AI-assisted English content review; independent teacher approval pending.
Practice scope and review
Original, AI-authored English practice with worked guidance. Selected topics only; independent teacher review is pending. This is not a complete syllabus, official past-paper collection or predicted exam score. Check your own exam-board specification: these concepts are not certified against a particular board or exam year.
151 original questions across 21 selected topic areas. No account or payment required to practise.
Cambridge International AS & A Level Computer Science (9618), examination year 2027: AI-assisted conceptual alignment to official sources, not teacher or exam-board certification. Syllabus mapping and remaining coverage.
5 questions in this topic
Exact English revisions checked in an AI-assisted academic review; independent teacher approval is still pending.
All topics
Original English exercises targeting selected 9618 2027-29 requirements. Conceptual alignment only; independent teacher approval, practical competence and complete teaching coverage are not established.
Subtopics in this topic (3)
Iterative versus waterfall development
Test planning and maintenance
Licence terms and AI privacy
5 of 5 questions shown
Loading saved responses
Question 1A school attendance prototype changes weekly after staff trials. Compare an iterative process with a one-pass waterfall approach for this situation. Specify one acceptance test for a minimum age16 and classify fixing an age-comparison bug versus adapting to a new operating system. A supplied library licence permits copying/modification/distribution only with attribution: may the developer remove its attribution because code was modified? Finally evaluate reusing identifiable attendance logs to train an unrelated AI system without informing families.6 indicative marks
Construct, trace and explain6 indicative marksNo calculator
A school attendance prototype changes weekly after staff trials. Compare an iterative process with a one-pass waterfall approach for this situation. Specify one acceptance test for a minimum age16 and classify fixing an age-comparison bug versus adapting to a new operating system. A supplied library licence permits copying/modification/distribution only with attribution: may the developer remove its attribution because code was modified? Finally evaluate reusing identifiable attendance logs to train an unrelated AI system without informing families.
Show indicative guidance
Worked guidance
Choose iteration for the changing trial findings: agree a small requirement set, design/code, test with staff, then revisit the next change. Frequent feedback can expose errors early, but scope creep and repeated rework still require management. A waterfall comparison must acknowledge the cost of changing a previously agreed stage, not call it universally useless.
Acceptance criteria make a test checkable: input15 is rejected and16 accepted. Changing >=16 accidentally to>16 would fail the latter, requiring corrective maintenance. Porting correct software to a new OS is adaptive; adding desired new functionality could instead be perfective.
The problem supplies the permission and condition; modification does not erase attribution. This is not legal advice about every licence. For AI reuse, separate the original operational purpose from the proposed training purpose, assess identifiability and permissions, and consider less revealing alternatives and accountable review before proceeding.
Self-marking points
View marking points
Iterative cycles can incorporate changing feedback and retest revisions; waterfall suits more stable agreed requirements but late changes may be costly.
Include analysis/design/coding/testing activities rather than equating a prototype with finished acceptance.
Test age15=>reject and16=>accept, with expected results recorded in the test plan.
Fixing the erroneous comparison is corrective maintenance; new-OS compatibility is adaptive.
The stated licence's attribution obligation still applies after modification; open availability does not mean absence of conditions.
Unrelated identifiable-log reuse raises privacy/purpose/consent and harm concerns; seek appropriate authority, minimise data and review governance rather than assuming an AI benefit justifies reuse.
Examiner tip
Indicative distinct points, not an official Cambridge mark allocation. State bounds, assumptions and intermediate states; practical teacher review remains required.
Question 2A booking accepts integer group sizes1..6 inclusive. Faulty pseudocode is Accepted<-(Size>1 AND Size<6). Trace sizes0,1,2,5,6,7; repair it and classify the maintenance. Explain why testing only size3 could miss the defect.4 indicative marks
Analyse4 indicative marksNo calculator
A booking accepts integer group sizes1..6 inclusive. Faulty pseudocode is Accepted<-(Size>1 AND Size<6). Trace sizes0,1,2,5,6,7; repair it and classify the maintenance. Explain why testing only size3 could miss the defect.
Cambridge 2027: Direct conceptual match
Enumerated six original/repaired outcomes and interior3 indistinguishability; original task is targeted boundary repair rather than the existing age test.
Strict comparisons wrongly exclude both stated endpoints. Sizes0 and7 are rejected correctly, while2 and5 are accepted correctly.
Accepted <- (Size >= 1 AND Size <= 6)
The repaired outputs are FALSE,TRUE,TRUE,TRUE,TRUE,FALSE. Size3 is inside both strict and inclusive ranges, so it cannot distinguish the versions; fixing this faulty rule is corrective maintenance.
Self-marking points
View marking points
Faulty outputs are FALSE,FALSE,TRUE,TRUE,FALSE,FALSE.
This is corrective maintenance; interior3 passes both versions.
Examiner tip
Test both boundaries and adjacent rejections, not a single normal case.
Question 3Checkout uses TotalPence<-Quantity*UnitPence returned by Price(). Price's documented contract returns integer pence, but a changed implementation returns pounds as REAL. For Quantity=3 and priceGBP2.50, give expected TotalPence and faulty result. Design a stub and integration test that exposes the mismatch even if both modules pass tests using their own assumptions.4 indicative marks
Analyse4 indicative marksNo calculator
Checkout uses TotalPence<-Quantity*UnitPence returned by Price(). Price's documented contract returns integer pence, but a changed implementation returns pounds as REAL. For Quantity=3 and priceGBP2.50, give expected TotalPence and faulty result. Design a stub and integration test that exposes the mismatch even if both modules pass tests using their own assumptions.
Cambridge 2027: Direct conceptual match
Checked3*250=750 versus3*2.5=7.5; stub and real-provider integration target an explicit units/type mismatch.
The stub exercises checkout against its promised interface and yields750. Supplying2.50 instead yields7.5, which is numerically in pounds rather than the required pence.
An integration assertion on the real Price plus checkout must require750 and the agreed integer-pence interface. Separate tests that silently use different units can both pass while the combined system fails.
Self-marking points
View marking points
Expected total is750 pence.
Faulty numeric result is7.5 in a variable intended for pence.
Stub Price returns250 as an integer with known units.
Integration asserts3 items produce750 and checks shared unit/type contract.
Examiner tip
Interface tests need units as well as plausible numerical values.
Question 4A fictional attendance export rule is Allow<-Authenticated AND IsTeacher. Teacher T is authenticated but assigned only classA; T requests named records for classB. Trace the existing rule, propose a corrected predicate requiring class assignment, and trace it for assigned/unassigned classes. Explain why retaining every export forever is a separate privacy concern even after this access fix.4 indicative marks
Analyse4 indicative marksNo calculator
A fictional attendance export rule is Allow<-Authenticated AND IsTeacher. Teacher T is authenticated but assigned only classA; T requests named records for classB. Trace the existing rule, propose a corrected predicate requiring class assignment, and trace it for assigned/unassigned classes. Explain why retaining every export forever is a separate privacy concern even after this access fix.
Cambridge 2027: Direct conceptual match
Checked role-only overpermission and resource-specific true/false cases; privacy evaluation is conceptual, not a legal compliance claim.
Authentication and teacher status are both true, so the original rule authorises classB despite no assignment. It confuses a general role with need-to-access for a particular resource.
Allow <- Authenticated AND IsTeacher AND AssignedToRequestedClass
For this teacher the final term is true forA and false forB, producing allow/reject respectively. Reducing access does not answer how long named exports should exist; purpose-limited retention and secure disposal still need a justified policy, without assuming a particular jurisdiction's law.
Self-marking points
View marking points
Existing rule allows T's classB request.
Adds AssignedToRequestedClass to the AND predicate.
Corrected rule allows classA and rejects classB for T.
Retention must be limited/justified separately; access restriction alone does not justify indefinite storage.
Examiner tip
Evaluate access scope and retention as two different controls.
Question 5A fictional library licence allows redistribution of modified code only if its copyright notice is retained. A release checker sets Ready<-TestsPass, ignoring NoticePresent. Trace (TestsPass=TRUE,NoticePresent=FALSE); propose a corrected release predicate and test both notice states when tests pass. Explain why source-code availability alone does not authorise removing the notice.4 indicative marks
Analyse4 indicative marksNo calculator
A fictional library licence allows redistribution of modified code only if its copyright notice is retained. A release checker sets Ready<-TestsPass, ignoring NoticePresent. Trace (TestsPass=TRUE,NoticePresent=FALSE); propose a corrected release predicate and test both notice states when tests pass. Explain why source-code availability alone does not authorise removing the notice.
Cambridge 2027: Direct conceptual match
Truth-tested the two-condition release gate and bounded interpretation to the fictional permission terms; no external legal facts invented.
Passing executable tests establishes neither the presence of a notice nor satisfaction of a reuse condition. The existing checker approves the stated failing licence case.
Ready <- TestsPass AND NoticePresent
With tests passing, changing NoticePresent from FALSE to TRUE changes Ready from FALSE to TRUE. This rule addresses this supplied condition only; it is not a complete copyright/legal clearance process or a claim that all licences use identical terms.
Self-marking points
View marking points
Existing checker returns TRUE despite the missing notice.
Corrects to Ready<-TestsPass AND NoticePresent under this supplied licence.
When tests pass, absent notice gives FALSE and present notice TRUE.
Availability/modification does not remove the supplied redistribution condition.
Examiner tip
Distinguish a supplied licence condition from a universal rule about all open code.
