Difficulty Verdict

This exam sits at a moderate-to-high difficulty tier (4 stars out of 5). While the system fundamentals and networking questions offered direct, syllabus-standard marks, Section B's algorithmic and program-construction tasks required impeccable logical precision. In particular, the two-pass array algorithm in Paper 1 and the Selection Sort implementation in Paper 2 Option D acted as major differentiators between grade 6 and 7 candidates.

Where the Marks Are

The bulk of the marks are concentrated in Object-Oriented Programming (45 marks) and Computational Thinking (21 marks). Mastering pseudocode mechanics, loop structures, and conditional logic alone accounts for nearly half of the total combined marks. In Paper 1, the trace table (Q5) and the double-loop pseudocode algorithm (Q11c) offered a substantial 13 marks, highlighting that runtime evaluation and design are paramount.

Examiner Pitfalls & Lost Marks

  • The Two-Loop Trap: On Q11c, weaker candidates incorrectly attempted to compute the class average and award grades in a single loop, leading to immediate logical failure since the average \( (\text{Sum} / 200) \) must be finalized before individual deviations can be assessed.
  • Logic Diagram Precedence: In Q4, students frequently placed the NOT gates after the AND/OR gates instead of before them, fundamentally changing the expression to \( \text{not } (B \text{ and } C) \) instead of \( \text{not } B \text{ and } not } C \).
  • Sorting Boundaries: In Paper 2 Q12b, many candidates wrote generic sorting templates instead of tailoring the selection sort to extract the five most expensive items, often neglecting the correct boundaries or missing the swap implementation details.

Success Strategy & Predictions

To secure top marks in future sessions, candidates must practice dry-running parallel arrays and writing robust, index-safe pseudocode. Additionally, pay close attention to precision command verbs: 'Explain' and 'Describe' require deep, systemic links rather than simple definitions. Looking ahead, since this paper focused heavily on parallel 1D arrays, we predict Two-Dimensional Arrays and reconstructive trace tables with nested logic are highly overdue and likely to form the core of the upcoming session's major algorithm design questions.