A research results section presents the findings produced by a study’s analysis. It reports relevant evidence objectively through text, statistics, tables, figures, themes, or quotations. The section normally explains what was found, while the discussion explains what the findings mean, why they occurred, and how they relate to previous research.
Introduction
The results section is where the evidence from a research project becomes visible. After explaining the research problem and methods, the writer uses this section to show what the data revealed.
Writing the section is not simply a matter of copying output from SPSS, R, Stata, Excel, NVivo, or another analytical tool. Researchers must select relevant findings, organize them around the study’s questions, report them accurately, and present them in a form readers can evaluate.
This guide explains how to write quantitative, qualitative, mixed-methods, and systematic-review results. It also covers statistical reporting, tables and figures, non-significant findings, reporting guidelines, digital research tools, artificial intelligence, common mistakes, and a reusable structure.
Key Takeaways
- The results section answers: What did the study find?
- Findings should normally be organized around research questions, hypotheses, outcomes, or themes.
- Quantitative reporting should include relevant estimates, measures of uncertainty, sample sizes, and test results—not p-values alone.
- Qualitative findings should present analytically developed themes or categories supported by appropriate evidence.
- Relevant null, negative, contradictory, and unexpected findings should not be hidden.
- Journal instructions and study-design reporting guidelines take priority over generic advice.
What Is a Research Results Section?
The research results section is the part of an empirical paper, dissertation, thesis, or report in which the researcher presents the outcomes of data collection and analysis.
It may contain:
- Participant or sample characteristics
- Response and retention rates
- Missing or excluded observations
- Descriptive statistics
- Estimates and confidence intervals
- Statistical test results
- Tables, charts, graphs, or other figures
- Qualitative themes, categories, patterns, or narratives
- Representative quotations or field evidence
- Integrated mixed-methods findings
- Study-selection and synthesis results in systematic reviews
The section should provide enough evidence for readers to understand what was observed and evaluate whether later conclusions are supported.
What Is the Purpose of the Results Section?
The primary purpose is to present the evidence that answers the study’s research questions or tests its hypotheses.
A strong results section performs five functions:
- Reports the findings accurately.
- Organizes evidence in a logical sequence.
- Shows the connection between analyses and research questions.
- Makes important patterns visible through appropriate prose and visuals.
- Creates an evidential foundation for the discussion and conclusion.
The results section should not be treated as a list of every calculation performed. It is a structured account of the findings that are relevant to the stated aims.
Where Does the Results Section Appear?
In a conventional IMRaD research article, the sequence is:
- Introduction
- Methods
- Results
- Discussion
A thesis or dissertation may use a similar structure, although the results may appear as one long chapter or several empirical chapters.
Not every discipline separates results and discussion. Some qualitative traditions, humanities-oriented studies, design research, case studies, and journal formats combine them under a heading such as Results and Discussion or Findings and Interpretation.
Always follow:
- The target journal’s instructions
- Departmental or university requirements
- The conventions of the research design
- The supervisor’s approved structure
Results vs. Findings vs. Discussion vs. Conclusion
The terms results and findings are often used interchangeably, but their use can differ by discipline.
| Section | Main question | Typical content | Interpretation |
|---|---|---|---|
| Results | What did the analysis show? | Statistics, estimates, tables, figures, observed patterns | Usually limited |
| Findings | What patterns or understandings emerged? | Themes, categories, narratives, quotations | May include method-appropriate analytical commentary |
| Discussion | What do the findings mean? | Interpretation, comparison with literature, explanations, implications | Central purpose |
| Conclusion | What is the overall answer? | Synthesis, contribution, final implications, recommendations | Concise and integrative |
A useful test is to examine the function of a sentence.
Results sentence:
“Students in the structured-feedback group obtained higher mean scores than students in the standard-feedback group.”
Discussion sentence:
“This difference may indicate that structured feedback helped students identify and correct conceptual errors.”
The first reports a pattern. The second proposes an explanation.
What Should You Do Before Writing?
Check the required reporting standard
Do not begin with a generic template alone. Identify whether your research design has an established reporting guideline.
Examples include:
- APA JARS for quantitative, qualitative, and mixed-methods psychology research
- CONSORT for randomized trials
- STROBE for observational studies
- PRISMA for systematic reviews
- COREQ for interview and focus-group research
- SRQR for qualitative research
- Discipline- or journal-specific reporting checklists
A reporting guideline is not a substitute for good methodology. It helps ensure that important information is not omitted from the report.
Confirm that the analysis is final
Before drafting:
- Check data cleaning decisions.
- Verify variable labels and units.
- Confirm sample sizes.
- Review exclusion rules.
- Reproduce key analyses.
- Check table values against analysis output.
- Confirm that primary and secondary outcomes are correctly identified.
- Distinguish prespecified from exploratory analyses.
Writing too early can create inconsistencies when the analysis changes.
Build a research-question-to-result map
Create a planning table before writing.
| Research question or hypothesis | Analysis | Main result | Evidence format |
|---|---|---|---|
| RQ1: What was the average satisfaction level? | Descriptive analysis | Mean and distribution | Text and table |
| H1: Did the intervention improve scores? | Independent-samples t test | Estimated group difference | Text and confidence-interval plot |
| RQ2: How did students describe the intervention? | Thematic analysis | Three themes | Subheadings and quotations |
This prevents irrelevant output from entering the section and helps identify unanswered research questions.
Audit completeness
Compare:
- Planned analyses
- Analyses described in the methods
- Registered or approved outcomes
- Tables and figures
- Results reported in the text
Every major method should lead to a result, even when the result is null, inconclusive, or unsuitable for the primary narrative.
How to Write a Research Results Section
Step 1: Begin with a brief orientation
Open with one or two sentences identifying the dataset, sample, or order of presentation.
Example:
“This section presents findings from the survey of 312 undergraduate students. Results are organized according to the three research questions, beginning with sample characteristics and descriptive statistics.”
Avoid repeating the full research background or methodology.
Step 2: Report participant or sample flow
Where relevant, state:
- How many individuals, cases, records, documents, or observations were assessed
- How many were eligible
- How many participated or were included
- How many withdrew or were lost
- How many observations were excluded
- Why exclusions occurred
- The final sample used in each analysis
Example:
“Of the 428 students invited to participate, 337 opened the survey and 306 completed all primary measures. Eleven cases were excluded because they failed the prespecified attention check, producing a final analytical sample of 295.”
A flow diagram may be clearer for multi-stage recruitment, clinical trials, longitudinal studies, and systematic reviews.
Step 3: Describe the sample and data quality
Provide descriptive information necessary to interpret the findings.
Depending on the research, this may include:
- Age, location, educational level, or other relevant characteristics
- Group sizes
- Baseline values
- Response rates
- Missing-data patterns
- Measurement reliability
- Distributional characteristics
- Data-quality problems
- Protocol deviations
Do not include every demographic variable merely because it was collected. Include variables relevant to understanding the sample, evaluating comparability, or interpreting the analysis.
Step 4: Present descriptive findings
Descriptive statistics summarize what was observed before inferential claims are considered.
Common descriptive measures include:
- Frequencies
- Percentages
- Means
- Medians
- Modes
- Standard deviations
- Interquartile ranges
- Minimums and maximums
- Rates
- Proportions
Choose statistics appropriate to the measurement scale and distribution.
Example:
“Participants studied for a median of 7.5 hours per week (IQR = 4.0–11.0). The mean examination score was 72.8 (SD = 10.6).”
A median and interquartile range may be more informative than a mean and standard deviation when a distribution is strongly skewed.
Step 5: Report primary findings first
Primary outcomes and prespecified hypotheses should usually appear before secondary or exploratory analyses.
For each result:
- Identify the research question, hypothesis, or outcome.
- Name the analysis when necessary.
- State the direction and magnitude of the finding.
- Report the relevant numerical evidence.
- Refer to a table or figure where appropriate.
- Avoid explaining why the result occurred.
Example:
“To test H1, post-intervention scores were compared between groups. The intervention group scored 6.4 points higher than the control group, 95% CI [3.1, 9.7], t(118) = 3.84, p < .001, Cohen’s d = 0.70.”
This sentence reports the estimated difference, uncertainty, test statistic, p-value, and standardized effect.
Step 6: Present secondary and exploratory analyses clearly
Secondary, subgroup, post hoc, and exploratory analyses should be labeled honestly.
Do not present an unexpected exploratory association as though it had been predicted from the beginning.
Useful labels include:
- Secondary analysis
- Prespecified subgroup analysis
- Sensitivity analysis
- Post hoc comparison
- Exploratory analysis
Detailed exploratory output may be placed in supplementary materials when it is not central to the paper.
Step 7: Select the right presentation format
Choose the format that communicates the result most efficiently.
- Use text for a few important values.
- Use a table when readers need exact values across several variables or groups.
- Use a figure when shape, trend, uncertainty, distribution, or relationship is important.
- Use an appendix or supplement for detailed supporting output.
- Use a joint display to integrate quantitative and qualitative findings.
Every table and figure should be mentioned in the text.
Step 8: Report null, negative, and unexpected findings
A relevant result does not become irrelevant because it fails to support the hypothesis.
Report:
- Non-significant tests
- Effects close to zero
- Wide confidence intervals
- Contradictory patterns
- Failed manipulations
- Unexpected adverse events
- Findings that conflict with predictions
- Important negative cases in qualitative research
Avoid saying that a non-significant result “proved there was no effect.” A non-significant result may reflect a small effect, imprecise estimation, insufficient information, or compatibility with several possible effect sizes.
Step 9: Check consistency and edit
Verify that:
- Sample sizes agree across text, tables, figures, abstract, and methods.
- Variable names remain consistent.
- Percentages use the correct denominator.
- Test statistics match the analysis output.
- Confidence intervals match the reported estimate.
- Tables and figures are numbered in order.
- Every prespecified primary outcome is reported.
- Exploratory analyses are clearly labeled.
- Interpretive sentences have been moved to the discussion.
How to Report Quantitative Results
Quantitative results describe numerical patterns, differences, associations, predictions, or model estimates.
Participant flow and missing data
Report the number of observations entering each major stage of the study.
Also explain:
- How much data were missing
- Which variables were affected
- Why values were missing, when known
- Whether cases were excluded
- Whether imputation or another missing-data method was used
The method itself belongs mainly in the methods section. The results should report its consequences for the analytical sample and findings.
Descriptive statistics
Report the sample size and appropriate summaries for major variables.
Categorical data:
“Of the 240 respondents, 156 (65.0%) selected online delivery, 61 (25.4%) selected blended delivery, and 23 (9.6%) selected classroom-only delivery.”
Continuous data:
“The intervention group had a mean post-test score of 81.3 (SD = 7.2), compared with 75.1 (SD = 8.4) in the control group.”
Estimates, uncertainty, and effect sizes
A good quantitative report emphasizes what was estimated and how precise that estimate is.
Where appropriate, report:
- Mean differences
- Regression coefficients
- Odds ratios
- Risk ratios
- Correlations
- Standardized mean differences
- Hazard ratios
- Marginal effects
- Credible intervals in Bayesian analyses
- Confidence intervals in frequentist analyses
A p-value does not communicate the size or practical importance of an effect. Report effect estimates and uncertainty whenever the design and analysis allow it.
Statistical tests
The exact details depend on the test and disciplinary style. Common elements include:
- Test name
- Sample size
- Estimate or group values
- Test statistic
- Degrees of freedom
- Exact p-value, except where convention uses a threshold such as p < .001
- Confidence interval
- Effect size
Do not write only:
“There was a significant difference.”
Report the difference and the evidence supporting it.
Statistical assumptions and model fit
Important assumption checks or model diagnostics may be summarized when they affect interpretation.
Examples include:
- Severe non-normality
- Heteroscedasticity
- Multicollinearity
- Poor model convergence
- Influential observations
- Non-proportional hazards
- Poor calibration
- Inadequate fit
Minor diagnostic output need not overwhelm the section. Put detailed diagnostics in an appendix or supplement when appropriate.
How to Report Qualitative Results
A qualitative findings section presents analytically developed patterns supported by evidence from the dataset.
It should not be reduced to a collection of quotations.
Organize findings around themes or analytical claims
Common organizing units include:
- Themes
- Categories
- Processes
- Cases
- Narratives
- Discourses
- Experiences
- Chronological stages
- Conceptual models
Each subsection should make a clear analytical point.
A useful sequence is:
- Name the theme or finding.
- Define what it represents.
- Describe its scope and variation.
- Present supporting evidence.
- Include exceptions or contrasting cases where relevant.
- Transition to the next finding.
Use quotations as evidence
A quotation should illustrate or complicate an analytical claim.
Weak approach:
“Participant 4 said, ‘I found the process difficult.’”
Stronger approach:
“Participants frequently described the application process as difficult to navigate, particularly when instructions changed between institutional webpages. One participant explained, ‘I completed the form twice because each office gave me a different list of documents’ (Participant 4).”
The quotation supports the theme but does not replace the analysis.
Protect confidentiality
Use identifiers that match the approved ethics and consent process.
Examples include:
- Participant 07
- Teacher B
- Interviewee 12
- Pseudonyms
- Role-based labels
Remove or generalize details that could unintentionally identify individuals.
Show variation and negative cases
Do not imply that every participant shared the same experience unless the data support that statement.
Use proportionate language such as:
- Most participants
- Several interviewees
- A small group
- Across all three sites
- Two participants offered a contrasting view
Avoid converting qualitative counts into claims of population prevalence unless the study design supports that inference.
Interpretation in qualitative findings
The boundary between findings and discussion is not identical across all qualitative traditions. Some approaches combine evidence and interpretation because interpretation is part of the analytical result.
The correct question is not simply, “Does this sentence contain interpretation?” Instead ask:
- Is this an interpretation produced by the stated analytical method?
- Is it grounded in the dataset?
- Is it appropriate in this discipline or journal?
- Does it belong with the finding, or does it compare the finding with wider literature and theory?
Comparisons with previous studies, broad implications, and recommendations usually belong in the discussion even when analytical commentary appears in the findings section.
How to Report Mixed-Methods Results
Mixed-methods reporting should show more than two parallel sets of findings. It should demonstrate how the quantitative and qualitative components relate.
Separate reporting
A paper may use:
- Quantitative results
- Qualitative findings
- Integrated findings
This is useful when each component requires substantial explanation.
Integrated reporting
Results may instead be organized by research question, with numerical and qualitative evidence presented together.
Example:
“Survey scores indicated improved confidence after the workshop, with a mean increase of 1.2 points on the five-point scale. Interview findings supported this pattern: participants described feeling more able to select appropriate statistical tests. However, several remained uncertain about interpreting interaction effects.”
The qualitative finding confirms and qualifies the quantitative pattern.
Joint displays
A joint display places findings from different components into one table.
| Research issue | Quantitative result | Qualitative result | Integrated observation |
|---|---|---|---|
| Confidence | Mean score increased | Participants described greater independence | Convergence |
| Software use | Completion time decreased | Some still relied on step-by-step instructions | Partial convergence |
| Advanced interpretation | Little improvement | Participants reported confusion | Convergence on continuing difficulty |
Use terms such as convergence, complementarity, expansion, divergence, or contradiction carefully and define them where needed.
The broader interpretation or meta-inference may appear in the discussion, depending on the chosen structure.
How to Report Systematic-Review Results
A systematic-review results section differs from that of a primary empirical study.
It commonly includes:
- Study selection: Number of records identified, screened, excluded, and included.
- Study characteristics: Designs, settings, participants, interventions, exposures, and outcomes.
- Risk of bias: Judgments for individual studies or evidence domains.
- Individual-study results: Effect estimates or key findings from each study.
- Synthesis results: Meta-analytic estimates or structured narrative synthesis.
- Heterogeneity: Statistical and clinical variation where relevant.
- Subgroup or sensitivity analyses: Clearly labeled as prespecified or exploratory.
- Reporting-bias assessment: When sufficient evidence is available.
- Certainty of evidence: Where the review framework requires it.
Use a PRISMA flow diagram for the selection process and follow the current PRISMA checklist or another guideline appropriate to the review type.
When to Use Text, Tables, Figures, or Appendices
| Format | Best used for | Avoid when |
|---|---|---|
| Text | A few central values or findings | Many repeated numbers make the paragraph unreadable |
| Table | Exact values across groups, variables, or models | The pattern is easier to see visually |
| Figure | Trends, distributions, uncertainty, relationships, or processes | Readers need many exact values |
| Appendix | Detailed material needed for assessment or coursework | The information is essential to understanding the main finding |
| Supplement | Extensive diagnostics, secondary analyses, codebooks, or additional models | It is used to hide an inconvenient primary result |
| Joint display | Mixed-methods integration | Quantitative and qualitative findings have not been meaningfully connected |
Do not reproduce every table value in prose. Highlight the most important pattern and direct readers to the table for detail.
Examples of Statistical Results
The following examples are fictional and are provided only to demonstrate reporting structure.
Descriptive result
“The 186 participants completed an average of 4.8 training modules (SD = 1.6). Completion was lower among part-time students (M = 4.1, SD = 1.5) than full-time students (M = 5.2, SD = 1.4).”
Independent-samples t test
“Students receiving structured feedback obtained higher final scores (M = 78.6, SD = 8.1) than those receiving standard feedback (M = 73.2, SD = 9.0), a mean difference of 5.4 points, 95% CI [2.1, 8.7], t(104) = 3.25, p = .002, d = 0.63.”
Paired-samples t test
“Knowledge scores increased from pre-test (M = 14.3, SD = 3.2) to post-test (M = 17.1, SD = 2.8), mean change = 2.8, 95% CI [2.1, 3.5], t(79) = 7.92, p < .001, dz = 0.89.”
Analysis of variance
“Mean satisfaction differed across the three delivery formats, F(2, 147) = 6.84, p = .001, η² = .085. Adjusted pairwise comparisons showed higher satisfaction for blended delivery than classroom-only delivery; the remaining comparisons were inconclusive.”
Correlation
“Weekly study time was positively associated with examination score, r(198) = .34, 95% CI [.21, .46], p < .001.”
Do not describe a correlation as evidence that one variable caused the other.
Chi-square test
“Course completion differed by enrollment status, χ²(1, N = 284) = 8.16, p = .004, Cramér’s V = .17. Completion was recorded for 82% of full-time students and 68% of part-time students.”
Linear regression
“After adjustment for prior achievement and study time, attendance remained positively associated with examination score, b = 0.42, SE = 0.11, 95% CI [0.20, 0.64], p < .001. The model explained 31% of the variance in scores, adjusted R² = .31.”
Logistic regression
“Students receiving weekly reminders had higher odds of completing the course than students receiving no reminders, OR = 1.74, 95% CI [1.13, 2.68], p = .012.”
Non-significant result
“The estimated difference in retention between the two groups was 1.8 percentage points, 95% CI [−4.5, 8.1], p = .58. The interval included both a modest reduction and a modest increase.”
This wording communicates the estimate and uncertainty without incorrectly claiming that the groups were proven equivalent.
How to Report Null, Negative, and Unexpected Results
Relevant findings should be reported regardless of whether they support the expected conclusion.
Null results
A null result may mean that the analysis did not provide sufficiently clear evidence of a difference or relationship. It does not automatically establish equality or absence.
Prefer:
“The analysis did not detect a clear difference between groups.”
Avoid:
“The hypothesis was proven false.”
Negative effects
The word negative may refer to:
- A result in the opposite direction
- An undesirable outcome
- A negative numerical coefficient
- Failure to support a hypothesis
State which meaning applies.
Unexpected results
Report the result objectively in the results section. Explanations belong in the discussion.
Results:
“Contrary to H2, the higher-frequency notification group completed fewer optional tasks than the lower-frequency group.”
Discussion:
“One possible explanation is that frequent notifications contributed to message fatigue.”
Should a Hypothesis Be Accepted or Rejected?
In many research contexts, it is better to state whether the result supported, did not support, or was inconclusive regarding a hypothesis.
“Accepted” can imply that a hypothesis has been established as true. “Rejected” can oversimplify uncertain evidence.
Preferred wording:
- “The result supported H1.”
- “H2 was not supported.”
- “The estimate was imprecise and did not permit a clear conclusion regarding H3.”
Follow the terminology required by the relevant discipline.
Results Section Writing Style
Tense
Past tense is commonly used for completed procedures and findings:
- “Participants reported…”
- “The analysis identified…”
- “Scores increased…”
Present tense is often used when directing attention to a table or figure:
- “Table 2 shows…”
- “Figure 3 illustrates…”
Conventions vary, so examine recent papers in the target journal.
Voice
Both active and passive voice may be acceptable.
Active:
“We identified three themes.”
Passive:
“Three themes were identified.”
Active voice often makes responsibility and action clearer, but journal and disciplinary conventions take priority.
Precision
Replace vague language with measurable information.
Vague:
“Many participants improved considerably.”
Precise:
“Forty-one of the 58 participants improved by at least five points.”
Objective language
Avoid emotionally evaluative terms such as:
- Fortunately
- Disappointingly
- Obviously
- Remarkably
- Surprisingly
These words communicate the writer’s reaction rather than the finding itself.
A genuinely unexpected result can be described as contrary to a hypothesis without calling it surprising.
Citations in the Results Section
Citations are less common in a separate quantitative results section because the section reports the present study’s findings.
Citations may still be appropriate when:
- A measure, classification, or externally defined threshold must be identified.
- A secondary dataset is being reported.
- The study design combines results and discussion.
- A qualitative approach requires engagement with texts or sources.
- A systematic review is identifying included studies.
- A reproduced or adapted figure requires attribution.
Comparisons with previous research usually belong in the discussion.
Digital Research Tools
Statistical and qualitative software can calculate or organize results, but software output is not a finished results section.
Quantitative tools
Common tools include:
- R
- Python
- SPSS
- Stata
- SAS
- Jamovi
- JASP
- Excel
Use scripts, syntax files, or saved analysis workflows where possible so that results can be reproduced.
Qualitative tools
Common tools include:
- NVivo
- ATLAS.ti
- MAXQDA
- Dedoose
- Taguette
These tools support coding, retrieval, memoing, and visualization. They do not independently establish the credibility of a qualitative interpretation.
Reproducible tables and figures
Where practical, generate tables and figures directly from the verified dataset and analysis code. This reduces transcription errors and makes revisions easier.
Also consider providing:
- Data-availability statements
- Code-availability statements
- Analysis scripts
- Codebooks
- De-identified datasets where ethically and legally appropriate
- Supplementary tables
- Preregistration or protocol links
Artificial Intelligence and the Results Section
AI tools may help with limited tasks such as:
- Checking grammar
- Improving sentence clarity
- Suggesting table labels
- Identifying inconsistent terminology
- Converting verified output into a preliminary reporting format
- Checking whether all research questions appear to have corresponding results
However, AI should not be trusted to:
- Invent missing values
- Calculate statistics without verification
- Decide which inconvenient results to omit
- Generate participant quotations
- Fabricate themes
- Create unverifiable references
- Interpret confidential data through an unauthorized public tool
- Replace methodological or statistical expertise
Every number and claim must be checked against the original data, code, output, or analytical record.
Researchers should also:
- Follow institutional data-security rules.
- Avoid uploading identifiable or confidential data to unapproved services.
- Keep a record of substantial AI use.
- Follow the target journal’s disclosure policy.
- Retain full human responsibility for the submitted work.
Common Mistakes in a Research Results Section
Interpreting instead of reporting
Incorrect for a separate results section:
“The higher scores occurred because students were more motivated.”
The data may show higher scores, but the proposed reason belongs in the discussion unless motivation was directly measured and analyzed.
Copying raw software output
Software tables often contain unnecessary diagnostics, labels, and formatting. Extract the information required for the paper and present it in the journal’s style.
Reporting p-values alone
A p-value does not show the magnitude or practical importance of a result. Include an estimate, effect size, and interval where appropriate.
Hiding non-significant findings
Omitting relevant results because they do not support a preferred conclusion creates a distorted account of the study.
Presenting every analysis
Not every exploratory calculation belongs in the main article. Include analyses that address the aims and label additional analyses transparently.
Repeating the same data
Do not present identical data in a paragraph, table, and figure. Each format should contribute something useful.
Using unclear denominators
A statement such as “40% responded positively” is incomplete when the reader cannot determine whether the denominator was all recruited participants, all respondents, or only complete cases.
Overloading qualitative findings with quotations
Quotations should support analysis. A long series of quotations without a clear analytical claim is not a developed findings section.
Treating frequency as importance
The most frequently mentioned qualitative theme is not automatically the most conceptually important theme.
Changing variable names
Use the same terminology in the methods, results, tables, figures, discussion, and abstract.
Making causal claims from non-causal designs
An association in a cross-sectional or observational study does not by itself demonstrate causation.
Reusable Results Section Template
Opening
“This section reports the findings from [sample or dataset]. Results are organized according to [research questions, hypotheses, outcomes, themes, or analytical stages].”
Sample and data quality
“Of the [number] cases assessed, [number] met the inclusion criteria and [number] were included in the final analysis. [Number or percentage] had missing data for [variable], and [brief consequence or handling].”
Descriptive results
“Participants had a mean [variable] of [value] (SD = [value]). [Category] was the most frequently observed category, accounting for [percentage] of the sample.”
Quantitative primary result
“To address [RQ/H], [analysis] was conducted. [Group or variable] was associated with [outcome], [estimate and direction], 95% CI [lower, upper], [test statistic and degrees of freedom], p = [value], [effect size].”
Qualitative theme
“Theme 1, [theme name], captured [brief analytical definition]. Participants described [pattern and variation]. As one participant explained, ‘[short quotation]’ ([identifier]). A contrasting view was expressed by [case or subgroup].”
Mixed-methods integration
“The quantitative and qualitative findings [converged/diverged/complemented one another] regarding [issue]. Whereas [quantitative result], interview participants [qualitative result].”
Closing transition
“Together, these findings addressed [research questions or outcomes]. Their interpretation and implications are considered in the discussion.”
A closing paragraph is optional. Do not add one merely to repeat every preceding result.
Final Results Section Checklist
Before submission, confirm that:
- Every research question, hypothesis, or primary outcome has a corresponding result.
- Every major result can be traced to a method.
- Sample sizes and denominators are clear.
- Missing data, exclusions, and attrition are reported where relevant.
- Descriptive statistics are appropriate.
- Estimates, uncertainty intervals, and effect sizes are reported when applicable.
- Relevant null and unexpected results are included.
- Exploratory analyses are labeled.
- Qualitative themes are supported by evidence.
- Contradictory or negative cases are not concealed.
- Mixed-methods findings are integrated rather than merely placed side by side.
- Tables and figures are numbered, titled, referenced, and non-duplicative.
- Statistical values match the original output.
- Terminology is consistent.
- Interpretation has been reserved for the appropriate section.
- The applicable reporting guideline has been checked.
- Any substantial AI assistance has been verified and disclosed as required.
Conclusion
A research results section should provide a clear, complete, and verifiable account of what the study found. Its structure should follow the research questions, outcomes, hypotheses, themes, or synthesis plan rather than the order in which software produced the output.
Strong reporting combines selective presentation with transparency. It includes relevant null and unexpected findings, communicates the magnitude and uncertainty of quantitative estimates, supports qualitative claims with evidence, and follows the reporting standards appropriate to the research design.
