The Results Section: How To Present Findings Without Letting Interpretation Slip In

Cluster Post 4 | Module 1: Understanding the Structure of Research Papers and Theses

Table of Contents

From Concept to Submission Series | 2026

← Back to Pillar Post: The Complete Guide to Research Paper Structure

The Results Section: How to Present Findings Without Letting Interpretation Slip In

Your pillar post explained the core rule: Results presents facts, Discussion interprets them. This post goes deeper — how to sequence your findings logically, how to report statistics correctly, how qualitative theme presentation works, how to use tables and figures well, and the specific language patterns that let interpretation creep in without you noticing.

Why This Section Is Harder Than It Looks

The Results section looks like the easiest part of a paper to write. You already have the data. You just report it. What is difficult about that?

The difficulty is discipline. You have spent months thinking about what your findings mean. Every instinct you have as a researcher is pushing you toward interpretation — toward explaining, contextualising, connecting. The Results section requires you to suppress that instinct completely and report only what you observed.

This is not arbitrary. It exists because science and scholarship depend on separating observation from interpretation. When you present your data cleanly — without telling readers what to think about it — you give them the ability to evaluate your findings independently, before your interpretation is applied. The moment you mix results with interpretation, you are doing some of their thinking for them. Experienced reviewers notice immediately, and it damages the credibility of everything else in your paper.

How to Sequence Your Results

One thing the pillar post does not cover is the order in which findings should appear. There is no single correct sequence, but there is a clear logic that works across most research types.

Start with descriptive statistics or background findings that give context before the main results. Then present your primary findings — the ones that directly answer your research questions or test your hypotheses — in the order your research questions appeared in the Introduction. Finally, present secondary or exploratory findings. This sequence means your most important results are never buried.

For a study with multiple research questions, the cleanest approach is to use your research questions as subheadings within the Results section. This makes it immediately obvious which findings answer which question, and it sets up your Discussion perfectly since the Discussion should follow the same order.

Example structure: Results → 3.1 Sample Characteristics → 3.2 Research Question 1: Does peer mentoring frequency predict retention? → 3.3 Research Question 2: What peer support mechanisms do students describe? → 3.4 Additional Analyses: Subgroup effects by gender and discipline.

Reporting Quantitative Results Correctly

Statistical results need four elements every time: the test statistic, degrees of freedom, p-value, and effect size. Most researchers report the first three and forget the fourth. Effect size is not optional — it tells readers whether a statistically significant finding is also practically meaningful.

Incomplete: “There was a significant difference in retention rates (p = .003).” Complete: “Intervention students showed significantly higher retention rates (82%) than control students (71%), χ²(1, N = 450) = 8.92, p = .003, φ = .14.”

The effect size (φ = .14 in the example above) is small by conventional standards — which is useful information. A finding can be statistically significant and practically trivial. Reporting effect size gives readers what they need to judge which it is.

Statistical test	What to report
t-test	t(df) = value, p = .xxx, Cohen’s d
ANOVA	F(df1, df2) = value, p = .xxx, η² or partial η²
Chi-square	χ²(df, N = total) = value, p = .xxx, φ or Cramer’s V
Pearson correlation	r(df) = value, p = .xxx
Multiple regression	β = value, SE = value, t = value, p = .xxx, R² for the model

Two other things to address in your quantitative results: missing data and outliers. How many data points were missing, and how did you handle them? Did you identify any outliers, and did you include or exclude them? These are not minor housekeeping details — they can affect your findings substantially, and reviewers will ask about them if you do not address them yourself.

Presenting Qualitative Results: Themes, Quotes, and Negative Cases

Qualitative results work differently from quantitative ones. You are not reporting numbers — you are presenting themes or patterns that emerged from your analysis, supported by evidence from your data.

The standard structure for each theme is: name it, define it in one sentence, explain how it manifested in the data, and then support that explanation with one or two direct quotes from participants. The quote is evidence, not decoration. It should illustrate the theme specifically, not just confirm that something was said.

Theme: Peer mentors as institutional navigators. Definition: Students described mentors as guides to processes and resources they would not have found independently. Evidence: Participants consistently described mentors helping them navigate administrative systems rather than just academic content. One student explained: “My mentor showed me how to apply for the fee waiver. I didn’t even know it existed. Without her I would have just dropped out” (P12, Arts). Another described a similar experience with hostel allocation: “He knew who to talk to, which office, what to say. I was completely lost” (P7, Commerce).

Notice what the example does: it defines the theme, describes the pattern across participants (“consistently described”), and uses two quotes from different participants to show breadth rather than one isolated instance. A single compelling quote is not evidence of a theme — it is evidence that one person said something interesting. Themes require pattern.

Negative cases: the part most researchers ignore

Negative cases are data points that do not fit your themes — participants whose experiences contradict the pattern you are describing. Most qualitative researchers quietly ignore them. Strong researchers address them directly.

Acknowledging negative cases does not weaken your analysis. It strengthens it, because it shows you have been rigorous rather than selective. A theme that holds for twenty-two out of thirty participants and is meaningfully different from the eight it does not hold for is a richer, more credible finding than one presented as universal without qualification.

“This theme was evident across most participants but was absent in the accounts of six students who had transferred from private colleges. These students already possessed the institutional knowledge that mentors provided to others, suggesting the navigational function of peer mentoring may be specifically valuable for first-generation or government-school-educated students.”

Using Tables and Figures Without Letting Them Do Your Writing

Tables and figures are tools for communicating data patterns that would take many words to describe in prose. Used well, they make your Results section cleaner and more readable. Used poorly, they create confusion and generate reviewer queries.

Every table and figure must be referenced in your text before it appears. Not after, not in a caption — in the body of the text, before the reader reaches it. This reference should also tell the reader what to notice, not just that the table exists.

Weak: “The results are shown in Table 2.” Strong: “Table 2 shows retention rates across the three colleges. Intervention colleges (A and B) show consistently higher retention than the control college (C), with College B — which used group rather than individual mentoring — showing the highest rate of all (87%).”

The second version is doing real work. It tells the reader what the table shows and which finding within it is most important. This is not interpretation — it is navigation. You are directing the reader’s attention, not telling them what to conclude.

Every table needs a clear descriptive title above it. Every figure needs a caption below it. Columns and axes must be labelled with units. If significance levels are indicated with asterisks, a footnote must explain what each asterisk means. These are not stylistic preferences — they are conditions for the table or figure to be readable.

The Language Patterns That Let Interpretation Slip In

Interpretation rarely enters the Results section as a whole paragraph. It enters word by word, through language that carries evaluative or causal meaning. Learning to spot these patterns in your own writing is one of the most useful editing skills you can develop.

Watch for these specific words and phrases in your Results section and remove or relocate them every time you find them:

Evaluative adverbs: “surprisingly,” “worryingly,” “encouragingly,” “importantly.” These are your reactions to the data, not the data itself.
Causal language: “because,” “therefore,” “as a result,” “which caused.” Causation is an interpretation. Results can show association; Discussion explains mechanism.
Suggestive phrases: “this suggests,” “this indicates,” “this demonstrates,” “this shows that.” All of these belong in Discussion.
Evaluative adjectives: “high retention,” “strong correlation,” “significant improvement.” Statistical significance is a result. Whether it is “strong” or “high” is a judgment.

Contains interpretation: “Surprisingly, the intervention group showed a high retention rate of 82%, which suggests peer mentoring is effective.” Cleaned: “The intervention group showed a retention rate of 82%.”

The cleaned version is shorter and feels almost too plain. That is exactly right. Plain is the correct register for Results. Save the evaluative language for Discussion, where it belongs and where it will have real effect because it is clearly your interpretation rather than a contamination of your data presentation.

🔱 For Law Students

Doctrinal legal research does not have a Results section in the IMRAD sense — but it does have a findings layer, and the same principle applies: present what you found before you argue what it means.

Presenting case analysis findings

When you have analysed a body of case law, the findings are the patterns you have identified: which principles the courts have established, how interpretation has evolved across time, where judgments are consistent and where they conflict. These findings should be presented before you move to critical analysis — before you argue whether the doctrine is coherent, adequate, or in need of reform.

The practical equivalent of the Results section in doctrinal research is a section that maps the legal landscape as it currently stands. Use IRAC or CREAC as your micro-structure for each case: Issue, Rule (or Conclusion, Rule), Application, Conclusion. Present this analysis systematically before you shift into evaluative mode.

Findings layer (equivalent of Results): “Analysis of fifteen Supreme Court judgments from 2017 to 2025 reveals three identifiable phases in the Court’s interpretation of Article 21 privacy rights. In the first phase (2017–2019), the Court applied Puttaswamy’s proportionality test broadly to all state surveillance. In the second phase (2020–2022), two judgments narrowed the test’s application to active surveillance, exempting passive data collection. In the third phase (2023–2025), the Court has not directly resolved this inconsistency, leaving lower courts without clear guidance.” Critical analysis layer (equivalent of Discussion): “This doctrinal inconsistency is problematic because passive data collection by AI systems constitutes the most pervasive form of contemporary surveillance…”

Keeping these layers distinct — findings before analysis — makes your argument cleaner and your methodology more credible. It shows that your critical argument is grounded in a systematic reading of the sources, not selected to support a conclusion you arrived at in advance.

References

Creswell, J. W., & Creswell, J. D. (2022). Research Design: Qualitative, Quantitative, and Mixed Methods Approaches (6th ed.). Sage Publications.
Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101.
American Psychological Association. (2020). Publication Manual of the APA (7th ed.).
Field, A. (2024). Discovering Statistics Using IBM SPSS Statistics (6th ed.). Sage.
Tinto, V. (2012). Leaving College: Rethinking the Causes and Cures of Student Attrition. University of Chicago Press.
K.S. Puttaswamy v. Union of India, (2017) 10 SCC 1. Supreme Court of India.

Next: Cluster Post 5 — The Discussion Section: Turning Findings Into Knowledge

← Back to Pillar Post: The Complete Guide to Research Paper Structure