The best IB Math AI EE topics use mathematics to investigate a focused real-world problem. Strong options for fall 2026 include climate modelling, transport networks, queue simulations, sports probability, population forecasting, voting systems, and renewable-energy prediction.
A Mathematics Extended Essay must be driven by mathematical reasoning, not simply an interesting context. You need to explain methods, justify assumptions, interpret results, and evaluate limitations. The ideas below provide practical starting points rather than questions to copy unchanged.
Is there officially a Math AI Extended Essay?
The IB does not define a separate Mathematics: Applications and Interpretation Extended Essay category. You submit a Mathematics EE, while your Math AI background may influence your choice of statistics, modelling, simulation, graph theory, or technology.
This distinction matters because the essay cannot become a discussion of climate science, economics, sport, or technology with a few calculations attached. The mathematics must shape the research question, method, analysis, and conclusion.
The official Math AI subject brief emphasizes mathematics in context, modelling, statistics, technology, and interpretation. These priorities make Math AI students well placed to compare models and examine what conclusions the evidence supports.
Students completing the EE for first assessment in May 2027 should follow the updated framework described in the IB Extended Essay curriculum update. The essay remains an independent research project of up to 4,000 words. Confirm your examination session and required guide with your coordinator, since school deadlines occur before the official assessment session.
What makes an IB Math AI EE topic suitable?
A promising topic should require more than calculating a correlation or producing one regression equation. It should create a mathematical choice that you must investigate and justify.
For example, rather than asking whether temperature affects electricity demand, compare several models, inspect their residuals, and test predictions using withheld data. This produces a sustained investigation with meaningful evaluation.
Before committing, check for:
- Mathematical focus: The central models, algorithms, or structures are identifiable.
- Manageable scope: The question uses one dataset, place, system, or defined period.
- Accessible evidence: Reliable data are already available.
- Comparative depth: You can compare methods, parameters, scenarios, or errors.
- Personal understanding: You can explain every important method and output.
- Evaluation potential: The topic allows discussion of uncertainty, bias, sensitivity, and limitations.
IB Math AI EE topics to scope for fall 2026
Adapt each idea to a location, dataset, sport, or system you can investigate responsibly. Discuss the final wording with your supervisor before beginning extensive work.
| Topic area | Possible research angle | Suitable mathematics | Scope warning |
|---|---|---|---|
| Urban temperature | Compare models of seasonal temperature at one station | Periodic functions, regression, residuals, RMSE | Do not model global climate change |
| Public transport | Measure how one route closure affects local accessibility | Graph theory, weighted networks, shortest paths | Select a small network |
| Waiting lines | Compare predicted and observed cafeteria waiting times | Distributions, expected value, simulation | Avoid identifiable personal data |
| Sports strategy | Estimate scoring probability from distance and angle | Logistic regression, probability, calibration | Do not claim causation |
| Population growth | Compare exponential, logistic, and piecewise models | Functions, parameter estimation, error measures | Study one city or country |
| Renewable energy | Predict solar output from daylight and weather variables | Multiple regression, correlation, validation | Limit the number of predictors |
| Disease modelling | Test how an SIR simulation responds to parameter changes | Recurrence relations, numerical methods, sensitivity | Avoid medical recommendations |
| Voting systems | Compare winners under simulated voter preferences | Monte Carlo simulation, matrices, ranking algorithms | Compare only 2-3 systems |
| Game fairness | Test how one rule changes expected value | Conditional probability, Markov chains, simulation | Keep the state space manageable |
| Resource allocation | Optimize a club schedule or purchasing plan | Linear programming, inequalities, sensitivity | Use realistic constraints |
Three especially promising directions
Climate and energy modelling
Public environmental data are abundant and suitable for model comparison. NOAA Climate Data Online provides historical measurements from individual weather stations.
A focused question could compare sinusoidal, polynomial, and moving-average models for monthly temperature. You could calculate RMSE, inspect residual patterns, validate predictions against unused observations, and evaluate how extreme events affect reliability. The mathematical issue is model suitability, not climate policy.
Transport networks
Network investigations combine discrete mathematics, algorithms, technology, and interpretation. If a transport authority publishes GTFS schedule data, stops can be represented as vertices and connections as weighted edges.
A manageable question might examine how closing one station changes mean shortest travel time within a selected subnetwork. Depth can come from comparing distance-based and time-based weights or testing several disruption scenarios. Explain how the algorithm works rather than treating software as a black box.
Population modelling
The World Bank Open Data platform provides downloadable indicators and supporting metadata. This makes it practical to compare exponential, logistic, or piecewise models for one population measure.
Avoid placing many countries into one regression. A stronger essay estimates parameters for a focused case, validates the model against withheld years, and evaluates whether structural changes make its predictions unreliable.
Turning an idea into a focused question
A practical Math AI research question normally identifies:
- A mathematical object, such as a network, time series, or probability distribution.
- A mathematical action, such as comparing, optimizing, simulating, or estimating.
- A bounded context, including a named dataset, location, period, or system.
- A judgment criterion, such as error, waiting time, robustness, or accessibility.
“How can mathematics improve public transport?” is too broad. A stronger version is: “To what extent does removing each of five selected stations change mean shortest-path travel time in the central section of a specified transport network?” This establishes what changes, what will be measured, and where the investigation ends.
A practical fall scoping plan
Use the first 6-8 weeks to test feasibility rather than write polished paragraphs:
- Identify three areas and locate one credible dataset for each.
- Draft a question and list the required mathematics.
- Complete a small pilot graph, calculation, or simulation.
- Choose the option offering the best balance of depth and feasibility.
- Define variables, assumptions, methods, and validation measures.
- Revise the question after mathematical reading and supervisor feedback.
If you cannot obtain the data, explain the method, or produce a meaningful preliminary result, the topic is not ready. Use RevisionDojo Mathematics EE exemplars to study scope and mathematical communication, not to copy a question. The RevisionDojo Mathematics EE guide provides additional planning guidance.
Common topic-selection mistakes
A fashionable context does not guarantee mathematical depth. Artificial intelligence, cryptocurrency, climate change, and social media remain unsuitable if the essay mainly explains the context.
Common problems include:
- treating one regression or a high coefficient of determination as a complete conclusion
- collecting more variables than you can explain
- reporting software output without explaining the method
- using data with unclear definitions or missing metadata
- assuming correlation demonstrates causation
- choosing mathematics so advanced that you can only paraphrase it
Technology should extend the investigation rather than replace your reasoning. Interpret every important parameter, graph, simulation, and error measure, and test significant assumptions where possible.
Conclusion
Strong IB Math AI EE topics combine an accessible context with a precise mathematical problem. Climate models, transport networks, queues, sports probabilities, voting simulations, and population forecasts work well because they support comparison, validation, and evaluation.
Begin with a pilot, secure reliable data, and select methods you can explain independently. RevisionDojo's Math AI Study Notes, Math AI Questionbank, Mathematics EE exemplars, and Jojo AI can help you strengthen the underlying mathematics while keeping the research decisions and final writing your own.
Sources and referenced URLs
- IB Mathematics: Applications and Interpretation subject brief
- IB Extended Essay curriculum updates
- NOAA Climate Data Online
- GTFS documentation overview
- World Bank Open Data
- RevisionDojo Mathematics EE exemplars
- RevisionDojo Mathematics Extended Essay guide
- RevisionDojo Math AI Study Notes
- RevisionDojo Math AI Questionbank