Science assignments
Lab report conclusion: connect results to evidence
Use your measurements to answer the experiment question, explain uncertainty, and write a conclusion that does not overclaim.
The experiment is finished, the table is full, and the last box on the worksheet says ‘conclusion’. It is tempting to write that the hypothesis was correct and the experiment went well. That sentence does not tell a reader what you measured or how the measurements answer the question. A useful lab report conclusion makes a smaller, testable claim: it states the pattern, identifies the evidence behind it, and marks where the data stop supporting certainty.
You can write one with your notebook, graph, assignment instructions, and a careful reread of the raw observations. This guide uses a simple pendulum as a running example, but the method works for chemistry, biology, and other practical tasks. The example numbers are illustrative, not an answer key. Follow your teacher’s required format if it differs, and never invent a result to make the ending sound stronger.
1. Recover the question before writing the lab report conclusion
Read the aim and the exact question again. A pendulum task might ask how length affects the time for one swing; another might ask whether a model predicts a measured value. Those questions require different closing sentences. Underline the independent variable, the measured outcome, and the conditions held steady. If the assignment asks for a short conclusion rather than a full discussion, keep the answer compact, but do not omit the evidence that makes it meaningful.
Write a private one-line prompt: ‘What does my evidence allow me to say about this question?’ That wording is safer than ‘Was my hypothesis right?’ because the data may support part of a prediction, show a weak trend, or remain inconclusive. If you did not control a variable well, the conclusion should reflect that. University of Toronto guidance also distinguishes interpreting results from merely repeating them; your final paragraph should explain the significance of the pattern within the experiment’s scope.
2. Audit the raw measurements and units
Before writing polished prose, compare the table with the original observations. Check whether you recorded time for one swing or ten, whether length is in centimetres or metres, and whether each trial used the same start angle. A beautiful graph cannot rescue a unit conversion made from the wrong column. Mark values you excluded and record the reason; do not erase an awkward reading simply because it makes the trend less tidy.
For a pendulum, you might time ten oscillations three times at each length and divide each total by ten. If one trial differs sharply, inspect what happened: a late stopwatch start, a miscounted swing, or a genuine observation you cannot explain. Keep the original value visible. Averages can summarize repeated trials, but they cannot replace the question of how consistent those trials were. Your conclusion is only as trustworthy as the path from recorded observation to reported result.
3. Choose the result that actually answers the question
Do not retell every row of the results section. Select the one or two comparisons that answer the aim. In the illustrative pendulum example, a longer string might correspond to a longer average period. A useful evidence sentence names both the direction and the comparison: the average period rose across the tested lengths. If your teacher expects values, add two representative endpoints with units and the number of trials. Check that the values match the table exactly before copying them.
When a graph shows scatter, describe that too. ‘The measured periods generally increased’ is more honest than ‘Every increase in length caused exactly the same increase in time.’ If the graph is nearly flat, say that the data did not reveal a clear change under the tested conditions. If you only compared two lengths, do not claim a universal law. The purpose is to make a claim at the same scale as the evidence, not to squeeze every measurement into one triumphant sentence.
4. Explain the pattern without confusing observation and interpretation
A result says what happened in your dataset. An interpretation suggests why it happened or how it fits the model studied in class. Keep the two visible in separate clauses: ‘The average period increased as length increased; this pattern is consistent with the model that a longer pendulum swings more slowly under similar conditions.’ The first part points to measurements. The second links them to a principle. Neither requires the claim that the experiment proved the model for every possible pendulum.
If your class introduced an equation, use it only if you understand its assumptions. A small-angle pendulum model, for example, is less suitable when swings start from very different large angles. You do not need to derive the whole equation again in the conclusion. State the relevant relationship in plain language and cite class material when the rubric requires it. If the results disagree with the model, compare the direction and size of that disagreement before deciding which explanation is plausible.
5. Treat the hypothesis as a prediction, not a grade
A hypothesis is useful because it lets you compare expectation with measurement. It is not a test that the student passes by getting the expected answer. Say whether the pattern supports, partly supports, or fails to support the prediction under the conditions you tested. Then give the measurement that justifies that judgment. ‘My hypothesis was correct’ without a result is too vague; ‘the measured period increased with length across our tested range’ tells the reader what the support consists of.
A surprising result deserves the same care. First check arithmetic, units, graph axes, and setup notes. If the mismatch remains, report it. You may suggest a cause, but label it as a possibility unless you have evidence. Changing the hypothesis after seeing the data makes the original comparison impossible. You can describe a revised question for a later experiment instead. Honest disagreement often teaches more than a forced match, because it shows where the method or model needs scrutiny.
6. Name a limitation and explain its effect
‘Human error’ is not an analysis. Identify a particular source of uncertainty, how it could influence the measured quantity, and whether its effect is likely small or large relative to the observed pattern. In a hand-timed pendulum, reaction time affects the start and stop of each trial. Timing ten swings rather than one can reduce the fraction of each period affected by that delay, though it does not eliminate it. A changing release angle may create another source of variation.
Compare the size of the limitation with the claim. If the difference between two average periods is tiny and repeated trials vary just as much, do not describe the trend as decisive. If the difference is much larger than the observed spread, you may say the pattern appears robust within those trials while still acknowledging the setup. The University of Vermont writing center notes that lab-report expectations differ by course, so use your teacher’s required uncertainty language and calculations when provided.
Check the University of Vermont's advice on lab report expectations
7. Propose an improvement tied to the limitation
An improvement should target a named weakness, not serve as a generic wish to ‘be more careful’. If manual start timing was the main concern, use a light gate if available or increase the number of oscillations timed per trial. If the length was measured inconsistently, define exactly where the measurement starts and ends, then use the same method for each setup. If release angles varied, mark a small standard angle and release without a push.
Explain what the change would reveal. Repeating more trials could show whether the apparent trend survives ordinary variation; testing more lengths could show whether a two-point pattern continues. Neither change guarantees an expected result. Some limitations cannot be fixed with classroom equipment, and it is better to name that boundary than to promise an impossible perfect measurement. A good next step is feasible, directly connected to the evidence, and proportionate to the time and tools available.
8. Build a concise paragraph in claim, evidence, meaning, limit order
Draft four moves before polishing style. First answer the experimental question. Second give a representative result with correct units. Third explain what the result suggests in relation to the model or hypothesis. Fourth mark a limitation and, if requested, a targeted improvement. This order prevents the conclusion from becoming a second methods section. It also prevents the final sentence from introducing a dramatic claim that the report never measured.
For an illustrative pendulum report, a cautious draft might say that average swing time tended to increase with string length in the tested range, which is consistent with the course model. It could then note that hand timing and inconsistent release angles limit precise comparison, and that repeated trials with a fixed release angle would test the pattern more reliably. Replace this template with your own values, scope, and teacher’s wording. Never copy a model paragraph as if it were a measured result.
9. Revise for traceability and honest language
Read each sentence and ask where a reader can find its support. A numerical claim should match a table, calculation, or graph. A causal explanation should connect to a model or controlled comparison. A limitation should connect to something that actually occurred or could plausibly affect this setup. Replace ‘proved’, ‘perfect’, and ‘definitely’ when the evidence only supports ‘suggests’, ‘is consistent with’, or ‘within the tested range’. Weakening an overstated verb usually strengthens the scientific writing.
Check that units, variable names, and trial counts are consistent across the report. Do not write ‘accuracy’ when you only compared the consistency of repeated measurements; precision and closeness to a reference are different questions. Confirm that you have not introduced a new result in the conclusion that never appeared earlier. Read the paragraph aloud: if you cannot explain how one sentence follows from the previous one, make that link explicit or cut the sentence.
Review how to attribute a model or outside source in your writing
10. Use a hint only after your own evidence pass
If you remain stuck, first circle the exact gap: choosing the most relevant result, explaining a surprising pattern, or deciding how a limitation affects confidence. You can ask a teacher or classmate about that question. If your school permits AI assistance, Lirno can be downloaded or started for free and used for a hint or a check of your own draft; some AI usage levels or advanced features may require Premium. Verify any response against your notebook, because a tool cannot know measurements you did not supply and can misread a photograph.
A narrow check might ask, ‘Which claim in this paragraph lacks a measurement?’ rather than ‘Write my conclusion’. Keep private names and grades out of any uploaded page, follow your school’s rules, and write the submitted paragraph yourself. Afterwards, cover your draft and explain the claim, evidence, and limitation from memory. If you cannot, return to the table or graph. The goal is a conclusion you can defend in the next lab, not just a polished ending to this one.