Troubleshooting Common IA Mistakes in IB Assessments

A calm truth about your IA

At some point, almost every IB student has the same quiet panic: you open your IA draft, scroll for a minute, and realize you’re not sure what the examiner is supposed to reward. The writing looks fine. The data exists. The citations are there. But the piece doesn’t feel sharp.

That feeling is usually not a work ethic problem. It’s a troubleshooting problem.

An IA is one of the few parts of the IB where the marks are lost in small, fixable places: an over-wide research question, a method that can’t be replicated, analysis that never turns into interpretation, or an evaluation that says “limitations” without explaining impact. This guide walks you through the most common IA mistakes, how to diagnose them quickly, and how to fix them before submission.

If you want subject-specific criteria clarity alongside these fixes, start with the IB IA Guides hub.

IA troubleshooting checklist (use this before you rewrite anything)

When your IA feels “off,” run this checklist first. It prevents you from polishing sentences while the structure leaks marks.

• Is the IA research question narrow, measurable/examinable, and worth analyzing?

• Can a stranger replicate your method without guessing?

• Do you have enough data (and enough variation) for a meaningful conclusion?

• Does your analysis explain why patterns happened, not just what happened?

• Does your evaluation explain how limitations affected results and what you’d improve?

• Is communication clear: labeled figures, consistent units, tidy structure, accurate citations?

For a rubric-aligned second opinion, you can run your draft through the IB Coursework Grader to see which criteria are underdeveloped.

The most common IA mistakes (and how to fix them)

Mistake: A research question that is broad, vague, or too obvious

A weak IA research question creates a domino effect. You collect unfocused data, write general background, and end up with conclusions that sound like common sense.

Fast diagnosis: If your question could be answered with a paragraph from a textbook, it’s too simple. If it contains two or three big ideas at once, it’s too broad.

Fix it:

• Reduce the scope: one relationship, one context, one clear lens.

• Make it testable/examinable with the tools and time you actually have.

• Make the wording “plain” and specific.

RevisionDojo’s guide on refining your IA research question is a helpful calibration tool, especially if you tend to overcomplicate the phrasing.

Mistake: “Personal engagement” that reads like decoration

Many students misunderstand this. Personal engagement in an IA isn’t about sounding emotional or inserting a dramatic origin story. It’s about intellectual ownership: making choices, justifying them, and reflecting on their consequences.

Fast diagnosis: If you could swap your name with a classmate’s and nothing would change, the IA likely feels generic.

Fix it:

• Explain why your research question matters to you academically (curiosity, local context, specific case).

• Show decision-making: why that method, why those sources, why that model.

• Briefly reflect on a turning point: a pilot result, a limitation you discovered, a choice you reversed.

If you want to see what “authentic engagement” looks like in real student work, browse the Coursework Exemplars library.

Mistake: A method that cannot be replicated (or doesn’t control what it claims)

A strong IA method reads like a careful recipe, not a memory of what happened. In sciences that means variables and controls; in humanities it means consistent criteria for selecting sources/cases; in math it means defined assumptions and a traceable process.

Fast diagnosis: If a reader would ask “how exactly?” more than twice, your method needs tightening.

Fix it:

• Write steps as if someone else must reproduce them.

• Explicitly define independent, dependent, and control variables (where applicable).

• Include a short pilot and explain what you changed because of it.

When you’re unsure what your subject rewards, open the relevant guide inside RevisionDojo’s IA Guides and map your sections to the criteria.

Mistake: Not enough data (or data that can’t show a pattern)

A common IA trap is collecting data that is technically “correct” but too thin to support analysis. Or collecting neat data that shows almost no variation, leaving you with nothing to interpret.

Fast diagnosis: If your graph looks like a straight line because everything is nearly identical, you may not have enough range, trials, or sensitivity.

Fix it:

• Increase trials or sample size where realistic.

• Record uncertainty, measurement precision, and conditions.

• Add qualitative observations when they explain anomalies (but don’t replace numbers with vibes).

Mistake: Description pretending to be analysis

This is the silent mark-loser. Students write long results sections, then summarize them again in the discussion. But analysis is not repeating. It’s interpreting: causes, implications, comparisons, and judgement.

Fast diagnosis: If your paragraphs mostly start with “This shows…” and end without explaining why, you’re still describing.

Fix it:

• Add the “because” sentence: explain mechanisms, theory, or reasoning.

• Compare to expectations or models.

• Address anomalies: what might explain them, and how they affect confidence.

RevisionDojo’s post on moving from description to analysis in an IB IA gives a clean paragraph-level test you can apply quickly.

Mistake: Evaluation that lists limitations but doesn’t discuss impact

Evaluation is where you sound like a scientist, historian, economist, or mathematician who knows what their work can and cannot claim. The biggest error is writing a bullet list of “limitations” without explaining how each one changes the reliability or validity of the conclusion.

Fast diagnosis: If your evaluation could be copied into any IA (small sample size, human error, time constraints), it’s not doing criterion work.

Fix it:

• For each limitation, explain impact: does it increase uncertainty, bias results, reduce generalizability?

• Suggest improvements that are realistic (not “buy a $10,000 sensor”).

• Tie evaluation back to the research question: how confident are you, and why?

Use How to Write a Compelling IA Evaluation Section as a model for making evaluation specific and marks-driven.

Mistake: Communication problems that quietly drain marks

A strong IA can still underperform if the communication is messy: unlabeled graphs, inconsistent units, unclear structure, or poor referencing.

Fast diagnosis: If a reader has to hunt for your research question, your method, or what a graph shows, the communication isn’t exam-ready.

Fix it:

• Use clear headings and consistent formatting.

• Label tables/figures with titles and units.

• Keep the logic linear: question -- method -- results -- analysis -- conclusion -- evaluation.

• Proofread specifically for ambiguity and missing links between claims and evidence.

Mistake: Overusing AI or sliding into academic misconduct

IB students are surrounded by tools that can generate text. The risk is letting those tools replace thinking, or worse, generating data or copying phrasing without proper citation. That’s not “efficient” -- it’s a credibility problem.

Fix it:

• Use AI for brainstorming and clarification, not as the author of your IA.

• Never fabricate data.

• Cite all sources carefully and keep your process transparent.

RevisionDojo’s platform is built to keep that boundary clear: you can use AI Chat for explanations, then use Grading tools for criterion feedback, while your writing stays yours.

A practical workflow: fix your IA without burning a week

Here’s a calm sequence that works when you’re close to submission:

• Step 1: Re-read the question. Put your IA research question at the top of the document and highlight every paragraph that clearly serves it.

• Step 2: Run a criteria scan. Use your subject’s rubric language from the IA Guides and check whether each criterion has evidence.

• Step 3: Get targeted feedback. Upload to the IB Coursework Grader and focus on the two lowest criteria first.

• Step 4: Calibrate with examples. Compare one section to a high-quality model in the Coursework Exemplars library.

• Step 5: Protect exam revision. Keep a short daily loop alive using the Questionbank so your IA doesn’t steal your exam momentum.

If you’re juggling multiple components, Which IA Should You Start First? A Calm IB Strategy can help you prioritize without panic.

Closing: turn IA troubleshooting into marks

A high-scoring IA is rarely the one with the fanciest topic. It’s the one where the research question is tight, the method is defensible, the analysis explains meaning, and the evaluation is honest about impact. In other words, it’s the one that’s been troubleshot.

If you want a single home base while you revise, RevisionDojo pulls the whole loop together: IA Guides for criteria, Coursework Exemplars for models, AI Chat for quick clarification, Grading tools for structured feedback, plus Study Notes, Flashcards, Questionbank, Predicted Papers, Mock Exams, and Tutors to keep exam prep moving while your IA improves.

Your IA doesn’t need to be perfect. It needs to be clear, analytical, and built to earn marks on purpose.