IB Biology HLStage 3 — Conclude & EvaluateIA & Paper 2Core skill~11 min read
Drawing Conclusions
A conclusion is short. It answers the question you asked, quotes the numbers that prove it, says whether your hypothesis survived, and admits how confident you are. Anything longer than a paragraph usually means you have started evaluating instead.
📚 What you need to know
A conclusion is a brief, focused summary of your findings and a direct answer to your research question.
It must be justified by your data. That is the single most important rule.
Quote key processed values as evidence — the final calculated value, the optimum from your graph, or the values at each end of the trend.
Never introduce new ideas in a conclusion. It only summarises what you already analysed.
State explicitly whether the results support or refute your hypothesis. Refuting it is a valid finding, not a failure.
Compare your outcome with accepted literature values and cite the source. This is what lets you comment on accuracy.
The spread of your data decides how confident you can be. Small, non-overlapping error bars mean a difference is likely significant; large or overlapping ones mean it may not be.
What goes into a conclusion
Four things, in this order. Get them all in and the paragraph almost writes itself.
Notice what is missing: no new explanations, no discussion of what went wrong. Those belong in the evaluation on the next page.
Part
What it does
Where the content comes from
1. Direct answer
Turns the trend you identified into a definite statement about the relationship
The description of your graph
2. Key values
Proves the statement with numbers — the optimum, the final calculated value, or the values at each end of the range
Your processed data table
3. Hypothesis verdict
States plainly whether the results support or refute what you predicted
Your original hypothesis from Stage 1
4. Confidence
Says how far the spread of your data lets you trust the finding
Your standard deviations and error bars
The rule that catches people is “no new ideas”. If a sentence in your conclusion contains a fact or an explanation that appears nowhere earlier in the report, it is in the wrong place. A conclusion collects things you have already shown; it does not introduce them.
Supporting or refuting the hypothesis
Say it in one sentence, and say it clearly. Something like: the rate peaked at pH 8 and fell either side, which supports the hypothesis.
If your results do not support the hypothesis, that is completely fine. It is a valid scientific finding, and the experiment has not “failed”. State it just as plainly, then save the reasons why for your evaluation.
Inconclusive is also an answer. If your error bars are so large that you cannot separate the conditions, the honest conclusion is that the data does not allow a firm answer. Say so. Claiming a trend your data cannot support is a much bigger problem than not finding one.
Comparing with the literature
A conclusion that stops at your own data is fine. A conclusion that also compares it with an accepted value is better, because that is the only way you can say anything about accuracy.
🧩 Making the comparison properly
State the expected pattern or published value — the known optimum pH of the enzyme, the expected direction of the trend.
Cite where it came from: the IB data booklet, a named textbook and edition, or a reliable scientific source.
Compare your result with it in words: higher, lower, or in close agreement.
If it is a specific number, quantify the gap by calculating the percentage error.
Say what that means for the accuracy of your outcome — and remember a small gap does not prove you were right, only that you were close.
Percentage error
|experimental value − literature value| ÷ literature value × 100
A percentage error turns “quite close to the book value” into a number the reader can judge for themselves.
How confident are you allowed to be?
The spread of your data sets the ceiling on your conclusion. This is where the standard deviations you calculated in Stage 2 finally earn their keep.
What the error bars look like
What you may claim
Sample wording
Small, and not overlapping between conditions
You can be more confident the difference you observed is significant
“The small standard deviations show the data is precise, and the non-overlapping bars between 60 and 70 °C suggest the increase is significant.”
Large, or overlapping between conditions
You can be less confident. The apparent difference may be down to random error rather than a real effect
“The overlapping bars between 20 and 30 °C indicate the difference may not be significant, and further repeats would help confirm the pattern.”
Confidence language is worth practising because it is easy to overclaim without noticing. “Temperature increases membrane permeability” is a bigger claim than your six data points can carry. “In this investigation, mean absorbance increased with temperature” is honest and still earns the mark.
WE 1
Writing a full conclusion
Write a conclusion for the beetroot investigation, in which mean absorbance rose from 0.08 at 20 °C to 0.86 at 70 °C, with standard deviations of 0.03 or below. (5 marks)
Part 1: answer the research question directly
Increasing temperature increased the pigment released from beetroot discs, so higher temperatures do increase membrane permeability.
Part 2: quote the evidence
Mean absorbance rose from 0.08 at 20 °C to 0.86 at 70 °C, with the steepest increase occurring above 40 °C.
Part 3: the hypothesis
This supports the hypothesis that pigment loss would stay low at first and then rise sharply once membrane proteins began to denature.
Part 4: confidence
All standard deviations were 0.03 or below, so the data is precise and the trend can be described as reliable.
One paragraph: answer, evidence, verdict, confidenceno new biology has appeared here — every claim was already made and justified in the interpretation
WE 2
Percentage error and accuracy
A student found the optimum pH of catalase to be 7.4. A textbook gives the accepted value as 7.0. Calculate the percentage error and comment on the accuracy. (3 marks)
Step 1: find the difference
7.4 − 7.0 = 0.4 pH unitsStep 2: divide by the literature value and multiply by 100
(0.4 ÷ 7.0) × 100 = 5.71…
Step 3: round and comment
The result is close to the accepted value, so accuracy is reasonably good. The small overestimate could point to a systematic error, such as an uncalibrated pH meter.
5.7 % error — reasonably accurate, but a small systematic shift is possibledivide by the literature value, not by yours — the accepted value is the thing you are measuring against
WE 3
Concluding when the data will not cooperate
A student’s results suggest light intensity increases the rate of photosynthesis, but the standard deviation error bars overlap at every intensity tested. Write a suitable conclusion. (3 marks)
Point 1: describe what the means appear to show
Mean oxygen production increased slightly as light intensity increased.
Point 2: be honest about the spread
However, the standard deviation error bars overlap at every intensity, so the differences between conditions may not be significant.
Point 3: state the limit of the claim
The data therefore does not provide firm evidence that light intensity affected the rate, and no reliable conclusion can be drawn without further repeats.
Inconclusive — and saying so is the correct answeran honest “inconclusive” scores; an overclaimed trend the data cannot support does not
💡 Exam tips
Open with a sentence that could stand as the answer to your research question on its own.
Put at least two numbers in. A conclusion with no data in it is just an opinion.
Always include your key final result — the optimum, the isotonic point, the final calculated value.
Use the words “supports” or “does not support” about the hypothesis. Do not leave the reader guessing.
Cite the source of any literature value you compare with.
Add one line on reliability, based on your standard deviations. It costs a sentence and strengthens the whole report.
⚠ Common mistakes
Writing an essay. A conclusion is a short paragraph, and length is not a substitute for evidence.
Introducing new explanations that appear nowhere in the analysis.
Claiming more than the data shows, especially when the error bars overlap.
Never mentioning the hypothesis, so the examiner cannot tell whether it was supported.
Treating a refuted hypothesis as a failure and apologising for it. It is a genuine result.
Starting the evaluation early by listing what went wrong. That is the next section.
Up next: Evaluating the Method — where all the things that went wrong finally get their say, sorted into systematic errors, random errors, limitations and assumptions, each with a realistic fix.
Want this explained one-to-one?
Book a free session with an experienced IB Biology tutor and get your trickiest topics made simple.
