Small Group Tutorials

Here to help students catch up, keep up, and move ahead. Book a consultation here.

Top 100 Secondary 4 G3 Vocabulary List for Punggol | Computing: Algorithms, Programming and Testing

Three learners review open books together at a classroom table, with stacks of textbooks, stationery and a whiteboard in the bright room.

Secondary 4 G3 Computing vocabulary helps when a student can explain a term, predict a result and test whether a solution works. This Punggol revision guide gives 100 terms with meanings and application checks, followed by worked examples and an independent checkpoint. Start with the words behind a current mistake: an incorrect condition, a loop that misses a value, a copied spreadsheet formula or an unexplained test result.

This is an original selection for revision towards the 2027 SEC G3 Computing examination, syllabus K349. It is not an official, ranked or compulsory SEAB hundred-word list. The revision includes upper-secondary foundations, including work first encountered before Secondary 4. Use your school’s current coverage and requirements to choose the next practice task.

  • Learn the meaning: explain the term without copying the table.
  • Predict the behaviour: decide what a short program or formula should do before running it.
  • Check the boundary: test a limit, an unsuitable input or another case that could reveal a hidden mistake.
  • Repair the cause: identify the instruction or assumption that produced the wrong result.
  • Retest independently: change the values or context and solve again without the worked answer.

Choose your next learning step

Use the route that answers your question today, then continue through the full guide.

Full chapter index · Independent practice · 200-term Computing reference

Full contents

Understand and learn
  1. Set the course boundary and choose a useful checkpoint
  2. The hundred terms: meanings and application checks
  3. Distinctions that prevent several mistakes at once
Apply and revise
  1. Five worked examples: predict, explain, repair and retest
  2. Eight independent questions and a diagnostic next step
  3. A manageable revision cycle for vocabulary and practical work
Plan support and check sources
  1. How parents and a small group can support the next decision
  2. Common questions about the hundred-term checkpoint
  3. Sources and further checking

CHAPTER 1 OF 9

Set the course boundary and choose a useful checkpoint

Back to contents

The official 2027 G3 directory identifies Computing as K349. Its five modules cover Computing Fundamentals, Algorithms and Programming, Spreadsheets, Networking, and Impact of Computing. The written paper assesses all five; the laboratory paper assesses programming and spreadsheets. These distinctions explain why coding vocabulary needs to sit alongside language about representation, formulas, networks and responsible decisions. The MOE G3 Computing teaching syllabus provides the broader teaching framework.

This page concentrates most of its entries on designing, understanding and testing solutions. It includes supporting terms from the other modules, but it does not claim to cover every examinable concept. Logic gates, further networking detail, software licences and other topics still need the teaching and practice specified for the course. A list of words is a starting point for finding gaps, not a replacement for the full programme.

Choose a small task before revising. For example, ask a learner to explain why a condition accepts an impossible value. If the student can define selection but cannot connect the condition to the branch taken, practising more definitions alone will not resolve the gap. The next task should involve predicting a branch, tracing the instruction and checking an alternative input.

Keep three kinds of evidence separate: an explanation given without help, a correction completed with support, and a fresh response completed independently. All three are useful, but they answer different questions. A corrected exercise shows that the explanation was useful; a later changed exercise shows whether the learner can now use the idea.

Contents · Next chapter

CHAPTER 2 OF 9

The hundred terms: meanings and application checks

Back to contents

Read one group at a time. Cover the meaning column, give an explanation, then attempt the final column. A complete response should connect the word to an operation, decision or observable result. If you can recognise a term but cannot use it, mark that row for a short practical task rather than copying it repeatedly.

Planning a solution

No.TermMeaningUse, check or example
1AlgorithmAn ordered set of unambiguous steps for completing a task or solving a problem.Explain the steps for finding a largest value before deciding which Python statements will implement them.
2Computational thinkingA way of analysing problems and developing solutions that can be represented as precise, executable steps.Identify the necessary information, divide the work and explain how the proposed solution will be checked.
3AbstractionKeeping the details relevant to a problem while leaving out details that do not affect its solution.A borrowing counter needs the number of books; the colour of each cover may be irrelevant.
4DecompositionBreaking a problem into smaller parts that are easier to understand, develop and test.Separate reading records, validating their contents, calculating results and displaying the answer instead of mixing every task.
5RequirementSomething a solution must do or a condition that it must satisfy.State whether the program must find the first matching item or every matching item; these require different behaviour.
6InputData supplied to a program or system for processing.A keyboard entry and a line read from a file are different input routes; both need suitable handling.
7OutputInformation or another result produced by a program or system.Specify whether the answer should be a count, a total, a position or a message before writing code.
8SequenceThe order in which instructions are carried out when control follows one step after another.Initialise the total before adding values to it; moving that step inside the loop can erase earlier work.
9SelectionChoosing which instructions to execute according to whether a condition is satisfied.An if statement can accept a permitted value or display an explanation when the value is outside the range.
10IterationRepeating a set of instructions, often while a condition holds or for items in a collection.Explain what changes after each repetition and how repetition ends; repeating code without progress can fail to terminate.
Terms 1–10: Planning a solution

Names, values and data types

No.TermMeaningUse, check or example
11VariableA name used to refer to a value that a program can use and, where appropriate, replace.After total = total + 4, the name total refers to the updated result rather than its previous value.
12IdentifierA name used to identify a program element, such as a variable or function.Use a meaningful valid name such as book_count; distinguish it from a quoted piece of text.
13LiteralA value written directly in source code rather than obtained by looking up a variable.In limit = 30, the number 30 is a literal and limit is an identifier.
14Data typeA category of values that determines which operations make sense and how values behave.Check whether a value is a number or a string before assuming that the addition symbol performs arithmetic.
15IntegerA whole-number value with no fractional part, including zero and negative whole numbers.A count of borrowed books normally uses integers; a negative count would need rejection under the stated requirements.
16Floating-point valueA numeric value represented using a finite-precision format that can approximate fractional quantities.A displayed decimal is not always stored exactly; distinguish numerical representation from the number of displayed decimal places.
17BooleanA truth value, represented in Python by True or False.The comparison score >= 50 produces a Boolean result; the result can control a branch.
18StringAn ordered sequence of characters used to represent text.The string “12” contains two characters; convert suitable text before using it as the integer twelve.
19ListAn ordered, changeable collection of values in Python, accessed using positions called indices.A list can contain repeated values; do not assume that finding one occurrence proves there are no others.
20DictionaryA Python mapping that associates unique keys with values.Use a member identifier as a key to retrieve a value; dictionary lookup is not the same as list indexing.
Terms 11–20: Names, values and data types

Expressions and operations

No.TermMeaningUse, check or example
21AssignmentAn operation that binds a value to a name or stores it in a selected location.In Python, count = 3 assigns a value; it does not ask whether count already equals three.
22ComparisonAn operation that checks a relationship between values and produces a truth value.Use == to test equality in Python, and inspect both operands when an unexpected branch is taken.
23ExpressionA combination of values, names and operations that can be evaluated to produce a value.The expression quantity * price produces a result; a complete assignment can store that result for later use.
24OperatorA symbol or keyword that specifies an operation on one or more values.Check the language: exponentiation uses ** in Python, while an ordinary spreadsheet formula uses ^.
25ConcatenationJoining sequences, such as strings or lists, end to end.“12” + “3” gives “123”; joining text is different from adding the numbers twelve and three.
26IndexingAccessing an individual item in a sequence by its position.Python starts ordinary forward indexing at zero, so the first character of “code” is at index 0.
27SlicingSelecting a portion of a sequence using start, stop and optionally step information.In Python, text[1:4] includes indices 1, 2 and 3; the stop position itself is excluded.
28ModuloAn operation that gives the remainder associated with division.For positive integers, 17 % 5 gives 2; use a remainder to test divisibility or repeated groupings.
29Floor divisionDivision that rounds the quotient down to the greatest integer not exceeding it.In Python, -7 // 2 gives -4; flooring is not the same as simply removing the decimal part.
30Membership testA check of whether a value occurs within a collection or sequence.The expression “a” in “gate” is true; it does not tell you the position or number of occurrences.
Terms 21–30: Expressions and operations

Control flow and functions

No.TermMeaningUse, check or example
31ConditionAn expression interpreted as true or false to control a decision or repetition.For an inclusive range, both the lower-limit test and the upper-limit test must be satisfied.
32BranchA possible route through instructions selected by a decision.Trace the chosen route for a boundary value; do not assume every branch executes during one run.
33For loopA repetition structure that visits successive items supplied by an iterable, such as a list or range.A loop over range(4) visits 0, 1, 2 and 3; it does not visit 4.
34While loopA repetition structure that continues while its condition is true.Check that the loop body can change the condition when necessary, so an invalid entry does not create endless repetition.
35Loop bodyThe block of instructions executed during each repetition of a loop.In Python, indentation identifies the block; moving an update outside it can change the entire result.
36AccumulatorA variable that combines successive values into a running result, commonly a total.Initialise total once, then add each qualifying value; explain why resetting it repeatedly loses the earlier contributions.
37CounterA variable that records how many times an event or condition occurs.Add one for each qualifying record; adding the record value instead would calculate a different quantity.
38FunctionA named unit of reusable code that performs a task and may return a result.Write a function with a clear purpose, then test calls with different inputs rather than one familiar example.
39ParameterA name in a function definition that receives a value when the function is called.In def double(number), number is a parameter; it lets the same function work with different supplied values.
40Return valueThe result passed back from a function call to the code that called it.Printing a result displays it; returning the result lets another expression use it in a further calculation.
Terms 31–40: Control flow and functions

Organising programs and data

No.TermMeaningUse, check or example
41ArgumentA value or expression supplied to a function when it is called.In double(6), the argument is 6; connect it to the receiving parameter before tracing the function body.
42Local variableA variable whose binding belongs to a function or another local scope.A temporary total created inside a function does not automatically become a variable available everywhere else.
43Global variableA variable defined at the module level and accessible according to the language’s scope rules.Avoid assuming that assigning the same name inside a function automatically changes the module-level binding.
44ScopeThe region or context in which a name can be accessed and interpreted.If two functions use the name count locally, trace their own bindings instead of merging the values.
45ModuleA unit of Python code that can provide definitions, functions and other reusable content.A suitable module can supply a mathematical operation, but using it still requires understanding the operation’s inputs.
46ImportMaking a module or selected definitions available for use in a program.Distinguish making a function available from calling it; importing a module can also execute its top-level instructions.
47File inputReading data stored in a file into a program.Decide whether to read the whole file, a line or a smaller part, then handle the actual format.
48File outputWriting program data to a file for storage or later use.Choose the correct file and writing mode; explain whether the operation replaces existing content or adds to it.
49DelimiterA character or sequence used to separate fields or items in data.If records use commas between fields, inspect how missing fields or unexpected commas affect the interpretation.
50CommentText in source code intended to explain the program and not executed as an instruction.A useful comment explains a decision or assumption; a correct comment cannot repair incorrect executable code.
Terms 41–50: Organising programs and data

Tracing and testing

No.TermMeaningUse, check or example
51Trace tableA structured record showing how selected values change as instructions execute.Record values after each relevant step; a final answer alone may hide where an incorrect update first occurred.
52Dry runFollowing a program or algorithm manually without executing it on the computer.Use a small input and apply each instruction literally; do not silently correct the code while tracing it.
53Expected outputThe result predicted from the requirements for a particular test input.Work it out independently before running the code, so the program’s answer does not become its own evidence.
54Actual outputThe result observed when the program or system is run.Record the actual result even when it is wrong, then compare it with the expected result to locate the mismatch.
55Syntax errorAn error caused by code that violates the language’s grammatical rules.A missing colon in a Python if statement prevents normal parsing; changing test data will not repair the grammar.
56Run-time errorAn error that arises while a program is executing.Accessing an unavailable list index can raise an error even when the statement follows valid Python syntax.
57Logic errorA mistake in the solution’s reasoning that produces behaviour inconsistent with its requirements.A program can run without an error message and still count the wrong records or calculate the wrong total.
58Test caseA specified set of inputs, conditions and expected results used to check behaviour.Name the purpose of the case: an ordinary success, a boundary decision or rejection of invalid input.
59Normal dataValid, representative input chosen to check ordinary expected behaviour.For permitted integers from 1 to 30, an input of 15 checks an ordinary value inside the range.
60Boundary dataInput at or close to a limit where the required behaviour may change.Testing 0, 1, 30 and 31 helps examine both sides of an inclusive range from 1 to 30.
Terms 51–60: Tracing and testing

Validation, repair and development

No.TermMeaningUse, check or example
61Invalid dataInput that fails one or more requirements for acceptance.A value can have the correct type but still be invalid; the integer 31 fails a maximum of 30.
62ValidationChecking data against specified rules before accepting or processing it.Passing a range check does not prove the data matches reality; an incorrect but permitted value can still pass.
63Range checkA check that a value lies within specified limits.Write down whether each endpoint is included, then choose comparison operators that express that exact rule.
64Length checkA check that the number of characters or items meets a stated requirement.An identifier with six characters can pass a length check while still containing characters forbidden by another rule.
65Presence checkA check that a required value has been supplied.Decide whether a blank or spaces-only entry counts as missing in the stated task; make the rule explicit.
66Format checkA check that data follows a required pattern or arrangement.An entry can look like a date yet describe an impossible day; format and meaningful validity are different questions.
67Existence checkA check that supplied data already exists in the relevant stored set or system.A member code may have the right length but fail because it does not match any registered member.
68Check digitAn extra digit calculated from other digits to help detect certain data-entry or transmission errors.Recalculate and compare it using the specified rule; a matching digit does not prove that every possible error is absent.
69DebuggingFinding the cause of a program fault and changing the program to correct it.Identify the first wrong value, repair its cause and rerun useful tests instead of changing unrelated lines.
70Iterative developmentDeveloping a solution through repeated cycles of building, checking and refining.Test a small working part before extending it, then check that the extension has not broken earlier behaviour.
Terms 61–70: Validation, repair and development

Computer systems and representation

No.TermMeaningUse, check or example
71BitA binary digit that can represent one of two values, usually written 0 or 1.State whether a capacity is measured in bits or bytes before comparing the numerical values.
72ByteA group of eight bits.Multiply a number of bytes by eight to obtain its size in bits; do not reverse that conversion.
73KibibyteA unit of digital information equal to 1024 bytes, abbreviated KiB.Distinguish 1 KiB from 1 kB, which uses the decimal prefix and equals 1000 bytes.
74BinaryA base-two number system using the digits 0 and 1.The binary number 1010 represents ten in denary; explain its place values instead of reading it as one thousand and ten.
75DenaryThe base-ten number system using digits from 0 to 9.When converting a binary value, add the contributions from occupied binary place values to obtain the denary result.
76HexadecimalA base-sixteen number system using 0–9 and A–F for its sixteen digit values.One hexadecimal digit corresponds to four binary bits; keep the representation distinct from the value being represented.
77Two’s complementA fixed-width binary representation used for signed integers, including negative values.State the bit width before interpreting the leading bit; the same written bits need a specified representation rule.
78ProcessorThe component that carries out instructions and performs operations needed to process data.Distinguish executing an instruction from storing a file permanently; these are different jobs within the system.
79Main memoryMemory used to hold data and instructions directly available for active processing.Explain why a program needs working data available during execution rather than describing memory only as a storage size.
80Secondary storageStorage used to retain programs and data beyond their immediate use in main memory.Compare a storage choice using the task’s capacity, durability, speed, portability and cost requirements rather than one feature alone.
Terms 71–80: Computer systems and representation

Spreadsheet reasoning

No.TermMeaningUse, check or example
81Cell referenceAn address identifying a spreadsheet cell, usually through its column and row.B3 identifies a location; the formula using it depends on the value currently stored at that location.
82Relative referenceA cell reference whose row and column positions adjust when the formula is copied.Copying =B2*C2 down one row normally changes it to =B3*C3, matching the next record.
83Absolute referenceA reference with both row and column fixed when the formula is copied.Use $F$1 for a shared rate that must remain in the same cell across copied formulas.
84Mixed referenceA reference with either its row or its column fixed while the other part can adjust.In $B2 the column is fixed; in B$2 the row is fixed. Explain which movement the task requires.
85FormulaAn expression entered into a spreadsheet to calculate a result from values and operations.A correct result in one row does not prove the formula will behave correctly after copying elsewhere.
86Spreadsheet functionA named spreadsheet operation that calculates a result from supplied arguments.Choose COUNT for numeric entries and COUNTIF for a stated condition; select according to the question being answered.
87Exact matchA lookup that requires the search value to match an entry under the function’s matching rules.When searching for a particular code, specify exact matching instead of accepting a nearby value.
88Approximate matchA lookup mode that selects a suitable entry according to ordering or threshold rules rather than requiring equality.For threshold bands, use the required table order and matching mode; an unsorted table can return misleading results.
89Conditional formattingAutomatic changes to cell appearance when specified rules are satisfied.Highlighting a low value can draw attention to it, but does not itself correct the value or reject its entry.
90Goal SeekA spreadsheet feature that changes one input cell to make a formula cell reach a specified target.Identify the target formula and the input allowed to change; other assumptions remain part of the model.
Terms 81–90: Spreadsheet reasoning

Networks, security and machine learning

No.TermMeaningUse, check or example
91NetworkConnected computers or devices able to exchange data through a communication medium.Explain which devices communicate and why; owning several unconnected computers does not by itself create a network.
92ProtocolAn agreed set of rules governing communication between systems.Devices need compatible rules for exchanging data; a shared physical connection alone is not the complete explanation.
93PacketA unit of data prepared for transmission through a network, with information needed for handling it.A larger message can be divided into packets; distinguish the message content from information used to deliver it.
94RouterA device that forwards data between networks using network addressing information.Distinguish connecting networks from forwarding data among devices within one local network.
95SwitchA device that forwards data between devices within a network, commonly using MAC addresses.A home box can combine several functions, but a combined device does not make router and switch interchangeable terms.
96EncryptionTransforming readable data into an encoded form that requires an appropriate key to recover its meaning.Protecting a transmission does not prove that the recipient is honest or that a requested action is sensible.
97PhishingDeceptive communication or imitation of a trusted service used to trick people into revealing information or taking harmful action.An urgent message and a familiar logo are not sufficient proof of authenticity; verify through a trusted independent route.
98Artificial intelligenceComputing methods that perform tasks associated with intelligent behaviour, such as recognising patterns or making predictions.Judge a system using its actual task and evidence; the label AI does not guarantee understanding, fairness or correctness.
99Machine learningAn approach in which a system uses patterns in data to produce a model for predictions or decisions.Training examples influence results; learning from biased or unsuitable data can produce unreliable conclusions.
100Nearest neighbourA method that predicts from the most similar stored example according to a specified distance or similarity rule.Compare the relevant distances using the same feature definitions; the first stored example is not automatically the nearest.
Terms 91–100: Networks, security and machine learning

Contents · Previous chapter · Next chapter

CHAPTER 3 OF 9

Distinctions that prevent several mistakes at once

Back to contents

A value, its representation and its type are different

The integer twelve, the string “12” and the binary representation 1100 are connected ideas, but they are not interchangeable descriptions. Ask which value is being represented, which representation is being used and which operation the program will perform. In Python, adding two suitable numbers is arithmetic, while adding two strings concatenates them. The official Python documentation for input() confirms that keyboard input is returned as a string. Reading digits does not automatically produce an integer.

A rule can be validly written and still express the wrong decision

Syntax concerns whether the code follows the language’s grammar. Logic concerns whether its behaviour matches the task. A condition using or can be syntactically correct yet fail to require both limits of a permitted range. Explain what each part means, and try a value that should be rejected. An ordinary value in the middle may not distinguish a correct rule from an incorrect one.

Validation checks a rule; testing checks behaviour

A validation rule may reject a blank entry or an out-of-range value. A test case checks whether the program applies that rule correctly. Neither proves that all accepted information is true. If a person enters 12 when the real number is 15, both values might satisfy the same range check. Use precise conclusions: “the entry passes this check” is narrower than “the information is correct”.

The same-looking operation can behave differently elsewhere

Python and spreadsheet notation share symbols but do not share every convention. The Python expression reference explains floor division; -7 // 2 is -4, because flooring moves down to the lower integer. In a spreadsheet, copying a formula also changes relative references. Name the language and the context before applying a remembered rule.

Contents · Previous chapter · Next chapter

CHAPTER 4 OF 9

Five worked examples: predict, explain, repair and retest

Back to contents

All records, values and situations in these examples are invented for learning. They are not school results, actual club rules or advertised tuition arrangements. Work through the prediction before reading the explanation. Running the code is useful after prediction because the difference between expected and actual behaviour creates something specific to investigate.

Example one: a range check that accepts everything

A fictional Punggol reading activity accepts a booking for between 1 and 30 places, including both limits. Assume places is already an integer. A student writes this condition:

accepted = places >= 1 or places <= 30

For 15, both comparisons are true, so the booking is accepted. That result looks reasonable. Now test 0: the first comparison is false, but the second is true. Because or needs only one true comparison, 0 is accepted. For 31, the first comparison is true and the second false, so 31 is accepted too. In fact, every integer satisfies at least one side of this particular condition.

The requirement is to meet both limits. The repaired condition is:

accepted = places >= 1 and places <= 30

Now 1 and 30 are accepted; 0 and 31 are rejected. The normal case 15 still passes. These tests reveal why boundary checks are useful: a change that preserves an ordinary success must also reject the unsuitable neighbouring values. They do not remove the separate need to handle non-integer text before using the numeric comparison.

Transfer check: change the requirement to integers from 5 up to, but not including, 12. The condition becomes places >= 5 and places < 12. Test 4, 5, 11 and 12. Merely replacing the two numbers while keeping both inclusive comparison signs would change the intended rule.

Example two: a largest-value algorithm that fails on negatives

Suppose a program must find the largest value in a non-empty list. Initialising best to zero looks harmless when all examples contain positive numbers. For [-7, -2, -9], however, no item exceeds zero. Such a program returns a value that was never in the list.

Use the first actual value as the initial candidate, then inspect the remaining values:

def largest(values):
    best = values[0]
    for value in values[1:]:
        if value > best:
            best = value
    return best

Initially, best is -7. When the program inspects -2, the comparison is true, so best becomes -2. The comparison with -9 is false, so -2 remains the return value. This explanation identifies both the initial state and the condition that changes it.

The assumption of a non-empty list matters. An empty list has no item at index zero and no largest member. The required response to empty input must be specified separately; inventing zero as an answer changes the meaning of the task. Test a one-item list as well: its only member should be returned even though there are no remaining items to inspect.

Transfer check: find the smallest value in [6, -4, 2] by adapting the comparison. The result is -4. Explain why choosing < changes the selection rule, while the first-item initialisation remains suitable for a non-empty list. If a task requires a manual algorithm, follow that instruction instead of substituting a built-in answer.

Example three: finding every matching position

The task is to return every position of a target character in a string. Here the output is a list of positions, not a count and not the matching characters themselves. That requirement determines what the program records.

def locations(text, target):
    positions = []
    for index in range(len(text)):
        if text[index] == target:
            positions = positions + [index]
    return positions

For locations("level", "l"), the loop visits indices 0 through 4. It records 0 when the first character matches, ignores the three middle characters, then records 4. The result is [0, 4]. The variable index describes the position being inspected, while text[index] supplies the character stored there.

If the loop used range(len(text) - 1), it would stop before index 4 and miss the final match. If it used range(len(text) + 1), it would eventually attempt an unavailable index. These are different faults: one silently produces an incomplete result, while the other reaches a run-time error.

Transfer check: locations("bell", "l") returns [2, 3]. Searching for "z" returns an empty list. The empty list is a valid result here because there are no matching positions. Contrast that meaning with the earlier task, where an empty input list had no largest member. “Empty” needs to be interpreted against the actual requirement.

Example four: a formula works once but fails when copied

An invented supplies spreadsheet records quantity in column B, unit price in column C and a shared discount rate in F1. Row 2 contains quantity 8 and price $1.50; row 3 contains quantity 5 and price $3.00. F1 contains the decimal value 0.10. The formula for row 2 is =B2*C2*(1-$F$1).

The first result is $10.80. Copying the formula down gives =B3*C3*(1-$F$1), producing $13.50. The quantity and price move to the next record, while the shared rate remains fixed. That is the reason for the combination of relative and absolute references; the dollar signs are not decoration or a currency instruction.

If the original formula used F1 without fixing it, copying down would refer to F2. A single correct result in row 2 would therefore provide weak evidence that the whole column is correct. Inspect the copied formula and deliberately change the shared rate to check its influence.

Transfer check: changing F1 to 0.12 should give $10.56 in row 2 and $13.20 in row 3. If only one row changes appropriately, investigate the references before changing the arithmetic. Displaying two decimal places makes the output readable, but display formatting and the underlying calculation are separate matters.

Example five: the nearest example is determined by a distance rule

Consider a small invented classification dataset with two numerical features measured on comparable scales. Reference A has coordinates (1, 1) and label A. Reference B has coordinates (4, 5) and label B. A new item is at (3, 4). This is a mathematical learning example, not a real-world classification policy.

Using ordinary straight-line distance, compare squared distances first. The squared distance to A is (3 − 1)² + (4 − 1)² = 13. The squared distance to B is (3 − 4)² + (4 − 5)² = 2. Taking square roots would preserve the order, so B is nearer and the one-nearest-neighbour prediction is label B.

That result is a prediction under this tiny dataset and distance rule. It does not prove that label B is correct in every real situation. The examples, feature choices and scale of measurement affect the decision. A tie would require an explicit handling rule rather than quietly choosing whichever result is more convenient.

Transfer check: move the new item to (1, 2). The squared distances become 1 to A and 18 to B, so the predicted label changes to A. Explain the changed evidence before naming the changed answer. A useful understanding of machine learning includes recognising these dependencies, not treating the output as an unquestionable fact.

Contents · Previous chapter · Next chapter

CHAPTER 5 OF 9

Eight independent questions and a diagnostic next step

Back to contents

Complete these without the worked solutions beside you. Write a prediction and a reason before using a computer. If you need a hint, record that fact and still finish the question; later use a changed example to check independent understanding. This checkpoint is a learning tool, not an official examination or grade conversion.

  1. Inclusive limits: a value must be an integer from 10 to 20 inclusive. Write the acceptance condition and state the expected decisions for 9, 10, 20 and 21.
  2. Types: predict what happens in Python when evaluating 4 + "2". Explain how str(4) + "2" differs from 4 + int("2").
  3. Accumulation: predict the output of the code below. Then predict the output if the condition changes to value >= 8.
  4. Slice boundaries: let text = "punggol". What does text[1:4] produce, and which positions contribute to it?
  5. Initialisation: a largest-value program starts with zero and checks [-8, -3, -5]. Explain the error and give an appropriate initial candidate for this non-empty input.
  6. Copied formula: a formula in D2 is =$B2*C$1. What formula results when it is copied one column right and one row down, to E3?
  7. Units: a file contains 3 KiB of data. How many bytes and how many bits is that? Explain why 3000 bytes would use a different unit interpretation.
  8. Security reasoning: an unsolicited message claims an account will close unless you enter a password on a linked page. The page uses an encrypted connection. Does encryption establish that the request is genuine? Give a reason and a sensible verification step.

Code for question three:

total = 0
for value in [3, 8, 2]:
    if value >= 3:
        total = total + value
print(total)

Answers with explanations

One: value >= 10 and value <= 20 expresses both inclusive limits. The decisions are reject, accept, accept and reject respectively. This assumes the value is already an integer. A program receiving unrestricted text needs a separate step to handle its type or conversion appropriately.

Two: 4 + "2" raises a type-related run-time error because these operands are not supported together for addition. str(4) + "2" joins two strings to produce "42". 4 + int("2") adds integers to produce 6. Conversion must be suitable for the data; not every string can be converted to an integer.

Three: the original total is 11. The values 3 and 8 satisfy the condition and are added; 2 is ignored. With the changed condition, only 8 qualifies, so the output is 8. The program adds qualifying values, rather than counting them. A count for the original condition would be 2.

Four: the result is "ung", using indices 1, 2 and 3. The character at index 4 is excluded because the stop position of this slice is not included. Check the positions explicitly rather than assuming the final number tells you how many characters to take.

Five: no listed value exceeds zero, so the faulty method retains a result that is not a member of the input. Start with the first value, -8, then inspect the others; the largest becomes -3. This repair depends on the non-empty input assumption. State the separate empty-input requirement before extending the program.

Six: the copied formula is =$B3*D$1. The first reference keeps column B but changes row 2 to row 3. The second keeps row 1 but changes column C to D. Explain each reference separately; moving the entire formula does not make every component move identically.

Seven: 3 × 1024 gives 3072 bytes, and 3072 × 8 gives 24,576 bits. The decimal unit kB uses 1000 bytes per kilobyte; KiB uses 1024. A conversion should identify the unit before applying a multiplication, because a plausible-looking number can still describe the wrong quantity.

Eight: no. Encryption protects data in transit under the connection’s security arrangements, but does not by itself establish that the receiving service or request is legitimate. Verify through an independently obtained official website or known contact route. The Cyber Security Agency’s phishing guidance explains how urgent messages and imitation websites can mislead users and recommends checking authenticity through official sources.

Use the response to choose the next action

Observed responseLikely issue to investigateRepair taskFresh independent retest
Defines a term but predicts the wrong outputThe meaning has not been connected to execution.Trace one instruction at a time and record the first changed value.Use a different short program with the same construct.
Passes an ordinary input but misses a limitThe requirement or comparison boundary is unclear.Write the accepted set in words and test both sides of each endpoint.Change one endpoint from inclusive to exclusive and predict again.
Confuses numbers, strings or positionsType and representation are being mixed.Label each value’s type and each index before calculating.Use new characters or values and explain the operation chosen.
Repairs the example but cannot explain the causeThe edit may have been copied without a model of the fault.Compare the faulty and repaired instruction using one revealing input.Give a related fault in a different position or context.
A copied formula works in one cell onlyReference movement has not been checked.Track each row and column independently while copying.Copy across and down using a different mixture of fixed references.
Gives an absolute claim from limited evidenceThe conclusion exceeds what the test or model establishes.Name what is known, the assumption and one remaining limitation.Change the dataset, recipient or test case and reassess the claim.
Diagnostic actions after the independent checkpoint

Do not turn this table into a label for the student. A response is evidence about a particular task at a particular time. Record the original answer, the repair and the retest separately. Improvement becomes visible when the learner can explain and use the idea again with less help, including in a context that looks different.

Contents · Previous chapter · Next chapter

CHAPTER 6 OF 9

A manageable revision cycle for vocabulary and practical work

Back to contents

First, choose a small group and one observable task

Pick five to ten related terms from a current school topic. For a loop task, that might include iteration, condition, accumulator, counter and trace table. Ask for meanings, but finish with a prediction or explanation that uses them. A session has a clearer purpose when its final question is something like “Can I explain why the last item was missed?”

Next, inspect the first mismatch

Compare expected and actual behaviour. Find the earliest point where the result stops matching the requirement, rather than rewriting the whole program immediately. An incorrect final total may arise from a bad initial value, the wrong condition, a misplaced update or an inappropriate operation. Those explanations require different repairs, so the final answer alone is insufficient evidence.

Then, vary one condition and retest

Change the size of the input, include a negative value, move a match to the last position or change a threshold. Keep the task small enough to predict manually. Once the revised solution behaves correctly, try a fresh task without the original answer. Do not keep adding new tests without purpose; select cases that examine a concrete uncertainty in the reasoning.

Balance screen work with explanations on paper or aloud. Practical familiarity helps a student save and run a solution, while a dry run helps reveal what the code actually means. The current K349 syllabus identifies both application and the development, testing and refinement of solutions. A revision routine should therefore produce evidence of use, not only evidence of remembered terminology.

Contents · Previous chapter · Next chapter

CHAPTER 7 OF 9

How parents and a small group can support the next decision

Back to contents

A parent does not need to know every Python command to ask a useful question. “What did you expect?”, “Which instruction changed that value?” and “What different input would test your explanation?” invite the learner to make their thinking visible. Ask for one short explanation and allow time to work; repeatedly supplying the next line can hide the part that still needs practice.

An illustrative small-group activity could give three students the same short faulty program. Each first predicts the output independently. One then explains the condition, another traces the values and the third proposes a revealing test. After discussion, all three complete a different question separately. The independent final response matters because a convincing group conversation does not show that every member can reproduce the reasoning.

When considering additional support, bring a recent piece of work and identify the actual difficulty. Is it understanding the task, recalling language rules, tracing control flow, selecting test data, using the software or explaining a conclusion? Ask how the teaching will check that specific gap and what the student will do independently afterwards. A large worksheet count is less informative than a clear example of changed performance.

Confirm the provider’s current subject coverage, software arrangements and lesson format directly. Families who want to discuss a learning need can use the eduKatePunggol first-step enquiry page, with the student’s course and a concrete example of the difficulty ready.

Contents · Previous chapter · Next chapter

CHAPTER 8 OF 9

Common questions about the hundred-term checkpoint

Back to contents

Are all these terms an official SEAB list?

No. They are an editorial selection designed to make revision more usable. Some name concepts, some name language features and others describe actions used when checking solutions. Read them alongside the actual syllabus and school materials. The word “top” in the title does not mean that SEAB has ranked these entries or promised their appearance in an examination.

Does knowing the definitions mean a student is ready?

Definitions are useful when they improve decisions. A learner should also be able to predict a result, give a reason, recognise a fault and test a repair. If those steps need support, use the diagnostic table to choose a small next task. A vocabulary result is not an examination grade or a guarantee of one.

Should students always use the shortest Python solution?

Follow the task’s requirements. A short built-in operation may be appropriate in one context, while another task asks the student to show an algorithm without that operation. Explain why the solution works and whether it meets the stated restriction. Brevity by itself is not evidence of correctness or understanding.

How does this differ from the wider vocabulary guide?

The broader SEC G3 Computing vocabulary guide is extended reading across the subject. This page provides a focused hundred-term checkpoint with independent predictions, explained answers and a repair-and-retest route. Use the broader guide to expand coverage after identifying the next learning need.

What if the school teaches these topics in a different order?

Follow the school’s teaching and assessment sequence. These groups organise ideas for end-of-course revision and do not prescribe a national Secondary 4 term plan. Return to an earlier foundation when a later task depends on it; revisiting types or comparisons can resolve a difficulty that first appears in a more complex program.

Contents · Previous chapter · Next chapter

CHAPTER 9 OF 9

Sources and further checking

Back to contents

The course references below were checked for this guide in November 2026. Use the official directory to locate the latest document for the student’s examination year. Definitions, invented scenarios, code examples and diagnostic activities on this page were written as teaching material; they are not reproduced examination questions or an official marking scheme.

Contents · Previous chapter · Continue to the wider vocabulary reference

Continue from here: Start Here · Tuition · Education · Pathways · Parenting 101 · All Site Routes

eduKate Punggol

Contact

83 Punggol Central, Singapore 828761

edu|Kate Bukit Timah

8 Fourth Avenue, Singapore 268674

By Appointment +65 8823 1234
admin@edukatesg.com

Email Us

When a child finally understands, school becomes less frightening and the future opens wider. Email us for the latest schedules and fees.

← 返回

感谢您的回复。 ✨