G3 Computing for the 2027 Singapore–Cambridge Secondary Education Certificate (SEC) has syllabus code K349, formerly O-Level 7155. All candidates sit two compulsory papers. Paper 1 is a two-hour written exam, 80 marks and 60%, assessing Computing Fundamentals, Algorithms and Programming, Spreadsheets, Networking and the Impact of Computing through mixed question types. Paper 2 is a 2-hour-30-minute lab exam, 70 marks and 40%, using spreadsheet software, Python and JupyterLab. It includes one spreadsheet question and four or five programming questions. Candidates must save and submit the required work electronically within the examination time, and a Python quick-reference guide is supplied.

For Punggol families, this is not simply a subject for teenagers who enjoy gaming or have tried a few programming tutorials. Computing requires understanding a problem, expressing it as an algorithm, testing the result, managing data responsibly and explaining trade-offs in systems. A student may write valid Python syntax yet fail because a boundary case is wrong or a spreadsheet formula copied with the wrong cell reference. This guide explains the entire 2027 K349 subject, what happens in the practical, how to study algorithms and spreadsheets from first principles, and what to check before choosing the elective. Students taking SEC in 2028 should follow the official 2028 document when available.

• 2027 SEC G3 Computing code K349; earlier O-Level 7155.
• Paper 1 Written: 2 hours, 80 marks, 60%, all questions compulsory.
• Paper 1 may use single- or multiple-answer MCQ, short answers, matching, cloze and structured questions.
• Paper 1 covers all five modules; relevant formulae are provided.
• Paper 2 Lab-based: 2h30, 70 marks, 40%, all questions compulsory.
• Paper 2 includes one spreadsheet question and four to five Python programming questions.
• Lab software includes spreadsheet software, Python and JupyterLab; a Python quick-reference guide is provided.
• Candidates submit softcopies; time for saving required files is inside the 2h30 duration.
• Five modules: Computing Fundamentals; Algorithms and Programming; Spreadsheets; Networking; Impact of Computing.
• Approved calculators may be used in both papers; the examination centre provides specified computer facilities.
• K349 is different from Electronics K353, Computer Applications at other levels and general ICT/CCA coding.

Quick answers for parents

• K349 2027 is 60% written theory and 40% lab practical.
• Paper 2 includes one spreadsheet task and four or five Python tasks.
• The lab uses approved software and requires timely softcopy submission.
• Paper 1 covers five modules, not only Python.
• Confirm school availability, software practice and exam year before investing in tuition or equipment.

Computing begins with problems, not keyboards

The subject develops a way of defining problems and expressing solutions precisely. Before coding, students identify inputs, required outputs, constraints and examples. A teenager who jumps into typing can spend much of the lesson repairing an incorrect interpretation of the task rather than an incorrect program.

For a fictional school-activity registration tool, ask what information is collected, what makes a registration valid, how duplicates are handled and what the output should report. Then write a small algorithm. This separates the problem definition from the programming syntax and trains judgement that survives changes in software.

G3 Computing and Posting Group 3 are different labels

Under Full Subject-Based Banding, a Posting Group relates to initial admission while a G3 subject has its own syllabus and national examination code. A student may have differing subject levels under school arrangements. K349 Computing does not mean every PG3 pupil can automatically add it to their Secondary 3 timetable.

Check the school’s current upper-secondary options, subject criteria, technical instruction and examination year. Nationally publishing K349 does not guarantee that every school in Punggol offers the class. The child should see an authentic example of problem-solving before choosing the subject solely from a career headline.

The 60–40 weighting requires both concepts and practical work

Paper 1 contributes sixty percent, which makes conceptual understanding essential. Paper 2 contributes forty percent and assesses work produced on a computer. A student who only memorises definitions is unprepared for the lab, while a child who only codes by imitation may struggle with networking, data representation or ethics in the written paper.

A balanced practice plan combines reading a technical explanation, tracing an algorithm and building a small tested solution. Use school-marked evidence to adjust emphasis. A learner who understands concepts but loses files in the lab needs workflow practice, not another pile of theory notes.

Paper 1 may use several answer formats

The two-hour written paper can include single- and multiple-answer selection, short answers, matching, cloze and structured questions. Students need to read instructions carefully: a multiple-answer question cannot always be solved by choosing one attractive option. Diagrams and precise technical terms matter.

A pupil might know the meaning of a binary number yet misinterpret a question asking for a hexadecimal representation, or know the word router without explaining its function. Teach the concept and then practise changed questions. Mixed formats test transfer rather than recall of one textbook sentence.

Paper 2 is a supervised lab examination, not a take-home project

The 2.5-hour Paper 2 is conducted at an examination centre using specified computer facilities. It includes one spreadsheet question and four or five programming questions. Work is saved and submitted electronically within the allotted time; files created after time expires cannot simply be added later as a personal homework submission.

Students must practise the actual school-approved software environment and instructions. A home laptop can support learning, but it may use different paths, versions or settings. Ask the teacher how files should be named and saved in school practice. The goal is confidence with the authorised workflow, not reliance on unapproved internet tools during a test.

Computing Fundamentals explains why computers work

Module 1 includes computer architecture, data representation and logic gates. Students need to understand the roles of processor, memory, storage and interfaces, not merely memorise a list of device names. These concepts explain performance and how data are processed through systems.

A simple distinction: main memory supports active processing, whereas secondary storage preserves files and programs for later use under normal conditions. A device with more storage is not necessarily faster at every task. Ask what performance bottleneck the student is trying to solve before recommending a hardware upgrade.

Bits, bytes and binary must be interpreted precisely

Digital systems represent information with bits. Eight bits make one byte. Binary notation uses powers of two, and the syllabus expects conversions involving binary, denary and hexadecimal, along with distinctions such as kilobyte versus kibibyte. Unit prefixes can change a numerical answer.

Example: the binary pattern 1011 represents 8 + 0 + 2 + 1 = 11 in decimal. The student should be able to explain the place values rather than use a conversion app without understanding. Likewise, a file size expressed in bytes cannot simply be re-labelled as bits without multiplying by eight.

Hexadecimal is a compact representation of binary patterns

Hexadecimal uses sixteen symbols, 0–9 and A–F, and maps conveniently to groups of four bits. It is useful for representing values compactly, but a hex digit is not a different kind of physical storage unit. Students need to move between representations while preserving the underlying value.

An original conversion: binary 1111 corresponds to decimal 15 and hexadecimal F. Binary 00010000 corresponds to decimal 16 and hexadecimal 10. Ask the pupil to show where each place value comes from. A leading zero can matter to a fixed-width representation without changing the ordinary unsigned numerical value.

Negative integers require a defined number format

The 2027 syllabus includes two’s complement for representing positive and negative integers using a fixed number of bits. Students should know that the bit width matters: the same apparent pattern can mean a different value under unsigned versus signed interpretation. A calculator display alone is not enough.

In a simple 8-bit two’s-complement interpretation, 11111111 represents −1, whereas the same bits interpreted as unsigned represent 255. A question must specify the representation. Ask the child to state the convention first, then compute. This prevents a common mistake where the bits are copied accurately but given the wrong meaning.

Logic gates turn conditions into testable outputs

AND, OR, NOT, NAND, NOR and XOR gates are covered. Their meaning can be shown through truth tables, Boolean statements and circuits. Students must explain when the output is true rather than guess from the gate symbol’s visual shape.

Imagine an invented safety indicator that lights only when two independent switches are on. That is an AND relationship in the simplified logic model. If the light activates when exactly one of two signals is present, XOR may be relevant. The real-world physical safety of a device involves much more than a classroom truth table; this is a conceptual model.

De Morgan’s theorem is about equivalent conditions

Boolean transformations can produce different-looking expressions that represent the same logic. De Morgan’s laws connect negated AND and OR expressions. Students should test these rules with truth tables rather than memorise one symbol substitution without understanding.

For two Boolean variables A and B, NOT(A AND B) is equivalent to (NOT A) OR (NOT B). The equivalence can be checked for all four input combinations. This matters when simplifying logic designs and writing conditions in code. Ask the pupil to verify equivalence explicitly rather than assert it because a diagram looks symmetrical.

Algorithmic thinking starts with inputs and outputs

Module 2 examines problem analysis, sequence, selection, iteration and programming. A student who writes code without defining valid inputs may make a program that works for one example and fails for others. Breaking a task into smaller steps makes it easier to reason about correctness.

For a fictional ticket-price calculator, specify the accepted age or category input and what happens if the input is invalid before programming discounts. The algorithm should produce the right amount for boundary values, not only for the sample case given in class. This teaches specification and testing from first principles.

Sequence, selection and iteration are different constructs

Sequence executes steps in order, selection chooses a branch based on a condition, and iteration repeats operations while the relevant condition or sequence holds. The three constructs can be represented in flowcharts and written in Python. A common confusion is treating every repeated check as one simple if statement.

For a fictional quiz, sequence might display an instruction; selection checks whether an answer is correct; iteration asks multiple questions. A clear flowchart shows which path returns to the repeated step. A learner should trace control flow for more than one input to see whether the program can terminate correctly.

Python variables hold values that can change

Variables name data used by a program. They are not an algebraic equation that always stays true; an assignment updates what the name refers to during program execution. Students should practise tracing changes through several lines of Python rather than guessing the final value from the first assignment.

Original example: set total to 0; add 4, then add 7. The final total is 11. If the program accidentally resets total to 0 inside the repeated step, it may produce only the last item. Understanding the state change explains the bug better than editing random characters until the program prints a plausible number.

An original Python loop to practise tracing

Consider a small list of prices. The program below adds each amount and prints a total. It is a learning illustration, not an examination question or a required library. The important idea is that the loop visits each item and updates the accumulator once per iteration.

A pupil should first predict the output without running the program, then test it. Next change the list to an empty list, one value and several values. Does the program still behave as intended? This process builds understanding of initial state, repeated execution and boundary cases.

prices = [4.50, 7.00, 3.50]
total = 0
for price in prices:
    total += price
print(total)  # 15.0

Inputs must be checked before use

Programs that accept data need to know what counts as valid. A string may not contain a number; a user might enter a negative quantity or leave a field blank. A robust solution considers these possibilities when the task calls for them, rather than assuming all users behave like the example input.

For a fictional school-order form, a quantity of −3 should not silently become an order for minus three items. The program can validate the allowed range and give a meaningful message. A K349 question will specify its own expected behaviour, so the pupil must follow that specification rather than add unrequested features.

Lists and iteration support repeatable solutions

Lists let programs store ordered collections and process data systematically. Index positions and loop boundaries can cause off-by-one errors. A solution that works for a list of four values may fail when the list is empty or contains a different number of items.

Ask the student to trace the list and count the iterations. If a program skips the first item, the reason may be its starting index. If it tries to access a position beyond the list, the boundary condition is wrong. Explain the specific cause rather than describing every such bug as carelessness.

Testing is not the same as checking one happy path

Testing compares actual behaviour with expected behaviour under specified inputs. A single example that returns the correct answer does not prove the program works for all valid and invalid cases. The syllabus emphasises development, testing and refinement, so deliberate test design matters.

Use a simple table with normal input, minimum valid input, maximum valid input and invalid cases where relevant. Record expected output before running the code. If a program fails, name the reason and change the logic. The error is useful evidence about the model, not merely a red message to be dismissed.

Debugging should be systematic

A programming error may be syntax, runtime or logical, and each requires a different investigation. An error message can identify where execution stopped, but a wrong final result may require tracing variables and conditions. Random edits often create additional problems.

Teach the pupil to reproduce the failure, reduce it to a small case, inspect the relevant state, make one change and retest both the original and a changed example. This method is more transferable than memorising a correction for one textbook question. It also makes learning calmer under timed lab conditions.

Functions and decomposition make code easier to test

A large program can often be broken into smaller functions that take defined inputs and produce predictable results. Decomposition helps the pupil reason about one step at a time and reuse tested logic. It does not mean splitting every line into a separate function.

For an invented booking tool, one function might validate a booking date while another calculates a total. Each can be checked with its own examples. A well-defined interface between them makes errors easier to isolate. The pupil should follow the school’s syllabus constraints and coding style during assessment preparation.

File input and output introduce another failure point

The Python syllabus includes file operations such as opening, reading and writing files. Programs that handle files must use the correct path, mode and format and should not accidentally overwrite important information. Lab conditions may specify exact source and output filenames.

An original practice program can read a teacher-provided list of simple fictional items and write a processed result to another file, then verify the contents. Keep personal student records or identifiable data out of casual practice examples. When practising for Paper 2, use the school’s official submission rules.

Spreadsheets are programs expressed through cells

Module 3 covers spreadsheet concepts and formulas, including relative, absolute and mixed references, functions, charting and data interpretation. A spreadsheet is not merely a digital notebook. Its cell formulas are repeatable rules; a reference error can spread across hundreds of rows while the table still looks tidy.

For a fictional budget, cell B2 could hold unit price and C2 quantity, with D2 calculated using =B2*C2. Copying the formula down should reference the new row when relative references are appropriate. The student should explain why the formula moves rather than memorise what cell was used in one sample file.

Absolute and relative references solve different problems

A relative reference changes when a formula is copied, while an absolute reference stays fixed. Mixed references fix either the row or the column. Choosing the wrong type can generate plausible-looking but incorrect totals. The $ symbol in a spreadsheet reference serves a special addressing purpose.

Imagine F1 contains an invented adjustment factor of 1.05. A formula =B2*$F$1 can multiply each row’s price by the same factor when copied downward. Without the absolute reference, the second formula may read F2 instead of F1. This is a conceptual reason for the syntax, not a requirement to memorise one sample column.

Functions are tools with conditions

The 2027 spreadsheet syllabus includes functions such as IF, AND, OR, COUNTIF, SUMIF, VLOOKUP, INDEX and MATCH, as well as statistical and text functions. A student should know the inputs and what the result means. Copying a function from an online answer without understanding the range or match behaviour can create errors.

For a fictional event attendance sheet, COUNTIF can count rows meeting a defined condition, but the result is only meaningful when data are correctly entered and the selected range includes the intended records. A blank row or inconsistent text label may alter the count. Data preparation matters as much as formula syntax.

Charts cannot rescue incorrect data

Spreadsheets can generate charts quickly, but the axes, categories and underlying data still need checking. A misleading scale or incomplete range can produce an attractive visual that supports the wrong inference. The syllabus develops both technical creation and critical interpretation.

Suppose an invented club survey has twenty responses but the chart source range includes only fifteen. The resulting percentage may be calculated perfectly for the wrong sample. Teach the pupil to inspect row counts, headers, source range and graph legend. Visual polish is a final step, not proof of correctness.

Paper 2 spreadsheet work must be saved properly

A pupil may complete the correct formulas but save the wrong workbook or fail to place the file where required. SEAB’s allotted lab time includes saving required softcopies, and the centre’s instructions govern naming and submission. Accurate file handling is part of practical readiness.

Practise a small workflow at school: open the supplied file, complete a change, check it, save in the specified format and location, reopen if instructed and verify the result. Do not make assumptions about which external storage or internet service is available during the real assessment.

Networking explains how devices communicate

Module 4 includes network types, physical and wireless media, client-server and peer-to-peer arrangements, protocols and security concepts. A pupil should distinguish the network’s scope and purpose rather than simply name Wi-Fi or the internet in every answer.

A local area network can connect devices within a limited environment, while wide area networks connect over much greater distances. A wired link may offer different reliability and installation considerations from wireless access. The correct choice depends on the case, not the belief that one medium is always superior.

Data privacy and cybersecurity are computing responsibilities

The Impact of Computing module addresses how digital technology influences people and organisations, including ethical, legal and economic questions. Students should understand that personal data has consequences beyond a database field and that technical feasibility does not alone justify collecting or sharing information.

For a fictional club registration, ask whether every requested personal detail is actually necessary, how access should be limited and how long data should be retained under the relevant rules. A student who proposes storing passwords in a public spreadsheet has not solved the problem responsibly even if the spreadsheet formulas are correct.

Computing is not the same as Electronics

Computing K349 centres on algorithms, software, data systems and networks. Electronics K353 examines electronic components and circuits under its own syllabus. The two can share logical reasoning, but the assessment tasks and practical skills differ. Neither can be chosen solely by which has a more modern-sounding title.

Ask the child whether they enjoy debugging instructions and data or physical circuit design and measurement, then check the actual school’s electives. A future computer-science or engineering pathway has its own admission requirements, so neither subject is a guaranteed ticket to one course.

A four-week Computing repair cycle

Week 1: gather marked written work and a practical exercise; classify errors in data representation, logic, Python, spreadsheet formulas, networking and file workflow. Week 2: repair the most repeated concept using small changed examples. Week 3: complete a Python program with deliberate test cases plus one spreadsheet task. Week 4: attempt timed selections from both papers, including file saving and checking.

This is a feedback cycle rather than a promise to learn the whole Computing syllabus in one month. A pupil in Secondary 3 may not yet have completed the programming curriculum. Follow the school software setup and teacher guidance, protect sleep and other subjects, and choose the next repair from evidence.

When extra tuition helps and when it is redundant

Good computing support should diagnose whether the pupil misunderstood the algorithm, coded the wrong branch, failed to trace data, used the wrong spreadsheet reference or mishandled files. These are different errors. A generic ‘learn Python quickly’ worksheet may not repair the specific examination difficulty.

eduKate provides small-group tuition in Punggol, but an extra class is not automatically necessary when the school and independent practice are working. Ask for a baseline problem and a new solution built independently after teaching. The child’s starting position should determine the support.

School availability and post-secondary choices need verification

K349 is a recognised SEC subject in the 2027 school-candidate list, but the school decides the combination it offers and its placement rules. A pupil who enjoys coding may still need to consider Mathematics, Sciences, English and the overall workload. Interest in software can also be developed through safe projects outside the examination.

Later computing diplomas, university courses and careers have separate admission criteria and practical expectations. The K349 result alone does not guarantee a place. Confirm actual requirements for the relevant year, then focus on clear problem analysis, reliable code and responsible handling of information now.

Home and school Computing practice checklist

• Define a problem’s inputs, valid range and expected outputs before coding.
• Trace an algorithm by hand on ordinary and boundary inputs.
• Test one new Python program rather than copying a finished solution.
• Use relative and absolute spreadsheet cell references correctly.
• Practise saving and reopening authorised lab files.
• Review one question each from networking, data representation and ethics.
• Keep personal information out of fictional practice datasets.
• Check the actual SEC year and school software instructions.

Frequently asked questions

What is the SEC G3 Computing code for 2027?

K349, previously O-Level 7155.

How many computing papers are there?

Two compulsory papers: written theory and lab-based assessment.

How long is Paper 1?

Two hours, 80 marks and 60%.

How long is Paper 2?

Two hours thirty minutes, 70 marks and 40%.

What software is used in the lab?

Spreadsheet software, Python and JupyterLab, following the centre’s authorised setup.

How many lab questions are on spreadsheets?

One spreadsheet question.

How many programming questions are in Paper 2?

Four or five.

Is a Python reference provided?

Yes, a quick reference guide is supplied for Paper 2.

Must files be saved before time ends?

Yes. The allotted lab time includes saving and submitting the required softcopies.

Does Paper 1 test only Python?

No. It covers all five Computing modules.

What are the five modules?

Fundamentals, Algorithms and Programming, Spreadsheets, Networking and Impact of Computing.

Can calculators be used?

Approved calculators are permitted for both papers under official rules.

Is K349 the same as K353 Electronics?

No. Electronics follows a different circuit-focused syllabus.

Does a child need their own expensive laptop?

The official lab centre must provide required facilities; ask the school what equipment is useful for ordinary learning.

Do students need to know how to debug?

Yes. The syllabus assesses design, testing and refinement of computing solutions.

Do I need tuition if my child can already code?

Not necessarily. Check written concepts, spreadsheet ability and lab workflow first.

Which syllabus applies for SEC 2028?

The official 2028 SEC Computing syllabus when published.

Official references and connected Punggol guides

SEAB official G3 Computing K349 2027 syllabus PDF · 2027 SEC G3 school-candidate subjects

What is G3 Computing in Punggol? · Computing versus Electronics · Computing versus Additional Mathematics · Preparing for Python and spreadsheet practical

Exam structure verified against SEAB’s 2027 K349 Computing syllabus on 14 November 2026. The code and spreadsheet examples here are original instructional exercises, not actual SEC tasks. The specific lab software setup and file instructions should be confirmed by the school or examination centre for the pupil’s cohort.