Official syllabus: 2026–2028 · version 6. Read the authoritative syllabus. Summary checked against its assessment overview on 5 October 2026. Administrative zone and centre-specific entry options remain unconfirmed.
Paper 1 Computer Systems and Paper 2 Algorithms, Programming and Logic are each 1 hour 45 minutes, 75 marks and 50%. Both are written examinations without calculators. The 2029 syllabus changes must not be substituted for this version.
For each proposed week: one 50-minute guided lesson, one 40-minute supported practice session and approximately 60 minutes of independent work. The appointed teacher must adjust pace from the initial diagnostic, map the complete syllabus and confirm accommodations before delivery. This sample is only the opening six-week sequence.
Each guided lesson uses: 5 minutes retrieval; 10 minutes explicit modelling; 15 minutes guided practice; 15 minutes independent response; 5 minutes exit check. Keep a plain-text copy and accept agreed low-bandwidth work. Collect only authorised learner records; no live learner records are included here.
Before week one, use the first unit’s questions without grading for certification. Record the method as well as the result. Identify a specific gap, teach one smaller example and retry a parallel question. After two unsuccessful supported attempts, the future teacher refers the learner to the academic coordinator with the work sample and the support tried. This is a proposed workflow; named staff are not yet appointed.
Convert small positive integers between denary and binary and explain the bit positions.
Draw an eight-bit place-value row 128,64,32,16,8,4,2,1. Represent 13 as 00001101 by selecting 8+4+1. Reverse the process with 00101010=32+8+2=42. Explain that a leading zero does not change the value and that this worksheet uses unsigned values.
Convert 25 to eight-bit binary, convert 00110110 to denary and state the largest unsigned value in eight bits.
00011001 with working (3 marks), 54 with working (3), 255 from 128+64+32+16+8+4+2+1 (2).
Provide the place-value row; extend by describing what happens when a result exceeds the available bits.
Retain the authorised work sample, annotate the specific criterion met or missed, give one actionable feedback point and check a parallel response. A mark alone does not demonstrate improvement. Log the next lesson’s adaptation.
Describe a simple system and distinguish roles from device names.
Use a fictional library borrowing kiosk. A scanner reads a book identifier; software checks the record; a screen displays an instruction; a storage system records a completed loan. Model a labelled flow and ask what would happen if the identifier is valid but the network is unavailable.
Draw or describe the data flow. Give one input, one processing action, one output and one stored item. Identify a failure and a safe response.
Four correctly classified roles (4 marks), plausible failure (1), response that avoids falsely recording a successful loan (2).
Use a text-only flow; extend by considering accessibility of input and output methods.
Retain the authorised work sample, annotate the specific criterion met or missed, give one actionable feedback point and check a parallel response. A mark alone does not demonstrate improvement. Log the next lesson’s adaptation.
Distinguish a rule about acceptable data from a check of accurate entry.
A fictional form accepts an age from 18 to 120. A range check can reject 150, but it cannot establish that 35 is the person’s true age. Verification compares an entry with its source or a second entry; it is not proof of identity. Model a test table with accepted, rejected and uncertain cases.
Explain why a valid value can still be wrong. Create four range-check tests, including boundary values. Describe one appropriate verification method.
Valid-but-wrong distinction (2 marks); suitable normal/boundary/invalid cases (4); source comparison or double entry explained with limitation (2).
Provide a test-table template; extend by distinguishing authentication from either check.
Retain the authorised work sample, annotate the specific criterion met or missed, give one actionable feedback point and check a parallel response. A mark alone does not demonstrate improvement. Log the next lesson’s adaptation.
Follow sequence and selection, including a boundary case.
Original pseudocode: INPUT score; IF score >= 50 THEN OUTPUT 'met threshold' ELSE OUTPUT 'below threshold' ENDIF. Trace scores 49,50,51. Show the difference between > and >=. Ask learners to record conditions as true/false before writing the output. This is a practice algorithm, not an examination question.
Trace all three values, explain the result at 50 and modify the rule for a threshold of 60. Add one test for each branch.
Correct outputs for 49/50/51 (3 marks), boundary explanation (2), correct >=60 rule (2), branch tests (2).
Supply a trace table; extend by designing a rule for two ordered thresholds and checking branch order.
Retain the authorised work sample, annotate the specific criterion met or missed, give one actionable feedback point and check a parallel response. A mark alone does not demonstrate improvement. Log the next lesson’s adaptation.
Write, trace and test a small program without confusing running with correctness.
Original Python example: def total_cost(count): return count * 12. Call the function with 0,1,5 and compare results with a hand calculation. Explain the assumed input: a non-negative integer count. A run that produces a number does not prove the program meets its requirements. Add validation or describe where invalid input is rejected.
Write a function or equivalent pseudocode for a R15 ticket price. Provide expected and actual/desk-traced results for 0,1,4, and a negative-count case. Explain the input assumption and invalid-case handling.
Correct calculation (3 marks), normal and boundary tests with expected outcomes (3), invalid case and explained handling (3). No software installation is needed for a desk trace.
Allow pseudocode and manual trace for low-bandwidth practice; extend by adding a fixed booking fee and separating it from ticket cost.
Retain the authorised work sample, annotate the specific criterion met or missed, give one actionable feedback point and check a parallel response. A mark alone does not demonstrate improvement. Log the next lesson’s adaptation.
Explain simple AND, OR and NOT conditions with a truth table.
A fictional door opens when a valid badge AND an approved access period are both true. Build the four-row truth table and compare it with OR. Explain why changing the operator changes the rule. Do not build a real access system from this exercise; it is a logic model.
Complete the AND table, compare the false/true row under OR, and describe NOT valid_badge in words. Propose one test that catches an accidental OR implementation.
AND outcomes F,F,F,T (4 marks), OR comparison (1), NOT meaning (1), test with just one input true (2).
Use plain-language true/false labels; extend by adding a third condition and explaining operator grouping.
Retain the authorised work sample, annotate the specific criterion met or missed, give one actionable feedback point and check a parallel response. A mark alone does not demonstrate improvement. Log the next lesson’s adaptation.
For the six-week review sample, combine one task from weeks 1–2, one from weeks 3–4 and one from weeks 5–6. State the purpose and local rubric before use. These are original practice tasks, not Cambridge examination questions, grade thresholds or official mark schemes.
All four remain required before formal teaching opens. School admissions and fees remain launch-gated.