📚 Year 13 CIE Computer Science: 2026 Exam Changes and Trends | Year 13 CIE 计算机：2026年考试变化与趋势
As the 2026 examination series approaches, students and teachers of CIE A-Level Computer Science (9618) are keen to understand the subtle shifts in assessment focus and syllabus interpretations. While the syllabus itself has not undergone a radical rewrite, examiners’ reports and recent paper patterns reveal evolving emphases on higher-order thinking, practical programming rigour, and the integration of contemporary computing concepts. This article dissects the key trends and expected nuances for the 2026 exams, offering a roadmap for effective preparation.
随着 2026 年考试季的临近，CIE A-Level 计算机科学（9618）的学生和教师都渴望了解评估重点的微妙变化和对大纲的解读。尽管大纲本身并未进行彻底改写，但考官报告和近年的试题模式表明，评估越来越重视高阶思维、严格的编程实践以及当代计算概念的融合。本文剖析了 2026 年考试的主要趋势和预期变化，为高效备考提供路线图。
1. Overview of CIE A-Level Computer Science (9618) | CIE A-Level 计算机科学（9618）概述
The CIE 9618 syllabus, introduced to replace the legacy 9608, has now matured through several examination cycles. It comprises four components: Paper 1 (Theory Fundamentals), Paper 2 (Fundamental Problem-solving and Programming), Paper 3 (Advanced Theory), and Paper 4 (Practical Project). The 2026 exams will continue to test the full range of topics from data representation to artificial intelligence, but with an increasing demand for application rather than rote recall.
CIE 9618 大纲取代了旧的 9608，现已经过多个考试周期的检验，趋于成熟。它包含四个部分：试卷1（理论基础）、试卷2（基本问题解决与编程）、试卷3（高级理论）和试卷4（实践项目）。2026 年的考试将继续涵盖从数据表示到人工智能的全部主题，但越来越要求应用能力而非死记硬背。
The syllabus aims to develop computational thinking, programming proficiency, and an understanding of the wider ethical and legal implications of technology. For 2026, candidates must be ready to demonstrate deeper analytical skills, particularly in Papers 3 and 4.
该大纲旨在培养学生的计算思维、编程能力以及对技术更广泛的伦理和法律影响的理解。对于 2026 年，考生必须准备好展示更深入的分析能力，尤其是在试卷3和4中。
2. Assessment Structure for 2026 | 2026年评估结构
The overall structure remains unchanged: Paper 1 (1 hour 30 minutes, 75 marks, 25% weight), Paper 2 (2 hours, 75 marks, 25%), Paper 3 (1 hour 30 minutes, 75 marks, 25%), and Paper 4 (coursework, 75 marks, 25%). However, subtle shifts in mark allocation within papers are evident. In Paper 1, more marks are being directed towards questions that require explanation and evaluation, not just simple definitions. Paper 2 now consistently includes a trace table question and a more complex pseudocode algorithm task. Paper 3 increasingly features extended writing on system design and emerging technologies. For Paper 4, the project requirements have been refined to demand a well-documented, modular solution with robust testing evidence.
整体结构保持不变：试卷1（1小时30分钟，75分，权重25%），试卷2（2小时，75分，25%），试卷3（1小时30分钟，75分，25%），试卷4（课程作业，75分，25%）。然而，试卷内部的分数分配有微妙变化。在试卷1中，更多分数倾向于需要解释和评估的题目，而不仅仅是简单定义。试卷2现在一贯包含跟踪表问题和一个更复杂的伪代码算法任务。试卷3越来越多地涉及系统设计和新兴技术的扩展写作。对于试卷4，项目要求已经细化，要求提供文档齐全、模块化的解决方案以及可靠的测试证据。
Candidates for 2026 should note that the coursework (Paper 4) must demonstrate genuine individual programming effort; collaborative code or over-reliance on auto-generated code is penalised. The practical project must solve a real problem for a clearly identified client.
2026 年的考生应注意，课程作业（试卷4）必须展示真实的个人编程成果；协作代码或过度依赖自动生成代码都会受到惩罚。实践项目必须为明确识别的客户解决一个真实问题。
3. Paper 1: Theory Fundamentals – Key Changes | 试卷1：理论基础 – 主要变化
Paper 1 covers information representation, communication, hardware, processor fundamentals, system software, security, ethics, and databases. Recent trends show a move away from simple binary conversion questions towards contextualised scenarios. For instance, a question on floating-point representation now often asks candidates to justify why a particular precision is insufficient for a given scientific calculation, linking to range and accuracy trade-offs.
试卷1涵盖信息表示、通信、硬件、处理器基础、系统软件、安全、伦理和数据库。近年的趋势显示，题目从简单的二进制转换转向情景化场景。例如，关于浮点表示的问题现在经常要求考生论证为何特定精度不足以满足给定的科学计算，并联系范围与精度的权衡。
Ethics questions are no longer just asking for a list of issues; they demand a reasoned discussion of the impact of a technology, such as the use of facial recognition, on different stakeholders. Candidates must refer to specific legislation like GDPR (General Data Protection Regulation) and the Computer Misuse Act where appropriate. Expect 2026 to include more ‘discuss’ and ‘evaluate’ command words in this paper.
伦理问题不再只是要求列出问题，而是要求对技术（如人脸识别的使用）对不同利益相关者的影响进行理性的讨论。考生必须在适当情况下引用具体的立法，如 GDPR（通用数据保护条例）和《计算机滥用法》。预计 2026 年该试卷将包含更多“讨论”和“评估”等指令词。
4. Paper 2: Fundamental Problem-solving and Programming Skills | 试卷2：基本问题解决与编程技能
Paper 2 is algorithm-heavy. Candidates write pseudocode, interpret code segments, and design solutions. The 2026 trend points to a stronger emphasis on efficiency analysis. Questions explicitly ask for the time complexity of an algorithm using Big O notation, for example O(n²) or O(log n), and require justification. The provided pseudocode syntax guide remains the authoritative reference; deviating from it in writing loops or conditionals can lose marks.
试卷2侧重于算法。考生需要编写伪代码、解释代码段以及设计解决方案。2026 年的趋势表明，对效率分析的强调将进一步加强。题目明确要求使用大 O 表示法给出算法的时间复杂度，例如 O(n²) 或 O(log n)，并需要说明理由。提供的伪代码语法指南仍然是权威参考；编写循环或条件语句时偏离该指南可能会丢分。
Structured English is tested as a way to decompose problems. Candidates should practice converting between flowcharts, pseudocode, and structured English. Another notable trend is the integration of file handling concepts: reading and writing text files using pseudocode, with careful attention to end-of-file detection. There is also a greater occurrence of abstract data types (ADTs) like stacks and queues in problem contexts, moving beyond simple array manipulations.
结构化英语被作为一种分解问题的方法进行考核。考生应练习在流程图、伪代码和结构化英语之间进行转换。另一个值得注意的趋势是文件处理概念的整合：使用伪代码读写文本文件，需特别注意文件结束检测。同时，栈和队列等抽象数据类型（ADT）在问题情境中出现的频率增加，超越了简单的数组操作。
5. Paper 3: Advanced Theory – Emerging Trends | 试卷3：高级理论 – 新兴趋势
Paper 3 extends into data representation (floating-point normalisation), communication protocols, CPU architecture (pipelining, SIMD), system software (virtual machine, translation software), security (encryption, digital signatures), and AI. The 2026 exams reflect a modernised take on these topics. For example, Boolean algebra and logic circuit questions now often incorporate real-world device control scenarios rather than abstract gate combinations.
试卷3扩展到数据表示（浮点规格化）、通信协议、CPU 架构（流水线、SIMD）、系统软件（虚拟机、翻译软件）、安全（加密、数字签名）和人工智能。2026 年的考试体现了这些主题的现代化趋势。例如，布尔代数和逻辑电路问题现在经常包含真实世界的设备控制场景，而不是抽象的门电路组合。
Artificial Intelligence is no longer a single peripheral topic; the syllabus asks for detailed knowledge of machine learning, neural networks, and expert systems. Candidates must be able to compare supervised and unsupervised learning, explain backpropagation conceptually, and discuss the societal impacts of AI bias. A typical 2026 question might provide a dataset and ask to identify whether it suits a classification or regression model. Depth of understanding is now expected, not just surface definitions.
人工智能不再是一个边缘主题；大纲要求详细了解机器学习、神经网络和专家系统。考生必须能够比较监督学习和无监督学习，从概念上解释反向传播，并讨论 AI 偏见的社会影响。2026 年的一道典型题目可能会提供一个数据集，并询问它是否适合分类或回归模型。现在期望的是深层次的理解，而不仅仅是表面的定义。
6. Paper 4: Practical Programming Project – New Emphasis | 试卷4：实践编程项目 – 新侧重点
The coursework component is internally assessed and externally moderated. For 2026, CIE has reinforced the need for authentic evidence of the development cycle. The project report must include a clear problem definition, a structured decomposition diagram, detailed algorithm design (often in pseudocode or flowchart form), and a fully commented program listing.
课程作业部分由校内评估、外部审核。对于 2026 年，CIE 加强了对开发周期真实证据的要求。项目报告必须包含清晰的问题定义、结构化的分解图、详细的算法设计（通常为伪代码或流程图形式）以及完整注释的程序清单。
The trend is towards valuing quality of software engineering over quantity. Examiners look for appropriate use of data structures (e.g., arrays of records, dictionaries), modular design with parameter passing, and robust input validation. Testing evidence must show a systematic approach using normal, abnormal, and boundary data, with annotated screenshots. Projects that rely solely on drag-and-drop interfaces without substantial custom code will struggle to achieve high marks.
趋势是重视软件工程的质量而非数量。考官看重是否恰当使用数据结构（如记录数组、字典）、带参数传递的模块化设计以及健壮的输入验证。测试证据必须展示使用正常、异常和边界数据的系统性方法，并附有注释的截图。仅依赖拖放式界面而没有大量自定义代码的项目将难以获得高分。
7. Programming Languages and Pseudocode | 编程语言与伪代码
CIE does not prescribe a specific programming language for Paper 4, but Python, Java, and VB.NET are commonly used. In Paper 2, the official pseudocode is used, which blends elements of Pascal, Python, and plain English. The 2026 exams continue to be strict about pseudocode syntax. For instance, OUTPUT, INPUT, WHILE...DO...ENDWHILE, and FOR...TO...NEXT must be written exactly as in the published guide. Array indices start at 1 unless specified otherwise, and the DECLARE keyword is expected for variable declarations.
CIE 未指定试卷4的特定编程语言，但 Python、Java 和 VB.NET 较为常用。在试卷2中，使用官方伪代码，它融合了 Pascal、Python 和通俗英语的元素。2026 年考试对伪代码语法仍然要求严格。例如，OUTPUT、INPUT、WHILE...DO...ENDWHILE 和 FOR...TO...NEXT 必须完全按照发布的指南书写。除非另有说明，数组索引从 1 开始，并且变量声明应使用 DECLARE 关键字。
Candidates should also develop fluency in converting pseudocode to their chosen high-level language, as this skill is implicitly assessed in the project and can be tested in Paper 2 through questions that ask for the output of a given algorithm. Minor syntactic differences are tolerated only when the logic is unambiguously correct.
考生还应熟练地将伪代码转换为自己选择的高级语言，因为这一技能在项目中会隐式评估，也可能在试卷2中通过要求给出特定算法输出的题目进行测试。只有在逻辑明确正确的情况下，才会容忍微小的语法差异。
8. The Growing Importance of Computational Thinking | 计算思维日益重要
Computational thinking — decomposition, pattern recognition, abstraction, and algorithm design — is the backbone of the 9618 syllabus. The 2026 papers are likely to feature more questions that explicitly target these skills. For instance, an abstraction question might ask a candidate to identify the essential features of a library system needed for a booking module, leaving out irrelevant details.
计算思维——分解、模式识别、抽象和算法设计——是 9618 大纲的支柱。2026 年的试卷很可能会包含更多明确针对这些技能的题目。例如，一道抽象题可能会要求考生识别图书馆系统中预订模块所需的基本特征，而忽略无关细节。
Algorithm design questions are moving beyond simple sorting and searching; candidates might be asked to design a greedy algorithm for a resource allocation problem or to apply Dijkstra’s algorithm on a weighted graph with a step-by-step trace. The ability to evaluate the suitability of different algorithms for a given task is becoming a key discriminator at the A* boundary.
算法设计题正超越简单的排序和搜索；考生可能会被要求为资源分配问题设计一个贪心算法，或在加权图上逐步追踪 Dijkstra 算法。评估不同算法对给定任务的适用性的能力正成为 A* 分界线上的关键区分因素。
9. Database and SQL: Evolving Expectations | 数据库与 SQL：不断变化的期望
Database theory appears in Papers 1 and 3, with a strong focus on normalisation, entity-relationship diagrams, and SQL. The 2026 trend indicates that mere knowledge of SELECT-FROM-WHERE syntax is insufficient. Candidates must handle complex joins (INNER, LEFT), aggregate functions (COUNT, SUM, AVG) with GROUP BY and HAVING, and nested subqueries. Data definition language (DDL) commands like CREATE TABLE, ALTER, and DROP are now regularly examined.
数据库理论出现在试卷1和3中，重点关注规范化、实体关系图和 SQL。2026 年的趋势表明，仅仅了解 SELECT-FROM-WHERE 语法是不够的。考生必须处理复杂的连接（INNER、LEFT）、聚合函数（COUNT、SUM、AVG）与 GROUP BY 和 HAVING 的结合，以及嵌套子查询。数据定义语言（DDL）命令，如 CREATE TABLE、ALTER 和 DROP，现在经常被考查。
Furthermore, examiners expect candidates to reason about database integrity, referential constraints, and the consequences of poorly designed tables. A typical question might provide a table with update anomalies and ask for the normalised forms up to third normal form (3NF). This requires a systematic approach to functional dependencies.
此外，考官期望考生能够推理数据库完整性、引用约束以及设计不佳的表格带来的后果。一道典型题目可能提供一个存在更新异常的表格，并要求将其规范化至第三范式（3NF）。这需要系统性地处理函数依赖关系。
10. Networking and Cybersecurity Updates | 网络与网络安全更新
Networking topics span the OSI model, TCP/IP stack, protocols, and hardware. For 2026, examiners are increasingly linking networking to real-world cybersecurity threats. Questions might ask for an explanation of how a man-in-the-middle attack exploits protocol weaknesses and how Transport Layer Security (TLS) mitigates it. Symmetric and asymmetric encryption (e.g., AES, RSA) must be understood at the algorithmic level, including key exchange and digital certificate validation.
网络主题涵盖 OSI 模型、TCP/IP 协议栈、协议和硬件。对于 2026 年，考官越来越多地将网络与现实世界的网络安全威胁联系起来。题目可能会要求解释中间人攻击如何利用协议弱点，以及传输层安全（TLS）如何缓解这种攻击。对称和非对称加密（如 AES、RSA）必须在算法层面理解，包括密钥交换和数字证书验证。
The role of firewalls, proxy servers, and VPNs is tested within scenario-based contexts. Candidates are expected to recommend appropriate security measures for a given network architecture, justifying their choices with reference to confidentiality, integrity, and availability (CIA triad). The 2026 papers will likely continue this applied approach.
防火墙、代理服务器和 VPN 的作用在基于场景的语境中进行测试。期望考生为给定的网络架构推荐适当的安全措施，并根据机密性、完整性和可用性（CIA 三元组）证明其选择的合理性。2026 年的试卷很可能会继续这种应用导向的方式。
11. Artificial Intelligence in the Curriculum | 人工智能在课程中的地位
The AI section of the syllabus has quietly become one of the most forward-looking parts. Beyond the basics of AI, candidates need to explain how machine learning models are trained, validated, and tested. Concepts like overfitting, underfitting, and the bias-variance tradeoff are fair game for examination. While the maths is not assessed deeply, a conceptual grasp of gradient descent and loss functions is beneficial.
大纲中的人工智能部分已悄然成为最具前瞻性的部分之一。在人工智能基础知识之外，考生需要解释机器学习模型如何训练、验证和测试。过拟合、欠拟合以及偏差-方差权衡等概念都可能出现在考试中。虽然数学不会进行深度评估，但对梯度下降和损失函数的概念性理解是有益的。
For 2026, expect questions that ask candidates to critically evaluate the application of AI in areas such as healthcare diagnosis or autonomous driving, discussing both technical limitations and ethical dilemmas. The ability to separate hype from realistic capability is a skill examiners are actively rewarding.
对于 2026 年，可以预期会有题目要求考生批判性地评估人工智能在医疗诊断或自动驾驶等领域的应用，讨论技术局限性和伦理困境。区分炒作与现实能力的能力是考官积极奖励的一项技能。
12. Exam Preparation Strategies for 2026 | 2026年备考策略
To excel in the 2026 CIE Computer Science exams, students should adopt a multi-pronged approach. First, master the official pseudocode; write algorithms daily and check them against the syntax guide. Second, for the project, start early, choose a well-defined problem with a real client, and document iteratively—do not leave the write-up until the end. Third, practice past papers from 2021 onwards, as they reflect the current standard most accurately. Pay special attention to mark schemes that reward precise technical language.
要在 2026 年 CIE 计算机科学考试中取得优异成绩，学生应采取多管齐下的方法。首先，掌握官方伪代码；每天编写算法，并根据语法指南进行检查。其次，对于项目，尽早开始，选择一个定义明确且有真实客户的问题，并迭代地记录文档——不要到最后才写报告。第三，练习 2021 年之后的历年真题，因为它们最准确地反映了当前的标准。特别要注意那些奖励精确技术语言的评分方案。
Fourth, build a solid conceptual framework for abstract topics like AI, encryption, and normalisation by using flashcards and mind maps. Fifth, stay updated with technology news; connecting syllabus theory to real-world events (e.g., a recent data breach) helps in application questions. Lastly, balance theory with hands-on coding—the best computer scientists are those who can both think and build.
第四，使用抽认卡和思维导图为 AI、加密和规范化等抽象主题建立扎实的概念框架。第五，关注科技新闻；将大纲理论与现实世界事件（例如，近期的数据泄露）联系起来有助于回答应用题。最后，平衡理论与动手编码——最优秀的计算机科学家既能思考又能构建。
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