Resource Guide

Research Project Ideas for High School Students

Twenty-four worked questions across six fields. Each one names the method that makes it feasible, the source the evidence comes from, and the mistake that most often sinks it.

Worked questions

Research methods

Subject guides

Need a lab

How to Use These Ideas

These are not finished project titles. They are starting questions, the kind that should change substantially once you begin reading. That change is the point. A research project that ends exactly where it began probably did not involve much enquiry.

Most project ideas fail for one reason: the student has no realistic way to answer them. So each question below is tagged with the method that makes it feasible and the source the material actually comes from. If you cannot name the evidence you would consult, you do not yet have a project, however interesting the question sounds.

Each entry also names the mistake that most often sinks it. Those are not hypothetical. They are the failure modes that show up when a student has done the reading but not yet worked out what the reading obliges them to conclude.

Start Here

Choose a method before a topic

The method is the decision that determines whether a project is finishable. The topic is secondary, and choosing it first is the most common way students end up with a question they cannot answer. Every question on this page is tagged with one of these six.

Literature review

Read six to ten papers on one narrow question and synthesise what they collectively show, including where they disagree. The closest thing to what academic researchers actually spend their time doing.

Google Scholar, a reading log, and a question narrow enough that ten papers can answer it.

Public dataset

Analyse data someone else collected and published. The skill on display is careful handling of real numbers and honesty about what they cannot show, not the collection itself.

A spreadsheet, one dataset, and every chart made by you rather than copied.

Policy analysis

Take one policy or intervention and evaluate it against its own stated goals, using published evaluations. Sits between evidence and politics, which suits applicants to law, economics, PPE and public health.

Government or NGO evaluation reports, which are free and public.

Ethics argument

Defend a position on a genuine moral problem, engaging the strongest counterargument rather than a convenient one. Undervalued by students and highly valued at interview, because it is where reasoning shows.

Core reading on the problem and a willingness to be disagreed with.

Archival research

Work with primary material: records, correspondence, parliamentary debates, contemporary press. Slower than it looks, and the only route that can produce a genuinely new observation at school level.

A digitised archive and a narrow window of time or place. Breadth is the enemy here.

Close reading

Build an argument about a text, building or work from its own formal features. The discipline is evidential: every claim has to point back to something on the page.

One text you can reread until you notice something that is not in the criticism already.

Medicine & Life Sciences

Explore this field →

Medical research at school level is almost always literature-based or data-based, which is appropriate rather than a compromise. Original clinical research needs ethical approval, a qualified investigator and clinical responsibility, none of which a school student has. What is open to you is the work that surrounds a trial rather than the trial itself.

Literature review Intermediate

Does sleep deprivation in adolescence affect long-term cognitive development?

The evidence genuinely conflicts, and the conflict is methodological rather than political, which makes it a clean introduction to weighing study quality.

What you produce: A 4,000-word review of eight to ten studies ending in a verdict on what the evidence can and cannot support.
Where the material is: PubMed Central and Google Scholar. The longitudinal cohort studies are the ones that carry the argument; the cross-sectional ones mostly do not.
Where it goes wrong: Almost every study here is observational. A review that treats correlation as settled causation fails on the first question an interviewer asks.

Ethics argument Intermediate

What are the ethical limits of genetic screening in prenatal medicine?

It forces the empirical question of what screening can actually detect to sit alongside the moral question of what should follow from detecting it.

What you produce: A 3,000-word position paper built on one real screening case, such as late-onset conditions, that states a principle and answers the strongest objection to it.
Where the material is: The Nuffield Council on Bioethics publishes its full reports free, and they are the standard reference point in UK practice.
Where it goes wrong: Surveying four ethical frameworks and choosing none. A position paper has to land somewhere and defend the landing.

Literature review Advanced

How does behavioural treatment compare with medication for adolescent depression?

Mental health research is where methodological disagreement is sharpest, because the outcome being measured is itself contested.

What you produce: A comparative review of the trial evidence with each trial assessed against a standard quality checklist rather than taken at face value.
Where the material is: The Cochrane Library, free to read in the UK under a national licence, is the cleanest starting point.
Where it goes wrong: Comparing effect sizes across trials that measured depression with different instruments, as though the numbers were interchangeable.

Policy analysis Advanced

What does antimicrobial resistance reveal about the limits of the antibiotic development model?

The causes are simultaneously biological, economic and regulatory, so the question rewards systems thinking that single-discipline projects cannot show.

What you produce: A 3,500-word analysis connecting resistance mechanisms to the economics of drug development, ending in one assessed policy proposal.
Where the material is: The O'Neill Review on Antimicrobial Resistance and the WHO priority pathogens list.
Where it goes wrong: Writing the biology and the economics as two essays stapled together. The argument has to be that one explains the other.

Artificial Intelligence & Technology

Explore this field →

AI projects work when they engage a specific, bounded question and fail when they generalise about the technology as a whole. The field moves quickly, which is an advantage: the disagreements are live, documented in public, and not yet settled into textbook consensus.

Literature review Advanced

What does fairness mean in machine learning, and can the competing definitions be reconciled?

Several formal definitions of fairness are provably incompatible with one another, so the question has a mathematical spine rather than only an ethical one.

What you produce: A paper working through three formal definitions, applied to one real deployed system, arguing for which should govern it.
Where the material is: The ProPublica investigation into the COMPAS recidivism tool and the technical replies to it are the most thoroughly documented case in the field.
Where it goes wrong: Treating this as a pure ethics essay. The impossibility results are formal, and a paper that avoids them has avoided the question.

Literature review Intermediate

What can large language models actually do, and where does apparent competence break down?

Public discussion collapses several distinct capabilities into one, and separating them is genuinely useful analytical work.

What you produce: A review setting published benchmark claims against documented failure cases, with a defended position on what the systems can be said to do.
Where the material is: arXiv for the technical papers, plus published benchmark results and the critiques of how those benchmarks are constructed.
Where it goes wrong: Using your own chat transcripts as evidence. Anecdotes about a model are not measurements of it, and an interviewer will say so.

Policy analysis Advanced

Should autonomous weapons systems be permitted under international humanitarian law?

There is an existing legal test to argue against, which turns a broad ethical worry into a tractable legal question.

What you produce: A legal analysis against the established requirements of distinction and proportionality, ending in a stated position.
Where the material is: The UN Group of Governmental Experts under the Convention on Certain Conventional Weapons publishes its full session reports.
Where it goes wrong: Arguing about machines in the abstract instead of against the specific legal tests the law already applies.

Public dataset Intermediate

How have recommendation algorithms changed the economics of the music industry?

Music is the best-documented creative industry of the streaming era, so the claims can be checked against published revenue data rather than asserted.

What you produce: A data-backed essay with your own charts on how streaming-era revenue is distributed across artists.
Where the material is: IFPI Global Music Report headline figures and the annual Luminate year-end reports.
Where it goes wrong: Confusing a change in how revenue is measured with a change in how it is earned. Reporting methods changed too.

Economics & Policy

Explore this field →

Economics projects are strongest when empirically grounded rather than theoretical. The subject has an unusual advantage for school-level research: a great deal of high-quality data is free, and the major disagreements are about identification and method, which is exactly what an admissions tutor wants to hear a student reason about.

Literature review Advanced

Does raising the minimum wage reduce employment among low-paid workers?

It is the most instructive methodological argument in applied economics, and following it is a genuine education in how economists establish causation.

What you produce: A review focused on why well-designed studies reach opposite conclusions, ending in a position on what the balance of evidence supports.
Where the material is: Card and Krueger's 1994 New Jersey study and the replies to it are where this argument actually lives.
Where it goes wrong: Counting papers on each side. The question is which research designs identify the effect, not how many authors agree.

Public dataset Intermediate

What were the measurable economic effects of Brexit on one specific sector?

Narrowing to a single sector converts an unwinnable political argument into a tractable empirical one.

What you produce: A focused analysis of one sector using published trade data, with your own charts and a stated conclusion.
Where the material is: HMRC trade statistics and the Office for Budget Responsibility's published assessments.
Where it goes wrong: Attributing everything after 2020 to Brexit when the pandemic sits in the same window. The project has to address that directly.

Policy analysis Advanced

What makes some carbon pricing schemes more effective than others?

Dozens of jurisdictions have implemented carbon pricing with different designs, which creates genuine comparative evidence rather than speculation.

What you produce: A structured comparison of two or three schemes, identifying which design features the evidence suggests actually matter.
Where the material is: The World Bank's Carbon Pricing Dashboard tracks every scheme in force and its price.
Where it goes wrong: Comparing headline carbon prices without accounting for how much of each economy the scheme actually covers.

Public dataset Intermediate

Why do cities become unaffordable, and what does housing supply have to do with it?

The subject is directly visible in students' own lives, and the data to test the competing explanations is published and free.

What you produce: An analysis of one housing market pairing price data with planning and completion data, with a policy-oriented conclusion.
Where the material is: ONS house price and affordability tables, alongside published planning and housing completion statistics.
Where it goes wrong: Treating supply as the whole story. Credit conditions and demand move prices too, and a good project says so.

Law & Politics

Explore this field →

Law and politics reward precision and primary sources. This is the field where the raw material is most completely free: judgments, statutes, parliamentary debates and treaty texts are all published in full, which means a school student can work from exactly the same documents an academic would.

Policy analysis Advanced

Does the European Convention on Human Rights adequately protect privacy against state surveillance?

There is now a substantial body of judgments testing surveillance powers against Article 8, so the question can be argued from decided cases rather than principle alone.

What you produce: A legal analysis of the standards the case law has established, and an assessment of whether they hold up against current surveillance capability.
Where the material is: HUDOC, the European Court of Human Rights' own case database, is free and full-text. Big Brother Watch and Others v United Kingdom is the anchor case.
Where it goes wrong: Summarising judgments in sequence. The essay needs a standard of its own to measure them against.

Archival research Advanced

What are the constitutional limits of executive power, and have they been exceeded?

Questions that were academic a decade ago have become live constitutional disputes with a full documentary record on both sides of the Atlantic.

What you produce: A constitutional analysis of one specific episode, working from primary sources, arguing a clear position on legality or legitimacy.
Where the material is: For the UK, the Miller prorogation judgment and Hansard. For the US, the Federal Register, which publishes every executive order in full.
Where it goes wrong: Conflating what is unlawful with what is merely unwise. The distinction is the entire argument.

Policy analysis Advanced

How should international law handle state responsibility for climate change?

Climate litigation is developing fast enough that the open questions are genuinely open, and a well-argued student position is not competing with settled doctrine.

What you produce: A legal research essay on existing frameworks and an argument for how accountability for climate harm should work.
Where the material is: The Sabin Center's Climate Change Litigation Databases catalogue cases worldwide, including the filings.
Where it goes wrong: Arguing that something should be actionable without engaging with why attribution and causation have been the obstacle.

Policy analysis Intermediate

How should liberal democracies regulate disinformation without undermining free expression?

Two major regulatory regimes are now in force with different designs, which turns a theoretical tension into a comparison of real texts.

What you produce: A policy essay comparing two regimes on specific provisions and arguing for an approach on principled and practical grounds.
Where the material is: The EU Digital Services Act and the UK Online Safety Act 2023 are both published in full and can be compared clause by clause.
Where it goes wrong: Restating the free speech dilemma in the abstract. The value is in what the actual provisions do differently.

Humanities

Explore this field →

Humanities projects make a specific claim and defend it with evidence from the text or the record. The common failure is not lack of ambition but lack of a claim: a project that describes a work thoroughly and argues nothing has not yet begun. Narrow the object, then say something contestable about it.

Close reading Intermediate

What do a novel's formal choices reveal about the assumptions of the period that produced it?

It puts formal analysis and historical context in tension, which is where the interesting disagreements in literary criticism actually happen.

What you produce: A critical essay arguing that specific formal features, not plot events, encode or resist the assumptions of their moment.
Where the material is: The text itself, reread, plus the critical tradition around it to establish what has already been claimed.
Where it goes wrong: Reading the plot for politics. The claim has to rest on form: structure, narration, and what the book cannot bring itself to say.

Archival research Advanced

How much of a specific historical famine was caused by policy rather than by nature?

The scholarly consensus has moved decisively on this, and testing that consensus against one case is real historical work.

What you produce: A research essay weighing environmental against political causation for one famine, defending a position on their relative weight.
Where the material is: Amartya Sen's Poverty and Famines is the argument to test. The National Archives holds much of the administrative record for British-era famines.
Where it goes wrong: Choosing a famine so large that the essay becomes a summary. One region and one year produces a better project than a country and a decade.

Ethics argument Advanced

Is there a coherent case for free will given what neuroscience has found?

The philosophical and empirical literatures genuinely engage each other here, and neither has won, which leaves room for a defended position.

What you produce: An analytical essay working through compatibilist and incompatibilist positions and arriving at a defended conclusion.
Where the material is: Libet's readiness-potential experiments and the four decades of methodological objections to them.
Where it goes wrong: Taking the neuroscience at face value. The experimental design has been contested for as long as the results have been cited.

Close reading Advanced

What is the significance of an undertranslated text within its own intellectual tradition?

For a student with a second language, this is the rare school-level project that can reach material genuinely under-examined in English.

What you produce: A close analytical essay working from the original text, situating it in its tradition and arguing for its significance.
Where the material is: The primary text in its original language, plus scholarship in that language where English criticism is thin.
Where it goes wrong: Claiming neglect without checking. Establish first that the English-language gap is real, because an examiner may know otherwise.

Arts & Architecture

Explore this field →

Arts and architecture projects combine formal analysis with historical context. The strongest questions ask what a building or a work assumed about the people who would use it, and what happened when that assumption met reality. Buildings are unusually good evidence because their failures are public and documented.

Close reading Intermediate

How did one artistic movement respond to the social conditions of its moment?

Formalist and contextual readings routinely reach different conclusions about the same works, so there is a real disagreement to enter.

What you produce: A critical essay analysing a specific group of works through both lenses and arguing for how the relationship should be understood.
Where the material is: The works themselves, exhibition catalogues, and the movement's own manifestos where they exist.
Where it goes wrong: Letting context do all the work. If the argument would survive with the images removed, it is a history essay, not an art one.

Archival research Intermediate

What did modernist housing assume about human behaviour, and what happened when it was built?

The gap between architectural intention and social outcome is documented in unusual detail, including the ways the standard story has since been corrected.

What you produce: An architectural history essay on one project, drawing on both design analysis and social history, arguing for where the gap actually opened.
Where the material is: The RIBA archive for UK schemes. For Pruitt-Igoe, the demolition record and the later scholarship correcting the popular account.
Where it goes wrong: Repeating the received story that modernist design alone caused these schemes to fail. Funding, maintenance and policy are large parts of the record.

Close reading Intermediate

How do curatorial decisions shape the meaning of the works a museum displays?

It is a question a student can research in person, by visiting, which is rare and produces observations that are genuinely their own.

What you produce: A critical analysis of one exhibition, examining specific decisions about sequence, adjacency and labelling, and their effect on interpretation.
Where the material is: A single exhibition visited more than once, its catalogue, and its wall texts recorded verbatim.
Where it goes wrong: Generalising about museums. The evidence is one room, in one order, with these labels.

Policy analysis Advanced

What does the law's treatment of sampling reveal about musical authorship?

Copyright decisions on sampling have forced courts to state, on the record, what they think musical originality consists of.

What you produce: An essay combining legal and musicological analysis, arguing what the case law assumes about creativity.
Where the material is: The leading sampling judgments, including Campbell v Acuff-Rose and Bridgeport Music v Dimension Films, are published in full.
Where it goes wrong: Treating the cases as a timeline. The argument is about what the reasoning assumes, not the order the rulings arrived in.

Worked Examples

What the finished work actually looks like

An idea is not a project. Before choosing a question, it helps to see the shape of the thing you are trying to produce. These are the three formats most high school research projects take, with what makes each one hold up under questioning.

Format 1

The critical literature review

A question like the sleep-deprivation one above, answered by reading systematically rather than experimenting. Typically 3,500 to 5,000 words: what the evidence shows, where studies genuinely disagree, and a defended position on what can be concluded.

What makes it defensible: the student can name the two or three studies that carry the argument, and say why one is stronger than another. That is the exact conversation an interviewer starts.

Format 2

The public-data analysis

A question answered with an open dataset: ONS releases, World Bank indicators, a published survey. Typically 2,500 to 4,000 words with the student's own charts, honest about method limits. The common failure is claiming causation from correlation; the strong version names that limit itself.

What makes it defensible: every figure in it was made, not copied, and the student can explain why the method fits the question, not just what the result was.

Format 3

The policy brief

A question in politics, law, economics or public health, written for a decision-maker rather than an examiner: 1,500 to 2,500 words plus a short presentation. Brevity is the discipline; the student must decide what the evidence obliges them to leave out.

What makes it defensible: a specific recommendation with a stated trade-off. "It depends" is the draft; the finished brief takes a position and prices it.

Whichever format fits, the guides on writing a literature review, when to start a research project and how to publish research as a high school student cover the mechanics end to end.

Subject Library

Twenty-four subject guides

Each guide narrows the questions above to one subject, with methods and reading specific to it. All twenty-four now carry the same treatment: twelve questions each, every one labelled by method, difficulty and the output it should produce.

Computing, Engineering & Quantitative Fields

Medicine, Life Sciences & Behaviour

Society, Economics, Law & Global Affairs

Humanities, Media & Creative Fields

Free PDF

Take the 50 strongest ideas with you

A free PDF with fifty of these questions tagged by method and difficulty, the output each should produce, and the feasibility checklist we apply before recommending any project. We email it to you, along with occasional guidance worth having. No obligation, and your details are not passed to anyone.

Related Guides

Guides on developing your research question, writing up your findings, and getting started with a project.

Frequently Asked Questions

How do I choose the right research project topic?

Choose the method before the topic. Most project ideas fail not because the question is uninteresting but because the student has no realistic way to answer it. Once you know whether you are reviewing literature, analysing a public dataset, evaluating a policy, or arguing an ethical position, the field of possible questions narrows to the ones you can actually finish. From there, pick the question you genuinely want the answer to.

Do I need lab access or special equipment for a high school research project?

No, and most strong projects do not use one. Literature reviews, public-data analyses, policy evaluations and argued essays all produce serious work with nothing but a browser and a reading log. Original laboratory or clinical research is rarely feasible at school level and is not what admissions tutors expect to see.

How long should a high school research project take?

A serious research project typically takes eight to twelve weeks of sustained work, including reading, analysis, writing, and revision. The total time commitment is usually six to ten hours per week. Shorter projects are possible but limit the depth achievable. Rushing a project to meet a deadline generally produces work that does not reflect the student's capability.

Can I use a research project for both my UCAS personal statement and US college essays?

Yes. A well-executed research project provides material for both UK and US applications. For UCAS, it demonstrates super-curricular depth and gives specific intellectual content for the personal statement. For US applications, it provides evidence of intellectual curiosity and sustained independent work that strengthens the essay and activities list.

What makes a research question too broad?

If you cannot name the specific body of evidence you would consult, the question is too broad. "How do we cure cancer" names no evidence. "What does the trial evidence say about screening intervals for one cancer in one age group" names exactly what you would read. The narrower question almost always produces the more impressive project, because it can be finished and defended.

ScholarBridge

Ready to start your research project?

Apply to ScholarBridge

Now placing students for the October start. Places in each field are limited by mentor availability. We review applications on a rolling basis.

ScholarBridge matches students with doctoral-level or equivalent research mentors across six academic fields. Every project is student-led and completed to a standard the student can stand behind in any university interview.

Explore all programmes

Research Project Ideas for High School Students

How to Use These Ideas

Choose a method before a topic

Medicine & Life Sciences

Does sleep deprivation in adolescence affect long-term cognitive development?

What are the ethical limits of genetic screening in prenatal medicine?

How does behavioural treatment compare with medication for adolescent depression?

What does antimicrobial resistance reveal about the limits of the antibiotic development model?

Artificial Intelligence & Technology

What does fairness mean in machine learning, and can the competing definitions be reconciled?

What can large language models actually do, and where does apparent competence break down?

Should autonomous weapons systems be permitted under international humanitarian law?

How have recommendation algorithms changed the economics of the music industry?

Economics & Policy

Does raising the minimum wage reduce employment among low-paid workers?

What were the measurable economic effects of Brexit on one specific sector?

What makes some carbon pricing schemes more effective than others?

Why do cities become unaffordable, and what does housing supply have to do with it?

Law & Politics

Does the European Convention on Human Rights adequately protect privacy against state surveillance?

What are the constitutional limits of executive power, and have they been exceeded?

How should international law handle state responsibility for climate change?

How should liberal democracies regulate disinformation without undermining free expression?

Humanities

What do a novel's formal choices reveal about the assumptions of the period that produced it?

How much of a specific historical famine was caused by policy rather than by nature?

Is there a coherent case for free will given what neuroscience has found?

What is the significance of an undertranslated text within its own intellectual tradition?

Arts & Architecture

How did one artistic movement respond to the social conditions of its moment?

What did modernist housing assume about human behaviour, and what happened when it was built?

How do curatorial decisions shape the meaning of the works a museum displays?

What does the law's treatment of sampling reveal about musical authorship?

What the finished work actually looks like

The critical literature review

The public-data analysis

The policy brief

Twenty-four subject guides

Computing, Engineering & Quantitative Fields

Medicine, Life Sciences & Behaviour

Society, Economics, Law & Global Affairs

Humanities, Media & Creative Fields

Take the 50 strongest ideas with you

Related Guides

Frequently Asked Questions

Ready to start your research project?

If you are applying to US universities