← Back to all articles

Competition Guides

Peterhouse Kelvin Sciences Essay Competition: How to Write a Research-Led Science Essay

A practical guide to eligibility, submission rules, question choice and the evidence-led writing seen in official winning entries.

11 Nov 202613 min read
Article cover image

Peterhouse Kelvin Sciences Essay Competition: How to Write a Research-Led Science Essay

The Peterhouse Kelvin Sciences Essay Competition asks students to do something that school science rarely leaves enough room for: take a question whose answer is uncertain, read beyond the curriculum and decide what the available evidence really supports.

That makes it less a test of how many facts a student can collect and more a test of scientific judgement. The strongest response is not the one with the longest bibliography or the most technical vocabulary. It is the one that gives the reader a clear argument, distinguishes evidence from speculation and knows where its conclusion must remain provisional.

A Kelvin essay should not read like an encyclopaedia entry. It should read like a young researcher examining a difficult question and making a careful, evidence-based decision.


Editorial overview

The Peterhouse Kelvin Sciences Essay Competition is built around open scientific questions rather than a fixed body of examinable knowledge. The current set ranges from species conservation and biological chirality to biosignatures, machine learning in chemistry and quantum computing. That breadth is not an invitation to summarise an entire field. It asks students to narrow a large problem, identify the evidence that matters and explain what can and cannot yet be concluded.

Peterhouse's official archive makes its expectations unusually visible. A winning essay on platypus fluorescence moved from molecular mechanisms to competing evolutionary explanations, tested whether preservation could distort the observation, acknowledged the limits of a small evidence base and proposed further research. Another winning essay used the idea of an adaptive landscape to organise an argument about the distribution of venom among animal groups. These essays succeed because they use scientific concepts to control the argument rather than decorating it.

Our view is that the Kelvin competition suits students who enjoy reading around a question and are willing to revise their conclusion as the evidence develops. It is a poor fit for someone looking for a quick credential or a topic that can be completed by compiling web sources over a weekend. The word limit is tight, individual feedback is not provided, and each school can normally forward only a small number of entries. The preparation burden is therefore real, but so is the value of learning how scientific uncertainty becomes a structured essay.

Quick Facts

FieldDetails
CompetitionPeterhouse Kelvin Sciences Essay Competition
OrganiserPeterhouse, Cambridge
Typical studentsStudents in their penultimate year of school who are educated in the United Kingdom or Ireland
FormatOne independently researched scientific essay submitted as a PDF, with teacher confirmation
Best forStudents who enjoy reading scientific research, comparing explanations and writing a concise argument
DifficultyA demanding combination of independent research, source evaluation and precise writing within a short word limit

Review Evaluation

Rated Advanced. The competition requires an independently researched scientific argument, careful use of evidence and sustained academic writing within a 2,000-word limit.

For the current competition status, organiser page and any updated dates, use the Peterhouse Kelvin Sciences Essay Competition detail page. Annual questions and arrangements can change, so this guide keeps the main advice evergreen and uses the current questions only as examples.


What makes a Kelvin essay different?

The official purpose is to let students explore scientific ideas beyond the classroom and engage with research. The rules reinforce that aim: the main focus should not simply reproduce material already studied in a school course, and the essay must not be work submitted to an examination board.

This does not mean every entrant needs university-level technical knowledge. It means the essay should go beyond explaining an accepted concept. A successful response usually has four layers:

  1. A precise interpretation of the question. Define what the important terms mean and decide what is actually being judged.
  2. A scientific framework. Use a mechanism, model or set of criteria to organise the evidence.
  3. Evaluation rather than accumulation. Compare sources, alternative explanations and limitations instead of placing one fact after another.
  4. A proportionate conclusion. State what the evidence supports without pretending that an unresolved research question has become certain.

That is why a topic such as quantum computing cannot be answered well by listing recent machines and possible applications. The word “useful” needs a definition. The student might distinguish scientific demonstration, commercial advantage and practical deployment, then test current evidence against those different standards.

Reading the current questions as research problems

The current official list contains six questions. Each one hides a more specific research decision inside its broad wording.

Question areaThe real analytical task
Saving a species from extinctionUse one conservation effort to show which missing or available information changed the outcome
Chirality in biologyConnect molecular handedness to biological structure, recognition, function and consequences
Why lifespans differChoose mechanisms or evolutionary trade-offs that can explain variation rather than catalogue long- and short-lived species
Biosignatures and life on exoplanetsDecide what would count as convincing evidence and how false positives could be excluded
Machine learning and experimental chemistryEvaluate which parts of experiment can be predicted or automated and which still require empirical testing
Whether quantum computing is currently usefulDefine present usefulness and separate demonstrated capability from future promise

These examples reveal a common pattern. The question is broad enough to invite curiosity but narrow enough to require judgement. A student who tries to cover every relevant subfield will lose space for analysis.

👉 Before researching, rewrite the chosen title as a smaller working question: “Which evidence would change my conclusion, and what comparison will allow me to judge it?”


What official winning essays reveal

Peterhouse publishes a selection of winning entries, which gives students better evidence than generic advice about “being original”. The 2024 winning essays show several recurring habits.

They explain a mechanism before making a claim

The winning essay on platypus fluorescence begins by explaining how biological fluorescence works at a molecular level. That mechanism gives the later discussion something to test. The essay can then distinguish different candidate fluorophores and ask whether the observed colour is consistent with them.

This is more persuasive than opening with a large claim about the evolutionary purpose of fluorescence. It first establishes what the phenomenon is and how it could arise.

They test competing explanations

The same essay considers whether the observation could be an artefact of preserving museum specimens, whether fluorescence offers an adaptive advantage, and whether it may instead be an inherited or non-adaptive trait. Its strength lies in comparison: the student does not choose an attractive explanation and ignore the rest.

Another winning essay, addressing why venom is distributed so differently across animal groups, uses the adaptive-landscape concept as a framework. Rather than producing separate descriptions of reptiles, mammals and birds, it asks how different evolutionary routes and existing adaptations may make venom more or less likely to develop.

They show where the evidence becomes weak

The platypus essay explicitly recognises the limited sample base and the difficulty of assigning a function to a trait that has not been studied extensively in living animals. It then identifies observations and laboratory work that could strengthen the conclusion.

That is scientific writing, not indecision. A good conclusion can say that one explanation is currently the most plausible while also specifying what remains unknown.

They research before they commit to a thesis

In Peterhouse's interview with a former winner, Sohan explains that he read a little each day for weeks, used reliable general material to find the field, and then followed citations toward primary evidence. Only after becoming familiar with the question did he begin drafting. He also recommends addressing caveats and criticism rather than hiding them.

This sequence matters. Writing a thesis first and searching only for evidence that agrees with it encourages confirmation bias. Research first, develop a provisional answer, and be prepared to change it.


How to choose the right question

The most familiar topic is not always the best choice. Nor is the question with the most available material. A good choice sits at the intersection of genuine interest, usable evidence and a manageable line of argument.

Use this test before committing:

  • Can you explain the central scientific idea in your own words?
  • Can you identify at least two plausible explanations, approaches or standards to compare?
  • Are reliable reviews and primary studies available at a level you can interpret?
  • Can you narrow the question to a few decisive examples?
  • Is there enough uncertainty for evaluation, rather than only settled facts to report?
  • Can you imagine a conclusion that answers the exact wording?

If two questions still look equally promising, spend one hour building a source map for each. The better topic is usually the one that produces a clearer controversy or comparison, not the larger stack of search results.

For a broader decision framework, see CompeteMap's guide to choosing an essay competition question.

A practical research and writing process

1. Turn the title into sub-questions

Break the prompt into four to six smaller questions. For a biosignatures essay, these might cover the definition of a biosignature, leading candidate signals, abiotic false positives, the limits of remote observation and the evidence threshold for a convincing claim.

2. Begin with orientation, then move toward primary evidence

Use reliable reviews, university material or major scientific organisations to understand the vocabulary and identify important debates. Follow their citations to research papers that directly support the claims you expect to use.

The aim is not to cite only primary research at any cost. It is to know which source supports which proposition and whether you have read enough context to represent it accurately.

3. Build an evidence table

Before drafting, make a simple record:

Claim or questionBest supporting evidenceLimitation or counterpointRole in the essay
Main mechanismReview plus key studyConditions or sample limitsExplain the scientific basis
Competing explanationStudy or documented caseConflicting resultTest the main interpretation
Practical implicationReal application or case studyScalability or uncertaintyAnswer why the issue matters

This prevents the bibliography from becoming detached from the argument.

4. Write the conclusion before the introduction

Draft a provisional answer to the question in three or four sentences. If it cannot answer the title directly, the research is still too broad. Use that answer to decide which evidence earns space.

5. Give each paragraph a job

A strong paragraph should normally make one claim, introduce the relevant evidence, explain why it matters and connect it to the overall judgement. Avoid paragraphs that exist only to display another interesting fact.

6. Revise for precision and intellectual honesty

Check every use of words such as “proves”, “causes”, “will” and “always”. Scientific evidence often supports probability, association or a mechanism under defined conditions rather than universal certainty.

Students who find essay competitions unexpectedly difficult may also benefit from Why Good Students Often Struggle With Essay Competitions.


Submission rules that matter

The official guidelines contain several rules that should shape the work from the beginning, not be checked five minutes before submission.

RulePractical implication
Maximum 2,000 words, including footnotes and appendicesDo not move excess argument into footnotes or an appendix
Bibliography excluded from the word countKeep full references in the bibliography and use a consistent standard system
One essay per studentChoose one question and develop it properly
Work must be the student's own and not produced by AIUse tools for ordinary file preparation only where permitted; the research, reasoning and prose must be the entrant's own
No work submitted to an examination boardThe essay must be separate from assessed school coursework
Main focus should go beyond current or previous school-course materialUse the curriculum as a foundation, not as the whole essay
PDF submission with numbered pagesFinalise formatting and check the file itself before submitting
Name and school on the first pageFollow the required identification and filename format
Teacher approval requiredSpeak to a teacher early and provide the correct contact details
Normally no more than four entries per schoolA school may need an internal selection process

The normal four-entry limit does not apply to sixth-form colleges whose year groups exceed the size specified by Peterhouse. Where several students want to enter, the school should decide its internal process early rather than wait until essays are complete.

Peterhouse welcomes diagrams that support the argument and states a preference for original drawings. A diagram should clarify a mechanism, relationship or comparison; it should not be included merely to make the essay look scientific.

Once an essay has been submitted, the official FAQ says it cannot be changed. The teacher then confirms the entrant's eligibility and compliance with the rules. Students should therefore leave time for a final PDF check and make sure their teacher knows the verification request is coming.

What the prize provides—and what it does not

Peterhouse states that the competition has a £750 prize pool shared among the winners and that a prizegiving event is held in Cambridge. The official winning archive also gives selected essays a public life beyond the submission itself.

The competition does not provide individual feedback because of submission volume. Students should enter because the research process itself is worthwhile, not because they expect a detailed assessment or a guaranteed admissions advantage.

A strong entry can demonstrate curiosity and disciplined independent work, but Peterhouse does not present the prize as a route into Cambridge. The former-winner interview is most useful as a picture of how one student's interest developed: sustained reading led to a stronger essay and further scientific exploration. That is a more realistic benefit than treating any single prize as an admissions credential.

Key Takeaways

  • The Kelvin competition rewards a research-led argument, not a broad summary of a scientific field.
  • The strongest official examples explain mechanisms, compare competing interpretations and state the limits of the evidence.
  • Choose a question that supports a clear comparison or judgement and can be narrowed within 2,000 words.
  • Research before fixing the thesis; use reviews to understand the field and follow citations toward the evidence that matters.
  • The essay must be the student's own work, go beyond ordinary course material and receive teacher confirmation.
  • Check the CompeteMap competition detail page and Peterhouse's official page for each cycle's questions, dates and submission arrangements.

Sources checked

Original CompeteMap editorial content. Brief quotations must clearly credit CompeteMap and link to this article. For complete reuse terms, see Copyright & Reuse.

EXPLORE NEXT

Not sure where to start?

Find the right competition
View all articles →

Loading comments…