Computer Science

1. What makes Computer Science distinct?

A Computer Science EE must:

• investigate a computational problem, algorithm, model or system

• apply concepts such as efficiency, optimisation, complexity or data handling

• analyse performance or behaviour through testing, modelling or comparison

• use programming, simulations or theoretical reasoning appropriately

• focus on computational processes rather than digital design or user interface

• interpret results through computational thinking and algorithmic logic

• evaluate performance, limitations and suitability of the chosen approach

Your essay should feel computational, not mathematical, design based or purely technical.

2. What counts as a suitable Computer Science topic?

A strong topic is:

• focused on a specific algorithmic, data or computational issue

• supported by measurable performance variables such as time, memory or accuracy

• grounded in computer science theory or algorithmic reasoning

• feasible using school level programming resources

• able to generate clear, analysable results

Examples of suitable topic types

• comparison of pathfinding algorithms under controlled constraints

• analysis of compression techniques on different data types

• evaluation of hashing or searching algorithms

• simulation of network behaviour or traffic flow

• optimisation studies using heuristics or genetic algorithms

Examples of unsuitable topic types

• software design projects focused on interface rather than computation

• essays describing how to build an app without algorithmic analysis

• broad digital technology topics with no measurable computational component

• pure mathematics topics such as proofs or number theory

• projects requiring advanced hardware or large scale datasets

Computer Science requires computational analysis, not software tutorials.

3. Using data in Computer Science

Your essay must draw upon computational data, such as:

• execution time, memory usage or computational complexity measurements

• accuracy or error rates for machine learning models

• behavioural outputs from simulations

• performance metrics comparing algorithms

• results from controlled experiment runs

Avoid

• subjective user feedback with no computational measurement

• online code snippets used without understanding

• datasets with no clear origin or structure

• performance tests conducted without consistent conditions

Evidence must be interpreted using computational reasoning and algorithmic understanding.

4. Research methods appropriate for Computer Science

Your methods may include:

• writing and testing programs to implement specific algorithms

• running controlled experiments to measure time or space complexity

• modelling systems using simulations or agent based environments

• analysing datasets through appropriate computational methods

• comparing algorithmic approaches under defined constraints

• interpreting published algorithmic performance data

The method must directly support evaluation of the computational question.

5. Expectations for analysis

Computer Science analysis should:

• interpret performance data through algorithmic principles

• explain behaviour using concepts such as recursion, heuristics or data structures

• account for differences in efficiency or resource use

• use tables, graphs and annotated code meaningfully

• remain tightly focused on the computational process under investigation

Analysis must go beyond reporting numbers and show genuine computational insight.

6. Evaluation in Computer Science

Evaluation may include:

• limitations of code implementation or testing environment

• impact of hardware constraints on results

• sensitivity of algorithms to input size or data structure

• comparison of theoretical and actual complexity

• validity of assumptions made in simulations or models

• extent to which results generalise to other contexts

A strong Computer Science EE evaluates both the strengths and limitations of the computational approach.

7. Common pitfalls in Computer Science EEs

Avoid these issues, as they frequently lead to weak outcomes:

• topics dominated by software design rather than computation

• insufficient or inconsistent performance testing

• projects requiring specialised servers or unavailable hardware

• code borrowed from libraries without understanding

• essays that drift into mathematics, design or engineering

• absence of clear, measurable computational variables

• algorithms tested with datasets that are too small to be meaningful

8. Model research questions

Here are high quality examples of Computer Science research questions:

1. How do different pathfinding algorithms compare in efficiency when applied to large grid based maps?

2. To what extent does file type influence the effectiveness of lossless compression algorithms?

3. How accurately can a simple neural network classify handwritten digits with varying training set sizes?

4. How does input ordering affect the performance of quicksort compared with mergesort?

5. To what extent can an agent based simulation model pedestrian movement in confined spaces?

Each question is focused, computational and suited to school level investigation.