To write the discussion in a lab report, work five moves in order: restate the headline finding with its numbers, compare it against your hypothesis, explain the pattern with a mechanism, name the limitations with their fixes, and close with what the result means beyond the bench. The builder below gives every move a box, a starter sentence and a live word count.
Honesty note: the builder structures your argument; it cannot supply the science. The mechanism and the limits must come from your own experiment's background reading. The Load example pours in an invented, illustrative lab purely to show the five moves filled — every figure in it is made up for teaching, and submitting it would be fabrication as well as plagiarism. For the report around this section, the how to write a lab report guide plans all six IMRaD sections before you draft them.
The discussion builder: five boxes, one argument
Most discussion drafts fail in the same way: the student pastes the results in different words, apologises for "human error", and stops. The section reads thin because interpretation never actually happens — and interpretation is a sequence, not a gift. Each move below has one job, and each job feeds the next: the comparison tells you how hard the mechanism has to work, the mechanism tells you which limitations matter, and the limits bound the implication so it cannot overreach. Work the boxes top to bottom and the section assembles itself into an argument.
The chips under each box do the cold-start work for you. Click one and it drops a sentence frame into the text at your cursor — "The results show that…", "This differs from the expected…" — and you finish it with your variables and numbers. A frame is scaffolding, not an answer: the grader reads your finding, your mechanism and your honesty, not the scaffold. Two boxes deserve a warning before you start. Compare is the move students skip, because admitting the results disagreed with the hypothesis feels like admitting the lab failed — it is the opposite: a named turning point with a mechanism is worth more marks than a clean result with no argument. And Limitations is where "human error" goes to die; each limit must name what it touched and what a repeat would fix.
Discussion builder
100% private — runs in your browserThe headline result in one or two sentences, numbers included, no explaining yet.
The verdict against the hypothesis first: agreed, partly agreed, or turned at a point — name the point.
The because — a mechanism from your background reading that makes the pattern make sense.
Two or three named limits, each with its consequence and the fix a repeat would carry.
What the finding means beyond the bench, bounded by the tested range — then stop.
Nothing drafted yet — work the five boxes top to bottom, or press Load example to see a filled set.
Copy joins the five moves into one flowing discussion — blank moves are skipped, nothing is uploaded. Your boxes save to this browser only, 300 milliseconds after you stop typing; Reset erases them. The example fill never saves: reload without typing and it is gone.
Once the draft reads right, two neighbouring jobs stay. Compress the whole report — discussion included — into its first paragraph and see whether the argument survives: the lab report abstract builder runs that drill with a live 150–250 word budget. And fill the boxes this section argues from in one pass with the lab report template, whose seven sections feed these five moves directly.
The frame bank: 24 sentence starters
The builder's chips give you six starters per box. The frame bank below is the heavier version: 24 full frames across four families, written out as complete sentences with the blanks in square brackets. Copy a frame, paste it anywhere, and refill it with your variables, values and mechanisms. Restating and Comparing frames carry the verdict; Explaining frames force a because; Limiting frames are shaped so a fix rides along with every fault — which is what separates analysis from apology.
Restating the finding
Put the headline result on the record — numbers first, meaning later.
- The central finding of this experiment is that [change] produced [direction] in [measurement] across every tested level.
- The mean [measurement] moved from [value] at the lowest level to [value] at the highest, and the trials within each group sat close together.
- The clearest single result was [observation], which no individual trial contradicted.
- Restated in one line: [what you changed] drove [what you measured] in the direction [direction] within the tested range.
- The data in Table 1 put the pattern on the record: [column] rose while [column] stayed flat.
- Two of the three levels behaved as the pattern predicted; the exception was [level], where [observation].
Comparing with the expectation
Give the verdict against the hypothesis, including exactly where it turned.
- The results support the hypothesis for the lower levels of the tested range, because [pattern] matches the predicted direction.
- This differs from the expected outcome at [level]: the hypothesis predicted [expectation], but the data show [observation].
- The prediction held up to [point], after which the measurements turned and [direction].
- Against the background claim that [claim], these results land [where] — [agreement or difference] with the expected direction.
- The outcome was the opposite of the expected effect at [level], and that turning point drives the explanation below.
- Part of the expectation survived contact with the data: [element] appeared exactly as predicted, while [element] did not.
Explaining the mechanism
Force a because — name the process that connects cause to measurement.
- A likely mechanism is that [process], which would explain why [observation] rather than [alternative].
- The turning point at [level] is consistent with [mechanism]: once [condition] passes a threshold, [effect] takes over.
- This mirrors the role of [factor] described in the background, where [mechanism] produces the same direction of change.
- The pattern makes sense if [cause] drives [measured effect], because [chain in one clause].
- An alternative reading is that [alternative cause], though the data collected here cannot separate the two.
- In other words, the results trace a single chain: [change] causes [intermediate step], which causes [measured effect].
Limiting the claim
Named limits with their consequences — every fault carries a fix.
- One limitation sits closest to the result: [instrument] is a coarse stand-in for [ideal measure], and [better tool] would measure it directly.
- The sample size (n = [number]) leaves room for [source of variation] to have shaped the means, so the differences should be read as indicative.
- Because [condition] was not controlled across trials, [possible influence] cannot be ruled out; a repeat would hold it fixed.
- The [time window] was short enough that [slower effect] had no chance to appear, which narrows the claim to [range].
- Only [one brand or variety] was tested, so the finding may not transfer to [other cases].
- A repeat with [improvement] would test the explanation directly: if [prediction], the mechanism gains support.
Every frame is a shape, not content — replace the bracketed blanks with your own variables and values before the sentence goes near a report.
What each move must do
A discussion is only as good as the job each paragraph does. The table below is that job description in one view — the move, its single responsibility, the word share a typical school-lab discussion gives it, and the frame that opens it. Shares are heuristics for a standard write-up, not a rubric; your lab sheet's word count outranks them every time.
| Move | Its one job | Typical share of the discussion | The frame that opens it |
|---|---|---|---|
| Restate | Put the headline finding on the record, numbers first | ~15% | "The results show that …" |
| Compare | Give the verdict against the hypothesis, turn point named | ~20% | "The results support the hypothesis because …" |
| Explain | Argue the mechanism that makes the pattern make sense | ~25% | "A likely explanation is that …" |
| Limit | Own two or three limits, each with its fix | ~25% | "One limitation was …, which the repeat would fix by …" |
| Imply | Bound the meaning to the tested range, then stop | ~15% | "Within the tested range, this suggests that …" |
Notice where the weight sits: the three interpretation moves (compare, explain, limit) carry roughly seventy percent of the section, because graders reward argued numbers over repeated ones. If your draft spends its words on move one, it is a results section wearing a discussion's heading. And watch the boundary at the end: the implication move hands the baton to the lab report conclusion, which answers the aim in two or three sentences and stops — it never re-argues the mechanism.
The example discussion, move by move
Press Load example above and an invented lab pours into the five boxes: The Effect of Sugar Concentration on the Fermentation Rate of Yeast. Four doses — 0, 5, 10 and 15 percent sugar — three trials each, with foam height after ten minutes standing in for carbon-dioxide volume. The hypothesis commits to a rise that plateaus once the yeast has more sugar than it can use. The data disagree with that second half, and that disagreement is exactly what makes the example worth studying: means of 4, 21, 34 and 12 millimetres rise beautifully, then fall off the edge of the prediction.
Watch what each move does with that awkward number. Restate puts the fall on the record without flinching — means next to their trials, units attached, no explaining. Compare gives the split verdict: supported up to 10 percent, contradicted above it, turn point named. Explain reaches for a mechanism from the background reading — osmotic stress pulling water out of yeast cells past a threshold — and frames it as testable, not proven. Limit owns three coarse spots: foam as a stand-in for gas volume, a single yeast batch, a room that warmed two degrees across the session, each with the fix a repeat would carry. Imply bounds the practical claim to the tested range and stops.
Every figure in that fill is invented for teaching, which is exactly what makes it safe to study and unsafe to submit — graders check numbers against lab notebooks, and these numbers live nowhere. Study the moves, not the values: the restate never says "because", the compare names the exact dose where the story changed, and the implication never claims more than 0 to 15 percent. If you want a complete lab read this closely with a grader's margin notes, the lab report example page annotates a finished write-up section by section.
When the results differ from the hypothesis
A disagreement between your data and your prediction is not a failed lab; it is the moment the discussion earns its name. Handle it in three steps, in this order. First, check the arithmetic and the units — a sign error manufactures more "surprising results" than biology ever has. Second, say exactly where the disagreement starts: not "the results differed somewhat" but "the means rose to 34 millimetres at 10 percent and fell to 12 at 15". A turn point with a number can be explained; a vibe cannot.
Third, offer a mechanism and mark it as a claim, not a confession: "osmotic stress at high sugar concentrations would slow the cells exactly when the mixture is sweetest — a repeat across 8 to 14 percent could map the peak". That sentence does three marks' work at once: it explains, it stays honest about certainty, and it proposes the test that would settle the question. What never helps: deleting the disagreeing trials, rewriting the hypothesis after the fact, or hiding behind "human error". Graders have read all three a thousand times, and none of them is interpretation. The annotated pages in the reports shelf show this repair in context — and the frames in the Comparing family above are written for precisely this moment.
Limitations: analysis, not apology
The limitations paragraph has a bad reputation because most drafts write it as a guilt list: "human error, time pressure, limited equipment". None of those names a cause, so none of them can carry a fix, so none of them earns marks. The earning shape is one sentence per limit with three parts: the coarse thing, what it could have changed, and what a repeat would do. "Foam height is a coarse stand-in for gas volume; a gas syringe would measure it directly." "One yeast batch was tested, so batch variation went unmeasured; a second brand would test the transfer." "The room warmed about two degrees across the session; a water bath would hold it fixed."
Two or three limits, chosen for closeness to the result, beat five chosen at random. The useful test: could this limit plausibly have produced the pattern you just explained? If yes, it belongs — and saying so connects your limitations to your mechanism instead of trailing them after it like a receipt. If no, cut it. A discussion that says "the biggest threat to these means is the foam measure, and here is the tool that would remove that threat" reads like a scientist wrote it, because one did.
Draft built, deadline tomorrow?
Five boxes filled is half the distance; a submission-ready discussion also needs the numbers argued cleanly and the whole report holding together. If the deadline is close and the section still reads thin — or the mechanism paragraph is where the evening went — a quoted, deadline-bound writer can turn your measurements and your manual into a report you would defend out loud. Your data, your structure, your voice. The form below is the way to ask.
What this builder cannot do
The honest limits, because a tool that hands out sentence shapes should survive an audit:
- It cannot supply the science. Frames hold the sentence shapes open; the finding, the mechanism and the limits must come from your experiment and your background reading. An explained pattern you cannot defend is worse than an unexplained one.
- It does not know your rubric. Five moves is the common shape; some sheets merge the discussion into Analysis or demand a separate error-analysis section. Where your sheet and the builder disagree, the sheet wins.
- The example is invented. The yeast, the doses and every millimetre are illustrative teaching material. Quoting them in a submission is fabrication as well as plagiarism.
- The copy is plain text. It carries your five paragraphs — no headings hierarchy, no tables, no citation formatting. Formatting and references happen in your editor; the free writing tools include citation generators for the reference list.
- It stops at the school lab. A term investigation or a coursework paper discusses its results against published literature, which needs a different compression at the front. The research paper abstract guide covers that species, and the study guides hub keeps the whole reports shelf one click away.
Quiz: five questions on discussion moves
Five questions on the calls a discussion makes — where it opens, how it handles disagreement, what a limitation must name, where the conclusion starts, and what the builder stores. Every wrong answer reveals the right one, and four options keep the guessing honest.
1. Which opening sentence belongs in the discussion?
2. Your results disagree with the hypothesis at the highest dose. What does the discussion do?
3. Which limitation line would a grader reward?
4. Where does the discussion end and the conclusion begin?
5. What does the builder on this page save?
Frequently asked questions
How do I write the discussion in a lab report?
In five moves, one short paragraph each: restate the headline finding with its numbers, compare it against the hypothesis including exactly where the results turned, explain the pattern with a mechanism, name the limitations with their fixes, and close with what the finding means beyond the bench. The builder on this page gives each move a box and sentence-frame chips that drop a starter into the text, so the discussion grows in order instead of arriving as one intimidating blank page.
What is the difference between the discussion and the conclusion?
The discussion argues; the conclusion answers. The discussion is where the results meet the hypothesis — comparison, mechanism, limitations, implications — and it can run to several paragraphs. The conclusion is two or three sentences that answer the aim, bound the claim to the tested range, and stop without new numbers. If your conclusion repeats every figure your discussion just argued, the boundary has moved: keep the arguing up top and only the verdict below.
What goes in the limitations part of a discussion?
Two or three named limits, each with its consequence and its fix. The shape that earns marks is one sentence per limit: the measure or control that was coarse, what it could have changed, and what a repeat would do about it. 'Foam height is a coarse stand-in for gas volume; a gas syringe would measure it directly' earns marks. 'Human error' and 'we could have done better' name nothing, so they fix nothing.
Does the discussion builder write the discussion for me?
No — the frames are scaffolds, not answers. Your findings, your hypothesis and your mechanism come from your own experiment; the builder holds the order and the sentence shapes. The Load example pours in an invented lab (sugar concentration versus yeast fermentation) so you can see all five moves filled — every figure in it is made up for teaching, and submitting it would be fabrication as well as plagiarism. What the tool does honestly: structure, counters, copy, and nothing leaving your browser.
Is what I type saved anywhere?
Only in this browser. The five boxes save to local storage under one key, csw-lrd-v1, 300 milliseconds after you stop typing, so a closed tab does not lose your draft; nothing is uploaded anywhere. The Load example fill never saves — reload without typing and it is gone. Reset erases the stored key and empties every box, which makes the builder safe on a shared computer. Quiz answers, poll clicks and copied frames are never stored.
Did the five moves make the discussion click?