Manuscript Writing

Results | Discussion — What Goes Where

The single most common writing error in submitted manuscripts is content in the wrong section — interpretation buried inside Results, raw data restated inside Discussion. Here is how to sort each sentence correctly the first time.

Research Ramp May 2026 6 min read

Reviewers do not always say it directly, but a remarkable proportion of the comments they make on Results and Discussion sections reduce to one underlying complaint: this sentence is in the wrong place. Interpretation has leaked into Results. Raw data has been restated inside Discussion. Citations are doing analytical work where they should be merely framing.

Understanding the difference between results and discussion is not a technicality. It is the cleanest signal of whether a writer has internalised what each section is for. Once the boundary is clear, the sorting becomes mechanical — and the manuscript tightens by roughly fifteen percent without losing a sentence of substance.

Results
what
A precise, ordered account of what the data show. Descriptive. Reports without interpreting. No citations except for the methods being referenced.
Discussion
so what
A reasoned argument about what those findings mean for the field. Interpretive. Engages with literature. Names mechanisms, contributions, and boundaries.

The fundamental distinction

Results report. Discussion argues. The simplest test is the verb: if the sentence does any work of interpreting, comparing, explaining, suggesting, refining, or recommending, it belongs in Discussion. If the sentence only states, presents, shows, indicates, or summarises what the analysis produced, it belongs in Results.

The table below sets out the contrast across the six dimensions that most consistently separate the two sections in well-written manuscripts. The companion guides on writing the Results section and writing the Discussion section go deeper into each side individually.

Results Discussion
Purpose Report what the data show, accurately and in order. Argue what those findings mean for the field.
Voice Descriptive, neutral, precise. Interpretive, comparative, claim-making.
Verb tense Past — describing the analysis that was performed. Present — making claims about what is now known.
Content Statistics, tables, figures, thematic patterns. Interpretation, comparison to prior work, mechanisms, contributions.
Citations Rare — only when explaining a measurement or technique. Frequent — engaging directly with prior literature.
Forbidden Interpretation, theoretical claims, recommendations. New data, restated tables, novel statistics not introduced earlier.

The four-question diagnostic

When you cannot tell which section a sentence belongs in, run it through these four questions. The answers almost always point you the right way.

Which section does this sentence belong in?

Take any ambiguous sentence and ask, in order:

1

Does the sentence make a claim that goes beyond what the data show?

If yes → Discussion. Interpretation, mechanism, generalisation, recommendation — all live in Discussion regardless of how confident the claim sounds.

2

Does the sentence reference prior published work?

If yes → Discussion. The Results section cites methods and instruments only. Comparative citations belong with comparative argument.

3

Does the sentence describe a statistical or thematic pattern from your own analysis?

If yes → Results. Even if the pattern feels important, the report of it goes in Results. The meaning of it goes in Discussion.

4

Could a reader of just this sentence tell whether you are reporting or arguing?

If no → the sentence is doing two jobs at once. Split it. The reporting half goes in Results; the arguing half goes in Discussion.

A quick sorting test

Below are six short statements drawn from manuscripts we have edited. For each, the correct destination is tagged on the right. Read each one and see whether the tag matches your instinct.

Where does each sentence belong?

Six real examples, six correct answers.

"The autonomy support scale showed acceptable internal consistency (α = 0.84)."
Results
"This finding extends Reeve's (2016) framework into vocational contexts."
Discussion
"Engagement scores were significantly higher in the autonomy-supportive condition (M = 4.3, SD = 0.6)."
Results
"One plausible mechanism for this effect is the reduction in cognitive load associated with choice provision."
Discussion
"Three themes emerged from the thematic analysis: ownership, recognition, and pace."
Results
"Teachers in larger classrooms should consider rebalancing autonomy provision toward task choice rather than topic choice."
Discussion

Results | Discussion

If you cannot tell which one a sentence belongs in, the sentence is almost always trying to do both at once — and the fix is to split it.

Three edge cases that confuse people

Once the core distinction is clear, a few specific edge cases still cause hesitation. Each has a clean answer.

Tables and figures — both sections?

Tables and figures live in Results. In Discussion, refer to them by number; never reproduce the data.

Unexpected findings?

The unexpected finding itself goes in Results. The interpretation of why it was unexpected goes in Discussion.

Effect-size interpretation?

Reporting the value (β, η², Cohen's d) belongs in Results. Calling it small, moderate, or large is interpretation — that belongs in Discussion.

Manuscript Health Checker

Check whether your sentences are in the right section

Upload your Results and Discussion sections. We flag interpretation buried in Results, data restated in Discussion, and citations doing the wrong job — sentence by sentence.

Check Structure →

Four sorting mistakes that recur

1. Interpretation hidden inside Results

The most common error. Phrases like "suggesting that," "indicating a strong relationship between," or "in line with previous work" drift into Results paragraphs and quietly do Discussion's work. Strip them out. The reader does not need your interpretation while they are still trying to read the numbers.

2. Statistics reappearing in Discussion

Writing "as Table 3 showed (β = 0.42, p < 0.001), autonomy support predicted engagement…" at the start of a Discussion paragraph is a tell. The Discussion's reader has already read Results. Refer to the finding by name ("the autonomy effect"), not by its statistics.

3. Citations doing work in Results

Comparative citations belong in Discussion. If your Results paragraph contains "unlike Chen (2019), our participants reported…", you have crossed the boundary. Move the comparison to the corresponding Discussion paragraph; leave the descriptive report in Results.

4. Section length imbalance

In well-balanced empirical papers, Results and Discussion are of comparable length. A Discussion that runs half as long as Results usually has interpretation hiding in the wrong place; a Discussion twice as long usually has data restatement padding it out.

A useful check: read your Results section ignoring all sentences that contain "suggests," "indicates," "in line with," "consistent with," or any citation other than a method reference. If the section still reads as a complete account of what you found, those sentences belong in Discussion. If the section becomes incoherent without them, you have buried interpretation that needs to be repatriated.

The bottom line

Sorting Results from Discussion is mostly a discipline, not a skill. Results report what the data show; Discussion argues what those findings mean. Get the verbs right, keep the citations on the correct side of the boundary, and the two sections will do their separate jobs cleanly. Most reviewer comments on Results and Discussion vanish when the sorting is done well — and the manuscript reads as the work of a careful writer rather than a hurried one.

Free in 60 Seconds

Check your structure before you submit

Upload your draft. The Manuscript Health Checker flags every sentence in the wrong section — interpretation in Results, data restatement in Discussion, citations doing the wrong job.

Check Structure