How many marks is the IB Math AI exploration worth?

The current IB Math AI exploration is marked out of 20 and contributes 20% of the final subject grade. Presentation and Mathematical communication are worth 4 marks each. Personal engagement and Reflection are worth 3 marks each, while Use of mathematics is worth 6. The five scores are added rather than converted into a separate IA grade before being submitted. Final subject grade boundaries are determined separately, so a particular raw IA score does not automatically correspond to a fixed IB grade.

Are the IB Math AI IA criteria the same at SL and HL?

Criteria A-D use the same names, maximum marks, and descriptors for SL and HL. Criterion E, Use of mathematics, has different expectations at the two levels. SL students can reach the top level through correct, relevant, course-appropriate mathematics that demonstrates thorough understanding. HL students must also demonstrate the sophistication and rigour expected at HL, with both required for the top mark. Students should therefore check the level-specific Criterion E descriptors rather than assuming the entire rubric is identical.

Can simple mathematics score highly in a Math AI IA?

Yes, provided the mathematics is appropriate to the course, relevant to the aim, and developed with sufficient understanding. A focused statistical or modelling investigation can outperform an unnecessarily advanced exploration that the student cannot explain. At SL, mathematics beyond the syllabus is not required for full marks. At HL, the work must reflect HL expectations and demonstrate greater sophistication and rigour in the upper bands. In both cases, the best method is the one that genuinely helps answer the research question.

How do I show personal engagement without using personal data?

Personal data is not required because engagement is demonstrated primarily through mathematical ownership. You can formulate an original question, make and test a prediction, justify your choice of model, or adapt a method when its limitations become clear. Comparing approaches and explaining why one suits the context can also provide strong evidence. A topic can use a public dataset and still feel distinctly yours if your decisions shape the investigation. Conversely, collecting your own data will not secure high marks if the subsequent analysis merely copies a standard procedure.

Should reflection appear only in the conclusion?

No, reflection can and usually should appear throughout the exploration. After obtaining an important result, explain what it means, whether it is reasonable, and how it affects the next stage of your investigation. If a model performs poorly, analyzing that weakness and refining the method creates stronger evidence than postponing all evaluation until the end. The conclusion should still synthesize the main findings, answer the research question, and identify important limitations. A final generic paragraph listing human error and sample size is rarely enough for substantial critical reflection.

Will the IB Math AI IA criteria change for future students?

Yes, the IB has published a redesigned Mathematics: Applications and Interpretation framework for first assessment in 2029. That framework uses four stages connected to the mathematical inquiry process rather than the familiar five-criterion, 20-mark structure explained here. Students assessed before the new course takes effect should continue using the current criteria provided by their school. Students in the 2029 assessment cohort must follow the new subject guide and should not use an older online checklist as their authoritative rubric. Your teacher and school's programme documentation are the safest sources for confirming which syllabus applies to your cohort.

Back to all posts

IB Math AI IA Criteria: How It Is Actually Marked

IB Mathematics Internal Assessment IB IB Math IA

• 1 Oct 2026•11

IB Math AI IA Criteria: How It Is Actually Marked

Join 650k+ students breezing through the IBReady to breeze through the IB?

IB Math AI IA Criteria: How It Is Actually Marked | RevisionDojo

Back to all posts

IB Math AI IA Criteria: How It Is Actually Marked

IB Mathematics Internal Assessment IB IB Math IA

• 1 Oct 2026•11

The IB Math AI IA criteria divide the mathematical exploration into five assessed areas worth 20 marks in total. Your teacher marks the exploration, and the IB may externally moderate the school's marking. The criteria are Presentation, Mathematical communication, Personal engagement, Reflection, and Use of mathematics.

This article explains the current framework for Mathematics: Applications and Interpretation, first assessed in 2021. The IB has announced a redesigned course for first assessment in 2029, so students in that cohort must follow the new guide supplied by their school.

The Math exploration mark scheme

The exploration is compulsory at SL and HL and contributes 20% of the final subject grade, according to the official IB Mathematics: Applications and Interpretation subject brief. It is an individual written investigation into an area of mathematics chosen by the student.

Criterion	Focus	Maximum marks
A	Presentation	4
B	Mathematical communication	4
C	Personal engagement	3
D	Reflection	3
E	Use of mathematics	6
Total	Complete exploration	20

Each criterion is assessed separately using a best-fit judgment. Attractive formatting cannot compensate for weak mathematics, while sophisticated calculations do not excuse unclear notation or superficial evaluation. Because moderation considers the submitted work, every achievement must be visible in the report rather than depending on what your teacher knows about your effort.

Criterion A: Presentation, 4 marks

Presentation concerns the exploration's organization, coherence, and concision, not decorative design. A top-level report has a clear aim, develops its mathematics logically, and reaches a conclusion that answers the original question.

Graphs, tables, and diagrams should appear near the relevant discussion. Appendices should not contain essential reasoning that the reader needs to understand the investigation.

Concision does not mean extreme brevity. It means removing material that does not advance the exploration. Repeated calculations, long textbook explanations, raw calculator output, and unexplained data tables may prevent an otherwise organized report from earning full marks.

A practical structure is:

Context, motivation, and research question
Data, variables, assumptions, and method
Mathematical development
Interpretation and evaluation
Conclusion

RevisionDojo's step-by-step Math AI IA guide explains how to move from an initial question to a structured report.

Criterion B: Mathematical communication, 4 marks

This criterion assesses whether the reader can follow the mathematics through correct and consistent notation, terminology, definitions, representations, and reasoning. Full marks require relevant and appropriate communication throughout, not merely in isolated sections.

Define variables when they first appear. For example, state that t represents time in minutes before using it in a model. Include units, label graph axes, identify table columns, and use symbols such as correlation coefficients or standard deviation accurately.

Technology output still requires explanation. If software produces y = 2.41e^(0.083t), define both variables, explain how the model was selected, and interpret its parameters in context. A screenshot alone does not demonstrate mathematical communication or understanding.

Frequent mistakes include using an equals sign when an approximate sign is needed, changing symbols midway, omitting units, and inserting spreadsheet output without discussion. The RevisionDojo Math AI IA checklist can support a final notation and formatting review.

Criterion C: Personal engagement, 3 marks

Personal engagement measures how successfully you make the mathematical exploration your own. It is not a direct measure of enthusiasm, effort, time spent, or how personal the topic sounds.

A short anecdote in the introduction is therefore insufficient for 3 marks. Stronger evidence includes designing a data-collection method, making and testing a prediction, justifying the choice between models, adapting an unsuccessful method, or examining the problem from an original perspective.

Suppose you investigate running pace and heart rate. Collecting personal data establishes context, but deeper engagement appears when you decide how to control warm-up effects, identify unreliable readings, compare possible models, and justify which model best addresses your question.

Statements such as “I found this interesting” add little unless connected to a mathematical decision. Math AI IA exemplars can show how engagement becomes visible, but another student's decisions or structure should not be copied.

Criterion D: Reflection, 3 marks

Reflection evaluates how well you interpret, analyze, and evaluate the investigation. Stating a result or listing generic limitations usually demonstrates only limited reflection.

Consider a regression with a high coefficient of determination. Calling the model accurate solely because that value is high is incomplete. Stronger reflection examines residuals, outliers, assumptions, the model's valid domain, whether the relationship is causal, and whether extrapolation is defensible.

Reflection is strongest when it develops the investigation. If a linear model creates a curved residual pattern, recognizing the problem and testing a nonlinear alternative evaluates the first method while improving the analysis. This is more effective than postponing every evaluative comment until the conclusion.

Ask throughout the report:

Does this result answer the research question?
Is the result reasonable in context?
Which assumptions affect it most?
What does the model fail to represent?
Would another method be more suitable?
How could the data or model be improved?

The RevisionDojo Math AI coursework guide provides a useful basis for auditing a draft against each criterion.

Criterion E: Use of mathematics, 6 marks

Use of mathematics carries the most marks. The mathematics must be relevant, appropriate for the course level, sufficiently accurate, and explained well enough to demonstrate understanding.

Relevant mathematics advances the investigation toward its aim. Adding calculus, another regression, or an advanced formula solely to appear sophisticated can weaken the report if the method has no investigative purpose.

Level	What is needed for high marks
SL	Relevant course-level mathematics that is mostly or fully correct, with good to thorough understanding
HL	Correct, relevant HL-level mathematics showing thorough understanding, with sophistication and rigour required in the highest bands

At SL, mathematics beyond the syllabus is not required. A well-explained investigation using regression, statistical testing, probability, trigonometry, or financial modelling can score highly when the methods suit the question.

At HL, sophistication may come from challenging concepts, recognizing mathematical structure, connecting different areas, or approaching the problem from multiple perspectives. Rigour means supporting methods and claims with clear logical reasoning. For 6 marks, the mathematics must be precise and demonstrate both sophistication and rigour alongside thorough understanding.

Technology is expected and valuable in Math AI, but pressing buttons is not evidence of understanding. Explain why each method was selected, show representative working, interpret outputs, and test whether results are reasonable. Targeted practice through RevisionDojo's Math AI resources is preferable to inserting advanced mathematics that you cannot explain.

How the criteria work together

The criteria are distinct, but one passage can provide evidence for several. A residual plot can support Presentation when integrated clearly, Mathematical communication when labelled correctly, Reflection when used to evaluate fit, and Use of mathematics when interpreted accurately.

However, impressive software output may still receive limited credit under Criterion E if the underlying mathematics is unexplained. Similarly, choosing a personal topic does not secure Criterion C marks when every mathematical decision follows a standard template.

Common misconceptions

More advanced mathematics always earns more marks. Relevance and understanding matter more than unnecessary complexity.
Personal engagement means liking the topic. Independent mathematical choices provide stronger evidence.
Reflection belongs only in the conclusion. Strong reflection normally appears throughout the investigation.
A longer IA looks more substantial. Repetition and irrelevant material can reduce concision.
Calculator output proves the mathematics. Results must be explained, interpreted, and evaluated.
Effort earns criterion marks. Only evidence visible in the submitted exploration can be assessed.

Final criterion-by-criterion check

Before submission, identify specific passages that provide evidence for every criterion:

A: Can the reader follow the aim, method, development, and conclusion?
B: Are variables, notation, units, graphs, and technology outputs explained consistently?
C: Do your choices, predictions, adaptations, or perspectives shape the mathematics?
D: Do you evaluate results and use reflection to develop the investigation?
E: Is the mathematics relevant, accurate, level-appropriate, and demonstrably understood?

Conclusion

The current IB Math AI IA criteria reward a coherent mathematical investigation, not a collection of impressive-looking calculations. Strong explorations combine accurate mathematics with clear communication, independent decisions, and critical interpretation of what the results can and cannot establish.

Use the rubric during planning rather than waiting until the final edit. RevisionDojo's IA guides, exemplars, Coursework Review, and IA Feedback can help identify missing evidence while keeping the mathematical decisions and explanations genuinely your own.