A conclusion is a short, focused answer to your research question — nothing more, nothing less. Its whole job is to say what you found, back it with your own processed data, tie it to the physics, and compare it honestly to the accepted value. Get this right and the examiner sees a student who understands what their results actually mean, not just someone who did the maths.
📚 What you need to know
A conclusion is a brief summary that directly answers the research question
Every claim must be justified by your data — quote your key processed values
State clearly whether your results support or refute the hypothesis
Compare to an accepted literature value, cite the source, and find the percentage error
Compare percentage error with percentage uncertainty to judge accuracy
Never introduce new ideas or explanations here — save those for the evaluation
Start with a direct answer
Open your conclusion by answering the research question head-on, using the trend you found. Then support it with the specific numbers you calculated — a physical constant, a gradient, or a material property. This is what “justified by your data” means: the reader can see the evidence sitting right there.
✎ What a strong conclusion contains
A direct answer to the research question, stated as a definitive finding.
The key processed values as evidence — e.g. the final constant, the gradient.
A clear statement that the results support or refute the hypothesis.
A comparison to the accepted value, with source and percentage error.
The golden rule here is: no new material. A conclusion is a summary of what you already analysed, not a place to suddenly explain why something went wrong — that belongs in the evaluation. If you catch yourself writing “this might be because…” you’ve drifted out of the conclusion. Keep it tight: what did you find, and does it match the theory?
Support or refute — both are valid
You must say explicitly whether your data supports or refutes your hypothesis. And here’s the reassuring part: if your results don’t support it, that is not a failed experiment. A clear refutation is a genuine scientific finding — you just state it plainly and save the “why” for later.
This is the part that separates a top conclusion from an average one. You have two numbers that describe how good your result is: the percentage error (how far you are from the true value) and the percentage uncertainty (how much wiggle room your measurements had). Comparing them tells you what kind of error dominated.
If the accepted value sits inside your ± range, random error explains the gap; if it sits outside, suspect a systematic error.
Think of your uncertainty as a “net” you cast around your result. If the true value lands inside the net, your experiment caught it — any small gap is just random noise, and you can honestly call the result accurate. If the true value falls outside the net, something is consistently pulling your readings off in one direction. That’s a systematic error, and it’s the headline for your evaluation.
Worked example: concluding a pendulum experiment
WE 1
A student determines g from the gradient of a T²-vs-L graph and gets g = 9.86 ± 0.14 m s−2. The accepted value is 9.81 m s−2. Write the key comparison for their conclusion.
Step 1 — percentage error
(|9.86 − 9.81| / 9.81) × 100 = 0.5%Step 2 — percentage uncertainty
(0.14 / 9.86) × 100 = 1.4%Step 3 — compare
% error (0.5%) < % uncertainty (1.4%), and 9.81 lies in 9.72–10.00
Result consistent with accepted valueBecause the accepted value falls inside the ± range, the result is accurate and the small gap is explained by random error — no major systematic error indicated.
Worked example: concluding a resistance experiment
WE 2
A graph of resistance R against length L for a constantan wire is a straight line passing close to the origin. Write a conclusion that answers the research question and links to the physics.
Direct answer
The resistance is directly proportional to the length of the wire
Evidence
A graph of R vs L gave a straight line of best fit through (near) the origin
Link to theory
Consistent with R = ρL / A, which predicts R ∝ L when ρ and A are constant
Data supports the hypothesisNotice it answers the question, quotes the evidence, and ties to an equation — without drifting into “why the intercept wasn’t zero.” That’s evaluation territory.
💡 Top tips
Open with a direct answer to the research question — don’t bury it.
Always quote your final result with its uncertainty (e.g. 9.86 ± 0.14 m s−2).
Cite your source for the literature value (e.g. “IB Physics Data Booklet, 2025”).
Quantify the comparison with a percentage error — never just say “close.”
⚠ Common mistakes
Introducing new explanations for errors (that’s the evaluation’s job).
Saying the result was “close” without a percentage error.
Forgetting to state support or refute explicitly.
Quoting the result without its uncertainty.
Treating a refuted hypothesis as a failure rather than a valid finding.
Quick recap: A conclusion directly answers the research question, justified by your data. State support or refute, quote your result with uncertainty, and compare to a cited literature value via percentage error. If % error sits inside your % uncertainty, the result is consistent; if not, suspect a systematic error. No new ideas here.
You’ve now stated clearly what you found and how well it matches the accepted physics. But a good scientist doesn’t stop at “here’s my answer” — they ask “how much should anyone trust it, and how could it be done better?” That critical reflection is the final piece of the inquiry cycle, waiting for you on Evaluating the Method.
Want conclusions that lock in the marks?
Book a free meeting and we’ll practise answering the research question directly, comparing error with uncertainty, and citing literature values properly — the moves that make an IA conclusion airtight.
