eduKate Singapore

VIEW THIS AS

Auto mode follows the Route Engine until you choose a viewpoint.

YOU ARE HERE

ROUTE CHECK

CONNECTED TO

WHAT NEXT

Use the canonical route for this room, or HELP if you are unsure.

Next useful routeCanonical hubHELP

Learn Everything about G3 Computing in the MOE and SEAB Syllabuses | Understand SEC G3 Computing

—

by

eduKate Asia
University learner using a computational notebook with visibly out-of-order cell counters and restart and run icons

Did you know 2027 SEC G3 Computing K349 combines a two-hour written exam and a two-and-a-half-hour computer-lab exam? Learning to code is important, but the syllabus also tests how computers represent data, how networks work, how spreadsheets process information and how technology affects people.

Paper 1 is worth 60% of the qualification and covers five modules. Paper 2 is worth 40% and requires spreadsheet and Python/JupyterLab solutions saved for marking. This guide explains both papers, the most common technical mistakes and how parents can help a learner become a more independent problem-solver.

Details come from SEAB’s official 2027 G3 Computing syllabus. All sample data and programming scenarios are illustrative educational exercises, not official SEC examination questions. Always confirm the candidate’s subject code and lab arrangements through the school or SEAB.

Find the answer you need

Did you know SEC G3 Computing is code K349?

The 2027 Singapore–Cambridge Secondary Education Certificate uses K349 for G3 Computing, succeeding the former 7155 O-Level reference. Computing combines principles of digital systems, problem-solving, algorithms, spreadsheets, networking and responsible technology use. It is not simply learning to type rapidly or use office applications.

Students need conceptual understanding and the ability to make a program or spreadsheet behave correctly under unfamiliar conditions. An impressive-looking screen is not evidence that an algorithm is accurate. Ask the learner to explain the intended input, output, conditions and tests before assuming a solution is ready.

Do you know the two exam papers have a 60/40 split?

Paper 1 is a compulsory two-hour written examination worth 80 marks and 60% of the subject grade. Paper 2 is a compulsory two-hour-30-minute lab-based examination worth 70 marks and 40%. Paper 1 covers all five syllabus modules; Paper 2 focuses on programming and spreadsheets.

The higher weighting of the written paper does not make practical programming optional. A learner who remembers definitions but cannot trace a loop, correct a broken formula or save the required files may struggle. Conversely, a skilled coder still needs to explain computer systems, network concepts and computing’s impacts in writing.

Did you know Paper 1 is not purely multiple choice?

The written paper may include single- and multiple-answer MCQs, short answers, matching tasks, cloze passages and structured questions. It tests knowledge, application and design reasoning across the entire course. Relevant formulae are provided.

A candidate should practise recognising question instructions. One multiple-answer item may require selecting several correct statements rather than one letter, while a structured answer needs an explicit chain of reasoning. Memorising an old question bank’s answer pattern does not guarantee readiness for the 2027 paper.

Do you know what the lab paper actually requires?

Paper 2 includes one question on spreadsheets and four or five on programming. Candidates work at a computer with spreadsheet software, Python and JupyterLab, and submit the required digital files for marking. A Python quick reference guide is provided during the examination.

The 150 minutes includes time needed to save work. A student can write correct code and still lose assessment opportunities if the required output is saved under the wrong name, in the wrong location or not at all. Rehearse the official school lab workflow rather than assume a personal laptop’s setup will be available.

Did you know the five modules reach beyond coding?

The syllabus modules are Computing Fundamentals, Algorithms and Programming, Spreadsheets, Networking, and Impact of Computing. The written examination assesses all five. The lab examination specifically tests the second and third modules.

A balanced revision week therefore includes data representation and logic; program design and debugging; spreadsheet formulae; networks and security; and social, ethical or legal implications. Doing only Python puzzles might improve fluency but leave important examination content untouched.

Do you know why binary numbers matter to everyday computing?

Computers represent information through binary states. The syllabus includes conversions among binary, decimal and hexadecimal, fixed-width two’s-complement representation of integers, and text encoding. These are not arbitrary mathematics exercises; they explain how values can be stored and interpreted in finite hardware.

For a simple example, binary 1011 equals decimal 11 because it represents 8 + 0 + 2 + 1. The same visible bit pattern may mean something else when a different representation scheme is specified. Students should identify the word length and encoding assumptions before converting.

Did you know kilobytes and kibibytes are not always the same?

Computing uses decimal prefixes such as kilo and mega alongside binary prefixes such as kibi and mebi. One kilobyte is normally 1,000 bytes in the decimal convention; one kibibyte is 1,024 bytes. Confusing them affects capacity calculations and comparisons.

A pupil should show the conversion factor chosen and why it fits the question. The official K349 syllabus explicitly includes units from bits and bytes through petabytes and pebibytes. A memorised “multiply by 1024 every time” rule is therefore unreliable.

Do you know what a CPU, memory and storage each do?

The processor performs computation and control; main memory holds information for active processing; secondary storage retains data for longer periods. These different roles explain why adding storage capacity does not automatically produce the same benefit as increasing memory or improving processing speed.

A school computer may become slow for several reasons. A student should identify which operation is constrained rather than attribute every delay to an “old CPU”. The syllabus also includes data and address buses, input-output interfaces and comparison of magnetic, optical and solid-state storage media.

Did you know an interface and a storage medium answer different questions?

USB, HDMI and PCI Express relate to connecting devices or components and transferring signals or data. Magnetic, optical and solid-state describe broad storage technologies. Calling a USB connector an SSD or a display cable a memory device confuses their functions.

Students should compare typical purpose, throughput, portability, capacity, durability and price where relevant. Do not claim a single technology is best in every setting. The correct choice depends on the application and the criteria supplied in the question.

Do you know what a logic gate’s truth table really means?

K349 covers AND, OR, NOT, NAND, NOR and XOR gates, together with simple multi-gate circuits and Boolean expressions with up to three inputs. A truth table records the output for each permitted combination of input values under the specified logical rules.

For example, XOR is true for two inputs when exactly one is true. AND is true only when both are true. This differs from a casual assumption that “OR always means only one”. A student should construct the table carefully and use it to check an equivalent circuit, not guess from the gate’s name.

Did you know De Morgan’s laws help simplify logic?

The syllabus expects manipulation of Boolean statements using associative and distributive properties and De Morgan’s theorem. These rules allow equivalent conditions to be expressed in different forms. They are useful when evaluating whether a circuit and an expression actually implement the same requirement.

An everyday interpretation might ask whether a security gate activates when neither sensor detects an authorised condition. The mathematical negation must apply to the whole expression correctly. A pupil should verify a proposed transformation against a truth table before declaring it equivalent.

Do you know why programming begins with a specification?

Before writing Python, a learner should identify inputs, valid input requirements, expected outputs and relevant constraints. The K349 syllabus explicitly includes abstraction: removing unnecessary detail to focus on what must be computed. Beginning with code before understanding the requirement often creates avoidable bugs.

An invented program might calculate the average of test scores while rejecting out-of-range values. The input specification must decide what happens when no valid scores are entered, not simply assume the list is non-empty. These boundary decisions are part of the solution, not late decoration.

Did you know a flowchart is not executable code?

Flowcharts help represent sequence, selection and iteration. They make control flow visible before the student commits to Python syntax. An algorithm can be correct conceptually even when the first implementation contains a syntax mistake; the two should be diagnosed separately.

Ask the student to trace one input through each decision branch. When a flowchart contains a loop, identify which condition eventually stops it. A drawing whose exit condition can never be met may reveal the same underlying logic error as an infinite loop in code.

Do you know why Python variables can change type?

Python assigns values to names, and operations depend on the current types of those values. Input read from the keyboard normally arrives as text, so adding two input strings may concatenate instead of numerically summing them unless appropriate conversion is performed.

An illustrative program intended to add two entered integers should validate the input, convert it and then perform arithmetic. Students must consider invalid input before assuming every keystroke is a valid number. Type reasoning often explains surprising program outputs more clearly than guessing at punctuation.

Did you know list indexing is a frequent source of errors?

Python list indices begin at zero. For a list with five elements, the valid positive indices run from 0 to 4. A loop that tries to access element 5 will fail. A slice, however, uses its own exclusive stop boundary. Confusing these conventions can create off-by-one mistakes.

Practise with a short list and write down which elements are visited by each loop or slice. Then change the list length. If the algorithm only works for the original example, it needs better boundary reasoning rather than another copied solution.

Do you know why a dictionary and a list solve different needs?

A list stores items in an ordered sequence and is often accessed by numerical position. A dictionary associates keys with values and is useful for structured lookups. K349 covers dictionary insertion, querying, lookup and deletion, alongside lists and strings.

A fictional attendance program might use a dictionary keyed by a student identifier rather than depend on remembering which list position belongs to which learner. The choice should be explained in terms of the problem and expected operations. Avoid storing real classmates’ personal details in unnecessary practice files.

Did you know a function helps isolate responsibility?

The syllabus covers user-defined functions, parameters, return values and the distinction between local and global variables. A function is not automatically good simply because a code fragment is wrapped in a def statement; it should perform a clear, useful task.

A program that repeatedly validates input may benefit from a dedicated validation function. A program that relies on hidden changes to global state can be harder to debug. Ask what data enters the function, what comes back, and which side effects, if any, are expected.

Do you know why testing should include unusual inputs?

K349 requires test cases for normal, error and boundary conditions. A program may correctly process ten positive values yet fail on an empty list, the largest allowed value, a negative number or malformed text. These are not obscure distractions; they show whether the solution satisfies its specification.

Create a small testing table with input, expected output, observed output and interpretation. The child should choose test cases purposefully, not merely run the program a few times with convenient examples. An identified failure provides evidence for refinement.

Did you know syntax, runtime and logic errors are distinct?

A syntax error prevents code from being parsed; a runtime error arises during execution; a logic error produces an incorrect result or behaviour even though the program runs. Treating all three as “Python not working” makes debugging slow.

For example, a loop that sums scores but accidentally omits the last item may run successfully and still compute the wrong total. A trace table or inserted print statements can locate where the state diverges from the intended method. Debugging asks what the program actually does before guessing at a fix.

Do you know what a trace table reveals about a loop?

A trace table records variables at successive execution steps. It is useful for examining counters, accumulators, loop conditions and branching. The official syllabus expects manual dry runs, not only watching an interpreter execute code.

With an illustrative list [2, 4, 6], a total accumulator should progress through 0, 2, 6 and 12 when values are added correctly. If the trace shows a skipped value or premature exit, the incorrect control step becomes visible. This is a powerful way to reason about unfamiliar algorithms without a computer.

Did you know some algorithm questions restrict convenient built-in methods?

K349 includes algorithms for finding minima or maxima without calling min or max, calculating sums or averages without sum, searching without index or find, and splitting text without simply using split. These tasks test algorithmic reasoning rather than memorised library names.

A student may write a useful practical program using built-ins in an appropriate context, but an exam question that explicitly requires manual algorithm construction needs the internal process. Pay attention to the question’s constraints. The skill is to create steps that remain correct when the input changes.

Do you know how a spreadsheet reference changes when copied?

A relative reference such as A2 changes with the destination when copied, while an absolute reference such as $A$2 stays fixed. Mixed references lock one coordinate. K349 requires these distinctions because a spreadsheet with a correct formula in one cell can silently give incorrect answers after copying.

An illustrative worksheet multiplies each row’s quantity by a fixed tax rate stored in D1. The row reference may need to vary, but the tax cell should remain anchored. Test the copied formula in a later row, not just the first one.

Did you know spreadsheet formulae must be checked against data types?

A column of numbers stored as text may produce surprising results in calculations or lookups. Similarly, a blank cell can behave differently from zero in counting functions. K349 includes COUNT, COUNTA, COUNTBLANK, COUNTIF and related statistical operations.

Ask the learner which values a formula includes and what it ignores. A spreadsheet that displays a convincing total can still omit incorrectly formatted rows. Reviewing one edge case is often more revealing than admiring a neatly formatted table.

Do you know why IF and lookup formulas are different tools?

An IF expression chooses between outcomes based on a condition. Lookup functions retrieve information from a reference table, using exact or approximate matching as specified. The K349 function list includes VLOOKUP, HLOOKUP, INDEX and MATCH alongside logical AND, OR and NOT.

A program deciding whether a pupil meets a threshold may use an IF formula, while retrieving a category from a table may need a lookup. Confusing exact and approximate matching can silently misclassify records. The student should explain which reference table and matching behaviour are appropriate.

Did you know Goal Seek solves a problem in the opposite direction?

Ordinary formulas calculate outputs from inputs. Goal Seek adjusts a selected input so a formula output reaches a specified target. The syllabus also includes Conditional Formatting, which changes appearance automatically under defined rules. Neither feature is a substitute for understanding the underlying calculation.

In a fictional savings worksheet, Goal Seek may help estimate a required contribution for a stated total under the model. Check that the result meets the constraints and assumptions. Conditional Formatting might flag negative balances, but the colour alone does not explain why a balance is negative.

Do you know why networking is more than a Wi-Fi password?

The Networking module covers wired and wireless transmission, local and wide area networks, client-server and peer-to-peer organisation, and network topologies. These terms describe different aspects of how devices are connected and communicate.

A household Wi-Fi network is an example of a LAN, while the internet interconnects networks worldwide. A student should distinguish device connectivity within a home from communication across the internet. Do not treat having a working browser as proof that every networking concept is understood.

Did you know MAC, IPv4 and IPv6 addresses have different roles?

The syllabus compares MAC addresses and IP addresses by purpose, bit length and typical formats. MAC addresses are associated with network interfaces and local network communication; IP addresses support routing across networks. A router and a switch therefore play different roles in the path data takes.

Teach the learner to trace how a packet reaches a different network and which device processes which address. A visually complicated diagram can often be simplified into local communication, routing and destination. The exact textbook conventions should be checked against the current syllabus.

Do you know why a checksum detects errors but cannot guarantee meaning?

K349 covers data-transmission checks such as parity, checksums, echo checks and automatic repeat requests. These methods can reveal certain transmission errors or trigger resending, but a valid checksum does not establish that a message is truthful, authorised or free from every possible problem.

This distinction is important when interpreting data systems. Integrity of transmission, security of access and accuracy of the underlying content are separate properties. A student should describe which problem a particular mechanism addresses and where its guarantees end.

Did you know cybersecurity and privacy are not identical?

Security protects data and systems against unauthorised access, modification or disruption; privacy concerns appropriate collection, exposure and handling of personal information. K349 includes anti-malware, firewalls, encryption, phishing, spyware, cookies and legal concepts including Singapore’s Personal Data Protection Act.

A password and encryption may reduce some risks, yet neither ensures that a company has collected only data it genuinely needs. Students should connect each protective measure to a plausible threat. Practising safe habits matters; creating phishing demonstrations aimed at real people is inappropriate.

Do you know what the Impact of Computing module asks?

The fifth module connects computers with communication, education, transport, retail, intellectual property, online information and emerging technology. Candidates examine benefits, limits and ethical implications. A digital tool can make a process faster while also introducing unequal access or privacy concerns.

A strong response explains the mechanism and a relevant trade-off instead of asserting that technology is always beneficial or always dangerous. It should distinguish a realistic risk from an unsupported claim about all users. General reasoning about impacts remains examinable in Paper 1 even though Paper 2 is lab-based.

Did you know copyright matters when sharing programs and assets?

Software and digital works are protected or licensed under different arrangements. The syllabus distinguishes proprietary software, freeware, shareware and free/open-source licences, alongside rules against unauthorised use and distribution. “Available free to download” is not equivalent to “public domain”.

Students should credit borrowed code or assets where required and follow applicable licence terms. For school assignments, the important academic issue is also authorship: submitting code generated or copied by someone else as independent work may violate school rules even when the underlying licence permits reuse.

Did you know AI is a topic but not a replacement for program reasoning?

K349 discusses artificial intelligence, machine learning, biased data and a simple nearest-neighbour classification task with two quantitative features. These help learners understand how some systems make decisions based on patterns and how biased input data can produce harmful or misleading results.

The syllabus does not require students to train large language models from scratch. Instead, a useful introductory exercise plots labelled points and identifies which known example is nearest to a new observation under a specified distance rule. The learner should discuss what the small model can and cannot establish.

Did you know that saving files correctly is an examination skill?

Paper 2 requires candidates to submit the specified electronic work, and the allotted time includes saving it. A functioning Python notebook stored in the wrong place or an unsaved spreadsheet may not be available for marking. Technical organisation therefore has direct assessment relevance.

Practise the official school examination environment and directions: which file types are required, what names are specified and how final submission is checked. Avoid assuming internet access, external packages or personal cloud storage are available. The centre provides the authorised computer and software.

Do you know how to repair a Computing grade without assigning endless coding?

A diagnostic review separates fundamental representations, algorithm design, Python syntax, debugging, spreadsheet functions, networking and ethical reasoning. Students may be competent at one while struggling with another. A total test score hides those differences.

Collect one marked written algorithm explanation, one Python task and one spreadsheet exercise. Identify the first incorrect assumption in each, then ask for an independently corrected solution using a fresh example. This is more valuable than copying an entire program that happens to run.

Did you know an eight-week K349 plan can include all five modules?

Weeks 1–2 can repair data representation, Boolean logic and simple algorithms. Weeks 3–4 can strengthen Python lists, functions, testing and debugging. Weeks 5–6 can practise spreadsheets, lookup logic, networks and privacy. Weeks 7–8 can integrate impact-of-computing reasoning with timed written and lab-based tasks.

This is an illustrative skills cycle, not a substitute for the school’s full course. Adjust the order to the learner’s current topics. The goal is to write, test and explain a correct solution under new conditions, not memorise a model answer or one fixed spreadsheet file.

Do you know the essential parent FAQs about K349?

Is Computing only a practical exam? No: written Paper 1 contributes 60%, and lab Paper 2 contributes 40%. Does the lab test Python alone? No, spreadsheets are included. Must students know every modern programming framework? No; follow the K349 scope and its Python quick reference. Is an approved calculator permitted? Yes, in both papers.

K349 appears in SEAB’s published 2027 G3 private-candidate catalogue, but individual eligibility and lab registration arrangements require current official confirmation. Any programs and numerical cases in this article are original teaching illustrations, not released examination questions.

Official K349 syllabus and related educational guides

Parent next action: confirm K349, collect one marked written response, a Python lab task and a spreadsheet workbook. Ask the child to justify the algorithm, explain one edge case and check that the correct files were saved. Repair the weakest step before assigning more exercises.