Write it up
Seven sections written at different times over several months, in slightly different voices. Nobody has read them one after another yet. And a dataset you downloaded a while ago: you should have written down its version at the time.
- One report that reads as though one person wrote it in one go, under the limit, fully referenced, and ready to hand over.
You cannot gain new marks here. You can still lose several.
Going over the limit, hiding evidence in an appendix nobody reads, or referencing so loosely that a source cannot be traced. You can fix all three in a few hours, and all three cost more marks than they should.
You have been carrying a figure on every page since step 1. Here they are together. Four follow published guidance for this format and three are ours. The IB sets none of them.
| Section | Crit | Marks | Words |
|---|---|---|---|
| Research question | A | 50 | |
| Environmental issue and background | A | 4 | 500 |
| Strategy, perspectives and tensions | B | 4 | 500 |
| Method | C | 4 | 500 |
| Treatment of data, prose only | D | 6 | 200 |
| Analysis and conclusion | E | 6 | 550 |
| Evaluation | F | 6 | 600 |
| Total | 30 | 2,900 |
Note Criterion D (Treatment of data): six marks for 200 words, because the tables and graphs that earn those marks are not counted at all.
Four things belong at the top of the first page: the title, your personal code, the codes of anyone you collaborated with on methodology, and the word count. There is no requirement for a cover page or a contents page.
The IB publishes no template. Lay the report out in the order of the criteria, A to F, with a clear subheading on each section. Then a marker working down the descriptors never has to search, and headings do not count as words.
Only these six are left out of the 3,000. Everything else you type counts, including the numbered steps of a method. Keep figure captions to a line or two, and do not hide analysis in them.
Everything below is how we suggest you actually do it.
You get one draft read, and it will say "develop this"
Your teacher may comment on one full draft: general remarks on strengths and weaknesses, annotations, advice. They may not correct it. You must make the improvements yourself.
This is the whole reason to aim at 2,900 rather than 3,000. Feedback on an ESS investigation almost always asks you to develop something, not to cut it. At 3,000 words, the only way to act on it is to delete something good. With a hundred words spare, you can improve the report instead of just rearranging it.
Hand over something complete rather than something polished. A finished but rough evaluation gets you useful feedback. A beautiful background with three missing sections gets you comments on the background only. You only get one draft read, so use it on the parts you are least sure about.
The IB publishes exemplar investigations with their marks and the examiner's commentary in the teacher support material. Ask your teacher for two: one strong, one weak. Mark them yourself against the descriptors, then read what the examiner said.
An hour of that teaches you more about what "explains" means than any amount of reading about it. Do it before your own draft, while you can still act on what you notice.
Four things the current criteria no longer ask for
Old templates are still around after the syllabus changes. If you got a structure from an older student, a website or a textbook printed before 2026, check it against this list before you fill it in.
Criterion B (Strategy) is about a real-world strategy. That is a different thing with the same word, and it does not assess your sampling. How your samples were placed still belongs in the method, in a sentence, because a method nobody could repeat fails Criterion C (Method). What has gone is the separate section justifying it.
No marks depend on a risk assessment. Say what was actually risky about this work, in two or three sentences that could only have been written about your investigation. Consent, where people were involved, still clearly belongs in the report.
Limitations, improvements, unresolved questions. Strengths were part of the old criteria and are not part of these. Every sentence praising your own design is a sentence taken from a strand that is marked.
The old course ended by recommending action. This one does not. A section called Application only partly overlaps Criterion B (Strategy), which is worse than not writing it. If you want to say what should be done, do it when you return to the strategy at the end of Criterion E (Analysis and conclusion).
Between them these four can take five hundred words out of a report that needed them elsewhere, and none can earn a mark.
Using AI, and saying so
The IB does not ban AI tools.
The question is whether you used AI to help you learn, or to make it look as if you did something you did not do. A tool can explain a statistical test, suggest search terms, or argue with a plan you have already written. It cannot choose your question, decide what your data means, or write your evaluation: those are the things being marked, and if a tool does them, you lose those marks.
Authenticity. The work is your own. A tool is not an author, and it cannot be listed as one.
Transparency. If you copy, paraphrase or modify anything a tool produced, name the tool in the body at the point you used it and add it to your bibliography. Where school policy and the IB's policy differ, follow the stricter one.
The declaration. The submission carries a declaration of AI use. It does not replace naming the tool in the body, and naming it does not replace the declaration. Do both, and be exact: "used to explain what a Mann-Whitney U test does" is a better answer than a tick.
Ask a chatbot for sources and it will sometimes invent them, with realistic-looking authors, journal titles and page numbers. Criterion A (Research question and inquiry) wants citations detailed enough to be traceable. So an invented reference loses a mark and also counts as academic misconduct.
The rule that prevents it: every source in your bibliography is one you have actually opened. If you cannot find it by searching its title, it may not exist.
Before you sign it off
You will confirm this is your own work. This can be checked against your proposal, your draft, your references and your usual writing. Where you worked with others on methodology, the writing must still be entirely yours. And the same piece of work cannot be submitted for both this and your extended essay.
Your work is moderated (checked) by somebody outside your school, so your identity and your school's must not appear anywhere in the report. The obvious half is names: yours, your teacher's, the school's, a classmate's. Your personal code is the only identifier the document carries.
The half people miss is images. A face, a name on a bag, a logo on a fleece, a minibus, a sports hall, a signed consent form photographed for the appendix, a screenshot with your account name in the corner of the spreadsheet. Crop it, blur it, retake it, or leave it out. Anything you did not take yourself is cited like any other source. Anything you did take is credited as "author's own photograph", which credits you without naming you.
Eight steps, one environmental issue, and a report that argues something about a real place.
The part worth keeping, long after the marks come back, is the idea in the middle. A decision about a hillside involves people who want things that cannot all happen together. Measuring the ground carefully can inform that argument without settling it. That is the whole subject, in 2,900 words.
Referencing a dataset
The IB does not set a referencing style; your school does. At the International School of Geneva - Campus des Nations it is Chicago footnotes, used across every IA and the extended essay, so you learn one system and reuse it everywhere.
Whatever style you use, the requirement is traceability. On this route that word matters more than usual, because your evidence is a file that can be revised, moved or replaced. A reference that gets somebody to the publisher is not enough. It has to get them to the version you used.
If the download offers a DOI (a permanent digital identifier), use it instead of the URL. It points to one specific version, and it still works if the publisher reorganises their website. A URL does not.
Two things in those cards are worth copying. The dataset has one author, because the programme that catches the animals also publishes the file. A compiled file has two, the body that collected the numbers and the body that tidied them, and both are credited, so check who did each job before you write the note. And the last card quotes no title, because I could not read one. A plain description is honest; a title in quotation marks claims those are the page's own words, and a guessed one is a reference nobody can find.
The last read-through
Read the whole thing once, in order, out loud if you can. Seven sections written over several months do not automatically agree with each other. You cannot see the disagreements while you are working on one section at a time.
On this route there is one specific thing to look for: a claim that is bigger than your rows. It gets in through ordinary words. A study of thirty monitoring stations becomes a study of a river; a study of what a programme caught becomes a study of what it killed; a study of four areas becomes a study of a whole state.
An early draft of the white shark study kept saying killed. "The share of the sharks killed by each gear that were white sharks." It is the natural word for a shark-control programme, and it is wrong: the file records animals caught, and 65 of the 147 white sharks were alive when the gear was checked. Whether an animal was released, died or was killed sits in a separate column, Fate, which my question never used.
The word had reached the title, the research question and the conclusion before anybody noticed. Each time it appeared, it made a claim the data could not support. Nobody had lied. One verb was claiming more than the data showed, and in a report about a programme people argue over, it was also quietly taking a side.
The fix was the same everywhere: say what the number is. The file counts sharks caught, so the word is caught. It costs nothing, and it makes a moderator trust the report instead of starting to check it.
Search your finished report for the verbs, not just the nouns. Killed, stopped, eliminated, cleaned up: each one claims an action, and your file almost certainly recorded a number instead.
It can respond to prose you have already written: where a paragraph is unclear, where you have said the same thing twice, whether a sentence means what you intended.
It cannot write or rewrite it for you. And whatever you did use it for must be named at the point you used it and listed in your bibliography. The section above explains how.
Ready to submit
Secondary data checklist0 of 15Once you submit the final version you cannot take it back, so go through this slowly.
The white shark investigation used on the secondary-data route of this guide is the author’s own analysis of a published dataset: the Queensland Shark Control Program’s record of every animal caught on its nets and drumlines, published by the Queensland Government under CC BY 4.0 and downloaded on 26 September 2026. The choice of the four areas, the analysis and the conclusions are the author’s, not the Queensland Government’s.