Original research does not require statistics. Archival history, close textual analysis, conceptual argument, comparative policy work and small-sample interview studies all produce genuinely new findings with no numbers beyond page references. The fifteen projects below each name a method and a free source of evidence. Several use archives containing hundreds of thousands of primary documents that almost no student has opened.
A lot of students assume original research means collecting numbers and running tests. It does not. Entire disciplines produce new knowledge with no statistics in them at all, and some of the richest primary sources available to a school student are archives of text that almost nobody has worked through.
Each of the fifteen below names a method and a source. That pairing is what makes a project doable, and it is what examiners and journals actually assess.
Archival history (4 projects)
Two free archives make this category unusually accessible.
The Proceedings of the Old Bailey Online is a fully searchable edition of 197,745 trials held between 1674 and 1913, amounting to around 127 million words, plus the Ordinary's Accounts — narratives of the lives and deaths of about 2,500 executed convicts published between 1676 and 1772. Historic Hansard covers Commons and Lords debates from 1803 to 2005, free to read and organised by sitting, with extracted indexes of speakers, constituencies, Acts, Bills and Divisions. The content remains under Parliamentary Copyright, so quote and cite rather than reproducing at length.
- How did defences offered in theft trials at the Old Bailey change across a defined fifty-year window? Take a consistent sample of trials, record the kind of defence offered, and argue about what shifted and why.
- How was a single piece of legislation argued for and against as it passed through Parliament? Follow one Bill through Hansard and reconstruct the argument. The historiographical payoff is that published accounts usually simplify what was a messy debate.
- How did the Ordinary's Accounts represent the condemned, and what does that tell us about the purpose of the genre? A close reading of fifteen to twenty accounts against each other.
- How did parliamentary language about a single social problem change across a century? Pick a term, trace its use in Hansard, and argue about what the change in vocabulary reveals.
The methodological move in all four: state your selection criteria explicitly. I examined every trial in the following years matching the following search terms, excluding the following categories, for these reasons. That sentence is what makes archival work rigorous.
Close textual and discourse analysis (4 projects)
- How do two translations of the same classical passage differ, and what interpretive commitments do the differences reveal? Requires no language expertise beyond what you can assemble with a lexicon and secondary commentary.
- How do university subject pages describe entry requirements, and are the descriptions actionable by a sixteen-year-old? A content analysis of twenty prospectus pages against criteria you define. Low glamour, genuinely useful, and nobody has done it.
- How is a single historical figure characterised across three school textbooks from different decades? Argue about what changed and what pressures produced the change.
- What claims do commercial education providers make about outcomes, and what evidence do they cite? Define a corpus, code the claim types, and report what the evidentiary basis turns out to be.
Conceptual and philosophical analysis (3 projects)
These produce an argument rather than a finding, which is a legitimate research output in philosophy, law and political theory.
- Does a widely used concept in a policy area survive examination? Take a term that does a lot of work in public argument — "merit," "resilience," "the public interest" — and test whether it is doing one job or several incompatible ones.
- Is a particular legal distinction defensible on its own terms? Pick a distinction drawn in statute or case law, reconstruct the justification, and assess it.
- What would follow if a prominent argument in your subject were correct? Working out the unstated commitments of an accepted position is a standard philosophical contribution and requires only careful reading.
Comparative policy and legal analysis (2 projects)
- How do three jurisdictions regulate the same thing differently, and what explains the divergence? Legislation is public. The contribution is the comparison and the explanation, not the discovery of the texts.
- How did one policy change between its proposal and its enactment? Compare the consultation document, the Bill as introduced, and the Act as passed. The gaps are the finding.
Interview and small-sample qualitative work (2 projects)
- How do people in a defined role describe a process that is usually described from the outside? Eight to twelve interviews, properly coded, is a real study. Teachers on university guidance, first-generation applicants on information gaps, student journalists on editorial limits.
- How do practitioners in a field account for a change they have lived through? Oral history, essentially, and a method with a long scholarly pedigree.
Interview work carries obligations the other categories do not. If your project is examined, read your centre's rules first: AQA's guidance on EPQ safeguarding requirements sets out that centres must check the topic and any primary research suit the student's age and maturity, must consult the Designated Safeguarding Lead before approving a title, and must complete a risk assessment where primary research involves vulnerable people or sensitive subjects. Topics grounded in personal experience of harm must not be approved. These rules exist for good reasons and they shape what interview projects are available to you.
What makes a non-statistical project rigorous
The absence of statistics removes a hiding place. In a quantitative paper a reader can be reassured by a method they recognise. Here your argument is on the surface, so three things have to be explicit.
| Quantitative paper | Non-statistical equivalent |
|---|---|
| Sample size and recruitment | Selection criteria for your documents or participants |
| Statistical test and why | Method of reading or coding, and why it suits the question |
| Confidence intervals, limitations | What your sources cannot show, stated plainly |
Say why those twenty documents. Say what you looked for and how you recorded it. Say what your evidence does not establish. Do those three and your project is methodologically sound; skip them and it reads as an extended opinion piece, however good the reading.
Where this fits with examined qualifications
All fifteen are viable EPQ, extended essay or Advanced Higher projects, and most are viable AP Research projects. The thing worth checking is how your rubric treats method.
AP Research in particular rewards this category more than students expect: its academic paper rubric awards seven points for "a rationale for the chosen approach, method, or process in relation to the research question" — and a justified archival or textual method scores there exactly as a justified experiment does. Nothing in the rubric privileges numbers. Our breakdown of the AP Research rubric covers each row, and our thirty AP Research topics include quantitative options if you want to compare.
For UK qualifications, our complete guide to the EPQ, EPQ ideas and extended essay ideas cover how the same projects translate.
Two practical notes
Reading volume is the real cost. You are trading method design for hours of careful reading. Budget for it. A project built on twenty archival documents needs those twenty documents read properly, twice.
Check publication scope early. Some student journals are STEM-only and will not consider humanities work however good it is. Read two or three recently published papers in any venue before you write for it — that tells you about fit faster than the guidelines do. Our guide to where high school students can publish notes which venues take non-quantitative work.
And if you were avoiding research because you thought it meant statistics: it does not, and the archives above contain more unexamined primary material than any of us will get through. Our guides to narrowing a research question and research without a lab cover the next steps.
Frequently asked questions
Can you do real research without statistics?
Yes. Most published research in history, law, philosophy, literature, classics and large parts of politics and anthropology contains no statistical analysis at all. Rigour in those fields comes from the quality of your evidence and the tightness of your argument, not from a p-value. Examiners and journals judge the method against the question, not against a preference for numbers.
Do student journals publish non-quantitative research?
Many do, though you should check scope before writing. Some high school journals are STEM-only; others take humanities and social science. Read two or three recently published papers in any venue before you start, because that tells you more about fit than the submission guidelines do.
Is a literature review a research project?
On its own, usually not — it reviews what others found rather than producing a finding. A historiographical analysis is different: it makes an argument about why interpretations changed, which is an original claim about the field. The distinction is whether you end with a conclusion of your own.
What counts as primary evidence in a non-statistical project?
Anything created by the thing you are studying rather than written about it: court records, parliamentary debates, newspapers, manuscripts, legislation, treaties, films, buildings, objects, interview transcripts you generated. Secondary sources interpret those; your contribution comes from working on the primary material yourself.
How do I make a qualitative project rigorous?
State your selection criteria and apply them consistently, so a reader knows why those twenty documents and not others. Say what you looked for and how you recorded it. Acknowledge what your sample cannot show. Those three moves do for qualitative work what reporting your test and sample size does for quantitative work.
Are these projects easier than quantitative ones?
Not easier, differently hard. You avoid method design and analysis, and you take on reading volume and argumentative discipline instead. A weak qualitative project is weak in an obvious way, because the argument visibly does not follow. There is nowhere to hide behind a technique.