What Is the Difference Between G1, G2 and G3 Computing Syllabuses? | 2027 SEC Guide

Parents comparing the G1, G2 and G3 Computing syllabuses are often looking for one practical answer: will the child be learning the same computer skills at three different speeds? No. In the 2027 Singapore–Cambridge Secondary Education Certificate (SEC), Computing is offered at three separately coded subject levels with different topic choices, tools, examination formats and balances between practical work and written reasoning.

The short answer is that G1 Computing K127 places strong emphasis on practical digital production, including documents, spreadsheets, media and block programming. G2 Computing K237 retains practical digital and computational work but changes the syllabus and assessment balance. G3 Computing K349 develops a more formal route through algorithms, data and Python programming. It would be misleading to describe the change as simply ‘more difficult Scratch’ or ‘more computer homework’.

The distinction matters when choosing a revision book, arranging access to software, discussing a subject-level change or planning the next school year. It also matters to students who enjoy computers but have not yet found out whether they prefer building useful digital artefacts, explaining computing systems or constructing and debugging algorithms. Those are related interests, but they are not identical skills.

This guide is for the 2027 SEC school-candidate specifications. Different graduation years may follow different documents, and a school decides which subjects and subject levels it can offer. The classroom examples are independent educational illustrations, not copies of SEAB examination questions and not assurances that a particular question will appear.

Throughout the article, look for one principle: choose the correct syllabus before judging the student’s progress. A program running once, a polished video or a correct-looking chart tells us something useful, but never the entire story. The evidence must match the intended learning outcome and the actual level.

Quick answer: G1 K127 emphasises applied digital production and Scratch, G2 K237 builds broader independence in computing tasks, and G3 K349 emphasises formal algorithms, Python and stronger written reasoning. The theory/practical weightings are respectively 40/60, 50/50 and 60/40.

Official sources: SEAB G1 K127 · SEAB G2 K237 · SEAB G3 K349. Checked 16 October 2026. Read the relevant document in full for assessable scope and permissions.

G1 vs G2 vs G3 Computing: the 2027 comparison

Level / code	Syllabus focus	2027 SEC examination structure
G1 K127	Seven modules: fundamentals, networks, impacts, document processing, spreadsheets, media software and programming; prescribed application software and Scratch.	Paper 1 e-Examination (1h 15m), 40%; Paper 2 lab (2h), 60%.
G2 K237	Six modules: fundamentals, networks, impacts, spreadsheets, media software and programming; stronger spreadsheet and block-programming reasoning.	Paper 1 e-Examination (1h 30m), 50%; Paper 2 lab (2h 15m), 50%.
G3 K349	Five modules covering computing foundations, data and algorithms, Python programming and spreadsheet work; more formal solution design and reasoning.	Paper 1 written (2h), 60%; Paper 2 lab (2h 30m), 40%.

A comparison table is a map, not the full syllabus. The actual learning outcomes determine the topic detail; the paper descriptions determine how evidence of learning is collected. These two documents should be read together. A learner with strong practical habits might still need explicit help with written explanation; one with strong theory could still lose marks by not saving, checking or submitting a required file correctly.

Contents: choose the question you need

• Twelve syllabus differences explained
• Fourteen worked everyday comparisons
• A practical eight-week preparation route
• Questions parents and students ask
• Official documents and related reading

Twelve syllabus differences explained

1. The central purpose of Computing

A computer is a tool for representing information, processing instructions and producing results that another person can use.

At G1: Recognise components, connect inputs and outputs, use applications safely and build functional digital artefacts. At G2: Use computing ideas with more independence across applications, spreadsheets, networks, media and programming. At G3: Reason about data, algorithms, software behaviour and formal programming solutions rather than relying on a successful demonstration alone.

Ask whether the student can explain what data entered the system, what transformation occurred and how the output can be checked. For a parent or teacher, the useful conversation is not ‘Which syllabus sounds better?’ but ‘Which performance would be observable after the child has genuinely understood this outcome?’ That question turns an abstract comparison into a fair assessment of actual readiness. A brief unaided attempt is more revealing than the amount of time spent at a computer.

When the student struggles, distinguish a knowledge gap from an execution gap. Ask them to explain the intended result before showing where they clicked or what they typed. Then give a slightly changed task to see whether the idea transfers. A mistake that returns under small changes calls for teaching, not simply another identical worksheet. Where the levels demand different topics or tools, teach the missing bridge explicitly rather than assuming practice at one level will automatically meet another.

2. What the subject codes actually tell you

Codes identify examinable specifications, not a child’s identity or personal worth.

At G1: K127 is the G1 Computing syllabus for the stated 2027 SEC cohort. At G2: K237 is the G2 Computing syllabus, with its own modules and paper structure. At G3: K349 is the G3 Computing syllabus, not an automatic extension of K237.

Read the code on the student’s actual registration information and check the assessment scheme inside the document. For a parent or teacher, the useful conversation is not ‘Which syllabus sounds better?’ but ‘Which performance would be observable after the child has genuinely understood this outcome?’ That question turns an abstract comparison into a fair assessment of actual readiness. A brief unaided attempt is more revealing than the amount of time spent at a computer.

When the student struggles, distinguish a knowledge gap from an execution gap. Ask them to explain the intended result before showing where they clicked or what they typed. Then give a slightly changed task to see whether the idea transfers. A mistake that returns under small changes calls for teaching, not simply another identical worksheet. Where the levels demand different topics or tools, teach the missing bridge explicitly rather than assuming practice at one level will automatically meet another.

3. The importance of application software

A familiar-looking screen can hide different expected skills.

At G1: Document processing, spreadsheets, charts and media tasks have explicit places in K127. At G2: Spreadsheets and media work remain central, while document processing is not a separate named K237 module. At G3: G3 focuses differently, including spreadsheet work within a data-and-algorithm-oriented course.

Avoid buying a guide on the strength of the word Computing alone; check the topics, software and practical format. For a parent or teacher, the useful conversation is not ‘Which syllabus sounds better?’ but ‘Which performance would be observable after the child has genuinely understood this outcome?’ That question turns an abstract comparison into a fair assessment of actual readiness. A brief unaided attempt is more revealing than the amount of time spent at a computer.

When the student struggles, distinguish a knowledge gap from an execution gap. Ask them to explain the intended result before showing where they clicked or what they typed. Then give a slightly changed task to see whether the idea transfers. A mistake that returns under small changes calls for teaching, not simply another identical worksheet. Where the levels demand different topics or tools, teach the missing bridge explicitly rather than assuming practice at one level will automatically meet another.

4. Scratch and Python are not synonyms

Students can learn computational thinking through more than one programming environment.

At G1: Use the syllabus-prescribed block-programming environment and explain sequences, decisions and repeated actions. At G2: Develop stronger independence in a block-programming environment while managing more demanding tasks. At G3: Use Python and the specified laboratory environment; work with written code and formal algorithmic reasoning.

A good transition plan protects concepts such as variables and testing, while explicitly teaching the changed syntax and tools. For a parent or teacher, the useful conversation is not ‘Which syllabus sounds better?’ but ‘Which performance would be observable after the child has genuinely understood this outcome?’ That question turns an abstract comparison into a fair assessment of actual readiness. A brief unaided attempt is more revealing than the amount of time spent at a computer.

When the student struggles, distinguish a knowledge gap from an execution gap. Ask them to explain the intended result before showing where they clicked or what they typed. Then give a slightly changed task to see whether the idea transfers. A mistake that returns under small changes calls for teaching, not simply another identical worksheet. Where the levels demand different topics or tools, teach the missing bridge explicitly rather than assuming practice at one level will automatically meet another.

5. Theory and practical examinations

The weighting can change how preparation should be organised even when a student likes the subject.

At G1: The 40% theory-oriented e-paper and 60% computer-based practical paper reward reliable application skills. At G2: The two papers each contribute 50%, with a longer written/e-paper component and a lab-based component. At G3: Written Paper 1 contributes 60% and computer-based Paper 2 contributes 40%; Python and algorithmic reasoning matter.

The weights are taken from the 2027 syllabuses; don’t treat software practice as a substitute for knowledge or the reverse. For a parent or teacher, the useful conversation is not ‘Which syllabus sounds better?’ but ‘Which performance would be observable after the child has genuinely understood this outcome?’ That question turns an abstract comparison into a fair assessment of actual readiness. A brief unaided attempt is more revealing than the amount of time spent at a computer.

When the student struggles, distinguish a knowledge gap from an execution gap. Ask them to explain the intended result before showing where they clicked or what they typed. Then give a slightly changed task to see whether the idea transfers. A mistake that returns under small changes calls for teaching, not simply another identical worksheet. Where the levels demand different topics or tools, teach the missing bridge explicitly rather than assuming practice at one level will automatically meet another.

6. What counts as a satisfactory result

A file can open correctly without proving that it follows the brief.

At G1: Check that the saved document, sheet, media item or program meets the task instructions. At G2: Add appropriate validation, organization and explanation rather than relying on output appearance. At G3: Demonstrate a defensible solution, appropriate reasoning and tests of algorithmic behaviour.

Show the test data, expected result, actual result and the reason any difference occurred. For a parent or teacher, the useful conversation is not ‘Which syllabus sounds better?’ but ‘Which performance would be observable after the child has genuinely understood this outcome?’ That question turns an abstract comparison into a fair assessment of actual readiness. A brief unaided attempt is more revealing than the amount of time spent at a computer.

When the student struggles, distinguish a knowledge gap from an execution gap. Ask them to explain the intended result before showing where they clicked or what they typed. Then give a slightly changed task to see whether the idea transfers. A mistake that returns under small changes calls for teaching, not simply another identical worksheet. Where the levels demand different topics or tools, teach the missing bridge explicitly rather than assuming practice at one level will automatically meet another.

7. Reading assessment instructions precisely

Missed instructions can erase marks despite sound technical knowledge.

At G1: Notice file names, saved locations, required formats, question constraints and the time left for submission. At G2: Interpret data conditions and task constraints while keeping work readable and verifiable. At G3: Identify assumptions, boundary cases, algorithm requirements and the difference between showing and proving.

Make a one-line checklist of deliverables before starting, then verify each item before the end. For a parent or teacher, the useful conversation is not ‘Which syllabus sounds better?’ but ‘Which performance would be observable after the child has genuinely understood this outcome?’ That question turns an abstract comparison into a fair assessment of actual readiness. A brief unaided attempt is more revealing than the amount of time spent at a computer.

When the student struggles, distinguish a knowledge gap from an execution gap. Ask them to explain the intended result before showing where they clicked or what they typed. Then give a slightly changed task to see whether the idea transfers. A mistake that returns under small changes calls for teaching, not simply another identical worksheet. Where the levels demand different topics or tools, teach the missing bridge explicitly rather than assuming practice at one level will automatically meet another.

8. Programming errors have different causes

A program may fail because the idea is wrong, not because the learner typed carelessly.

At G1: Separate mistaken blocks, wrong values and missed repeated steps. At G2: Diagnose the logic before moving blocks or changing a condition at random. At G3: Distinguish syntax errors, logic faults, runtime conditions and incorrectly specified algorithms.

Keep a small test log: input, expected output, observed output, likely cause and change attempted. For a parent or teacher, the useful conversation is not ‘Which syllabus sounds better?’ but ‘Which performance would be observable after the child has genuinely understood this outcome?’ That question turns an abstract comparison into a fair assessment of actual readiness. A brief unaided attempt is more revealing than the amount of time spent at a computer.

When the student struggles, distinguish a knowledge gap from an execution gap. Ask them to explain the intended result before showing where they clicked or what they typed. Then give a slightly changed task to see whether the idea transfers. A mistake that returns under small changes calls for teaching, not simply another identical worksheet. Where the levels demand different topics or tools, teach the missing bridge explicitly rather than assuming practice at one level will automatically meet another.

9. Data and spreadsheets

The hard part is deciding what the values mean and which operation answers the question.

At G1: Enter data accurately, select suitable formulas and use charts to communicate a pattern. At G2: Use more deliberate spreadsheet logic and interpret outputs in context. At G3: Apply spreadsheets alongside stronger data handling and computational problem-solving.

Always verify units, cell references and whether the chart actually answers the user’s question. For a parent or teacher, the useful conversation is not ‘Which syllabus sounds better?’ but ‘Which performance would be observable after the child has genuinely understood this outcome?’ That question turns an abstract comparison into a fair assessment of actual readiness. A brief unaided attempt is more revealing than the amount of time spent at a computer.

When the student struggles, distinguish a knowledge gap from an execution gap. Ask them to explain the intended result before showing where they clicked or what they typed. Then give a slightly changed task to see whether the idea transfers. A mistake that returns under small changes calls for teaching, not simply another identical worksheet. Where the levels demand different topics or tools, teach the missing bridge explicitly rather than assuming practice at one level will automatically meet another.

10. Cybersecurity and responsible computing

Security is a set of practical decisions rather than a memorised list of scary words.

At G1: Recognise scams, weak credentials, unsafe downloads and reasons for making backups. At G2: Apply security and responsible-use knowledge to more complex everyday scenarios. At G3: Explain wider ethical, social and legal implications of data and computing choices.

Don’t ask students to explore live malware or break accounts; use safe hypothetical cases. For a parent or teacher, the useful conversation is not ‘Which syllabus sounds better?’ but ‘Which performance would be observable after the child has genuinely understood this outcome?’ That question turns an abstract comparison into a fair assessment of actual readiness. A brief unaided attempt is more revealing than the amount of time spent at a computer.

When the student struggles, distinguish a knowledge gap from an execution gap. Ask them to explain the intended result before showing where they clicked or what they typed. Then give a slightly changed task to see whether the idea transfers. A mistake that returns under small changes calls for teaching, not simply another identical worksheet. Where the levels demand different topics or tools, teach the missing bridge explicitly rather than assuming practice at one level will automatically meet another.

11. Choosing revision material

Two workbooks can use the same title while preparing for very different papers.

At G1: Seek tasks that match document processing, spreadsheet, media and block-programming objectives. At G2: Seek problems matching K237 modules and both examination modes. At G3: Seek practice in Python, data, algorithms, written explanation and lab work aligned to K349.

Check the published specification and specimen documents first; old papers may provide practice without matching the new full examination. For a parent or teacher, the useful conversation is not ‘Which syllabus sounds better?’ but ‘Which performance would be observable after the child has genuinely understood this outcome?’ That question turns an abstract comparison into a fair assessment of actual readiness. A brief unaided attempt is more revealing than the amount of time spent at a computer.

When the student struggles, distinguish a knowledge gap from an execution gap. Ask them to explain the intended result before showing where they clicked or what they typed. Then give a slightly changed task to see whether the idea transfers. A mistake that returns under small changes calls for teaching, not simply another identical worksheet. Where the levels demand different topics or tools, teach the missing bridge explicitly rather than assuming practice at one level will automatically meet another.

12. What useful progress looks like

Confidence becomes more credible when it transfers from a familiar task to a slightly different one.

At G1: Complete a new small practical task without step-by-step prompting. At G2: Explain and adapt choices across more than one kind of data or programme. At G3: Defend an algorithm and its tests under unfamiliar conditions, not just after rehearsing one example.

A good tutor asks ‘what changed?’ after correction, and revisits the idea after a delay. For a parent or teacher, the useful conversation is not ‘Which syllabus sounds better?’ but ‘Which performance would be observable after the child has genuinely understood this outcome?’ That question turns an abstract comparison into a fair assessment of actual readiness. A brief unaided attempt is more revealing than the amount of time spent at a computer.

When the student struggles, distinguish a knowledge gap from an execution gap. Ask them to explain the intended result before showing where they clicked or what they typed. Then give a slightly changed task to see whether the idea transfers. A mistake that returns under small changes calls for teaching, not simply another identical worksheet. Where the levels demand different topics or tools, teach the missing bridge explicitly rather than assuming practice at one level will automatically meet another.

Fourteen worked everyday comparisons

These cases demonstrate how to think about different kinds of computing work. They are original teaching examples, not official specimen questions. An example can illuminate an idea used at several levels without claiming that identical questions appear in all three papers. The registered syllabus always decides which activities are examinable.

Case 1: A spreadsheet for a class outing

Imagine the following ordinary situation. A class records attendance, costs and transport payments; one total does not match the receipts. A helpful student can begin by stating the user’s goal in one sentence, then identifying the information needed and what a successful result would look like. This prevents the familiar mistake of reaching for a computer tool before deciding what problem it is meant to solve.

G1 learning lens: Check typed numbers, use suitable formulas and present a clear final table. G2 learning lens: Find whether the discrepancy comes from a formula, selected range or omitted category and explain the correction. G3 learning lens: Model the task carefully, test data handling, justify computed outputs and consider exceptional entries. These are differentiated learning illustrations; they should be checked against the actual specification rather than copied as predictions of examination questions.

The common trap: A pretty chart is not a reconciliation. Compare source totals with independently calculated values. A useful correction is to ask the learner to say what evidence would distinguish a good result from a lucky one. They should name one ordinary test, one changed condition and one reason for trusting or questioning the output. This adds a small amount of thinking to the practical activity and makes the learning visible.

For home practice, keep the case proportionate. The child might explain the plan aloud, show a short test record or correct a single flawed example. Do not turn every exercise into a full project. The goal is steady independence: clear instructions, accurate execution, a check of the outcome and an explanation of what to do if the result is not as expected.

Case 2: A game character stuck at the edge

Imagine the following ordinary situation. A pupil builds a simple movement game but the character escapes the screen. A helpful student can begin by stating the user’s goal in one sentence, then identifying the information needed and what a successful result would look like. This prevents the familiar mistake of reaching for a computer tool before deciding what problem it is meant to solve.

G1 learning lens: Check the Scratch sequence, boundary condition and repeated movement instructions. G2 learning lens: Trace conditions and block execution, reproduce the failure and correct the logic. G3 learning lens: Express the intended behaviour as an algorithm, test boundary values and implement a reliable written-code solution. These are differentiated learning illustrations; they should be checked against the actual specification rather than copied as predictions of examination questions.

The common trap: A single successful run with the character in the middle does not test the edge. A useful correction is to ask the learner to say what evidence would distinguish a good result from a lucky one. They should name one ordinary test, one changed condition and one reason for trusting or questioning the output. This adds a small amount of thinking to the practical activity and makes the learning visible.

For home practice, keep the case proportionate. The child might explain the plan aloud, show a short test record or correct a single flawed example. Do not turn every exercise into a full project. The goal is steady independence: clear instructions, accurate execution, a check of the outcome and an explanation of what to do if the result is not as expected.

Case 3: An online form asking for details

Imagine the following ordinary situation. A club registration form requests full names, addresses and emergency contact details. A helpful student can begin by stating the user’s goal in one sentence, then identifying the information needed and what a successful result would look like. This prevents the familiar mistake of reaching for a computer tool before deciding what problem it is meant to solve.

G1 learning lens: Identify which data is necessary, take care with privacy and choose a safe collection route. G2 learning lens: Explain data collection decisions and assess risks of accidental exposure. G3 learning lens: Reason about information security, data handling, permissions and ethical consequences of processing the collected details. These are differentiated learning illustrations; they should be checked against the actual specification rather than copied as predictions of examination questions.

The common trap: More data is not automatically better; collect only what the real task justifies. A useful correction is to ask the learner to say what evidence would distinguish a good result from a lucky one. They should name one ordinary test, one changed condition and one reason for trusting or questioning the output. This adds a small amount of thinking to the practical activity and makes the learning visible.

For home practice, keep the case proportionate. The child might explain the plan aloud, show a short test record or correct a single flawed example. Do not turn every exercise into a full project. The goal is steady independence: clear instructions, accurate execution, a check of the outcome and an explanation of what to do if the result is not as expected.

Case 4: A misleading chart

Imagine the following ordinary situation. Monthly temperatures have been plotted, but the vertical scale exaggerates a tiny change. A helpful student can begin by stating the user’s goal in one sentence, then identifying the information needed and what a successful result would look like. This prevents the familiar mistake of reaching for a computer tool before deciding what problem it is meant to solve.

G1 learning lens: Choose sensible chart labels, axes and units and notice the scale. G2 learning lens: Explain why the selected representation may mislead and rebuild it to match the question. G3 learning lens: Investigate how data transformation, selection and display choices influence conclusions. These are differentiated learning illustrations; they should be checked against the actual specification rather than copied as predictions of examination questions.

The common trap: A chart can be technically valid yet poor evidence for the claim beside it. A useful correction is to ask the learner to say what evidence would distinguish a good result from a lucky one. They should name one ordinary test, one changed condition and one reason for trusting or questioning the output. This adds a small amount of thinking to the practical activity and makes the learning visible.

For home practice, keep the case proportionate. The child might explain the plan aloud, show a short test record or correct a single flawed example. Do not turn every exercise into a full project. The goal is steady independence: clear instructions, accurate execution, a check of the outcome and an explanation of what to do if the result is not as expected.

Case 5: A slow computer

Imagine the following ordinary situation. A student assumes that deleting all files will fix slow application performance. A helpful student can begin by stating the user’s goal in one sentence, then identifying the information needed and what a successful result would look like. This prevents the familiar mistake of reaching for a computer tool before deciding what problem it is meant to solve.

G1 learning lens: Distinguish storage from memory and identify relevant hardware and software components. G2 learning lens: Compare plausible constraints instead of attributing every slowdown to one cause. G3 learning lens: Form a reasoned model of the bottleneck and decide which observations would support or refute it. These are differentiated learning illustrations; they should be checked against the actual specification rather than copied as predictions of examination questions.

The common trap: Never recommend destructive fixes before checking backups and understanding the problem. A useful correction is to ask the learner to say what evidence would distinguish a good result from a lucky one. They should name one ordinary test, one changed condition and one reason for trusting or questioning the output. This adds a small amount of thinking to the practical activity and makes the learning visible.

For home practice, keep the case proportionate. The child might explain the plan aloud, show a short test record or correct a single flawed example. Do not turn every exercise into a full project. The goal is steady independence: clear instructions, accurate execution, a check of the outcome and an explanation of what to do if the result is not as expected.

Case 6: A phishing email

Imagine the following ordinary situation. A realistic message asks the recipient to sign in through an unfamiliar link. A helpful student can begin by stating the user’s goal in one sentence, then identifying the information needed and what a successful result would look like. This prevents the familiar mistake of reaching for a computer tool before deciding what problem it is meant to solve.

G1 learning lens: Notice suspicious urgency and link cues, and practise safe verification. G2 learning lens: Identify risks and propose more than one appropriate defensive step. G3 learning lens: Assess the data/security context, the possible consequences and the limitations of superficial visual checks. These are differentiated learning illustrations; they should be checked against the actual specification rather than copied as predictions of examination questions.

The common trap: The safest exercise uses invented examples; do not open questionable links to prove a point. A useful correction is to ask the learner to say what evidence would distinguish a good result from a lucky one. They should name one ordinary test, one changed condition and one reason for trusting or questioning the output. This adds a small amount of thinking to the practical activity and makes the learning visible.

For home practice, keep the case proportionate. The child might explain the plan aloud, show a short test record or correct a single flawed example. Do not turn every exercise into a full project. The goal is steady independence: clear instructions, accurate execution, a check of the outcome and an explanation of what to do if the result is not as expected.

Case 7: An incorrect file format

Imagine the following ordinary situation. A slide deck has been saved under the wrong format and will not open on the school machine. A helpful student can begin by stating the user’s goal in one sentence, then identifying the information needed and what a successful result would look like. This prevents the familiar mistake of reaching for a computer tool before deciding what problem it is meant to solve.

G1 learning lens: Name files correctly, save to the required location and confirm that the output opens. G2 learning lens: Check format compatibility and manage the practical submission sequence. G3 learning lens: Treat environment constraints and reproducibility as part of the solution rather than afterthoughts. These are differentiated learning illustrations; they should be checked against the actual specification rather than copied as predictions of examination questions.

The common trap: Completing the edit is not the same as successfully delivering the requested file. A useful correction is to ask the learner to say what evidence would distinguish a good result from a lucky one. They should name one ordinary test, one changed condition and one reason for trusting or questioning the output. This adds a small amount of thinking to the practical activity and makes the learning visible.

For home practice, keep the case proportionate. The child might explain the plan aloud, show a short test record or correct a single flawed example. Do not turn every exercise into a full project. The goal is steady independence: clear instructions, accurate execution, a check of the outcome and an explanation of what to do if the result is not as expected.

Case 8: A repeated calculation

Imagine the following ordinary situation. A student manually adds the same set of figures every week and makes inconsistent mistakes. A helpful student can begin by stating the user’s goal in one sentence, then identifying the information needed and what a successful result would look like. This prevents the familiar mistake of reaching for a computer tool before deciding what problem it is meant to solve.

G1 learning lens: Replace unnecessary repetition with appropriate spreadsheet formulas. G2 learning lens: Use well-structured cell references and checks so the result updates when source values change. G3 learning lens: Consider a clearer computational procedure and test how it behaves for missing or unusual data. These are differentiated learning illustrations; they should be checked against the actual specification rather than copied as predictions of examination questions.

The common trap: An automated answer is only reliable when its source data and logic remain correct. A useful correction is to ask the learner to say what evidence would distinguish a good result from a lucky one. They should name one ordinary test, one changed condition and one reason for trusting or questioning the output. This adds a small amount of thinking to the practical activity and makes the learning visible.

For home practice, keep the case proportionate. The child might explain the plan aloud, show a short test record or correct a single flawed example. Do not turn every exercise into a full project. The goal is steady independence: clear instructions, accurate execution, a check of the outcome and an explanation of what to do if the result is not as expected.

Case 9: A video with missing credits

Imagine the following ordinary situation. A multimedia presentation uses other people’s images without identifying the sources. A helpful student can begin by stating the user’s goal in one sentence, then identifying the information needed and what a successful result would look like. This prevents the familiar mistake of reaching for a computer tool before deciding what problem it is meant to solve.

G1 learning lens: Learn responsible use of online materials and distinguish creating from copying. G2 learning lens: Select materials appropriately, document sources and explain practical restrictions. G3 learning lens: Explain digital ethics, intellectual-property obligations and how the solution could use appropriately licensed resources. These are differentiated learning illustrations; they should be checked against the actual specification rather than copied as predictions of examination questions.

The common trap: A school task does not give unlimited permission to reuse third-party creative work. A useful correction is to ask the learner to say what evidence would distinguish a good result from a lucky one. They should name one ordinary test, one changed condition and one reason for trusting or questioning the output. This adds a small amount of thinking to the practical activity and makes the learning visible.

For home practice, keep the case proportionate. The child might explain the plan aloud, show a short test record or correct a single flawed example. Do not turn every exercise into a full project. The goal is steady independence: clear instructions, accurate execution, a check of the outcome and an explanation of what to do if the result is not as expected.

Case 10: A Wi-Fi problem

Imagine the following ordinary situation. A laptop can connect in one room but not another, and somebody assumes the entire internet has failed. A helpful student can begin by stating the user’s goal in one sentence, then identifying the information needed and what a successful result would look like. This prevents the familiar mistake of reaching for a computer tool before deciding what problem it is meant to solve.

G1 learning lens: Identify devices, networks, local connections and common differences between wired and wireless links. G2 learning lens: Reason through possible points of failure and compare observed connection conditions. G3 learning lens: Analyse the situation using networking concepts and support the diagnosis with testable observations. These are differentiated learning illustrations; they should be checked against the actual specification rather than copied as predictions of examination questions.

The common trap: Connection to a router is not always proof of working external access. A useful correction is to ask the learner to say what evidence would distinguish a good result from a lucky one. They should name one ordinary test, one changed condition and one reason for trusting or questioning the output. This adds a small amount of thinking to the practical activity and makes the learning visible.

For home practice, keep the case proportionate. The child might explain the plan aloud, show a short test record or correct a single flawed example. Do not turn every exercise into a full project. The goal is steady independence: clear instructions, accurate execution, a check of the outcome and an explanation of what to do if the result is not as expected.

Case 11: An average hiding unusual values

Imagine the following ordinary situation. A canteen data set has several ordinary sales days and one unusually large purchase. A helpful student can begin by stating the user’s goal in one sentence, then identifying the information needed and what a successful result would look like. This prevents the familiar mistake of reaching for a computer tool before deciding what problem it is meant to solve.

G1 learning lens: Use a spreadsheet to check totals and suitable summaries with attention to units. G2 learning lens: Interpret how a surprising value changes a calculation and presentation. G3 learning lens: Analyse data and algorithm choices, selecting a representation justified by the problem. These are differentiated learning illustrations; they should be checked against the actual specification rather than copied as predictions of examination questions.

The common trap: Never change or drop an unusual value just to make the answer look cleaner. A useful correction is to ask the learner to say what evidence would distinguish a good result from a lucky one. They should name one ordinary test, one changed condition and one reason for trusting or questioning the output. This adds a small amount of thinking to the practical activity and makes the learning visible.

For home practice, keep the case proportionate. The child might explain the plan aloud, show a short test record or correct a single flawed example. Do not turn every exercise into a full project. The goal is steady independence: clear instructions, accurate execution, a check of the outcome and an explanation of what to do if the result is not as expected.

Case 12: A file lost before an exam

Imagine the following ordinary situation. A student has one working copy of a project and the device suddenly fails. A helpful student can begin by stating the user’s goal in one sentence, then identifying the information needed and what a successful result would look like. This prevents the familiar mistake of reaching for a computer tool before deciding what problem it is meant to solve.

G1 learning lens: Use regular backup and sensible file organisation to reduce the risk. G2 learning lens: Discuss appropriate local and cloud strategies and test whether saved material can be recovered. G3 learning lens: Evaluate reliability, integrity and the effect of failure cases on a complete computing workflow. These are differentiated learning illustrations; they should be checked against the actual specification rather than copied as predictions of examination questions.

The common trap: A backup is not verified until someone can actually retrieve a usable version. A useful correction is to ask the learner to say what evidence would distinguish a good result from a lucky one. They should name one ordinary test, one changed condition and one reason for trusting or questioning the output. This adds a small amount of thinking to the practical activity and makes the learning visible.

For home practice, keep the case proportionate. The child might explain the plan aloud, show a short test record or correct a single flawed example. Do not turn every exercise into a full project. The goal is steady independence: clear instructions, accurate execution, a check of the outcome and an explanation of what to do if the result is not as expected.

Case 13: A bug that happens only sometimes

Imagine the following ordinary situation. A quiz program gives a sensible score for most answers but fails on a blank response. A helpful student can begin by stating the user’s goal in one sentence, then identifying the information needed and what a successful result would look like. This prevents the familiar mistake of reaching for a computer tool before deciding what problem it is meant to solve.

G1 learning lens: Test ordinary and unusual inputs, then trace the relevant condition. G2 learning lens: Separate a reproduced test failure from a guess and make one controlled repair. G3 learning lens: Design boundary and invalid-input tests and explain how they support or limit confidence in correctness. These are differentiated learning illustrations; they should be checked against the actual specification rather than copied as predictions of examination questions.

The common trap: Repeating the same successful input five times is not five independent kinds of evidence. A useful correction is to ask the learner to say what evidence would distinguish a good result from a lucky one. They should name one ordinary test, one changed condition and one reason for trusting or questioning the output. This adds a small amount of thinking to the practical activity and makes the learning visible.

For home practice, keep the case proportionate. The child might explain the plan aloud, show a short test record or correct a single flawed example. Do not turn every exercise into a full project. The goal is steady independence: clear instructions, accurate execution, a check of the outcome and an explanation of what to do if the result is not as expected.

Case 14: A parent’s question about the next level

Imagine the following ordinary situation. The student enjoys computing and wants to know whether G3 would be the best next choice. A helpful student can begin by stating the user’s goal in one sentence, then identifying the information needed and what a successful result would look like. This prevents the familiar mistake of reaching for a computer tool before deciding what problem it is meant to solve.

G1 learning lens: Check whether practical foundations and independent work habits are stable. G2 learning lens: Look at the child’s readiness for more demanding reasoning as well as coding enthusiasm. G3 learning lens: Consider the formal content, Python requirements, written examination and the wider subject load. These are differentiated learning illustrations; they should be checked against the actual specification rather than copied as predictions of examination questions.

The common trap: No subject-level change should be decided by an interest label alone; ask the school about actual eligibility and future choices. A useful correction is to ask the learner to say what evidence would distinguish a good result from a lucky one. They should name one ordinary test, one changed condition and one reason for trusting or questioning the output. This adds a small amount of thinking to the practical activity and makes the learning visible.

For home practice, keep the case proportionate. The child might explain the plan aloud, show a short test record or correct a single flawed example. Do not turn every exercise into a full project. The goal is steady independence: clear instructions, accurate execution, a check of the outcome and an explanation of what to do if the result is not as expected.

An eight-week preparation route

The schedule below is a teaching illustration, not an official timetable. It should be adapted to the school’s actual topic sequence, weighted assessments, coursework instructions and the 2027 SEC paper that the student will take. The point is to protect steady practice while stopping exam preparation from becoming a mountain of disconnected tasks.

Week 1

Collect the syllabus, identify the actual code and software; inspect one piece of independent work and one corrected assignment. Finish with a short record of one thing the student can now do alone, one error that still occurs and one small change to try next. A record with specific evidence is more useful than a colour-coded timetable without learning checks.

Use work from the correct G-level as the reference, not a generic challenge list. The child should be able to close the notes and reproduce the key step on a fresh example. If accuracy collapses as soon as the surface details change, revisit the underlying concept before increasing speed or volume. A sustainable plan also leaves room for other SEC subjects and ordinary rest.

Week 2

Check computing fundamentals and the ability to explain input, processing, output and data storage using everyday devices. Finish with a short record of one thing the student can now do alone, one error that still occurs and one small change to try next. A record with specific evidence is more useful than a colour-coded timetable without learning checks.

Use work from the correct G-level as the reference, not a generic challenge list. The child should be able to close the notes and reproduce the key step on a fresh example. If accuracy collapses as soon as the surface details change, revisit the underlying concept before increasing speed or volume. A sustainable plan also leaves room for other SEC subjects and ordinary rest.

Week 3

Work through safe networking and cybersecurity case studies; explain each recommendation rather than memorising a list. Finish with a short record of one thing the student can now do alone, one error that still occurs and one small change to try next. A record with specific evidence is more useful than a colour-coded timetable without learning checks.

Use work from the correct G-level as the reference, not a generic challenge list. The child should be able to close the notes and reproduce the key step on a fresh example. If accuracy collapses as soon as the surface details change, revisit the underlying concept before increasing speed or volume. A sustainable plan also leaves room for other SEC subjects and ordinary rest.

Week 4

Practise a spreadsheet task with an intentionally wrong value and check the output against a manual reasonableness test. Finish with a short record of one thing the student can now do alone, one error that still occurs and one small change to try next. A record with specific evidence is more useful than a colour-coded timetable without learning checks.

Use work from the correct G-level as the reference, not a generic challenge list. The child should be able to close the notes and reproduce the key step on a fresh example. If accuracy collapses as soon as the surface details change, revisit the underlying concept before increasing speed or volume. A sustainable plan also leaves room for other SEC subjects and ordinary rest.

Week 5

Use the correct syllabus software for a practical task; verify filenames, result format and submission workflow. Finish with a short record of one thing the student can now do alone, one error that still occurs and one small change to try next. A record with specific evidence is more useful than a colour-coded timetable without learning checks.

Use work from the correct G-level as the reference, not a generic challenge list. The child should be able to close the notes and reproduce the key step on a fresh example. If accuracy collapses as soon as the surface details change, revisit the underlying concept before increasing speed or volume. A sustainable plan also leaves room for other SEC subjects and ordinary rest.

Week 6

Build a simple programming task and test ordinary, boundary and invalid inputs before declaring it finished. Finish with a short record of one thing the student can now do alone, one error that still occurs and one small change to try next. A record with specific evidence is more useful than a colour-coded timetable without learning checks.

Use work from the correct G-level as the reference, not a generic challenge list. The child should be able to close the notes and reproduce the key step on a fresh example. If accuracy collapses as soon as the surface details change, revisit the underlying concept before increasing speed or volume. A sustainable plan also leaves room for other SEC subjects and ordinary rest.

Week 7

Practise the paper format that carries the student’s greatest current weakness, not just the one they enjoy most. Finish with a short record of one thing the student can now do alone, one error that still occurs and one small change to try next. A record with specific evidence is more useful than a colour-coded timetable without learning checks.

Use work from the correct G-level as the reference, not a generic challenge list. The child should be able to close the notes and reproduce the key step on a fresh example. If accuracy collapses as soon as the surface details change, revisit the underlying concept before increasing speed or volume. A sustainable plan also leaves room for other SEC subjects and ordinary rest.

Week 8

Run a timed, syllabus-aligned review with new input data, correct the errors and keep a short reusable checklist. Finish with a short record of one thing the student can now do alone, one error that still occurs and one small change to try next. A record with specific evidence is more useful than a colour-coded timetable without learning checks.

Use work from the correct G-level as the reference, not a generic challenge list. The child should be able to close the notes and reproduce the key step on a fresh example. If accuracy collapses as soon as the surface details change, revisit the underlying concept before increasing speed or volume. A sustainable plan also leaves room for other SEC subjects and ordinary rest.

Questions parents and students ask

1. Are G1, G2 and G3 Computing merely different grades?

No. Each level has a distinct syllabus code, topic structure and examination requirements. Marks describe performance within a subject level; they do not turn one subject specification into another.

For a real decision, compare that answer with the learner’s most recent work and the official Computing specification. If uncertainty remains about offering, levels or assessment arrangements, the school is the appropriate source of confirmation. The aim is a decision made with evidence rather than a guess based on a title.

2. Is G1 Computing just typing?

No. It covers computer fundamentals, networking, responsible use, document processing, spreadsheets, media software and programming. Accuracy, problem solving and computational thinking matter throughout.

For a real decision, compare that answer with the learner’s most recent work and the official Computing specification. If uncertainty remains about offering, levels or assessment arrangements, the school is the appropriate source of confirmation. The aim is a decision made with evidence rather than a guess based on a title.

3. Is G2 Computing a miniature G3 Python course?

No. K237 has its own application, media, spreadsheet and block-programming orientation. Moving towards G3 requires checking and learning content that is not simply a harder version of the G2 practical tasks.

For a real decision, compare that answer with the learner’s most recent work and the official Computing specification. If uncertainty remains about offering, levels or assessment arrangements, the school is the appropriate source of confirmation. The aim is a decision made with evidence rather than a guess based on a title.

4. Does G3 Computing require Python?

Yes. The 2027 K349 specification names Python and JupyterLab among the Paper 2 laboratory tools. Check the current document and the school’s software arrangements.

For a real decision, compare that answer with the learner’s most recent work and the official Computing specification. If uncertainty remains about offering, levels or assessment arrangements, the school is the appropriate source of confirmation. The aim is a decision made with evidence rather than a guess based on a title.

5. Do all three practical papers have the same weighting?

No. G1 practical Paper 2 is 60%, G2 practical Paper 2 is 50%, and G3 practical Paper 2 is 40%. These are shares of the overall examination, not instructions to study only those topics.

For a real decision, compare that answer with the learner’s most recent work and the official Computing specification. If uncertainty remains about offering, levels or assessment arrangements, the school is the appropriate source of confirmation. The aim is a decision made with evidence rather than a guess based on a title.

6. Can a student revise using an older Computing paper?

Sometimes for shared skills, but never assume it is a full mock of the 2027 SEC syllabus. Use the exact subject code, current scheme and official specimen materials for examination rehearsal.

For a real decision, compare that answer with the learner’s most recent work and the official Computing specification. If uncertainty remains about offering, levels or assessment arrangements, the school is the appropriate source of confirmation. The aim is a decision made with evidence rather than a guess based on a title.

7. Does a correct computer output prove full understanding?

Not by itself. Ask whether the student can predict results, explain the method, test failure cases and reproduce the solution when the initial inputs are changed.

For a real decision, compare that answer with the learner’s most recent work and the official Computing specification. If uncertainty remains about offering, levels or assessment arrangements, the school is the appropriate source of confirmation. The aim is a decision made with evidence rather than a guess based on a title.

8. How can parents help without knowing how to code?

Ask the student to name the input, desired output, first test, observed result and next decision. These questions support reasoning without writing the child’s project or pretending to be the examiner.

For a real decision, compare that answer with the learner’s most recent work and the official Computing specification. If uncertainty remains about offering, levels or assessment arrangements, the school is the appropriate source of confirmation. The aim is a decision made with evidence rather than a guess based on a title.

9. Should we move immediately to G3 after one high mark?

Not solely on that evidence. Review teacher feedback, independent performance, programming readiness, software differences and overall subject load with the school.

For a real decision, compare that answer with the learner’s most recent work and the official Computing specification. If uncertainty remains about offering, levels or assessment arrangements, the school is the appropriate source of confirmation. The aim is a decision made with evidence rather than a guess based on a title.

10. Is cybersecurity knowledge enough without practical ability?

Both matter, but each syllabus weighs and examines practical and conceptual work differently. Build safe habits alongside hands-on tasks that the registered specification actually requires.

For a real decision, compare that answer with the learner’s most recent work and the official Computing specification. If uncertainty remains about offering, levels or assessment arrangements, the school is the appropriate source of confirmation. The aim is a decision made with evidence rather than a guess based on a title.

11. Is Computing the same as learning to use AI tools?

No. AI tools may be encountered in everyday life, but the assessment centres on the topics and outcomes of the registered Computing syllabus. Do not replace them with generic AI prompting.

For a real decision, compare that answer with the learner’s most recent work and the official Computing specification. If uncertainty remains about offering, levels or assessment arrangements, the school is the appropriate source of confirmation. The aim is a decision made with evidence rather than a guess based on a title.

12. Where is the official syllabus?

Use the relevant 2027 school-candidate G1, G2 or G3 directory on SEAB and open K127, K237 or K349. Check future cohorts separately.

For a real decision, compare that answer with the learner’s most recent work and the official Computing specification. If uncertainty remains about offering, levels or assessment arrangements, the school is the appropriate source of confirmation. The aim is a decision made with evidence rather than a guess based on a title.

13. Can schools offer different Computing subjects?

Schools determine which available subjects they offer, consistent with MOE rules. A national syllabus listing does not guarantee that every school runs every subject at every level.

For a real decision, compare that answer with the learner’s most recent work and the official Computing specification. If uncertainty remains about offering, levels or assessment arrangements, the school is the appropriate source of confirmation. The aim is a decision made with evidence rather than a guess based on a title.

14. What is the most efficient first revision step?

Find the correct syllabus code, compare it against two pieces of the child’s unaided work and choose one repeatable weakness to repair. Repeat the task after a short delay to check retention.

For a real decision, compare that answer with the learner’s most recent work and the official Computing specification. If uncertainty remains about offering, levels or assessment arrangements, the school is the appropriate source of confirmation. The aim is a decision made with evidence rather than a guess based on a title.

Official syllabus documents and related reading

• 2027 SEC G1 school-candidate subject directory
• 2027 SEC G2 school-candidate subject directory
• 2027 SEC G3 school-candidate subject directory
• eduKateSG 2027 all-subject directory
• Why One Successful Run Does Not Prove a Program Works

This guide compares 2027 school-candidate syllabuses only. A subject’s appearance in a national directory does not mean every school offers it. MOE’s Full Subject-Based Banding allows a mix of subjects at different levels; Posting Groups are not the same thing as the level registered for each examination subject. Requirements and assessment arrangements must be checked for the pupil’s school and cohort.

Last reviewed: 16 October 2026 · Purpose: independent parent-and-student explanation of the official syllabus differences. Course descriptions and weightings are sourced to the MOE/SEAB specifications; the scenarios and recommended study approaches are eduKateSG educational illustrations.