Exam Difficulty Verdict & Overview
The October/November 2025 sittings of Cambridge International AS & A Level Computer Science (9618) Paper 12 and Paper 22 presented a balanced, yet rigorous, evaluation of foundational computer science principles. Paper 1 (Theory) maintained its traditional focus on precise technical vocabulary and exact systems operations, while Paper 2 (Programming Skills) heavily rewarded structural accuracy, algorithmic modularity, and clean array-record parsing. Overall, we rate this series as a 3.5 out of 5—challenging enough to distinguish top-tier students but fair in its standard question distributions.
Where the Marks Were Won and Lost
In Paper 1, the hardware, networking, and database components held the lion's share of high-tariff questions. The 4-mark question on touchscreen technologies (resistive vs. capacitive) and the SQL queries on relational databases were significant mark-earners. Conversely, many candidates struggled with assembly language tracing, particularly with indexed addressing \( \text{LDX} \), which is a common stumbling block. In Paper 2, candidates who mastered record types and multidimensional logic thrived, capturing a massive block of marks across the library-loan pseudocode tasks. Those who failed to use dot notation (e.g., Loan[Index].StudentID) or struggled with efficient search termination criteria dropped crucial points.
Common Examiner Pitfalls to Avoid
- SQL Cartesian Products: In database querying, always write the join criteria explicitly. Simply listing tables without equating their primary and foreign keys is an immediate mark-loser.
- Confusing Freeware with Free Software: "Free Software" is about user liberties (run, study, modify, share), not monetary value. Refrain from calling it "free of charge."
- Inexact Pseudocode Assignment: Do not use the equal sign for assignment in pseudocode. The standard Cambridge format strictly expects the left-arrow operator (<-).
Preparation Strategy & Predictive Advice
Looking ahead to future sittings, we observe that assembly language instruction sets were tested only lightly via a short trace table in this series. This suggests a strong likelihood of a high-tariff assembly design or programming question in upcoming papers. Furthermore, State-Transition Diagrams and Finite State Machines (FSMs) were entirely absent, making them overdue and prime targets for revision. Candidates should focus heavily on writing robust pseudocode from scratch, practice complex string slicing, and ensure they can confidently implement nested loop searches.