SEC G3 Computing (SEAB K349) is the 2027 Singapore–Cambridge Secondary Education Certificate subject in which students learn how computer systems, algorithms, spreadsheets, networks and digital technologies actually work. Parents should know the examination format immediately: Paper 1 is a two-hour written examination worth 60%, marked out of 80; Paper 2 is a two-hour-30-minute lab-based examination worth 40%, marked out of 70. Both papers are compulsory. In the lab paper, candidates work at an individual computer with spreadsheet, Python and JupyterLab software, answer one spreadsheet question and four or five programming questions, and submit their required files within the allotted time. The subject is not simply about liking video games or being quick with a phone. It tests disciplined problem-solving, working code, testing, data understanding and responsible computing.
Families considering G3 Computing should first confirm that their secondary school offers the K349 syllabus and that their child meets the school’s subject-combination criteria. Next, evaluate mathematical reasoning, attention to instructions, willingness to debug rather than give up, and interest in both technical and ethical issues. This guide covers the five official modules, what happens in each paper, practical examples, what parents can do without writing code for the child, a full step-by-step study plan and realistic routes to JC, polytechnic, ITE and university. The earlier O-Level Computing reference was 7155. Computing is useful across many fields, but no SEC subject alone guarantees a software career or admission to a prestigious course.
- 2027 syllabus: G3 Computing K349 (earlier reference 7155); check separate G2 and G1 Computing codes before choosing resources.
- Paper 1: two hours, 80 marks, 60%; written questions across all five modules.
- Paper 2: two hours 30 minutes, 70 marks, 40%; one spreadsheet problem and four or five programming problems.
- Lab software: spreadsheet software, Python and JupyterLab; candidates save and submit the required softcopy files.
- Five modules: Computing Fundamentals; Algorithms and Programming; Spreadsheets; Networking; Impact of Computing.
- Assessment emphasis: knowledge, application, solution design, testing and refinement.
- First parent action: inspect a small program and ask the child how they know its result is correct for normal, invalid and boundary inputs.
- Authoritative source: SEAB/MOE/Cambridge K349 2027 syllabus.
What happens in a G3 Computing lesson?
Computing begins by turning a problem into an exact description. What information is given? What output is required? Which rules connect them? How will a computer represent the information? How can a solution be checked? The goal is to make a process reliable enough for someone else to use. A student who can write a few lines of Python but cannot explain the requirements or test an edge case has only learned part of the subject.
Consider an everyday example: a club wants to calculate its average attendance from a list of numbers. A child may know how to use a spreadsheet average function or Python’s sum and length functions. But what if the list is empty? What if one recorded value is negative? What if the list includes text? The deeper computing task is to decide valid inputs, process them correctly and handle cases that could cause wrong results. This combination of precise specifications, operations and validation is the foundation of software work.
SEAB identifies three assessment objectives: knowledge and understanding, application, and design/development/testing/refinement. Together they show why pure memorisation or purely hands-on coding is insufficient. The written paper asks students to reason about concepts across modules. The lab tests whether they can actually build and save working solutions within examination conditions. A balanced preparation teaches both.
Step 1: Check whether K349 is actually offered
The 2027 SEC G3 subject directory lists Computing K349 and its previous code 7155. The G2 Computing syllabus is K237 and G1 Computing is K127; these are different levels and should not be confused with K349. Not every school offers every subject or combination. Ask the school’s subject-selection team whether Computing is available, which other subjects it can be paired with, and what academic and timetable criteria apply.
A child who likes creating things, solving puzzles and investigating how systems behave may enjoy Computing. But enthusiasm for social media, gaming or generative AI tools does not by itself prove readiness for Python or networking theory. Parents should look for curiosity about logic and patience when a task fails. Equally, a child with little previous programming may still learn well because the syllabus teaches the foundations. Past access to expensive devices should not be mistaken for aptitude.
Discuss the total workload. G3 Computing requires time at a computer to practise and debug, plus reading and revision of written material. It can be compatible with Additional Mathematics, Sciences, Humanities, Art or other options when the school allows them, but no combination should be selected purely to impress admissions officers. Check actual prerequisites for plausible destinations and ask whether the student can sustain the necessary work with CCAs and rest.
Step 2: Build the foundations before upper secondary
Secondary 1: exact instructions and simple logic
Start with procedures rather than programming syntax. Ask the child to write steps for a simple task, then test whether the steps still work when conditions change. A recipe, a bus fare calculation or sorting books can illustrate sequence and decision-making. If the instruction ‘repeat until done’ lacks a clear stopping condition, a human reader may infer what to do but a program cannot. Such everyday examples help students see why computational instructions must be unambiguous.
Basic Mathematics supports the subject: positive and negative numbers, percentages, ratios, simple algebra, logic and interpreting graphs. Students need to distinguish a quantity from a label and understand that data types matter. The string “10” and the number 10 may look similar on screen but behave differently in computation. Teaching this distinction early prevents many later programming errors.
Secondary 2: learn decomposition and test cases
Decomposition means dividing a larger problem into manageable subproblems. Suppose a school event needs participant registration, attendance checking and a final count. Students can identify inputs, records, duplicate handling and reporting. The programme is easier to reason about when each part has a defined purpose. Encourage an explanation of the data flow before using a programming language.
Introduce testing as a separate activity. If a program calculates a fee for ages 12 to 16, test a normal age, the lower and upper boundaries, and values outside the allowed range. Testing only one convenient input cannot establish that the program is correct. A careful student learns to predict the expected answer independently and compare it with the actual result.
Secondary 3: learn systems, Python, spreadsheets and networks
Students enter a more formal curriculum. They study computer architecture, representation and logical circuits, then design algorithms using sequence, selection and iteration. They use variables, strings, lists, dictionaries and functions, and learn how to handle files at the prescribed level. Spreadsheets provide another problem-solving medium; networking explains how devices communicate and how security and privacy are protected. Ethical issues and the impact of technologies are examined alongside technical material.
The hardest shift for some students is accepting that errors are expected information. A program that fails is not proof that the student ‘cannot code’. The useful question is whether the failure comes from syntax, runtime conditions, data assumptions or the algorithm itself. Teach the child to isolate the smallest broken part and test it with known values. That habit is more valuable than copying a long working solution from someone else.
Secondary 4: connect the modules and prepare for the lab
The final year requires independent problem solving under time pressure. A problem may involve reading a text file, validating entries, processing a list, choosing a suitable algorithm and writing results. A spreadsheet question may require correct references, formulas, conditions and lookup functions. Candidates must also save the required files. A student can understand an algorithm yet lose performance if the file is misnamed or not saved according to the instructions. Practise an end-to-end workflow: read, plan, implement, test, save, verify and submit.
Step 3: Module 1 — Computing Fundamentals
This module includes computer architecture, data representation and logic gates. Students learn what processors, memory, storage and relevant interfaces do, and how information is represented in binary. They should understand the difference between a system component’s function and its advertised marketing description. More memory is not the same as more storage; a file capacity is not the same as processor speed. Ask the child to explain what happens when a program is stored, loaded and executed at a basic conceptual level.
Binary, hexadecimal and representation
A decimal number and a binary number can represent the same value using different positional bases. In decimal, 13 = 1×10 + 3. In binary, 1101₂ = 1×8 + 1×4 + 0×2 + 1×1 = 13₁₀. Students should be able to convert as specified by the syllabus and explain the method, rather than memorise strings of digits. They also study fixed-width signed representation using two’s complement and how text can be stored through character encodings.
The distinction between kilobytes and kibibytes and other capacity units matters where explicitly prescribed. Do not assume a computer manufacturer’s label and a binary storage calculation always use identical multiples. Ask which unit is meant in the question. This is a numeracy problem embedded in computing, and students benefit from writing units as carefully as they would in Physics.
Logic gates and truth tables
Students learn AND, OR, NOT, NAND, NOR and XOR, representing circuits with Boolean expressions and truth tables, including prescribed circuit combinations. For example, an AND gate outputs true only when both inputs are true. An OR gate outputs true when at least one input is true. Everyday language may use ‘or’ differently in some situations, so the student should follow the formal logic definition rather than intuition.
A good first-principles activity is to write the truth table for a circuit, then compare it with the circuit diagram. If the final output differs, trace the signals one gate at a time. This is also an introduction to debugging: find the earliest incorrect representation, not merely change the final answer until it matches an expected pattern.
Step 4: Module 2 — Algorithms and Programming
The largest source of anxiety for many families is programming. But the syllabus begins with problem analysis: identify inputs, valid values, required outputs and the conditions a correct solution must satisfy. Then design an algorithm using sequence, selection and iteration. Python is a language for implementing these decisions, not a substitute for understanding them.
Variables, selection and iteration
A variable stores a value that the program may use or change. Selection chooses a path based on a condition, such as if/elif/else. Iteration repeats a set of operations using a for or while construct. Students should be able to predict what happens to variables step by step. A trace table is useful when a loop behaves unexpectedly.
For an original illustration, consider a program that counts how many numbers in a list exceed a threshold. The student needs an initial count of zero, a loop through the list, a condition comparing each item with the threshold, and an increment when the condition is met. Test with no qualifying values, one qualifying value and several values. The generic algorithm is more important than memorising one precise line of Python. The same structure can count late arrivals or temperatures above a limit.
Lists, strings, dictionaries and functions
Lists let programs work with collections of items. Strings represent textual data and can be indexed or sliced. Dictionaries organise key–value relationships. Functions give a reusable part of a program a clear responsibility, and may accept parameters and return results. Students must understand data types and operations well enough to select a structure that matches the task. A student who stores every item in a separate variable may solve a tiny example but struggle to generalise.
Teach functions as contracts: what inputs are expected, what result is produced and what assumptions must hold. If a function claims to find the largest value in a list, what happens for an empty list? Does the problem guarantee non-empty input, or should the function validate it? This attention to specifications helps prevent hidden errors in larger programs.
Testing and debugging: where strong learners distinguish themselves
SEAB includes syntax, logic and runtime errors, tracing, print-based inspection, input validation, and test cases covering normal, error and boundary conditions. Parents should encourage students to predict the expected output before running a program. Otherwise they may accept an incorrect result because it looks plausible. A program that runs without crashing is not automatically correct.
For example, a program meant to accept examination marks from 0 to 100 should test 0, 100, typical values, negatives and values above 100. The student should know whether the task requires asking for a new input or ending with a suitable response. A hidden assumption about the range can cause a significant bug even when most school practice examples are ordinary numbers.
Do not let AI-generated code become a substitute for thinking. A tool may suggest working code, but the student must understand every relevant operation, test its output and abide by school examination and integrity rules. National examinations assess the candidate’s own capability in the prescribed environment. A copied answer that the student cannot debug is fragile preparation.
Step 5: Module 3 — Spreadsheets
Students use spreadsheets not merely to type numbers but to construct formulas that remain correct when copied, analyse data and solve tasks. The 2027 syllabus includes relative, absolute and mixed references, conditional formatting, Goal Seek, mathematical and statistical functions, logical functions, text operations and lookups. A spreadsheet can contain correct-looking outputs while referring to the wrong cells, so reference logic is a major learning target.
Relative and absolute references
Suppose each item price in column B must be multiplied by a fixed tax rate in cell F1. A suitable formula might use the item’s relative row reference and an absolute reference to $F$1. If the student copies the formula down without fixing the tax-rate cell, later rows may refer to F2, F3 and empty cells. Teach which relationship should move and which should remain fixed. The dollar signs represent a reasoned design choice, not decoration.
Logical and lookup functions
IF functions choose an output depending on a condition; lookup functions find related information in a table. The syllabus prescribes named functions, including VLOOKUP, HLOOKUP, INDEX and MATCH, as well as common statistical and text functions. Students need to know when exact matching is required and when an approximate lookup is appropriate. An approximate match in an unsorted or incorrectly arranged table may give a convincing-looking wrong answer.
For a parent-friendly exercise, create a fictional price list and ask the child to retrieve the correct product price from a code. Then deliberately add a nonexistent code. A robust solution must handle the possibility rather than silently report an unrelated item. This is the same validation mindset used in programming.
Spreadsheet quality is more than formula memorisation
Ask the student to explain the dataset, the intended output, the cell reference design and how the results can be checked. If a formula returns a large number, what range of values would be reasonable? Can a manual calculation verify one or two rows? A student who knows why their formulas are correct can adapt when the actual examination task changes layout or dataset.
Step 6: Module 4 — Networking
Networking covers how computers communicate, from home networks to the internet. Students distinguish wired and wireless media, LANs and WANs, common topologies, network protocols and the role of devices such as modems, switches, routers and wireless access points. It is easy to memorise device names without understanding the path of information. Teach students to draw a home network and trace how a message moves through the relevant components.
The syllabus includes MAC addresses, IPv4 and IPv6 addresses, and techniques for detecting transmission errors. The essential reasoning is that network communication involves identifying destinations, transmitting data and checking whether information arrives accurately. Different identifiers and devices have different roles. If a student treats a router, switch and modem as synonymous, introduce a diagram and ask what each actually contributes.
Security and privacy are distinct but connected
A system may be protected from unauthorised alteration yet still collect personal information in an intrusive way. The syllabus examines security tools, data privacy, malicious software, phishing, cookies, encryption and relevant legal concepts, including the Singapore PDPA. Students should understand that legal requirements can depend on specific circumstances; a secondary syllabus offers foundational literacy, not a complete legal interpretation.
Discuss safe daily habits: verifying sources, avoiding credential disclosure, updating devices appropriately and reviewing privacy choices. Explain the general purpose of encryption and firewalls without teaching the child to misuse networks or bypass security. Good Computing education includes responsible participation and respect for other people’s data.
Step 7: Module 5 — Impact of Computing
Computing affects communication, transport, education, retail and many other industries. K349 asks students to examine these impacts and issues such as intellectual property, software licensing, online falsehoods, artificial intelligence and machine learning. The purpose is not to teach that technology is always good or always harmful. Students should evaluate what changes, who benefits, what risks arise and what evidence supports the decision.
Copyright and software licensing
Students should know that digital software and other creations can be intellectual property, and that use depends on permissions and licences. Free-of-charge software is not necessarily identical to open-source software. A program’s licence may allow particular kinds of use, modification or distribution under conditions. Students should learn to inspect the terms rather than assume all downloadable files are unrestricted.
Social platforms and online information
Social platforms may promote material that attracts engagement, which can amplify both useful content and misleading claims. Discuss why a popular post is not automatically reliable evidence. The syllabus refers to Singapore’s POFMA framework; students should distinguish the law’s described objectives from a blanket claim that any disputed opinion is illegal. The healthy learning habit is to verify a claim, identify the source and understand how online incentives shape information.
AI and machine learning
The 2027 syllabus includes introductory concepts of AI and machine learning, including a nearest-neighbour classification example involving two quantitative features, as well as the problems that biased data or unethical use can create. A useful classroom example might classify an unfamiliar point by comparing it with labelled examples. Ask what features were selected, how distances are compared and whether the training examples represent the people or situations in question. A system can be mathematically consistent and still unfair or unreliable if the data are unsuitable.
A family interested in AI should resist the temptation to skip directly to advanced tools. Understanding inputs, representation, algorithms, validation, security and evidence will remain useful even as products change. A student should be able to explain when a computing system is appropriate and when human judgement or additional evidence is needed.
Step 8: Paper 1 — the written assessment
Paper 1 is a two-hour written examination carrying 80 marks and 60% of the subject. It covers all five modules, with mixed question formats: single- and multiple-answer multiple choice, short answers, matching, cloze passages and structured questions. Relevant formulae are provided. An approved calculator may be used. The paper therefore tests breadth and the capacity to apply computing concepts without necessarily running code.
A good preparation method is to predict the behaviour of a short algorithm on paper, then compare with a working implementation. Draw truth tables, explain network diagrams and solve spreadsheet-reference questions from the underlying logic. Students who only practise code in an editor may be weak at the written explanations. Conversely, those who memorise definitions but never debug working programs will be underprepared for Paper 2.
When reviewing the written paper, separate factual knowledge, logical reasoning and misinterpretation of the task. If a child incorrectly converts between binary and decimal, repair number representation. If they confuse a MAC address with an IP address, repair network layers at the syllabus’s level. If they misread a programming trace table, slow down and track variable updates one line at a time.
Step 9: Paper 2 — the lab-based examination
Paper 2 lasts two hours thirty minutes, carries 70 marks and contributes 40%. The official syllabus specifies one spreadsheet question and four to five programming questions. Candidates use a computer with spreadsheet software, Python and JupyterLab, and must submit softcopies of the required work. Saving the files is part of the allotted time. SEAB says a Python quick-reference guide is provided, but a guide is not a replacement for understanding programming structures.
Students should rehearse a reliable lab workflow: read the task fully; identify inputs and outputs; plan data structures and formula references; implement one small component; test with known cases; extend; test boundaries; save using the required naming and location; then verify the final files. A working solution that has not been saved correctly may not be assessable as intended. File organisation and checking are therefore genuine examination skills.
Parents should ask the school about the approved examination environment and practice arrangements. Do not assume any personal laptop configuration, third-party library, internet resource or AI assistant will be available during the examination. The syllabus defines the skills and environment; follow actual school and SEAB instructions. Home practice should be compatible with the prescribed Python and spreadsheet functions so that the student is not dependent on features outside the scope.
When a program fails, ask the child to reproduce the error with the smallest possible input. Compare expected and observed output, inspect variable values and find the first step where the logic diverges. Encourage calm testing rather than random edits. In Computing, debugging is not a sign that learning has failed; it is part of the learning.
Step 10: Diagnose the actual cause of poor results
- Problem analysis: cannot state correct inputs, outputs or validity conditions.
- Algorithm choice: copies a familiar loop without understanding the new task.
- Syntax: makes recurring Python structure, punctuation or indentation errors.
- Logic: program runs but produces an incorrect result for some cases.
- Data handling: confuses strings, numbers, lists, dictionaries or file contents.
- Spreadsheets: uses incorrect relative/absolute references or lookup matching.
- Systems knowledge: cannot explain hardware, logic gates or networking concepts.
- Digital judgement: memorises security terms without understanding purposes and trade-offs.
- Exam workflow: fails to test, save or verify the required files before submission.
A weak score can come from very different causes. A student who writes correct algorithms but mistypes syntax needs short language practice. A student whose code executes but fails on empty lists needs test-case design. A student who programs well but loses written-paper marks needs concepts and explanation. Diagnose the first failing step instead of giving every child a generic coding challenge.
Step 11: Six weeks of focused Computing improvement
Week 1 — evidence and a tiny working baseline
Collect a marked theory task, a program and a spreadsheet file. Ask the child to state what each was intended to do and how it was tested. Choose two specific weaknesses. Verify that an independently written small program can run and produce a predicted result.
Week 2 — variables, branches and loops
Practise simple transformations, then conditional and iterative tasks. Trace the values manually and compare with execution. Change an input or boundary to check whether the algorithm generalises. Encourage meaningful variable names and readable code.
Week 3 — collections and functions
Use lists, strings and dictionaries as appropriate to the current school sequence. Refactor a repeated calculation into a function with clear input and output. Test empty, typical and boundary inputs. Discuss how decomposition makes errors easier to isolate.
Week 4 — spreadsheets
Build a small fictional dataset with formulas and lookup tables. Copy formulas and inspect whether references behave correctly. Add conditions and check against a manually calculated result. Explain why each function was selected rather than memorising its name alone.
Week 5 — systems and ethical reasoning
Review binary representation, logic gates and a simple network drawing. Discuss one practical security or privacy situation and identify relevant safeguards. Study an AI or online-information example with evidence rather than hype. The child should be able to explain benefits, limits and responsible choices.
Week 6 — realistic lab workflow
Attempt a moderate spreadsheet and programming task with a time budget that includes reading, testing, saving and verification. Review the earliest error and re-test the original skills using unfamiliar inputs. Plan the next cycle based on observable independent performance.
Step 12: Is G3 Computing suitable for the child?
The subject may suit students who enjoy puzzles, problem decomposition, building things and learning from errors. Mathematical reasoning and the ability to read precise instructions help. Patience matters: many tasks require several attempts. A child does not need to own the newest computer or have studied programming since Primary 3. A supportive school and willingness to practise can be more important than early access to sophisticated tools.
It may be a poor fit if the child expects the subject to consist of entertainment, gaming or effortless use of apps. Discuss what actual lessons look like and whether the timetable allows regular independent coding practice. If the student is already overloaded, consider the overall combination rather than deciding the subject’s value from its popularity in the job market. Technology careers change; a coherent education requires more than an examination label.
Ask the school’s Computing teacher how the subject is taught, what prior Mathematics knowledge is assumed, how programming and written theory are assessed, and what support exists for students encountering coding for the first time. Arrange a small independent trial of algorithmic problem solving, not an expensive course commitment. The child should experience the work before making a choice.
Step 13: JC, polytechnic, ITE and later university routes
The 2027 SEC cohort enters revised post-secondary admissions in 2028. JC admission uses L1R4 with defined eligible subject groups and minimum requirements, while polytechnic diploma admission follows revised aggregate and course-specific rules. G3 Computing can be a useful subject, but it should not be assumed to replace required Mathematics or Science components in every aggregate or course. Always check MOE’s actual subject mapping and programme prerequisites.
Computer science, data science, engineering, information security and information-technology programmes may require strong Mathematics preparation, especially at higher levels. Taking K349 alone does not establish eligibility for a JC Mathematics combination, a computing diploma or a university degree. Some students enter digital fields through a polytechnic, ITE and later progression; others take A-Level or other routes. The admission conditions vary and should be checked for the specific intake.
Children may also use computing skills in design, commerce, biology, geography and creative work. It is sensible to take Mathematics and English seriously alongside Computing. A child who builds reliable algorithms but cannot explain them or interpret data critically is missing important transferable capabilities. Keep future pathway planning practical: identify real courses, verify criteria and maintain alternative routes.
Three common Computing learners
The enthusiastic app user: knows many digital products but struggles to write even a basic loop. Start with a small problem, clear inputs and a sequence of decisions. Explain that product familiarity is not the same as computational thinking.
The mathematical thinker: understands algorithms but makes syntax or file-submission mistakes. Use deliberate Python practice, trace tables and an explicit save-and-check workflow. Their reasoning strength is an asset that should be supported by accurate execution.
The fast coder: completes tasks quickly but neglects invalid inputs and ethics. Give boundary cases and require a test plan. Include privacy, copyright and responsible-use questions. Reliability and judgement matter even when a program appears to work.
Frequently asked questions
Does my child have to know Python before choosing Computing?
Not as a universal national requirement, but schools may set their own selection criteria. K349 includes instruction in Python and problem solving. Ask the actual school what preparation is expected. Prior familiarity can help but should not be equated with guaranteed success.
Is K349 the same as G2 Computing?
No. The SEC G3 subject is K349; the G2 Computing syllabus is K237, and G1 is K127. Their standards and assessment structures differ. Use the subject code and actual school registration to choose materials.
Can students use AI to complete the programming exam?
Do not assume access to any AI tool during the national examination. Follow SEAB and school rules for permitted software, internet use and candidate work. Preparation should develop independent coding, testing and explanation, not dependency on generated answers.
Why does a program that runs still lose marks?
It may produce wrong outputs for some inputs, fail to validate data, omit required work or not satisfy the task’s precise specification. Test normal, error and boundary cases and verify the saved output. Running without an error message is not proof of correctness.
Does the lab exam really require saving files?
Yes. The official syllabus states that candidates submit softcopies of the required work and that the allocated time includes saving. Teach a final verification routine and follow the precise examination instructions.
Will G3 Computing guarantee a software-engineering career?
No. It offers a strong introduction, but later qualifications, skills, projects, work experience and programme requirements shape opportunities. The subject may also be useful across non-computing careers.
What if the child has a low score in Mathematics?
Investigate the specific gap. Logic, numeracy, data representation and problem solving rely on mathematical thinking, but a general mark alone does not describe readiness. Talk with school staff, repair prerequisites and assess the total workload honestly.
Should parents install expensive coding tools?
Not necessarily. The official exam uses specified spreadsheet and Python/JupyterLab software. Ask the school for appropriate practice arrangements. A simple, reliable environment and carefully chosen tasks can be more educational than an elaborate setup.
Thirty-day parent action plan
Days 1–3: confirm the K349 entry and inspect one program, one spreadsheet and written theory work. Ask the student how outputs were validated and classify two recurring errors.
Days 4–10: practise variables, selection, iteration and trace tables through short original problems. Require predicted outputs before executing.
Days 11–20: build a small working program and spreadsheet, including an unusual input and a deliberate test plan. Correct errors at their first source.
Days 21–30: rehearse a lab-style task with time reserved for saving and checking, then revisit one networking or ethics topic. Review evidence, select the next priority and keep a sustainable school schedule.
Computing becomes meaningful when children learn to describe problems clearly, build solutions they understand, test them honestly and consider the people affected by the technology. That is the learning outcome worth pursuing beyond a single examination.
Official references
- SEAB 2027 SEC G3 directory — confirms Computing K349 and earlier 7155 reference.
- SEAB/MOE/Cambridge K349 2027 Computing syllabus — exact papers, five modules and lab software.
- SEAB 2027 G2 subject directory — distinct G2 Computing code.
- MOE 2026 announcements — PSE admissions transition.
- MOE 2025 announcements — JC L1R4 transition.
- eduKateSG immutable Clementi first-principles template.
Prepared using the official 2027 SEC information available on 14 November 2026. School combinations, examination software instructions and post-secondary prerequisites must be checked for the actual cohort.