📚 AS WJEC Computer Science: Exam Technique and Marking Criteria | AS WJEC 计算机：答题技巧与评分标准
Scoring high marks in AS WJEC Computer Science requires more than just knowing the content; you must understand precisely how examiners award marks and tailor your answers to meet the assessment criteria. This guide breaks down the exam structure, command words, marking principles, and practical techniques to help you avoid common pitfalls and maximise your score in both written and on-screen units.
在 AS WJEC 计算机科学考试中拿到高分，不仅需要掌握知识，还必须准确理解考官如何给分，并让你的答案契合评分标准。本指南将拆解考试结构、命令词、评分原则与实用技巧，帮助你避开常见陷阱，在笔试与上机考试中实现分数最大化。
1. Understanding the WJEC AS Computer Science Exam Structure | 理解 WJEC AS 计算机考试结构
AS WJEC Computer Science consists of two mandatory units: Unit 1 'Fundamentals of Computer Science' is a 2-hour written paper worth 70 marks, covering theory topics such as data representation, computer architecture, networks, and ethical issues. Unit 2 'Practical Programming to Solve Problems' is a 2-hour on-screen exam worth 60 marks, where you write, test, and refine programs using a high-level language like Python or Java.
AS WJEC 计算机科学包含两个必修单元：单元一“计算机科学基础”是一场 2 小时的笔试，总分 70 分，涵盖数据表示、计算机体系结构、网络与伦理议题等理论主题。单元二“实用编程解决问题”是一场 2 小时的上机考试，总分 60 分，要求使用 Python 或 Java 等高级语言编写、测试和改进程序。
Knowing the weight and style of each paper helps you allocate revision time effectively. Unit 1 demands precise definitions, clear explanations, and the ability to apply concepts to unfamiliar scenarios. Unit 2 rewards clean, working code that fulfills the specification, accompanied by evidence of systematic testing. The on-screen exam is auto-marked, so your program must produce exactly the required outputs or pass hidden test cases.
清楚每份试卷的分值与风格有助于你合理分配复习时间。单元一要求给出精确的定义、清晰的解释，并能将概念应用于陌生情境。单元二则奖励整洁且符合题目规格的运行代码，同时要有系统化测试的证据。上机考试采用自动判分，因此你的程序必须精确产生所需输出，或通过隐藏的测试用例。
2. Assessment Objectives and How Marks Are Awarded | 评分目标与给分方式
WJEC uses three Assessment Objectives (AOs) to evaluate your performance. AO1 tests recall and understanding of key concepts, typically through definitions and short-answer questions. AO2 asks you to apply knowledge to solve problems, such as tracing an algorithm or constructing logic circuits. AO3 requires analysis and evaluation, for instance discussing the pros and cons of a network topology or interpreting test results.
WJEC 使用三个评分目标 (AO) 来评估你的表现。AO1 考核对关键概念的识记与理解，通常通过定义题和简答题考查。AO2 要求运用知识解决问题，例如跟踪算法或构建逻辑电路。AO3 则需要分析与评估，比如讨论某种网络拓扑的优缺点，或解释测试结果。
Marks are not simply given for mentioning a keyword; they are awarded for showing a chain of reasoning. In the mark scheme, each point has explicit 'indicative content' but you can earn marks for any valid approach that demonstrates the relevant AO. For unit 2, marks map to functional code segments: partial credit may be given for correct logic even if syntax is slightly flawed, as long as the algorithmic intent is clear.
评分并非仅仅因为提到了某个关键词就给分；分数是奖励给展示出推理链条的答案的。在评分方案中，每个分数点都有明确的“指示性内容”，但任何能够展现相关 AO 的有效方法都能得分。在单元二中，分数对应功能代码片段：即使语法略有瑕疵，只要算法意图清晰，正确的逻辑仍然可能获得部分分数。
3. Mastering Command Words | 掌握命令词
Every question is built around a command word that signals exactly what the examiner wants. 'State' or 'Give' means a short, factual answer without explanation. 'Describe' requires you to give characteristics or a step-by-step account. 'Explain' demands a reason or cause-and-effect relationship, often using 'because' or 'therefore'. 'Compare' asks for similarities and differences, and 'Evaluate' means you must weigh up strengths and weaknesses with a justified conclusion.
每一道题目都围绕一个命令词构建，它明确指示了考官的要求。“State”或“Give”意味着给出简短的事实性答案，无需解释。“Describe”要求描述特征或者逐步说明。“Explain”则需要说明理由或因果关系，常常会用到“因为”或“因此”。“Compare”要求比较相似点与差异，“Evaluate”则意味着你必须权衡优缺点并得出有理有据的结论。
A common error is to provide a long description when the question only asks you to 'state' two advantages. This wastes time without gaining extra marks. Conversely, for 'explain', simply stating a fact like 'it is faster' without linking to a reason will score zero. Practise transforming a descriptive sentence into an explanatory one by adding 'because...' after every claim.
一个常见错误是当题目只要求“state”两个优点时，却给出一长段描述。这不仅浪费时间，还不会得到额外分数。相反，对于“explain”，仅仅陈述“它更快”这样的事实而没有联系原因将得到零分。练习在每个断言后加上“因为……”，将描述性句子转化为解释性句子。
4. Showing Your Working in Calculation and Tracing Questions | 在计算与跟踪题中展示步骤
. Write down each conversion clearly: for example, when converting 1101₂ to denary, list 1×2³ + 1×2² + 0×2¹ + 1×2⁰ = 13. If you make an arithmetic slip, you can still receive most of the marks.
对于二进制/十六进制转换、浮点表示或布尔代数题，考官会为中间的解题步骤给方法分。清晰地写下每一步转换：例如，将 1101₂ 转换为十进制时，列出 1×2³ + 1×2² + 0×2¹ + 1×2⁰ = 13。即使某个算术环节出错，你仍然可能拿到大部分分数。
Tracing an algorithm or a logic gate circuit is a process skill. Draw a trace table with columns for each variable and output, then update line by line. The mark scheme often allocates one mark for a correct initialisation, several marks for a row showing correct updates, and a final mark for the correct output. Never just give the final value.
跟踪算法或逻辑门电路是一种过程技能。画一张跟踪表，为每个变量和输出设置列，然后逐行更新。评分方案通常会为正确的初始化分配一分，为显示正确更新的若干行再给几分，最终输出正确再得一分。绝不要只给出最终值。
5. Algorithm Design and Pseudocode Excellence | 算法设计与伪代码优秀标准
WJEC does not prescribe a specific pseudocode syntax, but clarity and consistency are essential. Use indentation to show loops and selections, meaningful variable names, and standard constructs like WHILE ... ENDWHILE, IF ... THEN ... ELSE ... ENDIF. Write an algorithm that can be understood by a fellow student without ambiguity.
WJEC 没有规定特定的伪代码语法，但清晰和一致性至关重要。使用缩进来表示循环和选择结构，使用有意义的变量名以及标准构造，如 WHILE ... ENDWHILE、IF ... THEN ... ELSE ... ENDIF。书写一个能被同学无歧义理解的算法。
When designing an algorithm for a problem, first identify inputs, process, and outputs. Break the process into small logical chunks, and present them in a structured English or flowchart style. For Unit 1, you may be asked to write an algorithm to solve a sort or search problem. Never combine multiple steps in one vague sentence; each assignment, comparison, or iteration must be a separate line.
当为一个问题设计算法时，首先明确输入、处理过程和输出。将处理过程分解为小的逻辑块，并以结构化英语或流程图的形式呈现。在单元一中，你可能会被要求编写算法来解决排序或搜索问题。绝不要把多个步骤合并成一句模糊的话；每个赋值、比较或迭代都必须是独立的一行。
6. Programming Questions: How to Score Full Marks on Unit 2 | 编程题：如何在单元二拿满分
Unit 2 on-screen exam is marked by a combination of automated test runs and examiner inspection of your code. Read the scenario and the detailed requirements carefully. Most tasks ask you to open a file, process data, and output results. Identify exactly which functions you need: reading a CSV, performing calculations, sorting, and writing to a new file. Program incrementally, compiling and testing each small piece.
单元二的上机考试结合了自动化测试运行和考官对代码的检查来评分。仔细阅读场景和详细要求。大多数任务要求你打开文件、处理数据并输出结果。准确识别你需要哪些功能：读取 CSV、执行计算、排序和写入新文件。逐步编写程序，每完成一小部分就编译并测试。
Marks are awarded for functional outputs, use of appropriate data structures (lists, dictionaries), robust error handling, and annotation. Add comments to explain the purpose of complex sections: examiners look for evidence of clear logical thinking. If a specification says 'the program must not crash with invalid input', include a try-except block or validation checks. Even if your final output is slightly incorrect, structured annotation can earn method marks.
分数奖励给功能输出、合理的数据结构（列表、字典）使用、稳健的错误处理和注释。添加注释来解释复杂代码段的用途：考官会寻找清晰逻辑思维的证据。如果题意要求“程序在输入无效时不得崩溃”，就要加入 try-except 块或验证检查。即便你的最终输出略有偏差，结构化的注释仍能获得方法分。
7. Structuring Theory Answers for Maximum Impact | 构建理论答案以获得最大影响
For 6–10 mark extended theory questions, structure your answer like a mini essay. Start with a definition or a direct statement that addresses the command word. Then present arguments in a logical order, using connectives such as 'furthermore', 'in contrast', and 'consequently'. Finish with a concluding sentence if the question asks for a judgement.
对于 6 至 10 分的扩展理论题，将答案构建得像一篇小型论文。首先给出定义或直接回应命令词的陈述。然后按照逻辑顺序呈现论点，使用“此外”、“相比之下”、“因此”等连接词。如果题目要求做出判断，就以一个结论句收尾。
Use technical vocabulary accurately: instead of saying 'the computer slows down', say 'the increased number of fetches from main memory causes a bottleneck on the data bus, reducing throughput'. The more precise your terminology, the more you demonstrate expert understanding to the examiner. Always link back to the scenario given in the question stem.
准确使用专业词汇：与其说“电脑变慢了”，不如说“从主存取指次数的增加导致数据总线上的瓶颈，降低了吞吐量”。术语越精确，你向考官展示出的专家级理解就越充分。始终要与题目题干中给出的场景联系起来。
8. Common Pitfalls That Lose Marks Unnecessarily | 不必要失分的常见陷阱
One of the biggest traps is misreading the question, especially when it asks for 'one' benefit but you list three. Examiners only mark the first one. Similarly, in comparison questions, writing two separate descriptions without an actual comparative connective ('whereas', 'however') will not meet the 'compare' demand. Always mirror the question's language in your answer.
最大的陷阱之一是误读题目，尤其是当题目要求“一个”优点而你不加判断地列出三个时。考官只会对第一个评分。同样，在比较题中，写下两段独立的描述却没有使用真正的比较连接词（“而”、“然而”），将无法满足“比较”的要求。始终要在答案中呼应题目的措辞。
In algorithm and code questions, avoid missing initialisation of variables. In a trace table, forgetting to update a counter or index will cascade errors. For programming, not closing files or not following the exact naming convention for output files can result in zero marks, even if the logic is perfect. Finally, do not leave any question blank; an intelligent guess can often earn a mark.
在算法和代码题中，不要遗漏变量的初始化。在跟踪表中，忘记更新计数器或索引会导致错误级联。对于编程，不关闭文件或不遵守输出文件的精确命名约定，即使逻辑完美，也可能得到零分。最后，任何题目都别留空；一个有根据的猜测往往能拿到一分。
9. Time Management and Question Order Strategy | 时间管理与答题顺序策略
Unit 1 gives you roughly 1.7 minutes per mark. Begin by scanning the entire paper to identify high-tariff questions that need extended writing. Tackle the questions in order, but if you get stuck on a low-mark question, mark it, leave space, and return later. Never spend 10 minutes on a 2-mark definition.
单元一大约每一分有 1.7 分钟。开始时快速浏览整份试卷，找出需要扩展写作的高分值题目。按顺序答题，但如果在低分题上卡壳，就做个标记，留出空白，稍后再回来。绝不要在 2 分的定义题上花费 10 分钟。
For Unit 2, allocate time for reading the scenario and understanding the data files. Then split the remaining time based on the marks for each programming tasks. Start with the easiest functional component to build confidence and secure early marks. Save time at the end to run all in-built and extreme test cases, because unseen test data often carries heavy weighting.
对于单元二，要分配时间去阅读场景并理解数据文件。然后根据每个编程任务的分数分配剩余时间。从最简单的功能组件做起，建立信心并确保拿到早期分数。留出最后的时间运行所有内置和极端测试用例，因为隐藏的测试数据往往权重很高。
10. Review and Proofreading for Extra Marks | 复查与修改以获取额外分数
In the final minutes, quickly scan through your written answers. Check that you have not contradicted yourself, that all units are provided (e.g., seconds, bytes, hertz), and that graphs or truth tables are correctly labelled. Correcting a simple slip such as writing 1024 instead of 1024 MB when the unit matters can salvage marks.
在最后几分钟里，快速浏览你的书面答案。检查自己是否前后矛盾，所有单位是否都提供了（例如秒、字节、赫兹），以及图表或真值表是否都正确标注了。修正一个简单的笔误，比如在单位重要时写了 1024 而不是 1024 MB，就能挽回分数。
For programming, re-read the specification notes one more time: are you asked to output in CSV format, to round to two decimal places, or to exclude headings? Run a final test with the provided sample data and check every output character if necessary. Often a single misplaced newline or missing comma separates a high score from a middle one.
在编程部分，再读一遍规格说明：是否要求输出 CSV 格式，是否要四舍五入到小数点后两位，或者是否要去掉表头？用提供的样本数据运行最终测试，如有必要，检查每一个输出字符。往往一个错位换行符或缺失的逗号就会把高分与中分隔开。
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