Let's Meet John
John isn't particularly passionate about academic research. He plays video games, hangs out with friends, and has absolutely no desire to write his Extended Essay. He's already planning to get an AI to do it the night before his first draft is due.
John takes Maths AI HL and Psychology HL. One Sunday, he was debating with his friends about whether they would have voted for Trump or Harris in the 2024 US elections. He noticed patterns in how different people reasoned through their vote — and got curious. Could voter psychology be predicted? Could you quantify the influence of social media on the way his friends thought about politics?
He wondered if he could draw statistical conclusions between social media sentiment and voter behaviour. He noticed that the 2024 elections were heavily influenced by online commentators — and he started asking: to what extent were his friends's political opinions actually theirs?
John's Final Research Question
"To what extent can quantitative analysis of social media sentiment and engagement metrics explain variations in voter behaviour during the 2024 US elections?"
Notice how it's specific, analytical ("to what extent"), researchable with publicly available data, and connects directly to his subjects (Maths AI + Psychology).
John went from "I don't want to do this" to having a research question he was genuinely curious about — because it started with a real conversation he was already having. That's the process. Your EE should come from something you actually think about.
The Anatomy of a Strong Research Question
Every strong EE research question shares four properties. Use these as your checklist before you commit to yours.
Specific
Answerable within 4,000 words. If your RQ could be a PhD thesis, it's too broad. Narrowness is not a weakness — it's precision.
Analytical
Favours "to what extent" or "how" over "what." Allows for evaluation and argument, not just description of facts.
Researchable
You can actually find data, evidence, and sources to answer it. No dead-end topics where the data simply doesn't exist.
Original
Leaves room to present original analysis — not just a summary of existing work. You should be saying something, not just reporting what others have said.
Weak vs. Strong Research Questions
"How does social media affect politics?"
"To what extent can quantitative analysis of social media sentiment and engagement metrics explain variations in voter behaviour during the 2024 US elections?"
"What is ZARA's business strategy?"
"To what extent does ZARA's supply chain model constitute the primary source of its competitive dominance in the fast fashion retail market?"
"How does music affect the brain?"
"To what extent does exposure to classical music during study sessions improve performance on spatial reasoning tasks in adolescents aged 15–18?"
The Interdisciplinary Question
John's RQ works because it bridges two subjects: quantitative methods from Maths AI and behavioural theory from Psychology. This is called an interdisciplinary approach — and when it's well-executed, it can produce genuinely original analysis that subject-specific essays can't.
Interdisciplinary EEs are higher risk. You're essentially writing two analyses and integrating them — which requires significantly more methodological rigour. Only go interdisciplinary if you're strong in both subjects and have a clear reason why one discipline alone can't answer your RQ.
Finding Your Niche
Creativity is what turns a good EE into a great one. A unique angle on a common topic is worth more than a common angle on a unique topic. Examiners reward genuine intellectual curiosity — they can tell when someone actually cares about what they're researching.
Don't use AI to generate your research question from scratch. AI is not creative — it recycles and regurgitates. Your EE is a reflection of your intellectual curiosity. Let it be yours from the start. Once you have YOUR draft RQ, AI can help you stress-test it (see Module 13 for prompts).
The RQ Development Process
Draft your RQ
Start with your interests and academic strengths from Module 3. Write 3–5 possible RQs without censoring yourself. You'll narrow later.
Take it to your supervisor
Your supervisor can help you focus and refine. We consistently underestimate how focused an RQ needs to be — supervisors know the rubric.
Stress-test it
Can it be answered in 4,000 words? Is there sufficient data publicly available? Does it require analysis or just description? Would a counterargument be possible?
Lock it in after your EE Dump
After completing your EE Dump (Module 5), make final adjustments. Then freeze it. After that point — no more changes to your RQ.
Your research question is your contract with the examiner. Every word of your essay is a fulfilment of that contract.
Key Takeaways
- Your RQ must be specific, analytical, researchable, and original
- Use "to what extent" or "how" — avoid "what" questions which invite description
- Take your draft to your supervisor immediately — they know what gets marks
- Creativity in your angle separates good from great
- Lock in your final RQ after the EE Dump — no more changes after that