OCR A-Level · Analysis & Prediction

2023 OCR A-Level Computer Science - H446 Past Paper Analysis

The 2023 OCR A-Level Computer Science series represents a robust and programming-heavy iteration, testing core concepts through highly applied scenarios. Object-Oriented Programming (OOP) and advanced Data Structures like Trees, Stacks, and Linked Lists dominated H446/02, while H446/01 featured a tough mix of Systems Software, Web Technologies, and legal/ethical evaluations (featuring RIPA 2000). Success relied heavily on code-writing precision, formal trace methods, and structured evaluative prose.

Updated: 14 Jun 2026

Analysis Sample paper

Difficulty 3.8/5

Total marks

280

Duration

300 mins

Most tested

Programming techniques

Paper breakdown

H446/01: Computer Systems: 140 marks · 150 minsH446/02: Algorithms and Programming: 140 marks · 150 mins

Top chapters

Programming techniques · 59 marksData Structures · 54 marksAlgorithms · 28 marks

Overall Difficulty Verdict

The 2023 OCR A-Level Computer Science (H446) papers presented a moderately high difficulty level (4/5 stars), characterized by a heavy shift towards applied algorithmic design and robust object-oriented programming. Rather than relying on simple recall, the examiners required candidates to perform extensive technical dry-runs, write lengthy class definitions, and critically evaluate legal and technical trade-offs in essays.

Where the Marks Were Won and Lost

The vast majority of marks on this series were concentrated in two massive pillars: Programming Techniques and Data Structures. Together, these accounted for over 110 of the 280 available marks. In H446/02, the final multi-part scenario on Object-Oriented Programming (representing a text-based treasure hunt game) single-handedly carried 40 marks, testing encapsulation, subclass methods, grid initialization, and parameter passing. Candidates who mastered class structures and constructor logic secured high grades. Conversely, marks were heavily lost on floating-point conversions (especially negative normalized numbers) and complex dry-runs like Dijkstra's shortest path algorithm, where failure to demonstrate state-changing overrides in the trace table proved costly.

Examiner Pitfalls & Candidate Misconceptions

Examiners highlighted several persistent mistakes across both papers:

Case Sensitivity and Syntax: In the doCheck function call, many candidates lost marks for writing the function name incorrectly (e.g., lowercase 'c') or omitting quotation marks around the filename in file-handling statements.
The RIPA 2000 Snooper's Charter: In the 12-mark evaluation of the Regulation of Investigatory Powers Act, lower-band responses merely listed surveillance tools without analyzing the tension between national security and privacy, or falsely accused local councils of overreaching without referencing the actual legislative provisions.
Hashing vs. Encryption: Many candidates still treat hashing as a reversible cryptographic technique, failing to explain that hashing is a one-way mathematical function designed for verification, whereas encryption is two-way and requires key distribution.

Strategic Preparation Guide

To dominate future OCR exams, students must prioritize active coding practice over passive note-taking. You should treat the OCR Pseudocode Guide as a primary language—practice translating list traversals, file handling, and nested loops under timed conditions. When tackling Karnaugh maps and logic gates, always show intermediate Boolean steps to secure method marks even if your final simplified expression contains a slip. For the 9- and 12-mark essay questions, adopt a structured three-tier approach: define the technical terms (AO1), apply them directly to the scenario (AO2), and conclude with a justified, balanced evaluation (AO3).

Predictions for the Next Cycle

With Structure and Function of the Processor dropping to zero marks in this series, we predict a massive comeback of CPU-centric questions (including pipelining, CISC vs RISC, and GPU applications) in the upcoming paper. Furthermore, database normalization (2NF and 3NF) and complex SQL joins were highly under-represented, making them prime candidates for high-tariff structured questions in the next examination series, where comparing the efficiency of Merge Sort \(O(n \log n)\) with Bubble Sort \(O(n^2)\) remains a core theme.

Charts

Marks by chapter

See where the marks were concentrated so revision time goes to the highest-value topics.

Question type mix

Compare the mark share of each paper section and question type.

short_answer_and_structured175 marks · 48 questions
programming_and_algorithms48 marks · 10 questions
extended_prose_and_evaluation57 marks · 6 questions

Mark accessibility

Estimate which marks were basic, mid-level, or high-difficulty.

79%within easy or medium reach

135

easy: 85 marksmedium: 135 markshard: 60 marks

Difficulty trend

Compare difficulty across recent years.

Command word frequency

Spot common command words so answers match the expected response style.

Skill weighting

Shows the skill mix this paper tested most heavily.

Study ROI

Bigger bubbles recur more often; higher bubbles carry more marks, helping you rank revision priorities.

Programming techniques

59 marks · Difficulty 3 · recurrence 5%

Data Structures

54 marks · Difficulty 3.8 · recurrence 5%

Systems Software

22 marks · Difficulty 2.5 · recurrence 4%

Computing related legislation

15 marks · Difficulty 2 · recurrence 4%

Algorithms

28 marks · Difficulty 4.2 · recurrence 5%

Boolean Algebra

9 marks · Difficulty 1.8 · recurrence 3%

Time vs marks

Compare marks with suggested time allocation to plan exam pacing.

marksmins

Next-year prediction

Topics worth watching next year, with the reason shown directly below each bar.

Structure and function of the processor95%

This core topic dropped from 16 marks in 2022 to 0 marks in this series, leaving a massive gap in the CPU architecture assessment (such as the FDE cycle, register manipulation, pipelining, and CISC vs RISC structures) that is highly likely to be heavily tested in the next cycle.

Databases85%

Only 6 marks were awarded in this series, focusing on a basic delete query and 1NF. 2NF/3NF database normalization, transactional history, and complex SQL joins are highly overdue for a major, higher-tariff question.

Moral and ethical Issues80%

Replaced almost entirely by RIPA (legislation) in this paper, moral/ethical impacts of computer science (such as automated decision making, ethical AI, or environmental concerns) are highly likely to return as a 9 or 12-mark essay.

Cumulative marks ladder

The line is your running mark total question by question; dashed lines are the estimated grade cut-offs. See which question the line crosses your target grade at, so you know how far you must answer cleanly and which questions decide a band.

Topic-year heatmap

Compare topic weight by year to spot recurring and returning areas.

Jun 2022

Jun 2023

Programming techniques

Algorithms

Data Structures

Data Types

Structure and function of the processor

Databases

Software Development

Types of processor

Applications Generation

Moral and ethical Issues

Examiner insights

Official report

Common pitfalls

In programming/pseudocode, failing to match case sensitivity (e.g., doCheck is case-sensitive) and neglecting to put speech marks around the file name "storedvalues.txt".
In normalisation, confusing atomic attributes with unique attributes or mistaking data redundancy across different entities as 1NF violations within a single unnormalised table.
In essay questions, writing generic benefits like "faster" without explaining why or how the speed increases (e.g., multi-core or parallel execution).

Misconceptions

Believing that the BIOS is stored in RAM (it is read-only or flash memory).
Confusing hashing with encryption, specifically assuming that hashing is a two-way reversible process or that a private key is used to decrypt a hash value.
Thinking that a multi-core processor will always run any program faster, ignoring the dependence of instructions or the need for software to be explicitly written for parallel processing.

Where marks are lost

On the Karnaugh Map, failing to correctly loop/highlight the groups, or making careless errors in joining the simplified terms with an OR / "+" operator.
On Dijkstra's algorithm, neglecting to show the overwriting of temporary distances (e.g., replacing 19 with 14 at node G) in the working table, leading to lost marks even if the final path was correct.
On the floating-point normalized representation, failing to shift the binary point correctly or forgetting that the negative sign in two's complement requires a leading '1' for the mantissa.

Know what the examiners want — now go get those marks

Thinka is an AI practice app used by 100,000+ DSE students to drill questions and get instant feedback on every answer. You already know the high-frequency topics and where marks are lost — now practice those specifically on Thinka and turn this analysis into actual marks.

Start Practising Free See pricing

Know what's tested? Drill those topics on Thinka — AI questions, instant marking.

Start Practising Free