Step 1 · Topic and pathway
Choose something technical you can investigate, not something you will build.
By the end of this step
- A narrowed topic, three or more academic sources, and a pathway. Bring them to your first reflection session.
What a CS extended essay is
A technical investigation: how and why a computational system, algorithm or technology behaves the way it does. Three kinds of essay work well:
| Kind | What you do | Example |
|---|---|---|
| Experiment / benchmark | Implement or use two or more techniques and measure them under controlled conditions | Huffman vs arithmetic coding on texts of different sizes |
| Dataset / model study | Train or run models on a public dataset and compare them on defined metrics | Kernel choice in support vector machines on handwritten digits |
| Structured secondary study | Compare published results using clear selection criteria and your own analysis | How well published benchmarks support claims about post-quantum key exchange in TLS |
The IB guide saysLiterature-based essays are just as valid as experiments, but only if they follow a clear, consistent method (e.g. comparative analysis or technical evaluation). A descriptive or journalistic overview is not enough.
Is it computer science?
CS essay, or better in another subject? Self-check drill
Social, ethical, historical and business topics belong in other subjects. Sort these.
From interest to research question
The topic funnel Animation
Every good RQ starts as a vague interest. Watch how each step narrows it.
Real life: big results from small questions Edsger Dijkstra designed his shortest-path algorithm in about twenty minutes, sitting in an Amsterdam café in 1956, to answer one narrow question: the shortest route between two Dutch cities. And Python's sort, Timsort (Tim Peters, 2002), came from noticing that real-world data is often partly sorted already, which is a testable observation. Java and Android later adopted it.
A good topic
- You're curious about it, and you already know some of the theory (or can learn it).
- Narrow enough to lead to one focused question. “Encryption” is a field, not a topic.
- Testable or analysable with what you have: a laptop, free software, public datasets, published papers.
- Has academic sources: peer-reviewed papers, conference proceedings, textbooks. You need them to justify why the question matters, not just for background.
- Not your IA, and not something you could only answer with equipment or data you can't get.
Research it Open Google Scholar and search your area plus “benchmark” or “comparison”. Find three papers from the last ten years and note what each one measured. Where they disagree, or something hasn't been tested, there's a research question.
Idea bank40 CS EE ideas by syllabus topic and kind of essay, each with a sample RQ and metrics.I told my supervisor my EE topic was “computers”.
She said, “Lovely. Which one?”
Choose a pathway
| Subject-focused | Interdisciplinary | |
|---|---|---|
| What | Computer science only | Computer science and another DP subject, both visible in the essay |
| Choose it when | The question is answered by CS methods and theory | The second subject genuinely changes how you can answer the question |
| Registered as | Computer science | Two subjects + one of five frameworks |
| Marked with | The same criteria A–E | The same criteria A–E, applied to both subjects |
Which pathway suits you? Quick quiz
Three questions about your idea.
Most CS essays are subject-focused. Thinking of pairing CS with maths, physics, biology or economics? Read Interdisciplinary EE first.