The Extended Essay is capped at 4,000 words and assessed on your research process as much as your conclusion, so the question has to be narrow enough to answer and specific enough to argue about. Thirty are listed here by subject group, each with the source you would answer it from and each framed so that evidence could change your mind.
Most Extended Essay lists give you topics. "Climate change." "Artificial intelligence." Those are fields, and a student who takes one into February finds there is no way to finish inside 4,000 words.
These thirty are questions, each narrow enough to answer, each with the source you would answer it from.
What makes an EE question answerable?
You can say in one sentence what evidence would change your mind.
That is the whole test, and it is the same test a supervisor will apply in your first reflection session. "To what extent is social media harmful to adolescents?" fails it — no obtainable evidence settles it, and 4,000 words will produce a tour of other people's arguments. "To what extent does self-reported daily screen time predict self-reported sleep duration among US 12th graders in the 2023 Monitoring the Future survey, controlling for sex and region?" passes, because the data exists and the analysis is reachable.
Two constraints the Extended Essay imposes that the EPQ does not:
It sits inside a subject. Your question must be answerable using that discipline's methods and will be marked against that subject's criteria. A psychology EE answered with historical methods scores badly however good the writing.
It is capped at 4,000 words, and examiners stop reading at the limit. That cap is the single most useful constraint available to you, because it rules out any question that would need a literature review longer than your word count.
The practical rule: your question should force you to adjudicate a disagreement, in one discipline, using evidence you can obtain this term.
Thirty questions, by subject group
Group 1 and 2: Studies in language and literature
- How does the English translation of one named work alter the narrator's reliability compared with the original? (both editions, close reading)
- To what extent has the vocabulary of UK newspaper coverage of immigration shifted between two defined decades? (newspaper archive databases)
- How do two film adaptations of the same novel resolve the same ambiguity differently, and what does each resolution commit the viewer to? (the texts themselves)
- To what extent does the use of the second person in one author's short fiction change the reader's moral position? (close reading, single author)
- How has the frequency of a named grammatical construction changed in published English since 1900? (Google Books Ngrams)
Group 3: Individuals and societies
- To what extent did the 2018 UK Sugar Levy change the sugar content of soft drinks rather than their price? (ONS, manufacturer reformulation data)
- Does minimum wage variation across US states correlate with teenage employment rates once regional income is controlled for? (IPUMS, Bureau of Labor Statistics)
- To what extent do countries with higher female labour participation show lower fertility rates after income is controlled for? (World Bank Open Data)
- Has turnout in UK local elections diverged from general election turnout since 2010, and in which kinds of constituency? (Electoral Commission)
- To what extent did areas with greater manufacturing employment loss shift their voting behaviour in a defined election? (ANES, county returns)
- How far does a country's press freedom score predict its vaccination uptake, controlling for income? (RSF index, WHO Global Health Observatory)
- To what extent has the attainment gap between free-school-meal and other pupils in England changed since 2019? (DfE statistics)
- Do coalition governments pass fewer bills per session than majority governments in comparable parliamentary systems? (national legislative records)
- How did one named local authority's housing policy change rental prices relative to comparable neighbours? (ONS rental indices)
- To what extent does betting shop density predict personal insolvency rates in England and Wales? (Gambling Commission, Insolvency Service)
Group 4: Sciences
- To what extent is the urban–rural gap in life expectancy in England widening since 2010? (ONS)
- Does air quality measured near school locations correlate with recorded childhood asthma admissions? (DEFRA air quality data, NHS England)
- Has the flowering time of one common species shifted measurably in your region across the recorded period? (Nature's Calendar)
- To what extent do urban and rural bird survey counts show diverging species diversity trends? (BTO Breeding Bird Survey)
- Does the caffeine content stated on drink labels match published independent measurements? (literature review plus label survey)
- How far did counties adopting earlier public health measures show different excess-mortality curves in 2020–21? (CDC WONDER)
- To what extent has HPV vaccination changed cervical cancer incidence in the UK cohorts old enough to show it? (ONS, published surveillance data)
- Does water hardness across supply zones correlate with published appliance-scaling data? (water company quality reports)
Group 5: Mathematics
- How well does a simple epidemic model reproduce the published case curve of one outbreak, and where does it break down? (published case data)
- To what extent does Benford's Law hold across three different published financial datasets, and what explains the departures? (company filings, public accounts)
- How does the choice of averaging method change the published ranking in one named index? (the index's own methodology and data)
Group 6: The arts, and interdisciplinary
- To what extent has the instrumentation of charting popular music changed across three defined decades? (chart data plus systematic listening)
- How do two national galleries' wall texts for the same period differ in what they attribute to the artist versus the context? (the texts themselves)
- To what extent does the readability of Wikipedia medical articles vary with the prevalence of the condition? (Wikipedia API, readability metrics)
- How far do open-source projects with codes of conduct show different contributor retention? (GitHub REST API)
Where the data lives
Everything above is answerable from free, public sources.
| Source | What it holds | Groups it serves |
|---|---|---|
| IPUMS | Census and survey microdata, US and international | 3 |
| ONS | UK population, prices, health, labour market | 3, 4 |
| World Bank Open Data | Cross-country development indicators | 3 |
| CDC WONDER | US mortality, natality, surveillance | 4 |
| Monitoring the Future | Annual US adolescent survey | 3 |
| Google Books Ngrams | Word frequency across centuries | 1, 2 |
| GitHub REST API | Repositories, contributors, timings | 5, 6 |
Our fuller list of free datasets covers twenty-two sources with notes on which need registration.
How to turn a topic into a question in one afternoon
Four steps, in order, and the first inverts the usual advice.
Find the source first. Pick something you can actually open, spend an hour in its documentation, and look for a variable pairing nobody seems to have checked. Narrow questions come from data, not from brainstorming.
Write the question with its variables named. If your question does not contain the specific things being compared, it is still a topic.
Name the obvious confounder. If you cannot think of one, you have not understood the pairing. If it plainly explains everything, pick a different pairing.
Check that someone disagrees. Find two sources taking different positions. That disagreement is where your analysis lives, and a question answerable from a single textbook has nowhere to go.
What to avoid
Anything needing a lab you do not have. Not through lack of skill — most universities bar under-eighteens from BSL-2, radioactive and animal work and require constant supervision plus paperwork filed weeks ahead. Change the question rather than waiting for a reply. Research without a lab covers the alternatives.
Questions answerable in a paragraph. "What is the structure of a REIT" is a fact, not an essay.
Questions with no obtainable evidence. Counterfactual history is enjoyable and unassessable.
Surveying your own year group without thinking it through. Ninety respondents rarely supports the claims students want to make from it, and it raises consent questions your school will have a policy about.
If you want to publish it afterwards
Decide now, because it constrains the question.
Most student research journals want original findings and a paper structure — introduction, methods, results, discussion. An Extended Essay is a sustained argument with a reflection component. Converting one means narrowing the question, adding a proper methods section and cutting the reflective material: real work, not reformatting.
If publication matters, pick a question from groups 3, 4 or 5 above with a dataset attached rather than a purely argumentative one. Then read the venue's rules first. The Journal of Emerging Investigators wants original findings and refuses to expedite for any reason. NHSJS and the Columbia Junior Science Journal both accept literature reviews, which is unusual and useful if your EE turns out to be a synthesis.
Where we fit
The Oxford Centre for Advanced Research pairs students with PhD candidates and early-career researchers at Oxford, Cambridge, other leading UK universities and Ivy League institutions. The Oxford Scholar Programme is £2,000 for ten contact hours over ten to fourteen weeks; the Oxford Publication Fellowship is £3,400 and is built around submission.
The single most valuable thing a supervisor does for an EE happens in the first reflection session: refusing the question you arrived with and making you narrow it. Students who get that right finish over the summer. Students who do not are still reading in March.
You can get that from your assigned supervisor, from a teacher with a doctorate, or from us. The order is deliberate — the first two are free and built into the Diploma.
Before you commit to a question
Write it down, then write underneath it the single sentence describing what evidence would change your answer.
If you cannot write that second sentence, you do not have a question yet. Ten minutes there saves a term.
Frequently asked questions
What makes a good Extended Essay question?
You can say in one sentence what evidence would change your answer, and the evidence is something you can reach. 'To what extent is social media harmful' fails both tests. 'To what extent does reported screen time predict reported sleep duration in the 2023 Monitoring the Future sample' passes both, because the dataset exists and the analysis is within reach.
How is the Extended Essay different from the EPQ?
The EE sits inside a subject and is marked against that subject's criteria, so your question has to be answerable using that discipline's methods. The EPQ is subject-agnostic and can be a product rather than an essay. The EE is also capped at 4,000 words and requires formal reflection sessions with a supervisor.
Can I write the EE in a subject I do not take?
Generally your EE should be in a subject you are studying, and schools usually require it. The flexibility is in the angle rather than the subject — a history EE can be about science, an economics EE about sport, a English EE about a text nobody else in your cohort has chosen.
Do I need original data?
No. The EE does not require it and a well-argued treatment of existing evidence scores fully against the criteria. It matters only if you later want to submit to a research journal, since most student journals want original findings — so if publication is a goal, pick a question with a dataset behind it.
How narrow is too narrow?
Rarely a problem. Almost every failing EE is too broad, not too narrow. A question whose answer is a single number is thin, but widening a narrow question takes an afternoon; narrowing a vast one usually means starting again in February.
Can an Extended Essay be published?
Usually not as written. Journals want original data and a paper structure; the EE is an extended argument with a reflection component. Converting one means narrowing the question, adding a methods section and cutting the reflective material — real work rather than a reformat. NHSJS and the Columbia Junior Science Journal do accept literature reviews, which widens the options.