One belief quietly caps more IB Maths IA scores than almost anything else: the idea that an unusual or sophisticated-sounding topic will carry weak execution. Practitioner guidance on internal assessments is blunt about this. It notes that top-scoring IB Maths IAs, typically in the 17–20 range, succeed because they combine focus, strong structure, and precise communication: the scope is tightly controlled, the argument is clearly built, and the mathematics is explained, not just performed. The same guidance notes that mid-band scripts usually fall short through weak topic scope, descriptive writing, unclear maths, and missing criterion-level evidence rather than because the subject itself is ordinary.
The IB’s own 2026 examiner and moderation instructions confirm the pattern. Sample feedback comments repeatedly flag explorations that are too descriptive or treated superficially, that never establish a clear mathematical framework, or that rely on mathematics below the level of the course – even when the topic sounds ambitious. In IB Maths, the IA is graded against how effectively you construct and communicate the investigation. Between a 12–14 IA and an 18–20 IA, the difference is execution. Of those three failure types, descriptive or superficial treatment appears most consistently across the 2026 moderation feedback – suggesting it is usually the first structural problem an examiner encounters when a draft has not been built around the criteria.
What Each Criterion Rewards at the Top Band
Each IA criterion has a distinct behavioural jump between “solid pass” and “top band,” and Communication is usually the clearest place to see it. In the middle bands, writing is understandable and sections are labelled, but the exploration reads as a sequence of disconnected moves with graphs and tables simply inserted. At the top band, Communication forms a clear narrative in which each section advances the central question and every output is followed by concise interpretation that drives the next step.
For Mathematical Presentation, mid-band work usually has mostly correct mathematics but leaves reasoning implicit, with drifting notation and variables used before they are defined. At the top band, symbols are introduced precisely, notation is consistent, and each major step is justified and arranged so an examiner can follow the structured argument without guessing what was intended.
Personal Engagement is often misread as enthusiasm in the introduction, so many mid-band explorations open with an “I’m interested in…” paragraph then slide into generic processing of data or models. In top-band work, engagement appears in mathematical choices: posing follow-up questions when results are surprising, adjusting or comparing models when analysis reveals weaknesses, and commenting when outputs conflict with expectations. Curiosity is visible in how the mathematics is extended and questioned, not just in how the topic is introduced.
Reflection: The Criterion That Moves Scores the Most
Reflection is where many otherwise competent explorations stop climbing. Most middle-band IAs do include a final paragraph called “Reflection” or “Conclusion,” but it often summarises findings and lists limitations without analysing their mathematical consequences. Practitioner guidance emphasises building a coherent IA flow from a feasible topic with enough mathematical depth and personal engagement, then supporting it with justified modelling, calculations, and sustained reflection alongside those decisions, rather than trailing after them. High-scoring reports treat reflection as part of how decisions are made throughout the investigation, not as a post-hoc commentary.
The practical test is whether your IA uses descriptive or reflective language, and where that language appears. Descriptive sentences say “I found that…” and stop; reflective sentences push further, for example “This assumption limits the model’s validity because…”. If removing the final section would remove all such evaluation, reflection has been appended rather than woven. A further standard is whether you link important assumptions or choices to specific mathematical consequences, such as parameter sensitivity or restricted domains; without that, the Reflection criterion is unlikely to reach the top band.
Calibrating Mathematical Depth and Technology Use
‘Add more mathematics’ is familiar feedback on drafts, but it’s rarely actionable because depth is not about volume of techniques. The criterion concerned with mathematical content asks whether the mathematics genuinely supports the question being investigated and whether it’s commensurate with the level of the course. Depth comes from a focused chain of reasoning and analysis that the investigation actually requires, not from importing additional methods.
For Analysis and Approaches Higher Level (AA HL), that typically means working at or near the more advanced parts of the syllabus when the question genuinely requires it – demonstrating understanding rather than importing unfamiliar material for show. For Applications and Interpretation Standard Level (AI SL), depth comes from clear, well-justified use of statistics, modelling, and technology-informed analysis at course level, rather than forcing HL-style techniques into an otherwise data-driven project. In both cases, the question is the same: does this mathematics do real work in the investigation, or is it decorative?
Technology use is judged in the same spirit. A graphing calculator, spreadsheet, or statistics package can support the mathematics or replace it. A simple three-question test helps: can you explain what the tool computed, why that computation is appropriate for your question, and what the output means mathematically in context? When any of those answers is missing, the technology has become a black box, and criterion marks usually follow. The same boundary distinguishes real depth from artificial complexity: an extension counts as real depth when it grows from your own results, affects conclusions you actually use, and can be explained and interpreted in your own words. If it is added mainly because it sounds advanced and does not influence any later argument, it will usually read as unnecessary complexity and trigger the same concerns about superficiality, unclear framework, or level mismatch that examiners already warn about.
Evaluating Any Topic and Draft Against Key Criteria
Before committing to a topic, a short tractability check can prevent a lot of dead ends. First, ask whether the idea naturally generates a mathematical question you can explore with depth at your course level. Second, check whether the context leaves room for genuine personal engagement rather than simple data processing. Third, be honest about whether you are mathematically interested in the question itself, not just in how sophisticated it might look.
- How to use (fast): Mark each line Green / Yellow / Red. For every Yellow/Red, write one draft citation in your notes: section name plus what is missing or needs to change.
- Communication narrative (architecture test): From headings and paragraph openers alone – not screenshots or long calculations – could a reader state your central question, your main answer, and the chain of reasoning that connects them?
- Reflection thread (distribution test): After each major result, do you add evaluative interpretation about meaning, adequacy, and “so what?”, or is almost all evaluation confined to a final paragraph?
- Modelling and assumptions (consequence test): When you state an assumption, do you also state a mathematical consequence for validity or results – such as effects on fit, domain restrictions, or sensitivity – rather than just naming the assumption?
- Depth and level fit (necessity test): Can you write one sentence that honestly finishes “My question required this level of mathematics because…”, and does that match the methods you actually rely on to justify your conclusions?
- Technology (black-box test): For every software or calculator output you use to support a claim, can you explain in your own mathematical words what it computed, why it is appropriate, and what the numbers or graphs mean in context?
- Prioritisation rule (what to fix first): Fix Red items that break the narrative or make reasoning hard to follow (Communication and black-box technology) before Red items that weaken validity (assumptions, modelling, and missing reflection), and only then look for Yellow opportunities to deepen mathematics that still pass the “real depth” boundary from course-level standards.
This kind of scorecard does not predict a numeric mark; it highlights the structural ceiling-limiters that examiner comments return to most often. Focusing revision on those issues, rather than chasing ever more elaborate methods, is what moves a draft toward the behavioural targets of the top bands in each criterion.
Reframing How You Aim for Top-Band IB Maths IA Scores
The strongest IB Maths IAs are not defined by rare topics but by how deliberately their mathematics is chosen, structured, and interpreted. When students treat the criteria as behavioural targets and design a clear, well-justified mathematical narrative, they create the conditions for top-band performance from almost any tractable question. The most decisive step you can take before submission is therefore not to search for a flashier idea, but to audit your existing exploration against these standards and revise until its structure, modelling choices, and critical commentary clearly support the mathematical story you intend to tell.
