Did you know? Three students can all be taking Computing in Singapore and face completely different practical tasks. One may prepare a document, spreadsheet, graphic and small block-based game. Another may work with spreadsheets, media and visual programming. A third may need to write Python code, debug an algorithm and use spreadsheet data. The subject name stays similar, but the assessment and learning goals are not identical.
That is why families need a precise guide to G1, G2 and G3 Computing. Learning Computing is not the same as spending long hours on a computer or being familiar with apps. It means understanding what information enters a system, how instructions transform it, when a result is trustworthy and how to build a digital artefact that works for a purpose.
This guide compares the 2027 Singapore–Cambridge Secondary Education Certificate (SEC) syllabuses G1 K127, G2 K237 and G3 K349 using their official paper schemes. It explains the major shift from document and media production toward Python and computational problem solving, the correct theory–practical weightings and how parents can support responsible, independent learning.
The quick answer for parents
G1, G2 and G3 Computing share foundations in computational thinking, digital systems and responsible technology use. Each level nevertheless has different modules, software requirements and assessed task formats.
The valuable question is Can your child explain what the program, spreadsheet or digital file does, test it with a new example and improve it without copying the whole solution? That demonstrates more learning than merely showing a polished screenshot.
Why G1, G2 and G3 are subject levels
Full Subject-Based Banding allows students to study individual subjects at different levels where eligible and where school arrangements permit. A child’s Computing level does not determine the level of all other subjects or their ultimate potential.
An early interest in coding may grow through different routes. The most sensible choice considers school availability, the learner’s readiness and actual post-secondary entry requirements rather than a prestige label.
The 2027 subject codes matter
G1 Computing K127 and G2 Computing K237 are listed as new for 2027, while G3 Computing K349 has its separate syllabus. The exact code determines the papers and software that a student must prepare for.
Do not assume that a private introductory coding course aligns with all three. A familiar-looking game or program may illustrate a concept without preparing the student for the registered examination tasks.
The official theory–practical balance
G1 K127 has a 40% e-Examination and 60% lab-based paper. G2 K237 divides assessment equally: 50% e-Examination and 50% lab-based paper. G3 K349 has a 60% written theory paper and 40% lab-based examination.
This progression does not imply that practical ability matters less as concepts deepen. G3’s theory includes algorithmic analysis and design questions, while its laboratory practical tests coding and spreadsheet competence.
| 2027 Computing | Paper 1 | Paper 2 | Practical focus |
|---|---|---|---|
| G1 K127 | 75 min · 60 marks · 40% · e-Exam | 120 min · 90 marks · 60% | Documents, spreadsheets, media, block programming |
| G2 K237 | 90 min · 70 marks · 50% · e-Exam | 135 min · 80 marks · 50% | Spreadsheets, media, block programming |
| G3 K349 | 120 min · 80 marks · 60% · written | 150 min · 70 marks · 40% | Python, JupyterLab, spreadsheets |
G1 Paper 1: knowledge and application
G1 K127 Paper 1 is a 75-minute, 60-mark e-Examination worth 40%. It tests all seven modules through 20 multiple-choice questions for 20 marks and shorter structured items for 40.
Students must explain how computers, networks, applications, programming and digital responsibility work. Recognising menu icons or gaming terms is not enough to demonstrate conceptual understanding.
G1 Paper 2: practical creation
G1 Paper 2 is a 2-hour, 90-mark lab-based examination worth 60%. Its three task areas include media software (approximately 30 marks), document processing and spreadsheets (around 35) and programming (around 25).
Candidates must create and submit the required softcopy files. The time includes saving work, so following file names, output formats and checking an actual saved version are integral to competent completion.
G2 Paper 1: greater structured application
G2 K237 Paper 1 is a 90-minute, 70-mark e-Examination worth 50%. Section A contains 20 one-mark multiple-choice items; Section B contributes 50 marks of short structured questions across six syllabus modules.
Students need to explain relationships, apply concepts and interpret scenarios rather than simply recite definitions. More abstract understanding can improve their ability to work confidently across different tools.
G2 Paper 2: spreadsheets, media and programs
G2 Paper 2 is a 2-hour-15-minute, 80-mark lab-based paper worth 50%. Its approximate task marks are 30 for media, 25 for spreadsheets and 25 for programming.
This is not G1’s document-processing paper. Parents should check which modules are actually assessed before making a home revision plan or buying a software course.
G3 Paper 1: written computing concepts
G3 K349 Paper 1 is a 2-hour, 80-mark written examination worth 60%. It includes single- and multiple-answer MCQs, short-answer, matching, cloze and structured questions covering all five G3 modules.
Students study computer architecture and data representation, algorithms, programming, spreadsheets, networking and the impacts of computing. Accurate conceptual explanation and application are central.
G3 Paper 2: Python and spreadsheets
G3 Paper 2 is a 2-hour-30-minute, 70-mark lab-based paper worth 40%. It uses spreadsheet, Python and JupyterLab software, including one spreadsheet question and four to five programming questions.
The official syllabus states a Python quick-reference guide will be provided. This is a materially different assessment from G1 and G2 block-based programming and media tasks.
An accurate paper comparison
G1 and G2 both have e-Examination Paper 1 and a practical lab-based Paper 2. G3 has a written Paper 1 rather than the e-Exam description specified for G1 and G2, and a practical that focuses on Python and spreadsheets.
Parents should plan the correct examination experience. Practising only mouse clicks does not replace writing conceptual answers at G3; typing Python alone does not cover media tasks at G1 and G2.
The seven G1 modules
The G1 syllabus includes Computer Fundamentals, Networking, Impact of Computing, Document Processing, Spreadsheets, Media Software and Programming. The practical paper draws from the four production-focused modules.
The variety helps a learner build digital fluency. Creating a document, chart, image, video or game requires different skills, yet each benefits from clear instructions, checking and appropriate file handling.
The six G2 modules
G2 lists Computer Fundamentals, Networking, Impact of Computing, Spreadsheets, Media Software and Programming. There is no separate Document Processing module at this level within the official six-module specification.
That change should not be misread as saying document literacy is unimportant in life. It simply distinguishes what the 2027 K237 national examination specifically tests.
The five G3 modules
G3 K349 covers Computing Fundamentals, Algorithms and Programming, Spreadsheets, Networking and Impact of Computing. Algorithm design and programmatic problem-solving become more prominent.
Media production and presentation tools are useful outside the subject, but parents should not assume they form the official G3 lab-based Paper 2 when the specification focuses on spreadsheets and Python programming.
What is computational thinking?
Computational thinking includes identifying a problem’s essential requirements, separating a large task into manageable parts, choosing representations and developing a solution that can be tested.
It is not necessarily the act of using a computer. A learner can plan a program on paper, trace an algorithm with a table and identify a flawed assumption before writing any code.
Abstraction is purposeful simplification
Abstraction removes details that do not matter to the current problem while preserving the information needed for a correct result. It is not the same as forgetting important requirements.
For an invented school-canteen calculator, the price and quantity matter to the total; a customer’s favourite colour does not. Ask which inputs can be excluded safely.
Decomposition breaks a problem into jobs
A program might need to receive data, verify input, calculate a result and communicate it. Each step can be tested separately before the parts are combined.
This reduces debugging difficulty. If the final answer is wrong, the learner can locate whether the problem lies in the input, calculation or presentation instead of replacing everything.
A computer is a system, not one clever chip
Processing, memory, storage, inputs, outputs and software work together. Understanding their roles helps students interpret performance claims and troubleshoot ordinary problems.
A device with more storage does not automatically have more working memory. Ask the child what each component is responsible for in a specific task.
Memory and storage have different jobs
Main memory supports active processing, while secondary storage keeps files and other data persistently under appropriate conditions. The difference helps explain why an app can be open while its unsaved work remains vulnerable.
Teach the practical distinction between keeping a program running, saving a file and making a backup. These are different operations with different consequences.
Data size has meaning
Bits and bytes describe quantities of information. G3 deepens the treatment of digital units and representation, including different multiples and the structure of numerical data.
Students should estimate whether a stated file size is plausible and distinguish the numerical scale from a claim about quality. A huge video file is not automatically a better video.
Binary and hexadecimal at G3
The G3 K349 syllabus explicitly assesses conversion among binary, decimal and hexadecimal, and fixed-bit representation including two’s complement. These topics help explain how numbers are encoded and processed.
A pupil should be able to explain the meaning of a bit pattern with its defined width and representation, rather than memorise a conversion recipe detached from context.
Logic gates and truth tables
G3 Computing includes AND, OR, NOT and other logic gates and interpretation of Boolean relationships. These are precise models of conditions used in systems.
An effective learning exercise compares a statement in ordinary language with its truth table. A single word such as ‘or’ can be misunderstood if the underlying logic is not defined.
Networks: communication has dependencies
G1, G2 and G3 all consider networking concepts but at different depths. A network has devices, interfaces and routes for communicating or sharing resources.
Being connected to a local Wi-Fi network does not guarantee access to an external website. Ask what part of the route is working before guessing at the cause of a connection failure.
Internet and local network are distinct
A local area network connects devices across a bounded setting; the internet interconnects many networks. An intranet or organisation-only resource can operate under other access arrangements.
A child should be able to draw a simple conceptual diagram and explain where a message travels. Exact equipment choices depend on the real system.
Why cloud storage is not a magical backup
Cloud services can provide access to files and applications through networks, but permissions, synchronisation, failures and offline availability all matter.
Students should know where a file is stored, how to save changes and whether an independent backup exists. A cloud icon does not guarantee that all earlier versions are recoverable.
Responsible computing means protecting people
Cybersecurity, privacy, copyright and the social impacts of technology are important in the curricula. Tools that make work convenient can also create risks when poorly configured or carelessly used.
Teach secure habits without requesting passwords or private account details. Use fictional examples to discuss phishing, data collection and permissions.
An AI-generated answer may be convincing and wrong
AI systems can produce text or calculations that seem confident while containing errors or inappropriate assumptions. Understanding the output requires independent evidence and awareness of the tool’s limitations.
Students should comply with school academic-integrity rules and be able to explain their own work. Copying an AI-generated solution is not the same as acquiring programming competence.
Document processing remains valuable at G1
G1 includes document layout, text organisation, graphics, sections and other word-processing tools. These decisions affect how information is presented to a reader.
One learning task is a fictional school-event guide with a title, appropriate headings and clear time details. The result should be checked for readability, not merely decorated with shapes.
A spreadsheet is a model of a relationship
Spreadsheets appear at all three levels. Students use cells, formulas, functions and data displays to transform inputs into information.
A neatly formatted sheet can still be wrong if a formula references the wrong cell or an underlying quantity is mislabeled. Ask the learner to check one result manually before trusting a hundred-row table.
Relative versus absolute references
Cell references may adjust when a formula is copied or remain fixed when that is the intended logic. Confusing those behaviours can produce systematic errors across a table.
A small invented spreadsheet with two rows is enough to reveal the issue. The pupil should predict which references move and why before filling an entire column.
Choose the right chart for the data
Bars, columns and line charts represent different relationships. The choice depends on whether the task compares categories, shows change over time or communicates another data pattern.
Labels, scales, units and context determine whether a chart is honest. An attractive chart without a meaningful measurement is not evidence.
G3 spreadsheets and programming interact
A programming problem may ask for data to be processed consistently, while a spreadsheet provides a complementary way to organise or analyse numerical relationships. The two methods can be compared through simple examples.
The G3 practical requires the student to use authorised spreadsheet and Python tools, not assume that importing a spreadsheet automatically creates a correct algorithm.
G1 and G2 media are assessed, not optional decoration
Their practical papers include creation or manipulation of graphics, presentations, video or related media according to the specific syllabus. These outputs require a clear communication purpose and technical accuracy.
More transitions, sound effects or layers do not automatically improve a result. The student should be able to explain which audience need a chosen design serves.
Vector and raster images behave differently
Vector images represent shapes or paths, while raster images consist of pixels. Resizing affects them differently, and the correct choice depends on the intended output.
Ask whether a graphic needs to remain editable as shapes or whether photographic detail matters. Selecting the format should be a reasoned technical decision.
Presentation hierarchy helps a reader
A slide or graphic should make important content visible, use appropriate contrast and avoid unnecessary visual complexity. Design choices are part of making a useful digital artefact.
A learner can test whether the intended viewer understands the main point before explaining it. If not, the layout or message needs reconsideration.
Scratch develops block-based logic
G1 and G2 programming outcomes involve visual programming concepts such as sequences, variables, conditions, loops, events and sprites, with the relevant school software and syllabus requirements.
A block program is not automatically trivial. A child must still predict state changes, choose appropriate conditions and correct errors when inputs differ from expectations.
G3 Python requires text-based programming control
K349 explicitly includes Python, JupyterLab and algorithms. Students need to understand variables, conditions, iterations, functions, file interactions and the other constructs specified by the official syllabus.
Do not treat a successful copied script as proof of mastery. Ask the learner to predict its behaviour, explain a key line and modify it for a new requirement.
An algorithm can be correct before it is coded
A clear algorithm states what inputs are expected, what outputs are needed and which steps transform one into the other. Flowcharts and pseudocode-like descriptions can help reveal missing cases.
A student should be able to test the logic with small examples even before running a program. Coding makes the algorithm executable but does not automatically make its reasoning valid.
Conditions can fail at the boundary
An if statement may classify an input incorrectly when it equals the exact threshold. A program that works for obviously small and large values may still fail on that boundary.
Test a number below, at and above the threshold. The child should explain why the defined comparison treats each value as intended.
Loops can repeat too much or too little
A loop may terminate one step early, run forever or skip a needed case. The cause may be an incorrect condition or update rule.
Have the learner trace the values for three iterations on paper. That often reveals the error without blindly replacing blocks or lines of code.
Data validation protects the meaning of results
If a spreadsheet expects a non-negative quantity but receives text or an invalid value, its calculations may fail or produce misleading results. A program similarly needs to decide how to handle unsuitable input.
Students should identify what is considered valid, how errors are reported and whether the result remains meaningful after an exception. Robustness is part of computing competence.
Debugging is an evidence-based process
Debugging begins with a predicted result and an observed one. The learner then isolates the step where behaviour diverges, proposes a correction and tests again.
Randomly changing formulas or lines of code can sometimes appear to solve a problem but create new faults elsewhere. A precise diagnostic method is more reliable.
Test cases should cover more than the happy path
A solution should be checked with normal inputs, relevant boundaries and unexpected conditions that the task expects it to handle. The number of cases depends on the specification.
An ‘it worked once’ claim is weak. Students should be able to state which behaviours were tested and which limitations remain.
Why saved files matter in the lab exam
G1, G2 and G3 Paper 2 examinations require submitting softcopy work. The stated duration includes saving it on the candidate’s computer.
A functioning program that was not saved as directed is an incomplete submission. Practise checking file names, locations, formats and whether the saved file can actually be opened.
The hidden Mathematics in Computing
Percentages, ratios, coordinates, formulas, logical conditions and data representation all demand mathematical clarity. A programming result may be wrong because the model was misinterpreted, not because of syntax.
Identify the mathematical relationship first. Then check that the spreadsheet or code implements that relationship and uses units appropriately.
The hidden English in Computing
Technical tasks contain conditions such as ‘only if’, ‘at least’, ‘except’ and ‘before’. Missing one word can change a rule or invalidate the entire output.
Ask the child to restate requirements in plain language before creating the file. A beautiful artefact that solves the wrong task is still an incorrect solution.
A useful cross-level example: fictional canteen sales
G1 might prepare a readable sales document, chart and simple block interaction. G2 might build more developed spreadsheet and media outputs with visual programming. G3 might design a Python-based data-processing solution supported by a spreadsheet.
The example is illustrative, not an official examination task. All three levels can test accuracy, user needs, clear inputs and responsible handling of fabricated data.
How to avoid unnecessary personal-data collection
School projects do not need real classmates’ phone numbers, addresses, grades or account information to demonstrate technical competence. A fictional dataset can support meaningful practice.
Students should learn to minimise data collection, protect access and understand the consequences of publishing digital artefacts.
How parents can assess a finished project
Ask What problem does it solve? Which requirement did you test? What happens with a different input? Then let the child show a working file rather than only a screenshot.
A student who can repair a new error independently is showing more robust competence than one who can reproduce only the exact teacher example.
A weekly Computing study plan
Alternate a short explanation of a computing concept with an independently completed software task. Include one focused debugging exercise and a review of safe file-handling habits.
Match the software to the current syllabus: G1 includes document and media work, G2 focuses on spreadsheets, media and visual programming, and G3 needs Python and spreadsheet competence.
A four-week diagnostic cycle
Week one: check hardware, networks and data concepts. Week two: practise the practical software demanded by the level. Week three: apply programming concepts to unfamiliar test cases. Week four: complete a realistic timed task and analyse the most persistent error.
This is a suggested home structure, not SEAB’s timetable. School tools, assignments, required software and the student’s workload determine an appropriate pace.
When a student seems good at games but struggles in Computing
Playing a game involves interacting with a system, while Computing can require explaining its logic and creating a reliable one. Familiarity with an interface is not equivalent to understanding how it operates.
Ask the child to describe the game’s input, rules, state changes and output. That explanation can reveal a useful bridge toward programming.
When a student can code but struggles in the theory paper
They may lack models of data representation, hardware, networks or the ethical impacts of computing. These concepts cannot always be reconstructed by trial-and-error in a program.
Teach the specific missing idea with diagrams and plain explanations, then return to an applied scenario. A larger programming project is not automatically the best remedy.
How to recognise effective Computing tutoring
A capable tutor can distinguish software familiarity, conceptual gaps, mathematical errors, program design faults and debugging difficulties. They should know the actual K-code and official assessed tools.
Parents should look for independent task completion, clear testing and safe responsible technology habits—not simply a demonstration by the tutor while the child watches.
What Computing may lead to later
Computing can support interest in software, data, media, operations and technology-related learning, as well as general digital literacy useful in many roles. The subject does not guarantee a particular diploma or job.
Check current institutional prerequisites and school offerings. A course’s named subject requirements can matter more than the generic claim that all Computing qualifications are interchangeable.
Frequently asked questions: Are all three Computing levels new in 2027?
The SEAB directory marks G1 K127 and G2 K237 as new for 2027. G3 K349 has a separate published 2027 specification and is not described in the same way.
Frequently asked questions: Does G3 Computing use Scratch in Paper 2?
The official G3 K349 practical specifies Python, JupyterLab and spreadsheet software rather than G1/G2 block-based media-and-game tasks. Use the correct syllabus and school software.
Frequently asked questions: Which level has the greatest practical-paper weighting?
G1 K127 has 60% for its lab-based paper. G2 K237 has 50%, and G3 K349 has 40%. This does not make G3 less technical; its written paper tests deeper computational concepts and application.
Frequently asked questions: Are all examination questions compulsory?
The 2027 Computing K127, K237 and K349 paper schemes state that candidates offer both papers and answer compulsory questions in those papers. Follow the actual question paper’s instructions for individual tasks.
Frequently asked questions: Do students need an expensive personal laptop?
The official syllabuses specify examination-centre hardware and software arrangements. Families should ask the school about required practice tools before assuming an expensive computer purchase is necessary.
The lasting reason to study G1 G2 G3 Computing
Computing gives young people an opportunity to move from passive digital consumption to deliberate digital creation. They learn to question inputs, represent information accurately, build instructions and inspect the consequences of what technology does.
Read the official 2027 G1 Computing K127 syllabus, G2 Computing K237 syllabus and G3 Computing K349 syllabus. Compare Why Study G1 Computing, Why Study G2 Computing and Why Study G3 Computing. The unchanged Clementi Secondary 1 Mathematics tutorial reference shows a separate feedback-led teaching approach and does not imply that Computing classes are currently available.