Academic English revision

Your Lab Results Are Wrong. Write the Discussion Like an Autopsy.

The Discussion is not a defense of your data. It is a diagnostic of your process. Specificity kills the 'experimental error' shrug dead.

You are at 1:47 AM. Your data table stares back. The absorbance peak is at 520 nm, not the 620 nm your protocol promised. Your hypothesis just died inside a spectrophotometer. And now you are about to type 'This discrepancy may be attributed to experimental error' for the third consecutive time. Stop. That phrase is the academic equivalent of a white flag with a bow tied on top. Your professor has read it in every lab report since 1987. They do not want an apology. They want an autopsy. Give them the knife, not the tissue box.

Here is the thing the orientation slide deck never tells you: the Discussion section is not a defense of your results. It is not 'here is why I am still right even though the numbers say otherwise.' It is a diagnostic. The professor already knows the textbook answer. They know the hypothesis was supposed to work. What they are grading is whether your brain can take a busted result and dissect it into specific, defensible causes. A null result, handled well, scores higher than a perfect result wrapped in two lazy sentences of 'this supports the hypothesis.' The Discussion is where you prove you can think, not just record.

For a 'results do not match' discussion, use a four-beat structure. Beat one: state the discrepancy in one flat sentence. No drama. 'The observed absorbance was 520 nm, diverging from the predicted 620 nm.' Done. Beat two: generate two or three specific, mechanistic explanations. Not 'errors occurred.' Which variable drifted? Was the temperature two degrees above protocol? Did you draw the wrong buffer concentration? Was the reaction window 40 seconds short? Name the mechanism, not the vibe. Beat three: anchor to at least one external source. Has another group reported a similar shift? Is there a known edge case in the literature for this chromophore? Beat four: one sentence on what you would change if you reran it. Four beats. Not seven paragraphs of nervous circling.

Watch the language. Here are the three rewrites that actually land. ❌ 'The results did not match the expected outcome, which may be due to experimental error.' → ✅ 'The absorbance peak at 520 nm, rather than the predicted 620 nm, suggests incomplete oxidation, likely because the mixture was held at 37°C for 12 minutes instead of the protocol's 20.' ❌ 'This discrepancy could be because of human error in measuring the volume.' → ✅ 'A 0.5 mL overpour in the aliquot transfer would shift concentration by approximately 4%, consistent with the 3.8% deviation observed in well B4.' ❌ 'In conclusion, the hypothesis was not supported, but this may be due to limitations in the experimental design.' → ✅ 'The 12-minute incubation, 40% under protocol, is the most probable cause; extending it to 20 minutes in a repeat trial would isolate whether time, not technique, drove the discrepancy.'

The 'human error' trap is where students kill their own Discussion. The moment you write 'experimental error' for the third time, you have told your professor you stopped thinking. You have outsourced your analysis to a catch-all phrase that explains everything and proves nothing. Specificity is the entire game. 'I pipetted wrong' is a confession. 'The 10 µL tip was likely saturated beyond capacity at the 2:30 mark, introducing a 3 µL overshoot into well B4' is science. One is a shrug. The other is a mechanism. If you genuinely do not know what went wrong, say so — but say it as a hypothesis about your own process, not as a blanket alibi.

Practical move: draft your four beats as ugly bullet points first. Write them raw. Then, if your draft reads like a legal disclaimer — every sentence hedged, every claim buried under 'may,' 'could,' and 'it is worth noting that' — run it through easydue and let it flag where you have hidden behind qualifiers instead of committing to a specific cause. You do not need to sound certain. You need to sound like you formed a judgment and can defend it. The tool will not write your mechanism for you, but it will catch the fifth 'perhaps' where you actually meant 'I think the buffer concentration was off.'

You do not need your results to be pretty. You need your thinking to be visible. The student who writes 'my results were wrong' earns a B-plus. The student who writes 'my results were wrong, and here is the specific three-step chain explaining why, and here is how I would test it next week' earns the A — sometimes higher, because the professor just watched you think in real time. The hypothesis was never the point. Your diagnostic skill was. Now close the laptop. Go sleep. The data is not going anywhere, and neither is your ddl.