Results Section Examples for a Research Paper
The results section of a research paper, with length, structure, and results section examples, plus how results differ from the discussion.

The results section of a research paper is the account of what the study found. A useful section names the finding, points to the table or figure that holds the detail, and leaves the explanation for later. A results section example that only says the results were interesting does not do that work.
The results section examples in this guide were written for the guide. None of them is a finding from a published paper, a lab notebook, or a data file. Copy the shape, then use the outcome from your own draft. The passages leave out a real count and a real test. The phrase list includes a made-up test shape, and it says so. There is no study behind them.
When the result is already the one in your notes and only the wording is stiff, Rephrase AI can rephrase that sentence. Rephrase AI is designed to preserve the meaning. Set the new sentence next to the finding and keep it only if the finding is unchanged. Your instructor's rules on AI use still apply.
1. What the results section of a research paper is for
The George Mason University Writing Center describes the results section in an IMRaD report as the place where you present your findings. IMRaD means introduction, methods, results, and discussion. George Mason says the section typically contains only the findings, not an explanation of those findings and not a commentary on them. Results sections are usually written in the past tense, because the work is finished. Tables and figures are labeled and numbered separately. On that overview page, captions go above tables and beneath figures.
The same page draws the reporting line in one sentence. In the results section, you simply report your findings. In the discussion, you comment on them. George Mason's common-problems list names a results section that explains instead of reporting.
The UNC Writing Center describes the same job for a lab report. In UNC's table, the results section provides the raw, uninterpreted data, and it may express that data as a table, as a figure, or as percentages or ratios. The discussion is the section that considers whether the data support the hypothesis. UNC says the results text is where you report facts, and that nothing a reader can dispute should appear there. Conclusions stay in the discussion.
Purdue OWL calls the same four-part shape IMRAD. Purdue says it is common across most of the sciences and is often used in the humanities when the research is empirical. Purdue also says the shape is a poor fit for a proposal, and for exploratory writing in the humanities. On Purdue's page, the results section lists the data. It answers what the researchers found, or what they learned when the study was not an experiment, and how the research questions were answered.
A paper with no empirical finding does not need this section. This guide is for lab reports, surveys, interview studies, and other papers that have a finding to report.
The finding sentence inside a summary is drawn from this section, and it is much shorter. For that short form, see abstract examples.
2. How long a results section should be
Use the limit in the assignment or the journal. These guides do not share one word count.
| Guide | What the guide says about length | What that guidance is for |
|---|---|---|
| UNC scientific-reports handout | Often both the shortest section and one of the most important. Generally shorter than the other sections because it does not add interpretation. A simple experiment may need only a few lines of text. | A lab report. The text can be the whole section when there is little data to repeat. |
| APA qualitative reporting table | The findings section tends to be longer than in quantitative papers. The reason given is the writing needed to let a reviewer judge the analysis. The table does not give a word count. | A qualitative manuscript. The table is marked 2024. |
| UW–Madison science-report guide | No word count. Report the main results with selected data. Do not repeat the table, and do not interpret. | An IMRAD science report. Order several results from most important to least, from simple to complex, or by class. |
| George Mason results and discussion handout | No word count. Present and describe the outcomes in a systematic way. Tables and figures are a common way to keep that description clear. | An IMRaD report in the natural or social sciences. Organization still depends on the discipline, the journal, and the study. |
| Purdue OWL organization page | No word count. List the data and say how the questions were answered. | IMRAD writing, with the caveat that some genres change the shape. |
UNC's reason for the short lab section is selection. Before writing, look at the data and keep what relates to the hypothesis. Resist including every observation. Readers, UNC says, appreciate that selection more than a recital of every fact. A complex experiment can add tables and figures, and the text then points at each one. Do not put the same material in the paragraph and again in the display.
The UW–Madison Writer's Handbook makes the same cut in different words. Report main results, supported by selected data. Representative data means the most common case. Best-case data means the clearest example of the ideal or of an exception. Do not simply repeat the table. The handbook also says to avoid a sentence whose only job is to announce that a table shows a result. Name the result, and point to the table.
APA's qualitative table explains why a findings section can run longer than a quantitative one. The findings have to be written so a reviewer can judge the analytic procedure. The table calls that demonstrative rhetoric. A class lab that UNC would keep to a few lines is not the manuscript that table is describing. Neither pattern replaces the limit on the assignment in front of you.
3. How to structure the results section
Start with the outcome, in the past tense, then point to the display that holds the detail. UNC says most results sections have three parts, which are text, tables, and figures. In a simple experiment the text can stand alone. In a more complex one, refer to each table or figure directly.
APA Style's table setup puts the table number above the title, and the title above the body. Number tables in the order the paper mentions them. Notes, when a note is needed, go below the table. APA Style's figure setup puts the figure number above the title, and the title above the image. Notes go below the figure. Both pages allow two placements. Embed the display after the first mention, or put each display on its own page after the reference list. George Mason's overview still says captions go above tables and beneath figures. That is a different caption rule. Use the rule your assignment or your journal names.
APA's numbers and statistics guide tells writers not to repeat statistics in both the text and a table or figure. The text states the pattern. The display holds the values. In the text, use the word rather than the symbol, as in "the means were," and do not define the ordinary statistical abbreviations the guide lists, such as the abbreviations for mean, standard deviation, and probability. The guide says to report an exact probability to two or three decimals, and to report a probability smaller than .001 as p < .001. Those rules apply when the paper has statistics. They are not a reason to invent a statistic for a sentence that does not have one.
A February 2, 2026 APA Style post covers probability values inside tables and figures. The post says the Publication Manual prefers exact probabilities to two or three decimal places. When a table is already large or a figure is hard to read, the post says an alternative is to mark ranges with asterisks or daggers and to define those marks in a note. The ranges the post names as traditional are probabilities smaller than .05, smaller than .01, and smaller than .001, used alone or together. Use one of those two approaches in a given table or figure, not both. The asterisk approach is for the display, not for the paragraph. The post's own tables are labeled as fictional.
A quantitative study
APA Style's quantitative pages say the current quantitative standards were released in 2018. The 2018 report puts the results items below in its table for manuscripts that report new data collections.
- The flow of participants, including how many people were in each group at each stage, and a flow figure when one is possible.
- The dates that define recruitment and any repeated measurement or follow-up.
- Missing data, including how much is missing, what the report can say about why it is missing, and how the analysis dealt with it.
- Each primary and secondary outcome, for the whole sample and for each subgroup, including the number of cases and the measures that describe the data, such as cell means and standard deviations.
- The result of every inferential test that was run. When the test used null hypothesis testing, include the exact probability and the small set of statistics needed to construct the test.
- An effect-size estimate, and a confidence interval on that estimate, when possible, for each inferential test.
- A clear distinction between primary hypotheses and secondary hypotheses in the tests and estimates.
The line that asks for every inferential test is the line that keeps a non-significant result in the section. A test that did not meet the threshold the paper set is still a test that was run. Report it from the output. Do not drop it because the threshold was missed, and do not fill in a probability the analysis did not produce.
The procedure that produced those tests stays in the methods section. For that account, see methods section examples. The results section starts where that account stops.
A qualitative interview study
APA's qualitative table asks the findings to describe the themes, categories, or narratives, and the meaning the researcher drew from the analysis. A quotation or an excerpt can support that description. The table says the excerpt should not replace the description. The same table's note for reviewers says the findings section tends to be longer than in a quantitative paper. It also says that, depending on the approach, findings and discussion may be combined.
That combination is the qualitative exception to the lab-report rule. The same reviewer note also allows a more personal style that shows the researchers' part in the analysis. UNC, George Mason, and UW–Madison are writing about reports that keep interpretation out of the results. A qualitative paper that follows the table may put the researcher's reading of a theme in the findings. A paper that keeps a separate discussion still has a place to compare that reading with other studies. Use the split your instructor or your journal names.
Where the section ends
End the section on the last finding. UW–Madison says not to interpret in the results. George Mason says not to comment there. The next section is where the finding is explained, compared, and limited. A sentence that only points forward can close the results. A sentence that explains the finding has already become the discussion.
4. Results vs discussion
Results vs discussion is a difference in job, not a difference in heading size.
The George Mason results and discussion handout says the results section systematically reports the findings, and the discussion interprets them. If the paper has no separate conclusion, George Mason says the discussion includes elements of a conclusion, such as limitations and directions for future research. The handout allows three arrangements. Results and discussion can be separate, with the conclusion moves inside the discussion. Results and discussion can be combined, with a separate conclusion. Results, discussion, and conclusion can each be separate.
| Question | Results | Discussion |
|---|---|---|
| What question does the section answer? | What did you find? | What does it mean? |
| What does George Mason say to do? | Report the finding. State the main pattern, support it with the data, and note a secondary pattern or an exception when there is one. | Comment. Explain, compare with other research, say whether a hypothesis was supported, and state what the data mean. |
| What does UNC say to do? | Report uninterpreted data. Select what relates to the hypothesis. | The discussion headings ask whether the data support the hypothesis, and they name anomalous data, conclusions, earlier work, and implications. |
| What does UW–Madison say to do? | Describe the experiment in a sentence or two, without repeating the methods, and report the main results in the past tense. | Open with the most important findings, then patterns, agreement or conflict with earlier work, and a plausible account of those agreements and conflicts. |
| What does Purdue say to do? | List the data and say how the questions were answered. | Place the data in the larger conversation, including what the results might mean, whether they agree with other literature, and what later research could do. |
George Mason's results moves, which the handout draws from the University of Manchester Academic Phrasebank, include pointing back to the aim, pointing at a table or chart, stating a positive result, stating a negative result, noting a surprising result, and reporting a survey or interview finding. The phrase lists later in this guide were written for this guide.
UW–Madison asks the discussion to say whether the study did what the introduction promised. That promise is often the research question. The question is set before the results. For the shape of that question, see research question examples. The results section reports what was found. It does not rewrite the question.
UW–Madison also asks the discussion to avoid three things. Do not overgeneralize. Do not ignore a deviation in the data. Do not speculate in a way that cannot be tested in the foreseeable future. A deviation belongs in the results as a reported exception. The discussion is where that exception is explained.
When the discussion is also the ending of the paper, the limit and the next step live there. For that ending, see conclusion examples. The results section does not add the ending.
5. Results section examples
Each passage below was written for this guide. None of them reports a real lab, a real survey, or a real set of interviews. A published results section would use the counts and the tests from the study. The passages do not invent those values. The test frames in the phrase list are made-up examples of the shape.
Phrases for reporting a finding
Replace X and Y with the outcome in your notes.
- X was higher than Y.
- More of group A named X than group B.
- The repeated action was X.
Phrases for pointing to a table or figure
- Table 1 lists X for each group.
- Figure 1 shows the same comparison.
The table and the figure in these lines are not printed in this guide. A sentence that points at them is a shape, not a data claim.
Phrases for reporting non-significant results
The statistics in these frames are made-up examples. They are not results from a study. Replace them with the output from the analysis.
- X and Y did not differ significantly, t(df) = x.xx, p = .xx.
- The two groups did not differ significantly, t(df) = x.xx, p = .xx.
Phrases for handing off to the discussion
- The discussion takes up what X means.
- Whether the pattern agrees with earlier studies is left for the discussion.
When the finding is right and the sentence is still stiff, use the public paraphrasing tool to rephrase that sentence in your own words. Leave the finding, the named table or figure, and any number from your own output as they are. The tool changes wording only. It cannot supply the result.
Lab report
Strips of the same paper were dipped for a short time and for a longer time. The longer dip carried the dye farther up the strip than the short dip. Figure 1 shows the two strips. The two dip times did not produce a visible difference in the color of the dye. The comparison was the one named in the methods. The discussion takes up what the distance difference means for the dip time.
Why this example works. The finding is in the past tense. The figure is named and is not asked to repeat the sentence. A case with no visible difference stays in the section. The last sentence points at the discussion and does not give the explanation. A real lab report would add the distances and any test the methods named. This model leaves those values out.
Survey study
Students in a morning section and students in an afternoon section completed one questionnaire about where they study on campus. More students in the morning section named the library. Students in the afternoon section more often named a study room in the residence hall. Table 1 lists the place each section named most often. The two sections gave the same answer on whether they studied alone. The discussion takes up what the place difference means for library hours.
Why this example works. The passage reports two outcomes, points at a table for the detail, and keeps a no-difference outcome instead of dropping it. It does not say why the sections differ, and it does not propose a change in library hours. A real paper would add how many students answered, the dates, and any test that was planned. This model has no sample, so the model does not invent one.
Interview study
Tutors described the first minute of a session. One repeated action was to ask what the assignment was asking before looking at the draft. A second repeated action was to read one paragraph aloud with the student. The two actions are a reading of the transcripts. The discussion takes up how those opening actions relate to the question of how a session starts.
Why this example works. The passage names the repeated actions and calls them a reading of the transcripts, which is the kind of meaning the qualitative table asks the findings to carry. It does not put words in a tutor's mouth, and an excerpt is not asked to stand in for that reading. It does not count the interviews. A paper that combines findings and discussion might keep the comparison with other studies in this same section. This model leaves that comparison for a separate discussion.
A weak results paragraph and a revision
Weak version
The results were interesting and are explained later. Various findings were obtained. The tables show the data. The study was a success.
Revision
The longer dip carried the dye farther up the strip than the short dip. Figure 1 shows the two strips. The two dip times did not change the color of the dye. The discussion takes up what the distance difference means.
Why the revision works. The weak version praises the study, hides the finding, and sends the reader to unnamed tables. The revision names the pattern, names the figure, keeps the outcome that showed no change, and stops before the explanation.
A last check before you submit
- The section reports the finding. The explanation waits for the discussion, unless the journal combines the two.
- The text and the display do not repeat the same statistics. The paragraph states the pattern. The table or figure holds the values.
- Every planned test that was run is reported, including a test that did not meet the threshold, using the output from the analysis.
- Tables and figures are numbered in the order the paper mentions them, with the caption placement your style sheet requires.
- A qualitative finding can include the meaning you drew from the material. An excerpt supports that account and does not replace it.
- A rephrased sentence still matches the notes. The finding, the display it names, and any number from the output survive the wording change.


