The results section of a lab report shows the data and nothing else: every number you measured, in tables and figures that carry their own units, captions and numbering, described in a few past-tense sentences and never explained. Build a clean table below — title, unit, two to five rows, an optional mean — then apply the six number rules: numbers are numbers, units travel with every value, precision is earned, means carry their spread, trends are described rather than explained, and every table is called out in the prose.
Honesty note: results conventions vary by course. Some sheets want raw trials in the table and means worked outside it; some merge results with analysis into one section; some want every datum as a figure instead. Where your sheet disagrees with this page, the sheet wins. This page also teaches presentation, not science — it cannot tell you whether the numbers are any good, only whether the section carrying them is readable. Every sentence and table row here comes from an invented, illustrative potato lab, the same teaching lab the other guides on this shelf use.
The results table builder
A results table is the section's spine. The call-out sentences, the trend description and any figure all hang off the columns you set up, which is why it pays to build the table before writing a word of prose. The builder below works the way a results table thinks: name the table (the caption takes it), name the unit, then add one row per condition or per trial with the number you measured. The preview renders a real table with a numbered caption, and a mean row appears when you ask for it — rounded to the same precision as the readings you typed, the way the precision rule below demands.
The builder is deliberately strict in one place: a value cell holds a number. Type words into a value cell — 'about double', 'n/a', 'barely anything' — and the builder flags the row instead of guessing, because that is the exact habit the drill below rehearses. When the table reads right, copy it as Markdown for your lab notebook, or paste the shape into your document and restyle it there. One honest caveat: averaging makes sense when the rows are repeat trials of one condition. Across different conditions the mean is a checksum for your own eye, not a result to hand in.
Results table builder
100% private — runs in your browserOne row per condition or per trial: what you varied or measured, then the number itself.
Empty cells render as em dashes. Values keep the precision you type — the builder never invents decimals the instrument did not report.
Your table saves to this browser only — one storage key, csw-lrr-v1. Adding or removing a row saves at once; typing saves after a short pause. Reset erases the key. Quiz answers, drill clicks and poll votes are never stored.
A table that passes the builder still needs the sentences around it — the call-outs, the trend lines, the handoff to the analysis. Those moves have their own section further down, and the section that gives the whole report its shape is the how to write a lab report planner, which budgets words for results before you type any.
What a results table needs
Five parts do all the work in a results table, and each has one rule. Skip the caption and the table cannot be cited from the prose; skip the unit and every value in the column goes naked; let words into the value cells and the column loses its arithmetic. The fragments below come from the invented potato lab this shelf teaches with — study the shape, not the chemistry.
| Element | The rule | Illustrative fragment (invented lab) |
|---|---|---|
| Caption | Above the table, numbered, specific: quantity, unit, conditions | Table 1. Mean mass change (g) after 30 minutes at four salt concentrations |
| Column headings | One measured quantity per column, the unit in the heading | Salt concentration (percent) · Mass change (g) |
| Condition column | The one thing you varied, as levels or short phrases | 0 / 2 / 5 / 10 percent salt |
| Value cells | Numbers only, decimals aligned, precision consistent down the column | 0.12 · 0.37 · 0.58 · 0.83 |
| Mean row | Only when trials repeat, shown with the spread behind it | Mean of three trials: 0.41 g (range 0.40–0.42) |
Two of these rules do most of the marking work. The unit lives in the column heading so it is written once and inherited by every value underneath — that is the whole trick. And the caption is specific enough to be cited from another report: quantity first, unit in brackets, conditions last. The conventions that live outside the table — caption positions, figure numbering, page set-up — are the territory of the lab report format checker; this page stays inside the frame, on the numbers themselves.
Six number rules for the results section
Six habits cover almost every comment a grader writes into a results margin. Each rule below gets one paragraph and a place in the drill that follows — read them once, then let the drill test the reflexes.
- Rule 1 — Numbers are numbers. A value cell holds a number, full stop. 'About double the control', 'very small' and 'n/a' are sentences, not data, and they break the one thing tables exist for: a reader's ability to compare, average and plot. If a reading genuinely failed, say so in the prose or the method notes and leave the cell empty — an honest gap beats a word wearing a number's clothes.
- Rule 2 — Units travel with every number. A value without a unit is a rumour. In prose, the unit follows the number ('0.83 g'); in a table, the unit sits in the column heading and every value below inherits it. Doing it once per table the right way is fine; writing no unit anywhere is the single most common lost mark in results sections.
- Rule 3 — Precision is earned. The instrument sets the decimals: a balance reading to 0.01 g produces 0.41 g, and pasting 0.41000000001 from a calculator invents digits the measurement never had. Keep the decimal count consistent inside a column — 0.4 beside 0.417 makes the reader guess which one is the typo. This is the plain-language side of significant figures: match the instrument, then match yourself.
- Rule 4 — Means carry their spread. A mean of repeated trials is a summary, and a summary without its scatter is an answer without a question. Report the mean with the range or the trial values ('0.41 g; trials 0.40–0.42 g'), round the mean to the readings' own precision, and never average across different conditions — that erases the very pattern the section exists to report.
- Rule 5 — Trends are described, not explained. Results may say what moved: grew, fell, peaked at 10 percent, stayed flat between 2 and 5. The moment a sentence says 'because' or 'proves', it has left results — the discussion owns causes, mechanisms and verdicts, and it needs clean, unspent numbers to argue with.
- Rule 6 — Every table is called out. Number tables in the order the prose meets them (Table 1, Table 2) and mention each one before it appears: 'Table 1 shows…'. An uncalled table is furniture. The call-out sentence is also where the trend description lives, which is why results prose stays short — the table carries the data, the sentences carry the direction.
Where the border between reporting and interpreting actually sits is easier to feel than to define, and the drill below is the feeling practice. Each item is one invented results sentence cut into three fragments; exactly one fragment breaks its rule.
Number rules drill: find the faulty fragment
Click the fragment that breaks the rule-
Rule 5 — trends without interpretation. Which fragment crosses the line?
'Proves' is a verdict, and verdicts live in the discussion. Results may say the loss grew and where it peaked — the cause waits one section.
-
Rule 2 — units on every number. Which fragment leaves its numbers naked?
Four numbers, no unit. In prose the unit follows every number: 0.12 g, 0.37 g — or rephrase so the heading carries it and say 'in grams' once, plainly.
-
Rule 3 — precision is earned. Which fragment invents digits?
The balance earned two decimals, so the mean gets two: 0.41 g. The calculator's tail digits are precision the instrument never had.
-
Rule 4 — means carry their spread. Which fragment hides the scatter?
A mean without its range is half a result. '0.83 g (trials 0.79–0.88 g)' lets the reader judge the scatter — 0.83 g alone asks them to take it on faith.
-
Rule 1 — numbers are numbers. Which fragment puts a sentence in a value cell?
Words cannot be averaged, plotted or compared. The control gets its number; the comparison to the blank is one call-out sentence, not a cell entry.
-
Rule 6 — every table is called out. Which fragment leaves the table orphaned?
An uncalled table is furniture. One sentence — 'Table 1 shows the mass change at each concentration' — cites it, describes it and keeps the section prose-sized.
The drill never stores anything — a reload gives you a clean board. The sentences are invented for teaching and describe the same illustrative potato lab as everything else on this page.
Means, trials and honest averages
Averages earn their place in results when the rows are repeated measurements of one thing. Three trials at 5 percent salt are three looks at one condition; the mean is the honest summary of them, and the spread — the range across the trials, or the trials themselves — is what makes the mean believable rather than convenient. Put the mean in the table at the readings' own precision, keep the trials visible, and the reader can see both the summary and the evidence it stands on.
The rule bends the moment the rows are different conditions. The mean of 0.12, 0.37, 0.58 and 0.83 g across four salt concentrations is 0.475 g — a number that matches no measurement, hides the trend and would fail the report's own logic, because the pattern between conditions is the finding. Average trials, never conditions. The builder's mean row is built for trials: when your rows are conditions, read it as a checksum for your own eye and leave it out of the report.
| You measured | Report | Leave out |
|---|---|---|
| Three trials at one condition | The mean at the readings' precision, with the range | One tidy trial standing in for all three |
| One reading per condition | Each condition on its own row | A mean across unlike conditions |
| A trial that failed | The gap, plus a note in the method or prose | The word 'failed' inside a value cell |
| A zero result | The 0 with its unit, on its row | The row itself — a zero is a result |
The same numbers travel further than the results section. They feed the verdict in the lab report conclusion, and they get compressed to two sentences in the abstract — the lab report abstract builder handles the lab version, and the research paper abstract formula shows how findings shrink to a six-sentence frame for a full paper. Numbers that were ruled clean once do that travelling cheaply; numbers that were fuzzy get re-argued at every stop.
Trends without interpretation
Trend words are the results section's legal tender: they describe direction without spending a cause. The safe set is short — increased, decreased, peaked, plateaued, stayed flat, roughly doubled, varied little — and each is strongest with its numbers attached: 'peaked at 10 percent salt (0.83 g)' says more than 'peaked dramatically' and commits to nothing about why. The swap table below catches the phrases that wander into results drafts and hands back the reporting version.
| Interpretation creeps in | Reporting version |
|---|---|
| 'which proves the hypothesis' | 'the mass change grew with concentration (Table 1)' |
| 'because osmosis pulled water out' | cut it — the 'because' belongs to the discussion |
| 'a huge increase' | 'an increase from 0.12 g to 0.83 g' |
| 'the results are obviously correct' | 'the three trials sat within 0.02 g of one another' |
| 'the anomaly ruined the data' | 'trial 3 read 0.15 g against a 0.40–0.42 g range' |
Where does the interpretation go instead? One section later, with room to spread out. The lab report discussion builder words the analysis: causes, mechanisms, the verdict on the hypothesis, the comparison with sources. The border is worth patrolling because graders patrol it with a pencil — a 'because' in results does not just cost its own mark, it spends the discussion's opening argument and leaves the next section nothing fresh to say.
The prose around the numbers
Results prose is short on purpose. Three moves cover a school lab: call out every table and figure ('Table 1 shows…'), describe the shape in one or two sentences per table, and hand off — 'these results are examined in the discussion'. Past tense throughout, because the measurements happened before the sentence did; the tense that the rest of the report already uses wins if your sheet is silent. Numbers in prose keep their unit ('the mean change was 0.83 g'), and any number that appears twice — once in the table, once in the sentence — should match to the last decimal, because a mismatch reads as two different experiments.
Resist the urge to narrate every row. The table already shows the reader all four concentrations; the sentence that says 'the mass change grew with concentration, from 0.12 g to 0.83 g' has done the prose's whole job, and three more sentences re-reading the table row by row will read as padding. If the section feels too thin at two sentences per table, the fix is not narration — it is making sure every table deserves its place and every figure has a caption, which the format checker covers in its twelve-point pass.
Five results-section mistakes graders circle
The same five faults produce most of the red ink in results margins, and none of them require bad data — only rushed presentation:
- The interpreted result. 'Which proves…' inside results spends the discussion's argument before the discussion opens. Report the growth; save the proof for the section that argues.
- The naked column. No unit in the heading, no unit in the prose — every value in the column is unreadable to a stranger, and to the grader. One heading fixes the whole column at once.
- The borrowed precision. Calculator digits pasted past the instrument's last honest decimal, or a column that mixes 0.4 with 0.417. Round to the readings; a consistent decimal count is the tell that you understood the instrument.
- The orphan table. Captioned, numbered, never mentioned in the prose. If the table mattered enough to build, it matters enough for one call-out sentence.
- The averaged-away trend. One mean across different conditions, erasing the pattern the section exists to show. Average trials; keep conditions on separate rows.
Every one of these is a ten-second fix once someone names it — which is what the drill above is for. Run a finished results section against the six rules the night before submission, the same way the format page runs its checklist.
Numbers ruled, section still thin?
A clean table is not a finished report. If the results are measured and ruled but the discussion will not come, and the deadline is tomorrow morning, a quoted, deadline-bound writer can take your data and your manual from results table to formatted references while the experiment stays yours. Your data, your structure, your voice. The form below is the way to ask.
What this page cannot do
The honest limits, because a page that rules on numbers should survive an audit:
- It cannot check your data. The builder formats what you type; whether the readings are right, the trials enough and the balance calibrated are questions for your lab notebook, not this page.
- It does not know your sheet. Some courses want raw trials only, some merge results with analysis, some want every datum as a figure. Where the sheet and the six rules disagree, the sheet wins.
- The mean row is arithmetic, not advice. It averages whatever numbers pass validation — including, if you let it, conditions the page itself tells you not to average. Treat the toggle as a checksum and stay the editor.
- It will not write the section. The call-out and trend patterns here are frames to fill; the numbers, the conditions and the direction are yours to supply.
- It is one shelf-mate among many. The study guides hub keeps the whole reports shelf, and the free writing tools cover the mechanical neighbours — word counts, citations, readability — that the finished report also owes.
Quiz: five questions on results numbers
Five questions on the calls a results section makes — what belongs in it, where units live, how means travel, what a value cell holds and which sentences are legal. Every wrong answer reveals the right one, and four options keep the guessing honest.
1. What belongs in the Results section?
2. Where does the unit belong in a results table?
3. Three trials of one condition read 0.41 g, 0.40 g and 0.42 g. What goes in the table?
4. A value cell reads 'about double the control'. What is wrong?
5. Which sentence is legal inside Results?
Frequently asked questions
How do you write the results section of a lab report?
You report the data and stop there. Put the numbers in tables and figures — one measured quantity per column, the unit in the column heading, a numbered caption above every table and below every figure — then write a few past-tense sentences that call each one out: 'Table 1 shows the mean mass change at each salt concentration.' Describe the shape of the data in words (which condition read highest, where the trend flattened) but save every 'because' and every verdict for the discussion. The builder at the top of this page assembles a clean table, and the six number rules below cover the habits graders check first.
Do you put averages in a results table?
Yes, when the rows are repeat trials of the same condition: the mean belongs in the table, next to the spread it came from. Report the mean at the same precision as the readings — three trials of 0.41 g, 0.40 g and 0.42 g give a mean of 0.41 g, not 0.41000000001 g — and keep the range or the trial values visible so a reader can judge the scatter. When the rows are different conditions, do not average them into one number: each condition gets its own row, because a mean across unlike conditions hides exactly the pattern the section exists to show.
Should the results section explain why the data look that way?
No. Results report; the discussion interprets. The results section may describe the trend — 'the mass change grew with salt concentration, from 0.12 g to 0.83 g (Table 1)' — but the cause, the mechanism and the verdict on the hypothesis live one section later. If a result contradicts the hypothesis, it goes in results anyway, stated just as plainly; a results section that quietly drops inconvenient readings is a bigger problem than a failed hypothesis. The words 'proves', 'confirms' and 'because' are the tell that a sentence has wandered out of results.
How many decimal places should lab report numbers have?
As many as the instrument earned, and the same count everywhere in a column. If the balance reads to 0.01 g, values carry two decimals — 0.41 g, not 0.4 g and not 0.413 g — and the mean rounds to the same two. Copying extra digits out of a calculator adds precision the measurement never had, and graders read it as not understanding the instrument. Report the unit with the number or, in a table, in the column heading, so no value ever travels naked.
Does this page save my table?
In this browser only. The table you build — title, unit, the two to five rows and the mean toggle — writes to one local storage key, csw-lrr-v1: adding or removing a row saves immediately, typing saves after a short pause, and reopening the page restores the table. Reset erases the key and returns the empty frame. Quiz answers, drill clicks and poll votes are never stored, and nothing leaves the device — there is no account and no server copy.
Did the number rules survive the drill?