级应答”评分法，尤其是在伦理或系统评估题中。考官关注的是：逻辑结构、连贯的论点、相关例子的运用以及平衡的结论。得分凭借的是质量，而不仅仅是数量。
To achieve the highest level, you must address the question from multiple perspectives. In a question about the impact of AI, discuss benefits like efficiency and personalization, but also drawbacks like job displacement and bias. Use phrases like ‘On the one hand… On the other hand…’.
要达到最高等级，你必须从多个角度回答问题。在关于人工智能影响的题目中，既要讨论效率和个性化等好处，也要讨论工作替代和偏见等弊端。使用诸如 “On the one hand… On the other hand…” 的表达。
Always conclude with a justified judgment. A strong conclusion ties together the points made and gives a clear, reasoned opinion. The examiner wants to see that you have evaluated the information, not just listed it.
务必以有理有据的判断作结。一个有力的结论能将所提出的要点串联起来，并给出清晰、有根据的观点。考官希望看到你评价了信息，而不仅仅是罗列。
12. Final Revision Tips | 最后复习建议
Work through past papers under timed conditions, then mark them yourself using the official mark scheme. Pay special attention to the ‘Additional Guidance’ section, which shows acceptable alternative answers and common errors that will not be credited. This is the best way to internalize examiners’ expectations.
在限时条件下完成历年试卷，然后使用官方评分标准自行批改。特别留意 “Additional Guidance” 部分，它展示了可接受的替代答案以及不予给分的常见错误。这是内化考官期望的最佳方式。
Create concise revision cards for each topic, but also practice writing out full ‘explain’ and ‘discuss’ answers by hand. Many students understand concepts but cannot articulate them fluently within exam pressure. Write and write again until your pen flows naturally.
为每个主题制作简洁的复习卡片，但也要练习用手写下完整的“explain”和“discuss”答案。许多学生理解概念，但在考试压力下无法流畅表达。一遍遍动笔，直到下笔自然。
The night before the exam, review command words, key definitions, and trace table technique. Rest well. A clear head and a strategic approach to the paper are your greatest tools for turning knowledge into marks.
考试前一晚，复习指令词、关键定义和追踪表技巧。好好休息。清晰的头脑和对试卷的策略性应对，是你能将知识转化为分数的最有力工具。
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