Writing it up
All six sections drafted, every figure made, and your references collected as you went. If you have not been keeping references, that is the first job on this page and it will take longer than you expect.
- One report under 2,500 words that reads as though one person wrote it
- A word count broken down by section
- Every figure numbered, integrated and referred to
- Consistent references, with access dates
- Any AI use cited where it happened
- A single PDF, ready to submit
2,500 words, and a surprising amount of your report does not count towards them.
Examiners are advised to stop reading at the limit, so anything over it is simply not marked. The good news is that the exclusions list is long, and almost everything you built in steps 4 and 5 sits outside the count.
Title page · acknowledgments · contents page · titles and subtitles · references · footnotes up to 15 words each · map legends and keys · labels and annotations of 10 words or fewer · tables of statistical or numerical data, categories or class names · calculations · appendices containing only raw data and calculations.
Spend freely here. This is where the evidence lives.
All the main text: the research question, the introduction, the methodology, the prose justifying your treatment and presentation of the data, the written analysis, the conclusion and the evaluation. Plus any annotation over 10 words, and any footnote over 15. Text inside a box or a shape counts too, wherever it sits on the page: putting a paragraph in a text box beside a figure does not turn it into a label.
2,500 words in total, across everything above.
| Section | Criterion | Marks | Words |
|---|---|---|---|
| Introduction: question, hypotheses, syllabus link | A | 3 | 300 |
| …and geographic context, with the locational map | A | (included above) | |
| Methodology | B | 3 | 300 |
| Treatment of information collected | C | 6 | 500 |
| Written analysis | D | 8 | 850 |
| Conclusion | E | 2 | 200 |
| Evaluation | F | 3 | 300 |
| Total | 25 | 2,450 |
These per-section counts are ours, from the support booklet, the IB sets only the 2,500 total. The 50 words in hand are deliberate: feedback on a draft almost always asks you to develop something, and at the ceiling the only way to do that is to delete something good.
Everything below is how we suggest you actually do it.
Six sections, one report
You have written this in pieces, over weeks, and it will read that way unless you deliberately fix it. The commonest symptoms are a variable that changes its units between sections, a site that is "site 7" in one place and "the bridge site" in another, and figure numbers that do not run in order because two were added late.
Read the whole thing start to finish in one sitting, on paper or on a screen you are not editing. You are not looking for typos on this pass; you are looking for the joins.
One deliberate repetition is worth keeping. Your fieldwork question should appear in the introduction, be visibly answered in the analysis, and be answered again in the conclusion. That is not redundancy; two separate criteria ask whether the question gets answered.
Put the word count at the end of each section as well as the total. It is a top tip rather than a requirement, and it makes it obvious to an examiner that you are inside the limit, which is worth more than it sounds when the alternative is them wondering.
Your teacher may comment on one full draft: what is strong, what is weak, how it could be improved. What they may not do is correct it or heavily annotate it, and the next version you hand over is the final one.
So hand over something complete rather than polished. A rough-but-finished evaluation gets you useful feedback; a beautiful introduction and three missing sections gets you comments on the introduction.
After the final deadline the report does not come back for tweaking. Whatever is in it is what gets marked.
Figures, in order, in the body
Number every figure in the order it appears and refer to each by number in the text beside it. If a figure is never referred to, either the reference is missing or the figure is.
Everything that matters goes in the body. Appendices may hold raw data and calculations, but they are not read as evidence, so a graph exiled to the back is a graph that does not exist. Since tables of numerical data do not count towards your word limit wherever they sit, there is no reason to move them out of the body in the first place.
Remember the whole thing is submitted as a single PDF and marked on a screen. Check it as a PDF, at 100%, before you submit: figures that were crisp in a document sometimes come out unreadable, and an A3 map reduced to a page can lose its labels entirely.
The class process page allows one exception: a single A3 map, and the locational map is what it is for. So a map that would be illegible shrunk onto a portrait page does not have to be shrunk, but it is one sheet, not a habit, and check with your teacher before relying on it. The support booklet asks only for a single PDF and does not mention A3.
Everything else lives inside the PDF at page size, which is the constraint to design your figures against.
Axis labels and legends still legible at page size. Map text still readable. Photographs not compressed into mush. Nothing running off the edge of a page. Figure numbers still matching the text after everything reflowed.
This is a ten-minute check that regularly saves a mark in Criterion C (Quality and treatment of information collected), where conventions are explicitly assessed.
Referencing, including the awkward sources
The IB does not prescribe a 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 the style, the requirement is the same: every source of secondary information referenced, consistently, from the first note to the last, with a URL and a date of access for anything online.
Most of your sources here will be awkward ones. A style guide covers books and journal articles well; it says much less about a cantonal mapping portal, a commune's flood-protection scheme or an interpretation panel screwed to a post beside the river, which is exactly what you will be citing. When the guide runs out, fall back on the principle underneath it: traceability. Author or organisation, date, title, and where you got it, in that order, with enough detail that a reader could stand where you stood.
A caption like that is worth copying: it tells a reader what they are looking at, exactly where it is, and whose photograph it is. Note that it is a caption, not an annotation, at eighteen words it counts against your limit. The ten-word exemption is for text on the figure itself, so wherever a point can be made by annotating the image rather than describing it underneath, annotate it.
Being honest about AI
The IB's position is not a ban. Two requirements, and they are both about honesty rather than abstinence.
Authenticity. The work has to be yours. A tool is not an author, and the writing, the justification of your methods, the analysis and the conclusion must be your own; the guide says so explicitly about group fieldwork, and it applies just as much here.
Transparency. If you use text or any other output from an AI tool, copied, paraphrased, or an image you modified, you must cite it at the point of use, in the body of the work, and add it to the bibliography. A reference at the end alone is not enough.
There is one failure mode worth naming, because it is specific to these tools and it is serious. Ask an AI for sources and it will sometimes produce references that look perfect and do not exist. Fabricated references collide head-on with the requirement that every source be traceable, and an examiner who cannot find a source you cited has a problem that is much larger than a mark.
Write the report in a Google Doc shared with your teacher from the start. Its version history shows the work being built; a record of authorship that no finished PDF can give you.
That protects you. In an investigation where the question and the data are shared with the whole class, being able to show your own report taking shape over weeks is the strongest evidence there is that the writing is yours.
Not through detection. Through blandness.
Generated prose about river fieldwork describes the Bradshaw model competently and says nothing about your nine sites, your anomalous reading, or the reach the renaturation scheme is reshaping. It reads like the middle bands because that is what it is: general, plausible, unattached to any particular river.
The sentences that earn the top bands are the ones only you could write.
What gets submitted
The final version goes in as a single PDF, uploaded to ManageBac, complete with every map, graph and photograph in place. The draft is handed in on paper with the figures included, and uploaded as a PDF too.
Your report may be sent to an examiner for moderation, and you will not be told whether yours was in the sample. Write it for someone who has never met you, has never seen the Asse, and is reading it on a screen.
The deadlines for your cohort are on the class page and in the support booklet, not here; this guide is written to be true whichever year you are in.
The last useful thing you can do is read the analysis aloud. Sentences that sounded fine on screen turn out to have no verb, or to explain nothing, or to repeat the paragraph above.
It is also the fastest way to catch the one thing a spellcheck never will: a sentence that describes where you meant it to explain.
It can respond to prose you have written. Paste a paragraph in and ask whether the explanation is clear, whether the link to the hypothesis is visible, or whether you are describing where you meant to explain.
It cannot write it. And if you use a tool at this level, say so, a citation at the point of use plus an entry in the bibliography. The IB does not ban AI; it requires that the work is yours and that any tool is acknowledged.
Ready to submit?
0 of 14The last four boxes are the ones people skip when they are tired, and they are the ones that catch the errors nobody else will find for you.
Anything marked “our river” is specific to the fieldwork we do together on the River Asse at Nyon: the sites, the equipment and the way we collect the data. Site coordinates and altitudes are from our own fieldwork records and the exact list can change from year to year, so check them against the sheet you are given on the day. Photographs are the author’s own, taken at the river.