How to Discuss Unexpected Results in a Lab Report

Unexpected results do not make a lab report a failure. Report the pattern honestly, compare it with your hypothesis and relevant evidence, evaluate the quality of the data, and offer only explanations that your observations can support. Then propose a specific follow-up test. This article is for undergraduate science students who have usable data that did not match a prediction and need to write a clear Results and Discussion section. You will learn a five-step method, see a worked example, and get language that is cautious without sounding evasive.

The Short Answer: Treat Surprise as a Result

A hypothesis is a testable expectation, not a result you must force the data to confirm. Your report should preserve what happened, including observations that contradict the prediction. Purdue University Libraries explicitly advises students to present all results, including those that do not support the hypothesis, and to analyze errors rather than hide inconvenient findings.

The safest structure is simple: state the expected pattern, state the observed pattern, judge what the evidence does and does not support, compare the result with relevant literature, and name the next test that could distinguish among explanations. This turns “the experiment went wrong” into a scientific argument a reader can evaluate.

Key takeaways

• Keep data and interpretation separate. Put measured values and visible patterns in Results; explain meaning and limitations in Discussion.
• Describe the mismatch precisely. Replace “unexpected” with the direction, size, timing, and consistency of the difference.
• Rank explanations by evidence. Start with what your notes, controls, instrument records, or literature actually support.
• End with a discriminating test. A useful next step should help decide between at least two plausible explanations.

First, Separate Results From Discussion

Many weak lab reports mix observation and explanation in the same sentence. That makes it hard to tell where the evidence ends and speculation begins. The conventional IMRAD structure—Introduction, Methods, Results, and Discussion—gives each job a distinct place. The University of North Carolina Writing Center explains that Results provides the collected data, while Discussion considers whether those data support the hypothesis and evaluates implications and design limitations.

What belongs in Results

Report the outcome neutrally. Identify the comparison, include units, point readers to the correct table or figure, and describe the direction of the pattern. If you ran an approved statistical analysis, report it in the format required by your discipline and course. Do not remove a value merely because it looks surprising; follow your instructor’s stated rules for exclusions.

Example: “After 48 hours, the mean dissolved oxygen concentration was 6.1 mg/L in the shaded treatment and 5.4 mg/L in the illuminated treatment (Figure 2). This direction was opposite to the prediction.” That sentence tells the reader what changed without inventing a cause.

What belongs in Discussion

Interpret the pattern by returning to the research question and hypothesis. Explain the extent to which the evidence supports the prediction, compare the outcome with course concepts or published research, discuss limitations, and suggest a next experiment. University of Minnesota Libraries recommends supporting each conclusion with experimental evidence, comparing results with expectations and literature, and stating what the experiment cannot establish.

Avoid saying that a hypothesis was “proved.” One classroom experiment usually offers evidence that is more or less consistent with a prediction under the conditions tested. Your verbs should match that scope: “supported,” “did not support,” “was consistent with,” or “suggests” are usually more accurate.

Use the TRACE Method for an Unexpected Result

TRACE is a drafting sequence, not a scientific law. Use it to build one coherent discussion paragraph or expand each step into a short subsection when the assignment is longer.

1. Tie the result to the question. Restate the relevant prediction in one sentence. Do not repeat the entire introduction.
2. Report the discrepancy precisely. Name the actual direction and magnitude, the comparison groups, the time point, and whether the pattern appeared across replicates.
3. Assess the evidence quality. Check controls, sample labels, calibration records, raw observations, missing values, and any documented departure from the method.
4. Compare plausible explanations. Consider mechanism, measurement, sampling, assumptions, and design. Link each explanation to a specific observation or source.
5. End with the next test. Propose one change or measurement that could separate the strongest explanations rather than simply saying “repeat the experiment.”

A TRACE paragraph moves from evidence to inference. It does not begin with a long list of everything that might conceivably have gone wrong. Readers should be able to see why each explanation deserves attention and what evidence would raise or lower confidence in it.

Build Explanations Without Blaming “Human Error”

“Human error” is too broad to be useful. It neither identifies a mechanism nor tells a reader whether the conclusion is threatened. Replace it with a documented event and its likely direction of influence. If you do not know whether an issue pushed a measurement up or down, say so.

1. Measurement or procedure

Ask whether the instrument range, calibration, timing, reagent condition, or observation method could explain the pattern. Use the lab notebook: a temperature excursion or delayed reading is evidence; a vague memory that the instrument “seemed odd” is not. Explain which measurement could be affected and in what direction.

2. Sampling and natural variation

Small or uneven samples can produce unstable patterns, while biological and environmental systems can vary even when procedures are consistent. Describe the actual spread, replicate pattern, or sampling imbalance. Do not use variability as a universal excuse; connect it to what the data show.

3. Design or assumption

A comparison may be confounded, a control may be missing, or an analysis may depend on an assumption the data do not meet. These issues limit the conclusion rather than erase the observations. Name the affected comparison and state what alternative interpretation remains possible.

4. A genuine effect

Sometimes the expected theory does not apply under the tested conditions, or an unconsidered mechanism matters. Compare your outcome with relevant studies before making that argument. Reed College’s laboratory report guidance tells writers to discuss expected and unexpected findings in light of hypotheses, predictions, and cited literature, then ask what new questions the results raise.

Use Stronger Scientific Language

Cautious writing is specific, not timid. These substitutions keep the claim proportional to the evidence:

• Weak: “The experiment failed.” Stronger: “The observed direction did not support the predicted increase under the conditions tested.”
• Weak: “The data were wrong.” Stronger: “One reading fell outside the instrument’s calibrated range, so that value was retained and flagged according to the course protocol.”
• Weak: “Human error caused it.” Stronger: “The second sample was read 12 minutes later than planned, which may have lowered its measured temperature; the effect on the group comparison cannot be determined from these data.”
• Weak: “The hypothesis was disproved.” Stronger: “The results did not support the hypothesis in this experiment.”
• Weak: “More research is needed.” Stronger: “A follow-up experiment should measure light intensity at each vessel and randomize vessel position to distinguish an illumination effect from a location effect.”

Notice that the stronger versions name a boundary. They say which result, condition, or comparison is affected. That precision helps the reader judge the claim rather than asking for trust.

Worked Example: From Surprise to Discussion

Consider this hypothetical experiment: a student predicts that radish seeds kept at 25 °C will germinate more often than seeds kept at 15 °C after four days. There are 20 seeds in each treatment. The observed counts are 11 of 20 at 25 °C and 15 of 20 at 15 °C. During the second day, the student records that the warm tray’s paper towel felt dry and adds water. These numbers are invented solely to demonstrate the writing process.

A weak version

“The hypothesis was wrong because human error made the warm seeds too dry. The experiment should be repeated for better results.”

A stronger version

“The results did not support the prediction: after four days, germination was observed in 11 of 20 seeds at 25 °C and 15 of 20 seeds at 15 °C. The temperature effect cannot be separated from a possible moisture difference because the warm tray was recorded as dry on day two and received additional water. Evaporation may therefore offer one explanation for the lower count in the warm treatment, but moisture was not measured. A follow-up study should maintain equal starting water, record tray mass daily, and randomize tray positions while keeping the two temperatures unchanged.”

The stronger passage does four things: it reports the mismatch, cites a documented observation, refuses to claim more than the design allows, and proposes measurements that could test the explanation. It preserves the data instead of rewriting the outcome.

A 20-Minute Revision Checklist

1. Minutes 0–3: mark the claim. Underline every sentence that interprets the result. Ask whether the preceding sentence contains the evidence for it.
2. Minutes 3–7: quantify the mismatch. Add the direction, relevant values, units, comparison, and time point. Verify them against the raw record.
3. Minutes 7–11: audit explanations. Delete unsupported guesses. For each remaining explanation, add the observation, design feature, or source that makes it plausible.
4. Minutes 11–15: state limitations. Write one sentence beginning “This design cannot determine whether…” and finish it with the main alternative interpretation.
5. Minutes 15–18: sharpen the next test. Specify what to hold constant, what to vary, what to measure, and how the outcome would distinguish explanations.
6. Minutes 18–20: check tone and scope. Replace “proved,” “failed,” and “caused” unless the evidence truly justifies them. Confirm that the conclusion answers the original question.

If you organize lab notes in Snitchnotes, turn each claim–evidence pair into a short review card: put the interpretation on one side and the observation, value, or figure reference on the other. This makes unsupported claims easier to spot before submission.

The discussion is often the longest and most interpretive part of a student lab report. Connecticut State Community College suggests asking whether the evidence supports the hypothesis, how the result compares with research, what it signifies, how unexpected findings influenced the data, and what future work should change.

Frequently Asked Questions

Should I include an unexpected result in a lab report?

Yes. Include valid observations even when they contradict your prediction. Report them neutrally in Results, then interpret them in Discussion. If a value may be invalid, follow the course rule for exclusions and explain the decision transparently. Never delete or alter data merely to make the hypothesis look correct.

Does an unexpected result mean my hypothesis was wrong?

It means the result did not support the hypothesis under the conditions you tested. That is narrower than proving the hypothesis universally wrong. A measurement issue, limited sample, confounded design, or genuine exception could contribute. State the scope of the experiment and identify what a follow-up test would clarify.

How many possible explanations should I discuss?

Usually two or three well-supported explanations are stronger than a long catalog of guesses. Rank them using your notes, controls, data pattern, and relevant literature. Explain how each could produce the observed direction, then propose a measurement or design change that could distinguish the leading alternatives.

Can I say “human error” in a lab report?

Avoid the phrase by itself. Name the specific documented event, such as a late reading, incorrect dilution, mislabeled tube, or calibration problem. Then explain which measurement it could affect and whether the direction is known. If you cannot establish an effect, describe it as a limitation rather than a cause.

What if I cannot explain the unexpected result?

Say that the cause cannot be determined from the current data, then identify the strongest constraints on interpretation. You can still compare the pattern with published work and propose a targeted next test. Honest uncertainty is scientifically useful when the report makes clear what evidence would resolve it.

Conclusion: Turn Surprise Into Evidence

To discuss unexpected results in a lab report, preserve the observations, describe the mismatch precisely, separate Results from Discussion, weigh explanations against evidence, and finish with a testable next step. A surprising outcome is not a reason to manufacture certainty; it is a chance to show disciplined reasoning. Use the TRACE method and the revision checklist before submitting. If Snitchnotes is part of your study workflow, review your claim–evidence cards once more so every interpretation has a visible basis.

Sources

• University of Minnesota Libraries — Lab report: guidance on reporting unexpected results, supporting conclusions, comparing evidence, and stating limitations.
• Purdue University Libraries — Parts of a Lab Report: guidance on presenting all results and explaining errors, contradictions, and exceptions.
• University of North Carolina Writing Center — Scientific Reports: explanation of IMRAD roles and the purpose of Results and Discussion.
• Reed College Online Writing Lab — Laboratory Report Instructions: section-level guidance for interpreting expected and unexpected findings.
• Connecticut State Community College — Lab Report: prompts for analyzing significance, unexpected outcomes, future work, and experimental improvements.