Final Deadline for Winter 2026 Programme is 21 November. Apply Now →
HomeOCAR AcademyOxford Scholar ProgrammeOxford Publication FellowshipOCAR Student Research SymposiumOxford Essay PrizeOur AlumniOutreachPodcastsResearch FellowsBlogAboutFAQsFor Parents & CounselorsContactAdmissions Portal
A student's notebook showing a research question and data notes
How-To / Research Skills

What Counts as Original Research for High Schoolers?

In short

Originality in research means your work adds something to what was known, not that nobody has worked on the topic. New data on an old question counts. A replication in a new population counts. A null result counts. What does not count is summarising other people's findings, repeating a study with no change, or restating a known result with new wording. Most students set the bar far too high and conclude wrongly that their project is not original.

Students abandon good projects over this word. "It's not really original — people have studied this before." Almost always, that conclusion is wrong, and it rests on a definition of originality that nobody in research actually uses.

What originality actually means

Not: nobody has worked on this topic.

But: your paper adds something to what was known.

If the first definition were right, almost no research would be publishable. Essentially every paper in every journal sits in a literature where people have worked on the topic before — that is what a literature is. What makes a paper publishable is a stateable contribution.

The practical test is a sentence:

"This paper shows that ___, which was not previously established because ___."

Complete it and your work is original. The gap in the second blank can be as modest as "this had not been tested in a UK school population." That is a real reason, and it is enough.

This is also precisely how examined research frames it. The AP Research academic paper rubric reserves its top band for work that identifies a gap in the research base — not work that invents a field.

Six things that count

ContributionWhy it countsEffort
New data on an existing questionThe evidence did not exist beforeModerate
Replication in a new populationTests whether the finding generalisesModerate
Replication in a new periodTests whether it still holdsLow
A null or contrary resultCorrects the recordModerate
Secondary analysis asking a new questionThe analysis is newLow
Reconciling conflicting findingsExplains why studies disagreeLow

New data on an existing question. The commonest route and the most straightforward. The question is known; your evidence is not. A survey you designed, interviews you conducted, a corpus you coded.

Replication in a new population or period. Underrated and genuinely valuable. Most psychology and education findings were established in particular places at particular times; whether they hold in a UK sixth form in 2026 is an open question with a real answer. You get to state your contribution with total clarity, which makes the paper easy to write and easy to review.

A null or contrary result. Publication bias means findings that failed to replicate are systematically under-reported, so a well-designed study that found nothing contributes more than its author usually believes. The requirement is that the design was sound — a null result from a broken study tells nobody anything.

Secondary analysis. Using a dataset somebody else collected and asking a question of it they did not ask. This is how an enormous amount of professional social science is done, and it is entirely available to a school student with a laptop. Our guide to free datasets for student research covers where to find usable data.

Reconciling conflicting findings. Two studies disagree; you work out why — different measures, different populations, different operationalisations — and argue for a resolution. Low cost, high value, and it sits exactly where the AP rubric's Argument row asks you to explain how works with different perspectives relate to one another.

Three things that do not count

Summarising what others found. A literature review reports; it does not contribute. This is not a judgement on the work — reviews are useful and often hard — but most student journals want original evidence. The test from our readiness checklist: if you deleted every secondary source, would anything of yours remain?

The exception worth knowing: a historiographical or critical review that argues about the field — why interpretations changed, why two literatures conflict — ends with a conclusion of your own and can be original. The dividing line is whether you are reporting or arguing.

Repeating a study with nothing changed. A replication is original because something differs: the sample, the period, the setting. Re-running an identical study on an identical population adds nothing.

Restating a known result in new words. Paraphrasing an established finding is not a contribution, however well written.

The bar is lower than you think, and the trap is elsewhere

Here is the thing most students have backwards. The originality bar is low and easy to clear. The bar that actually catches people is whether your claims match your evidence.

A modest, true contribution clearly stated beats an ambitious claim the data cannot support — every time, with every reviewer and on every rubric. "This study found no association between X and Y in a sample of 62 Year 12 students, which is inconsistent with Smith's finding in a US undergraduate sample" is publishable. "This study proves X does not cause Y" is not, from the same data.

So the sequence is: find your modest contribution, state it precisely, and resist making it bigger. Our guides to narrowing a research question and building a question worth publishing cover the front end of this.

Originality without a lab

A practical note, because lab access is what students most often think stands between them and original work.

It largely does not. Survey research, interview studies, content analysis, archival work and secondary data analysis all produce original evidence with no equipment. And lab access is genuinely hard to get for under-18s — the University of North Carolina's minors in laboratories policy illustrates the age and supervision rules that apply across many institutions, and one Stanford lab's public page on high school placements records taking an intern roughly once every four to five years against tens of requests a year.

Choosing a method that does not need a lab is not a compromise. It is the realistic route to original work, and it is what the majority of publishable student research uses. Our guides to research without a lab and research project ideas without statistics cover the options.

Whose original contribution is it?

One question that follows directly from originality, and that students in paid programmes should think about before submission.

If your work is an original contribution, the authorship should reflect who made it. The ICMJE authorship criteria require all four of: substantial contribution to conception, design, or the acquisition, analysis or interpretation of data; drafting the work or critically revising it for important intellectual content; final approval of the version to be published; and agreement to be accountable for all aspects of the work.

And explicitly: acquiring funding, general supervision of a research group, and general administrative support do not qualify someone for authorship on their own. Nor do writing assistance, technical editing or proofreading. Those contributors should be acknowledged — with their written permission, since acknowledgement can imply endorsement.

That matters at school level because supervisory arrangements vary enormously, and a mentor whose involvement was genuine supervision belongs in the acknowledgements, not the byline. Our guide to student research paper authorship covers how to have the conversation, and who can supervise a student research project covers what good supervision looks like.

The short version

Your project is probably original. Topics are rarely new; contributions usually are.

Write the sentence — this paper shows that ___, which was not previously established because ___ — and if you can complete it honestly, stop worrying about originality and start worrying about whether your conclusions match your data. That is the one that decides whether the paper gets published.

Frequently asked questions

Does original research mean nobody has studied it before?

No. That standard would rule out almost all published research, including most professional work. Originality means your paper adds something — new evidence, a new population, a new period, a reconciliation of conflicting findings. Topics are rarely new; contributions are.

Is a replication study original research?

Yes, and it is a recognised contribution in every field that takes reproducibility seriously. A direct replication with a new sample tests whether a published finding holds, which is genuinely new information. It also clears the originality test by construction, because you can state exactly what your work adds.

Is a literature review original research?

Generally not, because it reports what others found rather than producing a finding. A review that makes an argument about the field — explaining why interpretations diverged, or reconciling two conflicting literatures — is a different thing and can be an original contribution. The distinction is whether you end with a conclusion of your own.

Can I use someone else's dataset and still be original?

Yes. Secondary analysis of existing data is how a great deal of professional social science works. The originality is in the question you ask of the data and the analysis you perform, not in who collected it. Cite the source properly and describe what you did to it.

Does a null result count as original?

Yes, and often more usefully than a positive one, because publication bias means null results are under-reported. A well-designed study that found no effect tells the field something. What matters is that the design was sound and you report the result honestly rather than reframing it.

Who gets credit for the original contribution?

Whoever meets all four ICMJE authorship criteria: substantial contribution to conception or data, drafting or critical revision, final approval, and accountability for the work. Acquiring funding or generally supervising a group does not qualify someone on its own. If a mentor's involvement was supervisory, they should be acknowledged rather than listed as an author.