IBDP Subjects • Internal Assessment • Written by the PrepSeven Editorial Team, reviewed by certified IB Examiners

Common IB IA Mistakes That Stop Students Scoring a 7

There’s a specific kind of frustration that comes with an IA mark that doesn’t match the effort behind it — hours in the lab, dozens of pages of data, a genuinely interesting topic, and still a mark that lands in the middle rather than the top. In almost every case PrepSeven’s examiner-tutors review, the gap isn’t a lack of effort or intelligence. It’s one or two specific, repeatable mistakes made early in the process that quietly cap the ceiling of the entire investigation, long before the final write-up. This guide breaks down the mistakes we see most often across subjects, why each one costs marks against the real criteria, and exactly how to fix them while there’s still time.

Mistake 1: A Vague Research Question That Leads to Irrelevant Data

This is the root cause behind more disappointing IA marks than any other single factor. A vague or overly broad research question doesn’t just cost marks under the formulation criterion directly — it cascades through every later stage. If your question doesn’t clearly specify what you’re measuring, comparing, or investigating, you’ll end up collecting data that’s only partially relevant, discovering gaps in your evidence only once you sit down to analyse it, and forcing your discussion section to work around a mismatch between what you actually gathered and what your title promises.

Mistake 2: Insufficient Data or Sample Size

Especially common in the sciences, this mistake shows up as too few repeated trials, too small a sample, or reliance on a single source or data point where the investigation genuinely calls for more. The issue isn’t just that small datasets are ‘less impressive’ — it’s that they genuinely undermine your ability to draw a reliable conclusion, which then weakens everything you say in your analysis and evaluation sections. A conclusion built on three data points, phrased with the same confidence as one built on thirty, reads as a lack of understanding of your own method’s limitations.

• In experimental sciences: plan for enough repeated trials from the outset to allow meaningful statistical treatment (means, standard deviation, error bars), not just enough to produce a single graph.
• In Economics and Business: ensure your chosen articles or case data are specific and detailed enough to support real application of theory, not just a general news summary.
• In Humanities subjects: cross-reference more than one source where your argument depends on a specific factual claim, rather than relying on a single account.

Mistake 3: Presenting Data Without Analysing It

A results section full of neatly formatted tables and graphs feels like solid work, but if it isn’t followed by genuine interpretation, it does very little for your mark under the analysis criterion — which, in most subjects, carries the heaviest weighting of all. The test is simple: for every table, graph, or piece of evidence in your report, ask whether you’ve written at least one sentence explaining what it actually means in relation to your research question. If the answer is no, that section is currently pure description, however well formatted it looks.

Mistake 4: Ignoring Uncertainties and Limitations

This mistake is particularly costly in the sciences and mathematics, where error analysis and methodological limitations are explicitly and heavily rewarded, but it shows up across every subject in a slightly different form. A science IA that never discusses measurement uncertainty, a Business IA that never questions whether its data sources might be biased, or a History IA that never acknowledges the limitations of its available sources all read as less sophisticated than one that engages honestly with these questions. Ironically, many students avoid discussing limitations because they worry it will make their work look weaker — in reality, the opposite is true: honest engagement with limitations is exactly what separates a sophisticated investigation from a naive one in an examiner’s eyes.

Mistake 5: Poor Writing and Presentation

Unclear structure, missing labels on diagrams and tables, inconsistent or absent citations, and visuals that are hard to read all fall under this category. While typically the lowest-weighted criterion, presentation problems have an outsized practical effect: they make it harder for an examiner to locate and credit the genuinely strong analytical work that might be present elsewhere in your report. A brilliant piece of analysis buried in a disorganised, unlabelled wall of text risks simply not being found and credited properly.

• Use clear section headings that match the logical stages of your investigation (question, method, results, analysis, evaluation)
• Label every graph, table, and diagram with a clear title and axis labels where relevant
• Apply one citation style consistently from the first page to the last
• Keep your results and your analysis in clearly distinguishable sections rather than blending them

Mistake 6: Using the Wrong Markband Examples as a Template

Many students look at exemplar IAs (sometimes ones that scored well, sometimes not) and copy their structural approach without understanding why that structure worked, or without checking whether it fits their own specific research question. A common version of this mistake is writing a conclusion that reaches further than the data actually supports, simply because an exemplar essay confidently stated a strong conclusion. Your conclusion should only claim what your specific evidence can actually justify — overreach here is one of the more common ways an otherwise solid IA loses marks under evaluation-related criteria.

Mistake 7: Starting Too Late to Allow Real Revision

As with the Extended Essay, most of the mistakes on this list are symptoms of time pressure as much as anything else. A student who begins serious data collection only a few weeks before the deadline has little room to discover that their research question needs narrowing, that their data is insufficient, or that their first draft’s analysis is too descriptive — all issues that are easy to fix with time, but very difficult to fix in the final week.

1. Finalise your research question early: ideally with a feasibility check completed, several weeks before serious data collection begins.
2. Build in a full draft review: aim to have a complete first draft at least three to four weeks before your school’s internal deadline.
3. Treat the first draft as a diagnostic tool: its purpose is to reveal exactly which criteria need strengthening, not to be your final submission.

Mistake 8: Not Reading the Actual Subject-Specific Guide

Generic IA advice (including much of this guide) describes shared patterns across subjects, but every subject’s official guide contains specific requirements — word or page limits, required sections, subject-specific conventions — that generic advice can’t fully replace. Students who rely purely on general study advice sometimes miss a specific structural requirement unique to their subject (a required abstract, a specific citation convention, a mandated section order) that costs marks under presentation-related criteria despite otherwise strong content.

A Quick Self-Diagnostic Before You Submit

4. Could a stranger tell exactly what I investigated from my research question alone?
5. Is every table or graph followed by genuine interpretation, not just a caption?
6. Have I discussed real, specific limitations of my own data or method?
7. Does my conclusion only claim what my actual evidence supports?
8. Have I checked my subject’s specific official guide for required structure and conventions?
9. Did I leave enough time for at least one full round of revision after a complete first draft?

How PrepSeven Helps You Avoid These Mistakes Before They Cost Marks

PrepSeven’s tutors are certified and former IB examiners across every DP subject group, and we catch exactly these mistakes during draft review — often while there’s still time to genuinely fix them, rather than after submission when it’s too late. Because we’ve personally marked real IAs against the actual criteria, our feedback is specific: not ‘this needs more analysis’ in the abstract, but exactly which sentences are descriptive rather than analytical, and what would move a specific section into the next markband.

• Full-draft review flagging exactly which criteria are underdeveloped
• Feasibility checks before you commit significant time to data collection
• Subject-specific guidance matched to your actual current official IA requirements
• Support tightening analysis, evaluation, and presentation before final submission

Frequently Asked Questions

Q1. Can I redo data collection if I realise my sample size was too small?

This depends on how much time remains before your internal deadline — if caught early enough, expanding your dataset is usually worthwhile, since insufficient data undermines the reliability of everything built on top of it.

Q2. What if I run out of time and can’t fix every issue before submission?

Prioritise fixing the analysis and evaluation sections first, since these typically carry the most weight — a technically imperfect but well-analysed dataset will usually outscore a large dataset with thin or missing interpretation.

Q3. Is it better to have a shorter, focused IA or a longer, more comprehensive one?

Within your subject’s page or word limit, a shorter, tightly focused investigation with genuine depth of analysis almost always outperforms a longer one that spreads its attention across too many angles without developing any of them fully.

Q4. How do I know if my conclusion is ‘overreaching’ beyond my data?

Check whether every claim in your conclusion is something your specific results could actually support — if you find yourself making a broad general claim that your particular sample size or method couldn’t really establish, scale the claim back to match what your evidence actually shows.

Q5. Should I include raw data in the main body or in an appendix?

Most subjects expect processed data (tables, graphs, summary statistics) in your main body, with extensive raw data or lengthy transcripts placed in an appendix — always check your specific subject’s formatting conventions to be certain.

Q6. Can a strong idea still result in a weak IA?

Yes — as with the Extended Essay, execution matters more than the initial idea. A genuinely interesting research question paired with insufficient data, purely descriptive analysis, or missing evaluation can still land in the middle markbands despite strong underlying potential.

Q7. What’s the fastest single fix if I only have a few days left before submission?

Go through your analysis section and add a genuine interpretive sentence after every piece of data or evidence that currently lacks one — this single pass typically produces the largest mark improvement for the smallest time investment.