Probability and statistics appear in both IB Mathematics AA and AI, but the balance differs. AA keeps statistical reasoning within a broader analytical course; AI places stronger emphasis on data, modelling, interpretation and technology.
Shared foundations
- Descriptive statistics
- Probability and conditional probability
- Random variables and distributions
- Correlation and regression
- Sampling
- Statistical interpretation
The main danger: calculator without interpretation
A graphing calculator can produce a regression line, probability, test statistic or distribution value. The student still needs to know what was computed, what assumptions were made and what the result means in context.
2029 transition
For first assessment in 2029, the revised AA course removes bivariate data at SL and HL. The revised AI course removes several higher-level statistical topics, including the Poisson distribution and selected hypothesis tests. Students taking final examinations through November 2028 remain on the current course generation.
Use BTT’s Probability and Statistics and Data owners for the stable Mathematics.
World Mathematics route: return to the World Mathematics Atlas for the wider map across examinations, curricula, competitions, mathematical objects and university routes.
IB Mathematics AA and AI statistics and probability: model uncertainty explicitly
Define variables and populations
Before calculation, identify what is measured, the population of interest and how the sample was obtained. Statistical meaning depends on design.
Descriptive statistics summarise, not explain
Mean, median, spread and graphs describe data. They do not by themselves establish cause. Compare summaries with distribution shape and outliers.
Probability models require assumptions
Independence, conditional probability and distribution choices should be justified from the scenario. Formula selection follows the model.
Technology handles computation, students handle interpretation
GDC functions can calculate summaries and probabilities quickly. The learner must enter correct data, choose the correct model and explain the output.
Correlation is not causation
A strong association can arise from confounding, common causes or chance. Regression can support prediction within a sensible domain but does not prove a causal mechanism.
Sampling affects inference
Convenience, voluntary response and non-response can bias results. Randomisation and representative sampling matter because conclusions extend beyond the observed data.
Hypothesis testing is evidence under a model
State hypotheses, significance level and decision rule correctly. A p-value is not the probability that the null hypothesis is true.
Confidence and uncertainty
Where confidence intervals are used, interpret them according to the course framework and context rather than as guaranteed ranges for individual observations.
Common errors
Students confuse mutually exclusive with independent events, overstate correlation, use technology without defining variables and write conclusions detached from the question context.
Transfer
Mix data interpretation, probability, distributions and inference so students practise choosing the statistical model rather than executing isolated calculator routines.