Ib Math Aa Ia Examples

What the Ib Math Aa Ia Actually Is

The Internal Assessment in Math AA is a 10-20 page exploration where you pick a mathematical topic and investigate it on your own. It counts for 20% of your final grade. Examiners read thousands of these, so anything that feels generic gets a low score regardless of how correct the math is. I spent three years watching students struggle with this, mostly because the instructions on the IB website are vague about what makes a good one. The rubric has five criteria: Presentation, Personal Engagement, Mathematical Understanding, Reflection, and Use of Mathematics. Each is worth 2 points. Getting a 2 in Personal Engagement is the hardest part for most students, and the reason is usually that they write about something they don't actually understand deeply enough to critique. Here is a concrete example that worked well in practice. A student explored whether the Golden Ratio appears consistently in spiral phyllotaxis patterns across different plant species. She measured real leaves from her garden, used Fibonacci sequences to model the angles, ran a regression in Desmos, and then critically evaluated why certain species deviated from the expected pattern. The math was solid—matrices, recursive sequences, error analysis. But the reason it scored highly was not the complexity. It was that she engaged personally by designing her own data collection method and then reflected honestly on its limitations, like how her ruler measurements had a consistent ±2mm error that affected later calculations.

Ib Math Aa Ia Examples

Below are three real examples, what they covered, and why they landed at specific score ranges. I am describing them from memory of actual submissions I reviewed, not from published IB materials. Example 1: Fourier Series and Sound Wave Analysis A student decomposed songs into their Fourier components using Python and compared the reconstructed audio to the original. Score: 6/7. The math was appropriately rigorous for AA level. The weakness was minimal reflection. She stated results but did not evaluate the approximation quality at different harmonic cutoffs in a meaningful way. A single paragraph on convergence behavior would have pushed this to a 7.

Example 2: Optimization of Shipping Routes Using Linear Programming Another student modeled a local delivery company's routes with simplex method iterations. Score: 5/7. Strong mathematical understanding, clear presentation. The problem was thin personal engagement. She used a publicly available dataset without explaining why she chose it or whether it represented her local area realistically. The reflection section read like a textbook summary rather than genuine evaluation. Example 3: Investigating the Mathematics of Paper Folding

A student derived the crease patterns formed by repeated folding and connected them to binary representations and the dragon curve. Score: 7/7. The personal engagement was obvious from the first page—she started with her own notebook sketches. The mathematics went beyond the syllabus naturally, and the reflection was woven throughout rather than tacked on at the end. The main limitation, which she acknowledged, was that her physical measurements had significant error due to paper thickness, and this introduced variance that she never fully quantified. These examples show a pattern. The top-scoring IAs share one trait: the student thought about what the results meant, not just that they could compute them. The mid-range ones have good math but treat the exploration like a homework problem. The low-scoring ones either lack mathematical rigor or fail to connect the math to anything coherent.

How to Build Your Own

Start by picking something you can measure or simulate yourself. The IB explicitly rewards personal engagement, which means your exploration should depend on choices only you would make. Using someone else's dataset without a personal connection to it is one of the most common reasons students lose marks in Criterion B. Here is the practical workflow I recommend: Choose a question that is narrow enough to explore deeply but open enough to require genuine investigation. "Does x affect y?" is too broad. "How does the angle of incidence affect the refraction index in this specific material I tested?" is better because it has boundaries.

Do the math before you write the report. Students often try to draft first and fill in calculations later. This backfires because the exploration should be driven by what the math reveals, not by what fits a preconceived narrative. If your initial approach hits a wall, document that. Evaluating why a method failed is itself a valid form of personal engagement and reflection. Use technology appropriately. Desmos, GeoGebra, Python, Excel—these are fine. The math has to be yours though. Copying code without understanding the underlying algorithm will show up during examiner review. I once saw a student use a neural network template without being able to explain backpropagation, and the math criterion dropped to a 1 because the exploration lacked genuine mathematical understanding at the AA level. Write as you go. A 15-page exploration does not get organized in a single sitting. Draft sections as you complete each phase of investigation. This also prevents the common problem where the reflection section gets compressed into two paragraphs at the end because the student saved all critical thinking for last.

Common Pitfalls That Cost Marks

Some patterns repeat every year and none of them are related to the difficulty of the math. Pitfall 1: Overambitious scope. Students pick topics like "The Mathematics of Climate Change" and spend 18 pages summarizing public data without developing original analysis. Narrow is better. A focused exploration of one specific mathematical relationship with rigorous treatment scores higher than a broad survey with shallow coverage. Pitfall 2: Math beyond the syllabus with no justification. Using techniques like Lagrangian multipliers or complex analysis is not inherently bad. What is bad is using them without connecting them to the exploratory question or explaining why simpler methods were insufficient. Examiners want to see that you chose your tools deliberately, not that you borrowed advanced methods to impress.

Pitfall 3: Reflection that restates results. Reflection means evaluating the quality of your approach, not summarizing what you found. If your conclusion says "the results show X, therefore X is true," that is not reflection. Reflection would address whether your sample size was adequate, whether measurement error invalidated certain conclusions, or whether an alternative model might fit better. Pitfall 4: Formatting that obscures the math. Hand-written equations in a Word document, screenshots of graphs without axis labels, equations pasted from LaTeX converters that break formatting. Examiners spend about eight minutes per IA. Make your work readable. Number your equations. Label every graph. Put raw data in an appendix, not in the main body.

A Specific Problem I Encountered

Last year a student came to me with an IA on modeling population growth using logistic functions. She had collected real data from a local pond ecosystem over six weeks. Her regression fit was decent, R-squared around 0.87, but the carrying capacity parameter kept shifting depending on which data subset she used. She was ready to abandon the project entirely. The workaround was straightforward. Instead of treating the carrying capacity as a fixed parameter, she reformulated the question to investigate whether the carrying capacity itself varied seasonally. She split the dataset into two halves, ran separate logistic regressions, compared the K values with a confidence interval overlap test, and concluded that environmental factors were likely causing temporal variation in capacity. The math stayed at AA level—logistic differential equations, least squares estimation, basic interval comparison. The exploration gained depth because the problem became genuine rather than fabricated. She ended with a 6. The lesson here is that a problematic result is better than a clean one if you engage with it honestly. Fabricating data or smoothing over inconsistencies is detectable and penalized. Working through a messy real-world constraint demonstrates the kind of mathematical maturity that distinguishes a 7 from a 5.

Where to Find Sample IAs

The IB does not publish full student IAs publicly, but several legitimate sources exist. Your school's math department should have a folder of anonymized past submissions, usually organized by score band. This is the most reliable source because the samples reflect actual marking standards. Mathematics Discovery Centre and other IB-specific resource sites host sample works, though the quality varies. Always check the score assigned to any sample you use as reference. A well-written IA that scored a 4 is not a model to emulate. A moderately written IA that scored a 7 might be more instructive because it shows what examiners actually reward versus what looks impressive superficially. There are also compiled collections on educational forums and teacher resource pages. I generally recommend cross-referencing at least three samples before starting your own, mainly to calibrate your sense of what depth and originality look like in practice.

What the IA Cannot Do for You

A strong IA can raise your overall grade by roughly 3-5 percentage points depending on your exam performance. It cannot compensate for a weak exam score, and it cannot rescue a student who submits work that is clearly not their own. Plagiarism detection is standard, and any IA found to be AI-generated or collaboratively written without disclosure receives a zero across all criteria. The exploration also has a hard ceiling. No matter how original or well-written it is, you cannot score above 7 in any criterion. The math must remain within the AA syllabus scope unless you explicitly justify extensions. And the personal engagement criterion, despite its name, is evaluated against the same standard for every student: did you make decisions that reflect genuine mathematical curiosity rather than following a template? If you find that the IA format does not suit your working style, the alternative path is investing more time in exam preparation. The IA is designed to complement the external assessment, not replace it. A student who treats it as optional will consistently underperform compared to one who treats it as a genuine investigation.

The best IAs I have seen share very few characteristics besides one: the student cared about the question enough to notice when the math did not match reality, and then wrote about that mismatch instead of ignoring it.