Check the final version criterion by criterion, avoid the commonest reasons for low marks, and understand how your teacher's marks are moderated by the IB.
The final version of the exploration is the one that is marked, and once it has been officially submitted it cannot be withdrawn or replaced. The last week before the deadline is therefore worth planning as carefully as the first. Many of the marks lost at this stage go on small, fixable problems: an undefined variable, a graph with no axis labels, an aim the conclusion forgets, a source that is never cited.
A systematic check, one criterion at a time, catches most of them. It works best a few days before the deadline, with time to act on what it finds, and after at least a day away from the draft, so that you read what is on the page rather than what you meant to write.
Before looking at quality, make sure the report meets the basic requirements. Any of these missing can cause problems regardless of how good the mathematics is.
Check | Requirement |
|---|---|
Cover page | States the title of the exploration and the number of pages |
Length | About 12–20 pages, double-spaced, including graphs and tables, excluding the bibliography |
Bibliography | Present, detailed, in one consistent style |
Citations | Every source cited where it is used, including data, formulae, images and any AI output |
Data | Own data collected individually; any group data combined and described |
Appendices | Only large data sets and supplementary material; nothing the argument needs |
Authenticity | Your school's declaration or form completed, confirming the work is your own final version |
Format | Whatever your school specifies (file type, page numbers, file name) |
Presentation rewards a report that is coherent, well organised and concise. Read the introduction and the conclusion one after the other: they should state and answer the same aim, in the same terms.
Question to ask | If the answer is no |
|---|---|
Does the introduction state a precise aim? | Rewrite it with the situation, variables, kind of result and scope |
Does the conclusion answer that aim directly? | Add a first sentence that does |
Does every section have a clear heading and a linking first sentence? | Add them |
Are all key graphs and tables in the body, next to the text that uses them? | Move them out of the appendices |
Is any calculation, graph or description repeated or irrelevant? | Show one instance and tabulate the rest; delete the irrelevant |
Is the length within about 12–20 pages? | Cut background and repetition first |
Mathematical communication rewards correct, consistent notation and clear representations. Go through the report with a highlighter and mark every variable at its first appearance; each must be defined there.
Question to ask | If the answer is no |
|---|---|
Is every variable defined in words, with units, at first use? | Add a defining sentence |
Is each symbol used for one thing only, from start to finish? | Make the notation consistent |
Is there any typed calculator or computer syntax ( | Replace it with mathematical notation |
Does every graph have a number, caption, labelled axes with units and a key? | Add what is missing |
Does every table have a number, title and units in the headings? | Add what is missing |
Is rounding consistent, and is used for rounded values? | Apply one stated rule |
Personal engagement must be visible in the work. Find three places where a decision of yours changed the direction of the exploration; if you cannot, the report probably reads as a textbook exposition.
Question to ask | If the answer is no |
|---|---|
Is there evidence of your own thinking beyond the introduction? | Explain the reasons for decisions where they were made |
Did you make and test a prediction anywhere? | State the prediction you had, and what the result showed |
Is the mathematics presented in your own way (own data, diagrams, examples)? | Replace copied explanations with your own |
Would the report be different if someone else had written it on the same topic? | Add your own perspective, comparisons or extensions |
Reflection should run through the report, not appear only at the end. Check that every major result is followed by a comment on what it means, how reliable it is and what it changes.
Question to ask | If the answer is no |
|---|---|
Is each major result followed by an interpretation in context? | Add one or two sentences after it |
Are limitations specific to your work, with their likely effect? | Replace generic limitations with specific ones |
Are different methods or models compared, with reasons for the choice? | Add the comparison |
Does the conclusion say how far the aim was achieved and suggest a specific next step? | Add both |
Use of mathematics rewards relevant mathematics at the level of the course, correct, with understanding demonstrated throughout. Recheck every calculation (ideally by a second method) and every claim the argument depends on.
Question to ask | If the answer is no |
|---|---|
Is all the mathematics relevant to the aim? | Remove what is not, however impressive |
Is it at the level of the course, not only prior learning? | Add course-level analysis where it serves the aim |
Is each method justified before it is used, with its conditions checked? | Add the justification and the checks |
Is each technology result explained, with what was entered? | Add the method and one instance in full |
Have you checked every calculation, and do results make sense in context? | Recompute; investigate anything unreasonable |
(HL) Is every claim justified, and is there evidence of sophistication or rigour? | Add justification; compare methods or link areas of mathematics |
(HL) Is every value error-free and given to a suitable accuracy throughout ("precise", needed for level 6)? | Recheck values and apply one stated rounding rule |
Your teacher's comments on the draft are the only outside advice the final version will receive, so each comment should become a specific action. Teachers point out that an error exists without correcting it, and they suggest rather than rewrite, so a comment such as "check your residuals in section 3" means that you have to find and fix the problem yourself.
A simple method is to copy every comment into a list, write the action it requires next to it, and tick it off only when the change is made and checked. Comments about one place often apply elsewhere too: if one graph lacks units, check every graph; if one method lacks a justification, check every method.
Draft comment | Action |
|---|---|
"The aim is broad." | Narrow it to one question; rewrite the introduction and conclusion to match |
"There is an error on page 7." | Recompute page 7 step by step with a second method; follow the corrected value through every later result |
"Why this model?" | Add the reason from the context and the evidence from the graph or residuals |
"Results are described, not evaluated." | After each result, add what it means, how reliable it is and what it changes |
Mathematical errors cost marks on criterion E, and an error that leads to an unreasonable conclusion also undermines the reflection built on that conclusion (D). Rechecking is most effective when it uses a different route from the original calculation.
Recompute by a second method. A regression from the GDC can be checked in a spreadsheet; a hand-solved system of equations can be checked by substituting the solution back.
Check units and orders of magnitude. A car that loses 150% of its value, a probability above 1 or a population of 3 billion in a small town shows an error somewhere earlier.
Check the follow-through. When one value changes, every result that depends on it must change too, including numbers quoted in the conclusion.
Check that graphs match the text. The equation in the caption should be the one plotted, and the values read from a graph should agree with those calculated.
The same problems appear across many explorations. Each one is listed with the criterion it hurts most and how to avoid it.
Problem | Criterion | How to avoid it |
|---|---|---|
Interesting context but mathematics at prior-learning level | E | Use course-level mathematics the aim needs |
Textbook-style exposition of a known result | C | Add your own data, angle or extension |
Results described but not evaluated | D | Interpret, compare and state specific limitations after each result |
Key graphs in appendices; unlabelled axes | A, B | Place graphs in the body with full labels and captions |
Undefined variables; calculator syntax in the text | B | Define every variable; use proper notation |
Pages of repeated calculations | A | One full instance, then a table |
Mathematics far beyond the student's understanding | E | Use mathematics you can explain fully |
Regression equations or p-values quoted without explanation | E | Justify the method, check conditions, interpret the output |
Chi-squared test with expected frequencies of 5 or less, unremarked | E, D | Combine categories or acknowledge the limitation |
Missing citations for data, formulae or images | Academic integrity | Cite every source where it is used |
Group data presented as one's own | Academic integrity | Collect your own data; describe any combining |
Your teacher marks every exploration in the school against the five criteria and confirms that each is the student's own work. The IB then asks for a sample of the explorations, which a moderator, an experienced teacher trained by the IB, re-marks against the same criteria. If the moderator's marks are close to the teacher's, the school's marks stand. If the teacher's marks are consistently too high or too low, the IB adjusts the marks of all the school's students in that group, not only those in the sample.
That has two consequences for you. First, the mark your teacher gives is provisional until moderation is complete; it can go up or down. Second, the safest way to keep a mark through moderation is to make the evidence for it obvious on the page. A moderator never meets you and sees only the report, so a level that depends on something you said in class, or on a step you did not write down, cannot be credited.
Once the final version has been submitted, it cannot be withdrawn, and your teacher cannot give further advice on it. Any academic integrity concern found after submission is handled under the IB's procedures. Keep your notes, data files and drafts until results are published, in case questions about authenticity arise.
A realistic plan for the final week leaves time for each kind of check and avoids last-minute changes.
Day | Task |
|---|---|
1 | Read the whole report as a stranger; list every problem without fixing anything |
2 | Fix the mathematics: recheck calculations, conditions and interpretations (E) |
3 | Fix the argument: aim, conclusion, headings, linking sentences, conciseness (A, D) |
4 | Fix the notation, graphs and tables (B); add any missing reflection (D) |
5 | Check citations and bibliography; complete the cover page and declaration |
6 | Final read-through; small corrections only |
7 | Submit |
The submitted final version is the one that is marked and cannot be withdrawn, so plan the final checks a few days before the deadline.
Check the requirements first: cover page with title and page count, about 12–20 pages, a full bibliography, citations, own data and the authenticity declaration.
Then check each criterion in turn: aim and conclusion, conciseness (A); notation, definitions and labelling (B); visible decisions of your own (C); reflection after every result (D); relevant, correct, justified mathematics (E).
The commonest losses are prior-learning mathematics, textbook expositions, results not evaluated, graphs in appendices, calculator notation, repetition and unexplained technology output.
Moderation: the IB re-marks a sample and can adjust every student's mark in the school, so the teacher's mark is provisional.
Evidence for every level must be on the page, easy to find, because the moderator sees only the report.
My cover page gives the title and the number of pages.
My report is about 12–20 double-spaced pages, excluding the bibliography.
My introduction states a precise aim and my conclusion answers it in the same terms.
Every key graph and table is in the body, numbered, captioned and referred to in the text.
Every variable is defined with its unit at first use, and no calculator syntax remains in my own text.
I can point to at least three decisions of my own that moved the exploration forward.
Every major result is followed by an interpretation and a comment on its reliability.
My limitations are specific to my work and say how they affect the conclusion.
Every method is justified, its conditions checked, and one instance of each calculation shown in full.
I have rechecked every calculation, ideally by a second method.
Every source, including data, formulae, images and any AI output, is cited in the text and listed in the bibliography.
Any data collected with others is described, and my own contribution is clear.
I have completed my school's authenticity declaration and kept my drafts and data files.
With a free account, Augment remembers where you stopped in the notes and which questions you have marked.