IB Extended Essay guide (2027): criteria, word count and examples

What the IB Extended Essay is in 2027

The Extended Essay (EE) is a piece of independent research on a question you choose. You write it in one subject (or, on the interdisciplinary pathway, by integrating two), with a school supervisor to guide you, and the IB marks it externally. The essay is at most 4,000 words. On top of that you write a reflective statement of at most 500 words on the Reflection and Progress Form (RPF), and that statement does not use up any of the 4,000. The IB's own estimate is about 40 hours of your work, and it recommends 3 to 5 hours of supervision, which includes three reflection sessions you must attend.

The first students assessed on the new model sit in May 2027. November 2026 was the last session on the old 34-mark model. So if an older sibling, a website or an old sample essay talks about 34 marks, it describes the previous scheme. This guide covers the 2027 model only.

The rules that matter most

• Marking is external. Examiners outside your school read the essay and mark it by best fit: they pick the description that matches your work most closely, not a tick-list where every point must be met.
• Examiners stop at the limit. Nothing after the 4,000th word is considered. For many students that is where the conclusion and half the evaluation sit.
• Criterion E uses the RPF only. The reflective statement is the only evidence for those 4 marks, so it has to be good by itself.
• It is your work. Your supervisor guides you but does not write or rewrite your essay, and every source you use must be acknowledged.

What changed from the old model

The old model had 34 marks. The 2027 model has 30, in five criteria, with the biggest share (8 marks) on discussion and evaluation. The IB also reshaped the choices you make at the start.

The practical effect is that reflection and evaluation now carry more of the grade, and the essay is judged as one piece of research rather than a bundle of separate skills. We say what we think this means for you in the sections below, and we mark clearly where something is our own reading rather than an IB rule.

The five criteria (the rubric) in plain words

The EE is marked out of 30 on five criteria. The IB calls the set of criteria and their descriptions the assessment rubric, and students often search for it as “the Extended Essay rubric”. Criteria A to D are judged on the essay as a whole. Criterion E is judged on the reflective statement alone.

Two things in that donut deserve a second look. First, D is 8 marks, more than any other criterion, so evaluating your own sources, methods and conclusion is worth more than a clean introduction or a tidy bibliography. Second, criterion E is only 4 marks, but it is the easiest place to lose marks through neglect, because the examiner has nothing else to go on.

We describe each criterion below in our own words. We do not reproduce the IB’s markband wording, and you should read the official 2027 descriptors alongside this guide (your coordinator has them). What we give you is how to show each criterion on the page and where marks usually go missing.

Criterion A: Framework for the essay (6 marks)

What it rewards. A research question, a method and a structure that fit together so the essay works as research from the first page.

How to show it. Write a question that can really be answered in 4,000 words and that invites analysis rather than a yes or no. Say in the introduction why the question matters, then name your method and give the reason for it. Choose enough sources of the right kind. If you run an experiment or survey, describe it so someone could repeat it. Make each heading a step in the argument, not a label.

Where marks go. A question so broad you can only survey it; a “how” or “what” question that leads to a narrative; sources picked because they were easy to find; no statement of method at all; headings such as “Background” and “Case 1”. In the sciences the usual loss is too little detail to repeat the work. In interdisciplinary essays it is two subjects sitting side by side with no integration.

Criterion B: Knowledge and understanding (6 marks)

What it rewards. Relevant research material turned into knowledge, shown through accurate terms and concepts that you explain in your own words.

How to show it. Use varied sources of suitable quality. Define the terms a reader would need and use them consistently. Explain a concept and then use it in your analysis. Keep quotations short and purposeful. Show where your question sits in the wider subject.

Where marks go. Pages of background that the argument never uses (the commonest problem in science and maths essays); terms used loosely; long block quotes in place of understanding; leaning on encyclopaedias, study guides or AI summaries; a concept introduced and then dropped; factual or formula slips.

Criterion C: Analysis and line of argument (6 marks)

What it rewards. Turning your research into findings, and connecting question, findings and conclusion into one clear thread.

How to show it. Make every paragraph a point linked to the question and support it with evidence. After each piece of evidence, say what it shows. Use the first and last lines of sections to connect them. Write a conclusion that follows from your findings and answers the question, with no new evidence in it.

Where marks go. A long summary of a text, event or theory with one line of analysis; data or a quote dropped in and left to speak for itself; paragraphs that could swap places; topic sentences that name a topic instead of making a claim; a conclusion that repeats the introduction.

Criterion D: Discussion and evaluation (8 marks)

What it rewards. A balanced discussion of what your findings mean, supported by evidence, and an honest evaluation of your own research: its sources, its methods and its limits, and how those limits affect the conclusion.

How to show it. Discuss the significance of the findings from more than one angle. State the strongest rival interpretation fairly and say why you prefer yours. Judge your sources where you first use them (author, date, purpose, method). Evaluate your method where you apply it. For every limitation say which way it pushes the result and by roughly how much.

Where marks go. One evaluation paragraph squeezed into the last 200 words; “more research is needed” with nothing specific; limitations listed but never tied to how far the conclusion can be trusted; all sources treated as equally reliable; confident words such as “proves” on thin evidence.

Criterion E: Reflection (4 marks)

What it rewards. A reflective statement that evaluates your decisions along the way and shows how you grew and what you will take into other work.

How to show it. Build the statement on notes from your three sessions. Name two or three real decisions (you changed the question, dropped a source, switched method), say what you weighed and what happened next. Give concrete examples of skills you have already used elsewhere.

Where marks go. A timeline of tasks (“first I… then I…”); thanking the supervisor; “I learned time management” with no example; no decision named, as if everything went smoothly; writing it all in one sitting at the end.

Three habits help across A to D: give reasons, not lists; put the argument in the main text, not in notes or appendices; and check that the same question runs from the title page to the conclusion.

Subject-focused or interdisciplinary: choosing your pathway

From 2027 there are two pathways. In the subject-focused pathway you research a question inside one Diploma subject, using that subject’s methods and conventions. In the interdisciplinary pathway you integrate the lenses of two subjects to answer a question that neither could answer alone. The old world studies essay now sits inside the interdisciplinary pathway, and the language A and B categories no longer exist. The same five criteria apply on both pathways.

What changes on the interdisciplinary pathway

In our reading, the criteria stay the same but you need to show a little more in each. In A, the question names or clearly needs both subjects, and the essay explains why. In B, your knowledge is grounded in both, though the balance need not be equal. In C, the findings and argument show both lenses, and the integration develops through the essay and not only in the introduction. In D, you judge how well the integration worked: what it added and what it cost. In E, you can draw on the challenge of working across two subjects.

Warning signs are two halves joined by a heading, one subject used only for background, and a conclusion that speaks about only one subject. A useful test is to ask: what could I not answer with one subject? If you cannot say, stay subject-focused. Some subjects, and some pairings, may not be permitted on this pathway, so confirm with your EE coordinator before you commit.

Examples of pairings that can work (our illustrations, not a list of approved pairs) are a mechanical model of an animal’s movement that needs both physics and biology, or a kinetic model fitted to chemistry data that needs both maths and chemistry.

How to choose a subject and topic

Most students lose time here, so slow down. The strongest essays come from a question where reasonable approaches could give different answers, where the student can actually get the evidence, and where there is something to weigh. Start from the subject, then narrow.

Pick the subject first

Choose a subject you know well and enjoy, ideally one where your grades suggest you can handle the method. Your EE is registered under that subject, and your supervisor will normally teach it. Ask your coordinator which subjects your school offers for the EE before you fall for a topic.

Then narrow, step by step

Six qualities make a topic workable. It is a real puzzle, not a settled fact. The evidence is within reach, because data that does not exist in usable form is a fast way to lose weeks. It goes a little beyond the course in one place, with you explaining that part. The scope fits 4,000 words: one system, one model, one comparison. It has a built-in way to evaluate, such as two models to compare, an experiment with error sources, or sources that disagree. And it interests you, because the reflection and the viva both depend on real decisions you made.

Fit by subject (our reading; confirm with your supervisor)

Mathematics. Mathematics should be the subject, not just a calculator step. Good questions compare two models or methods against a measure, ask when a result holds, or test a model on real data. Be wary of pure exposition of a known topic. Physics. A measurable relationship or a testable model, with variables you can control and measure in your school, or a public data set. Questions that only verify a textbook formula tend to be thin. Chemistry. A controlled investigation with a measurable outcome (rate, yield, concentration, enthalpy, pH) that is safe in a school laboratory, or analysis of reliable published data. Biology. A specific biological question with a measurable variable and a realistic sample size. Plants, microbes and open data are often easier to run than animal or human work, which normally needs approval. For all four, ask your supervisor about the 2027 subject-specific rules and the school’s ethics and safety policy.

In the humanities and social sciences, the natural evidence is texts, documents, published statistics or interviews you are allowed to run. The same funnel works: one case, one period, one lens, one judgement word.

Finding Extended Essay topics

A topic is an area you find interesting. A research question is something you can answer with evidence in about 4,000 words. The two are easy to mix up, and the gap between them is where most of the early work happens.

We keep a separate page of Extended Essay topic directions, sorted by subject, with the kind of evidence each one suits and the pitfalls to avoid: Extended Essay topics and ideas. For the four subjects we teach most, our IA idea lists also work as sources of measurable systems you can build an EE around: Maths, Physics, Chemistry and Biology.

A word of caution about copying. A published research question is a pattern, not a gift. If you take someone else’s question you take their evidence base, their limits and their problems, and you risk an integrity problem too. Use ideas to see the shape of a workable question, then rebuild it around something you can reach and care about.

How to write a strong research question

The research question drives criteria A, C and D at once. A vague question makes your framework weak, an unfocused question makes analysis hard, and a question with nothing to weigh leaves you nothing to evaluate. Fix it early and test it often.

Look at what changes on each rung. A case appears. A limit appears. A measure appears. Finally, a verb that forces a judgement appears. Nothing was added by making the question sound grander. Words such as “describe”, “outline” and “discuss” invite lists; “to what extent”, “how far” and “how well does” invite weighing.

Ten tests for your question

Your question will change. That is normal and healthy, and a real change of question, with the reason for it, is excellent material for your reflective statement. What matters is that by the time you start serious drafting the question is stable, and that your introduction, headings and conclusion all point at it.

Extended Essay examples: questions, outlines and a strong paragraph

People searching for Extended Essay examples usually want a finished essay with a mark attached. We do not publish student essays, and the IB publishes its own marked sample essays, which your EE coordinator can show you. Ask for them: reading a marked sample next to the criteria is one of the best things you can do.

What we can show is the part that is hard to see in a finished essay, which is its shape. Everything below is invented, generic and unmarked. It is there to teach a pattern. Do not copy it, because your own question, sources and argument are what the examiners are reading for.

Weak to strong research questions in eight subjects

In every stronger version a case, a limit and a measure appear, and the question carries its own way to judge: “to what extent”, “how far”, “judged by”. That built-in judgement is what makes criterion D easy to earn later.

An outline skeleton for a science essay

This is a headings-and-budget skeleton only. A practical science EE spends more on method and data, and keeps evaluation running through the results as well as in its own section.

Where the criteria are earned: A in the first two sections (a narrow question and a method chosen to test it); B in the background, where only the science the analysis uses appears; C in the results, where each graph leads to a claim about the model; D in the evaluation, which ranks error sources by size, and in comments inside the results; E in a note you keep about a choice you changed after the first session.

An outline skeleton for a humanities essay

Here A is the case, the period and the source plan, stated near the start; B is accurate context and correct terms in your own words; C is each argument section moving from evidence to explanation; D is the rival explanation and the reliability of sources weighed throughout, not just at the end; E is a note on a real decision, for example moving from one source to a mix after the first session.

What a strong analysis paragraph does

A strong paragraph is not longer than a weak one. It does five jobs in order. Here is the pattern as a schematic, with no subject attached.

Test any paragraph in your draft by finding each job. If you can find Claim and Evidence but not Analysis, you have a summary. If you cannot find Evaluate anywhere in your body, criterion D will be thin however good the ending is.

Sources, the literature base and judging quality

Every EE needs secondary research, which means using existing sources such as articles, books and data to build your knowledge base. Many subjects also use primary research, which is new data you gather yourself through an experiment, survey or fieldwork. Check your subject’s rules for how the two combine, and what is expected for each.

Think of your source list as the evidence the examiner can use to judge criterion B (how well you understand the field) and criterion A (whether you chose good sources for the question). Sources picked because they came up first in a search are easy to spot.

Four questions for every main source

• Who wrote it, and why? A researcher, a company, a campaign group, a student.
• When? A source can be sound and still out of date for your question.
• What is the evidence for the claim? Data, argument or opinion?
• Where else did you check it? A claim you can find in only one place needs a flag.

Write your answers as one or two sentences next to the first use of each main source. That is evaluation that counts toward D, and it also keeps your bibliography honest. Pick one referencing style (APA is common), use it throughout and keep every claim traceable. Use your own words and short, purposeful quotation. Do not copy the wording of a source, even with a citation, as a substitute for explaining it.

Structure and a 4,000-word budget

The EE needs a title page, a clear introduction, a body, a conclusion and a reference list. Figures and tables should be numbered, labelled and referred to in the text. The title page and layout details come from the IB’s guide and your school, so check them with your coordinator. A good structure is mostly a set of jobs, each done once.

Here is a working budget for a first plan. It is our own heuristic, not an IB rule, and you should adapt it: a practical science essay spends more on method and data, a literature essay more on close reading. The aim is to plan to about 3,700 words and keep a cushion for edits.

Extended Essay word count: what counts and what does not

The essay is capped at 4,000 words. This is a hard limit: examiners do not consider anything after it. The reflective statement is a separate document of at most 500 words and is not squeezed out of the 4,000. The table below follows the IB’s counting rules as we read them. Confirm the details against your school’s copy of the 2027 guide, and ask your coordinator how your school counts.

Three points that catch students out

• The prose around tables still counts. A table or graph is free, but the sentences interpreting it are text. Do not use tables to smuggle in argument.
• Quotations count. A long block quote spends your words on someone else’s writing.
• Do not hide argument in notes or appendices. A short footnote that only gives the original-language version of a translated quotation is an exception, but a note that argues a point counts, and examiners need not read an appendix.

What happens if you go over

Because examiners stop reading at the limit, going over does not just lose you a penalty. It removes the end of your essay. In a normal essay that is the conclusion and much of the evaluation, which hurts criteria C and D together.

Aim for a first full draft of 3,700 to 3,900 words so that edits do not push you over. Always count by the table above and not by your word processor’s total. If you are still over, ask which section carries the least argument. Cut description before analysis, and long quotes before short ones.

Analysis, line of argument and evaluating as you go

Criterion C asks whether your essay has one thread: a question, findings that come from analysing evidence, and a conclusion that answers the question. Criterion D asks whether you can judge your own work: how far your findings can be trusted and what they mean. Together they are 14 of the 30 marks, which is why a clean introduction and a long bibliography are not enough.

A line of argument you can test in two minutes

Print the essay and read only the first sentence of every paragraph. If those sentences, in order, tell the story of your answer, your argument is visible. If they read like a list of topics, the argument is hidden and a reader will struggle to follow it. Then read the introduction and conclusion side by side. The conclusion should answer the question in the same words and add nothing new.

After every table, graph, quote or result, ask yourself what it shows, why it matters and what it does not show. Those three small questions are the difference between description and analysis.

Evaluation belongs in the body

Discussion and evaluation is the largest criterion, and examiners look for it throughout the essay. Judge a source where you first use it. Judge a method where you apply it. Weigh competing explanations as you meet them. Then use the closing section to pull it together: which limitation matters most, which way it pushes the result, and how confident you are in the answer.

Make limitations specific. “Time was limited” tells an examiner nothing. “My sample came from one school, so the result may not hold for older students, and the true effect is probably smaller than I found” says which way the limit pushes and why. In the sciences and maths, include limits of the model (its assumptions) as well as the equipment. In the humanities, include limits of the sources (who wrote them, who was left out).

One more point. A conclusion that changes after a fair re-test is not a failure. It is often the best material for both discussion and your reflective statement, provided you report it honestly. Reporting the result you got beats reporting the result you hoped for.

Reflection: the three sessions, the RPF and the viva voce

The EE includes three mandatory reflection sessions with your supervisor: the first, the interim and the final session, which is the viva voce. After the sessions you write a single reflective statement of at most 500 words on the Reflection and Progress Form (RPF). That statement is the only evidence for criterion E, which is worth 4 marks. If the RPF is written entirely in a language other than the essay’s, one IB instruction says criterion E must receive 0, so write it in the language of your essay.

Take notes on the day

The strongest reflective statements are built from notes. After each session, spend ten minutes writing down what you decided, what you weighed and what you would change. By the end you will have far more than 500 words, and cutting a full set of notes to 500 is much easier than inventing specific examples at the last minute.

What examiners want in the statement

In our reading, a strong statement is evaluative, not descriptive. It picks two or three decisions that changed the essay and says what you weighed. It describes a change in your own thinking. It names concrete skills you have already used elsewhere. It is honest about what did not work. It does not thank the supervisor or repeat the essay’s conclusion.

The viva voce

The viva is a short final interview with your supervisor. It is a chance to reflect on what went well and what was hard, and it is part of how the school confirms the essay is your own work. It is not the place to revise the essay. How long it lasts is set by your school and the IB’s 2027 guide, and sources we have seen disagree, so confirm the length and format with your supervisor. Prepare by re-reading your essay, your notes from the first two sessions and your reference list, and by practising a calm two-minute account of your question, your method and your main finding.

The viva also sits next to your other core work. The EE is part of the Diploma core alongside Theory of Knowledge, and your coordinator can show you the current table for how the two are reported together.

Academic integrity and AI

The EE is your own work, and the IB expects every source you use to be acknowledged. Your supervisor guides you but does not write or rewrite your essay, and the supervisor is asked to confirm that the essay is yours and that your sources are acknowledged. We are not in a position to quote the 2027 guide on the supervisor’s exact duties, so ask your supervisor what they will check and when.

Using AI tools

• AI must not write your essay or your reflection. Passing off AI text as your own writing is dishonest.
• If you use AI output as a source, acknowledge and cite it. Treat it as a starting point you must check against a real source. It is not evidence.
• The IB’s precise wording for the EE is something to check. Read the IB’s academic integrity policy and the referencing section of the 2027 guide, and follow your school’s AI policy where it is stricter.
• If an AI tool flags a weakness in your draft, you decide what to do. Write every change yourself. Never paste suggested wording into your essay.

Other things that count as problems

• Someone else writing, rewriting or editing your sentences. This includes a tutor, a parent or a friend. Feedback and questions are fine. Supplying wording is not.
• Copying text from a source, even with a citation, in place of explaining it.
• Using the same work twice. Reusing your Internal Assessment as your EE counts as duplication of work. You can build the EE on an interest you met in an IA, but the question, the depth of sourcing and the write-up have to be new.
• Invented or untraceable sources. Every source must exist and say what you claim it says.

At Photon Academy we give feedback and ask questions. We never write, rewrite, paraphrase or supply wording for any part of a student’s essay or reflection, and we never replace the school supervisor, who holds the formal role.

A suggested Extended Essay timeline

Schools set their own dates for the EE, so your coordinator’s calendar comes first. What follows is our suggested order, and a sensible gap between steps for about 40 hours of work. Roughly half a year suits most students, and many schools use longer.

Two habits protect your timeline. Start a short reflection log on day one, because it turns criterion E from a scramble into a summary. And freeze your research question early, because every week spent on the wrong question is a week you do not get back.

Extended Essay or Internal Assessment: what is the difference?

Students often treat the EE as a long IA. It is not, and writing it like one is a common mistake. An IA-style write-up (method, data, result, conclusion) is not enough for an EE, which needs a literature base, an argued line and evaluation across the whole piece.

A student who has done an IA often has the experiment but not the framing, so check criteria A and D first. And never lift an IA straight into an EE: the topic, the question and the depth of sourcing are different, and duplication of work is an integrity problem. For the IA itself, our guides cover Maths, Physics, Chemistry and Biology.

The mistakes we see most often, and the fix

These patterns come from teaching and giving feedback on many drafts. Each one is fixable by the student, and none needs anyone to rewrite a sentence.

Extended Essay checklist before you submit

Tick each box when you can point to it in your essay. Your ticks are saved in this browser so you can come back to them.

How Photon Academy helps with the Extended Essay

We give feedback and ask questions. We never write, rewrite or edit a student’s essay or reflection, and we do not supply topics-as-answers, data or wording. Your school supervisor stays in charge of the process, and the work, and the marks, are yours.

Extended Essay help in Photon Academy lessons

Our students can get Extended Essay help as part of their lessons: talking through a topic and research question, planning the method and sources, preparing for the reflection sessions, and feedback on drafts. Our tutors ask questions and point at gaps, and you decide what to change and write every word yourself. Ask us on WhatsApp or book a consultation and tell us your subject and your school’s dates. Subject details are on our tuition pages for Maths, Physics, Chemistry and Biology.

Your school’s own rules come first, so ask your supervisor what outside help they are happy with.