O Level Computer Science: Algorithm Design and Problem-Solving — Revision Notes

Topic 7 of Cambridge O Level Computer Science (2210) appears on Paper 2 in a handful of recurring question types: identify a life-cycle stage, decompose a problem, state an algorithm’s purpose, choose test data, complete a trace table, find and correct errors, and write or amend an algorithm. These notes are organised around those question types, with the facts and templates each one needs.

Condensed for the final weeks. For the full teaching, use the algorithm design study guide; then practise with the algorithm design practice questions.

Syllabus coverage (2210, 2026–2028)

Ref	What you must be able to do
7.1	Program development life cycle: analysis, design, coding, testing
7.2	Sub-systems; decomposition into inputs, processes, outputs, storage; structure diagrams, flowcharts, pseudocode
7.3	Explain the purpose of a given algorithm
7.4	Linear search, bubble sort, totalling, counting, maximum, minimum, average
7.5	Validation (range, length, type, presence, format, check digit) and verification (visual, double entry)
7.6	Normal, abnormal, extreme and boundary test data
7.7	Trace tables recording variables, outputs and user prompts
7.8	Identify and correct errors in algorithms
7.9	Write and amend algorithms in pseudocode, program code and flowcharts

Question type 1: “Which stage of the life cycle?”

• Analysis — abstraction, decomposition, identifying the problem and requirements.
• Design — structure diagrams, flowcharts, pseudocode (decomposition again).
• Coding — writing code with iterative testing.
• Testing — running the code with test data.

Question type 2: “Decompose the problem”

Name one item for each of input, process, output, storage, using the words of the scenario. Abstraction removes details that do not affect the solution.

Question type 3: “State the purpose of the algorithm”

1. One sentence: what does it achieve? (“It finds the cheapest item.”)
2. Then the processes: what is input, what the loop does, what decisions are made, what is output.

Don’t translate it line by line — that describes the code, not its purpose.

Question type 4: “Choose test data”

Range 1–50 inclusive, whole numbers:

Normal	Abnormal	Extreme	Boundary
25 (accept)	51, 0, “ten” (reject)	1, 50 (accept)	0/1 and 50/51

Question type 5: “Complete the trace table”

• One column per variable plus OUTPUT; record user prompts in the output column too.
• Write a value only when it changes; move down a row each time.
• Check loop conditions carefully — a < versus <= decides whether the last pass happens.

Question type 6: “Find and correct the errors”

Typical planted errors:

Planted error	Fix
Counter or total starts at 1	Start at 0
Maximum starts at a large value (or minimum at a small one)	Start maximum low / minimum high, or at the first item
AND where OR is needed in a loop condition	Swap the operator
Output inside the loop	Move it after the loop
Wrong variable incremented (Count ← Index + 1)	Count ← Count + 1
Comparison the wrong way round (< for a maximum)	Use >

Question type 7: “Write the algorithm”

Templates to adapt (2210 pseudocode):

Validation with re-entry

REPEAT
    OUTPUT "Enter the number of guests (1 to 6)"
    INPUT Guests
    IF Guests < 1 OR Guests > 6
      THEN
        OUTPUT "Invalid, try again"
    ENDIF
UNTIL Guests >= 1 AND Guests <= 6

Minimum and average from an array of 15 prices

Cheapest ← Price[1]
Sum ← 0
FOR Item ← 1 TO 15
    Sum ← Sum + Price[Item]
    IF Price[Item] < Cheapest
      THEN
        Cheapest ← Price[Item]
    ENDIF
NEXT Item
OUTPUT "Cheapest: ", Cheapest, " Mean: ", Sum / 15

Linear search — loop while not found and not past the end; set a Boolean flag when found; output the result after the loop.

Bubble sort — repeat passes comparing List[i] with List[i + 1]; swap through a temporary variable; stop when a pass makes no swaps.

Validation and verification in one table

Check	Catches
Range	Values outside limits
Length	Too many or too few characters
Type	Wrong data type
Presence	Missing data
Format	Data not matching a pattern
Check digit	Transcription errors in codes such as barcodes
Visual (verification)	Entry differs from the source document
Double entry (verification)	Two entries differ

2210 and its sibling syllabus

Topic 7 learning outcomes are the same in 2210 and Cambridge IGCSE Computer Science (0478), including the pseudocode format published in section 4 of each syllabus.

Exam traps

Drawn from our practice page, which paraphrases the Cambridge 2210 June 2025 mark schemes and lists where marks are usually lost:

• For an inclusive range check, extreme data was credited only for the two limits themselves; boundary data needed a limit and the value just outside it.
• The credited reason for verification was making sure data was not altered as it was entered.
• Each error was credited for identifying the faulty line (by number or by quoting it) and giving a correct replacement.
• When explaining how a loop stops, name the variables and the condition — “when it is done” scores nothing.
• A range check with < instead of <= wrongly rejects a limit value.

Frequently asked questions

Will I have to draw a flowchart? You may be asked to write or amend an algorithm as a flowchart, so know the standard symbols in section 4 of the 2210 syllabus: terminator, process, input/output, decision, subroutine and flow lines.

Is a check digit validation or verification? Validation. The 2210 syllabus lists the check digit with the other validation checks.

How precise does my pseudocode need to be? Precise enough that there is only one meaning. Use operators such as >=, not words such as “is at least”.

Self-test

1. Which life-cycle stage includes abstraction?
2. Give the extreme data for a whole number from 5 to 25.
3. Name the two verification checks.
4. In a bubble sort, what variable lets the algorithm stop early?
5. Why should an average be calculated after the loop?

Answers: 1. Analysis. 2. 5 and 25. 3. Visual check and double entry check. 4. A Boolean flag (such as Swapped) that records whether any swap was made in a pass. 5. All the values must be added to the total first; inside the loop the average would be calculated, and possibly output, many times using an incomplete total.