IB Physics SLInquiry 3 — Concluding & EvaluatingInternal AssessmentPercentage error vs uncertainty~8 min read
Drawing Conclusions
After all the measuring, processing, and graphing, the conclusion is where you finally say what it all means. It’s a short, focused paragraph — not an essay — that answers your research question using only the evidence you’ve already produced. Do it well and you tie the whole investigation together; the trick is being direct, quoting your key numbers, and comparing them honestly to the accepted science.
📘 What you need to know
Open with a direct answer to your research question, justified by your data
Quote your key processed result as evidence — a constant, a gradient, a material property
State plainly whether the results support or refute your hypothesis
Compare to an accepted literature value: state it, cite the source, and find the percentage error
Comment on your uncertainty by comparing percentage error with percentage uncertainty
Never introduce new ideas — a conclusion only summarises what you’ve already analysed
Answer the question, backed by data
The single most important rule: your conclusion must be justified by your data. Start with a direct answer to the research question, using the trend from your interpretation to make a definitive statement, then back it with the actual numbers you calculated.
Good evidence to quote is your key processed output: the final value of a physical constant (like g), the gradient of your linearised graph, or an experimental material property (like resistivity). And a hard rule — never introduce a new idea or explanation here. The conclusion is a summary, not a place for fresh analysis.
🧭 The anatomy of a strong conclusion
Direct answer — state the relationship you found, in one clear sentence.
Evidence — quote the key processed value with its uncertainty (e.g. g = 9.72 ± 0.14 m s−2).
Hypothesis — say explicitly whether the data supports or refutes it.
Comparison — state the literature value, cite it, and quantify the difference as a percentage error.
Support or refute the hypothesis
You must say, in plain words, whether your results support or refute your hypothesis — for example, “the data shows T2 is directly proportional to L, which supports the hypothesis.”
A result that refutes your hypothesis hasn’t “failed” — that’s a myth worth unlearning early. A clear, honest finding that contradicts your prediction is a perfectly valid scientific outcome. State it plainly, then save the “why” for your evaluation. Examiners are marking your reasoning, not whether nature agreed with you.
Compare to the accepted value
A high-level conclusion always compares your experimental result to an accepted literature value — from the IB data booklet, a textbook, or another reliable source. That comparison is what lets you comment on the accuracy of your outcome.
Whenever you compare, do three things: state the literature value, cite your source (e.g. “IB Physics Data Booklet, 2025”), and quantify the gap with a percentage error rather than a vague “close to”.
Here’s the move that separates a top-band conclusion: don’t just report the percentage error — weigh it against your percentage uncertainty. The comparison tells you what kind of error dominates.
% error > % uncertainty
→ points to →
systematic error
% error < % uncertainty
→ result is →
consistent & accurate
If the percentage error is larger than the percentage uncertainty, the deviation is too big to blame on your equipment’s random error alone — a systematic error in your method is the likely culprit. If instead the literature value falls within your uncertainty range (percentage error smaller than percentage uncertainty), your result is consistent with the accepted value, and any deviation is explained by the random errors you already accounted for.
When the accepted value lands inside your uncertainty band, the result is consistent; when it falls outside, a systematic error is likely.
Quick recap: a conclusion answers the question, quotes the key result with uncertainty, states support/refute, cites a literature value, and compares percentage error against percentage uncertainty.
WE 1
A pendulum investigation gives g = 9.72 ± 0.14 m s−2. The accepted value is 9.81 m s−2. Write the comparison part of the conclusion.
Step 1 — percentage error
% error = |9.72 − 9.81| ÷ 9.81 × 100
= 0.92%Step 2 — percentage uncertainty
% unc. = 0.14 ÷ 9.72 × 100
= 1.44%Step 3 — compare them
0.92% (error) < 1.44% (uncertainty)
accepted value lies within rangeSo the result is consistent with 9.81 m s⁻² — deviation is explained by random error, with no sign of a significant systematic error.
WE 2
A resistivity experiment gives ρ = 4.60 × 10−7 Ω m with a percentage uncertainty of 2%. The literature value is 4.94 × 10−7 Ω m. What does the comparison tell you?
Step 1 — percentage error
% error = |4.60 − 4.94| ÷ 4.94 × 100
= 6.9%Step 2 — compare with the uncertainty
6.9% (error) > 2% (uncertainty)
points to a systematic errorThe gap is far too big for random error alone — something in the method (e.g. contact resistance, or an over-measured wire diameter) is skewing every reading. Save the specific cause for the evaluation.
💡 Top tips
Be concise and direct — a conclusion is a short paragraph, not a long essay. Answer the question first.
Only conclude what your data shows — if it’s inconclusive, say so; don’t overreach.
Always quote your final result with its uncertainty — it’s essential for discussing reliability and accuracy.
Quantify the comparison — calculate a percentage error instead of saying your result was “close”.
⚠ Common mistakes
Introducing new explanations in the conclusion — that analysis belongs earlier, and the “why” of errors belongs in the evaluation
Saying a result is “close” to the accepted value without a percentage error
Quoting the final value without its uncertainty
Treating a refuted hypothesis as a failed experiment instead of a valid finding
Up next: Evaluating in Physics — the critical reflection where you dig into your method’s weaknesses, classify random vs systematic errors, discuss limitations and assumptions, and propose realistic improvements that close the loop.
Want this to actually click before the exam?
Book a free meeting and let’s work through the tricky bits together.
