A free, examiner-style walkthrough of the Internal Assessment for the IB Diploma sciences (Physics, Chemistry, Biology and Environmental Systems & Societies): the four marking criteria, a section-by-section structure outline, the numbers examiners expect, and the mistakes that cost the most marks. Written by Dr. N. Rathankar Rao, IB Physics educator with 25+ years in the classroom.
For first assessment in 2025 the IB rebuilt the sciences and, with them, the Internal Assessment. The individual investigation is now marked against four criteria of six marks each — Research Design, Data Analysis, Conclusion and Evaluation — giving 24 marks in total, roughly 20% of your final grade. The old standalone Communication criterion is gone; clear communication is now rewarded inside the other four. Crucially, Conclusion and Evaluation together carry half the marks, so a tidy method with weak reasoning no longer reaches the top bands. The same four-criterion structure applies across Physics, Chemistry and Biology.
| Criterion | Marks | What examiners look for |
|---|---|---|
| Research Design | 6 | A focused research question that names the independent and dependent variables (or two correlated quantities), a clear description of the system and relevant background theory, and a method whose range and repetition can actually answer the question. |
| Data Analysis | 6 | Raw data recorded with units and uncertainties, processing that carries those uncertainties through (not just averages), and presentation — tables and graphs — that lets the trend speak for itself. |
| Conclusion | 6 | An answer to the research question that is justified by your processed data, interpreted with its uncertainty, and compared honestly against accepted scientific context or a published value. |
| Evaluation | 6 | A judgement of the method in light of the results: limitations ranked by how much they actually affected the data, plus realistic, specific improvements and a sensible extension. |
There is no compulsory template, but a report that a top-band candidate produces almost always flows through these sections. Aim to stay within the 3,000-word limit — tables, graphs, captions, calculations and references do not count toward it.
Open with the question you are answering, stated precisely enough that the variables are obvious. Give the reader the background theory they need — no more — and explain why the question is worth asking. This is where a personal angle or real-world hook earns its place.
Name the independent, dependent and controlled variables, and say how each control is held. Describe your apparatus and method with enough detail to be repeated, and justify your chosen range and number of repeats — this is what separates a design that can answer the question from one that cannot.
Present raw measurements in a titled table, with consistent units and an honest estimate of measurement uncertainty for every instrument. Note qualitative observations too — they often explain anomalies later.
Show one worked example of each calculation, then tabulate the processed results with propagated uncertainties. Graph the relationship with error bars and, where appropriate, a line of best fit with its gradient and intercept interpreted physically.
State the answer to your research question in one clear sentence, supported by the numbers. Interpret the trend, quote your key result with its uncertainty, and compare it with theory or a literature value — explaining any difference rather than hiding it.
Weigh the method against the data. List the limitations that mattered most, say why they mattered and in which direction they pushed the result, then propose specific, feasible improvements and one genuine extension of the investigation.
Cite every source and any data set consistently in one style. References sit outside the word count.
| Item | Working guide |
|---|---|
| Independent variable values | At least 5 distinct values across a sensible range |
| Repeats per value | Around 5, so a mean and spread are meaningful |
| Uncertainty | Estimated for every instrument and carried through processing |
| Word count | 3,000 words maximum (excludes tables, graphs, calculations, references) |
| Weighting | 24 marks · about 20% of the subject grade |
Database and simulation investigations are explicitly accepted when a live experiment is impractical — provided you process the raw data yourself and reach a real scientific conclusion, rather than summarising someone else's findings.
This guide is free and always will be. When you want criterion-by-criterion feedback on your actual IA, the IBHighway IA Project Diary walks you through each section above and marks your draft against these four criteria — and the free IA Mentor can get you started.
This guide is an independent educational summary of the current IB Diploma sciences Internal Assessment approach, written in the author's own words for study support. It is not affiliated with, endorsed by, or connected to the International Baccalaureate Organization (IBO), and it does not reproduce IB assessment documents. Always follow the official subject guide and your teacher's instructions for the definitive criteria and word count.
