Results chapter by the numbers
- 15 to 25% — typical proportion of dissertation word count for results chapter.
- 2,000 to 4,500 words — master’s results chapter; 10,000 to 25,000 for PhD.
- 5 to 15 tables and figures typical for quantitative results; 3 to 8 themes for qualitative.
- 2 to 4 verbatim quotes per theme in qualitative work — enough to ground claims without overwhelming.
- 43% of dissertations returned for revision cite “weak results presentation” — usually mixing results and discussion (UKCGE Examiner Reports, 2024).
The critical distinction: results vs discussion
| Results chapter | Discussion chapter |
|---|---|
| What you found | What it means |
| Numerical / textual data | Interpretation, theoretical engagement |
| Minimal interpretation | Substantial interpretation |
| Few literature citations | Heavy literature engagement |
| Reports significance levels, themes, patterns | Explains why those patterns occurred |
The most common results-chapter error is starting to interpret findings within the results chapter — saying “this finding is consistent with Smith (2020) and suggests that…” in the results chapter instead of the discussion chapter. Discipline yourself: results = data, discussion = meaning.
Quantitative results structure
| Section | Content |
|---|---|
| 1. Sample characteristics | Demographics table, response rate, completion rate |
| 2. Descriptive statistics | Means, SDs, distributions for all variables |
| 3. Validity and reliability | Cronbach’s alpha for scales, factor analysis if applicable |
| 4. Assumption checks | Normality, homogeneity of variance, multicollinearity |
| 5. Inferential tests | Organised by research question or hypothesis |
| 6. Effect sizes | Cohen’s d, eta-squared, R² — not just p-values |
| 7. Summary table | Hypothesis status (supported / not supported) at chapter end |
Need help with your stats analysis + results chapter?
SPSS, STATA, R, Excel — full analysis with annotated output, ready-to-drop tables and viva-defendable interpretation.
Quantitative results — worked example paragraph
A multiple regression was conducted with purchase intention as the dependent variable and sustainability claim exposure as the independent variable, controlling for age, gender and prior brand engagement. The model was statistically significant (F(4, 318) = 12.84, p < .001), accounting for 14% of variance in purchase intention (R² = .14). Sustainability claim exposure significantly predicted purchase intention (β = .27, t = 4.91, p < .001, 95% CI [0.16, 0.38]). Hypothesis 1 was supported.
Hypothesis 2: Perceived authenticity will mediate the relationship between sustainability claims and purchase intention.
The mediation model was tested using the PROCESS macro v4.2 (Hayes, 2022) with 10,000 bootstrap resamples. The indirect effect of sustainability claims on purchase intention via perceived authenticity was significant (b = 0.42, BootSE = 0.06, 95% BootCI [0.31, 0.54]). The direct effect of sustainability claims on purchase intention became non-significant when authenticity was included (β = .04, p = .47), indicating full mediation. Hypothesis 2 was supported.”
Notice: numbers reported precisely; effect sizes given; confidence intervals included; hypothesis-by-hypothesis structure; minimal interpretation (the “why” goes in the discussion chapter).
Qualitative results structure
| Section | Content |
|---|---|
| 1. Sample overview | Demographic table with pseudonyms; interview duration |
| 2. Analytical approach reminder | Brief recap of analysis method (e.g. Braun & Clarke 6-phase) |
| 3. Theme overview | List of 3–8 themes with sub-themes; thematic map figure |
| 4. Theme-by-theme presentation | Definition, evidence, 2–4 participant quotes per theme |
| 5. Negative cases | Counter-evidence within themes; acknowledged not hidden |
| 6. Summary | How themes answer the research question |
Qualitative results — worked example
Across the 12 interviews, participants consistently described authenticity as observable congruence between the influencer’s everyday content and the sustainability claims they made. Authenticity was not assessed from the sustainability content itself; participants looked for confirming signals in the influencer’s wider feed.
P3 (female, 22, Year 3 undergraduate) articulated this directly: “I don’t trust someone who suddenly starts posting sustainability stuff — they’re obviously paid. What I trust is when someone whose feed has always been about minimalism or thrifting then partners with a sustainable brand. That’s when the message lands for me.”
P7 (male, 24, recent graduate) extended this: “It’s almost easier to detect inauthenticity than authenticity. The minute you see someone whose style screams fast fashion posting about Patagonia, it’s over. Authenticity is everything they posted before the sustainability claim, more than the claim itself.”
This theme appeared in 11 of 12 interviews; the single exception was P9, who explicitly stated that they did not check the influencer’s prior content but assessed authenticity from the partnership content alone.”
Notice: theme is named clearly; participant quotes are verbatim with pseudonym + brief demographic; counter-cases are acknowledged (P9); the analytical claim is supported by quote evidence; interpretation is restrained (“participants described”, not “this suggests”).
Conventions for participant quotes
- Verbatim with minimal editing. Preserve speech features that matter; you may remove false starts and “um/er” if your method permits.
- Italics or block-quote formatting distinguishes quotes from your analytical prose.
- Pseudonym + brief demographic identifier per quote — “P3 (female, 22)” or “Maria (nurse, 8 yrs experience)”.
- Square brackets for clarifications — “I told [my supervisor] that I felt…”
- Ellipses for omitted text — “I really thought… well, it was complicated.”
- Don’t over-quote. 2–4 quotes per theme is sufficient; more dilutes analytical depth.
- Each quote should earn its place — it must support a specific analytical claim, not just illustrate the theme generically.
Mixed-methods results structures
Two common organisations:
- Convergent parallel — quant first, qual second, then integration. Each method gets its own subsection. Final section integrates findings. Common when both methods carry equal weight.
- Interleaved by research question. Each research question gets a subsection containing both quant and qual evidence. Common when the qual was designed to explain the quant (explanatory sequential design).
Whichever you choose, signpost clearly so examiners know which method produced which finding. A figure showing your mixed-methods design at the start of the chapter helps.
Tables and figures — when and how
| Type | Best for | Numbering |
|---|---|---|
| Table | Precise numbers; comparison across multiple variables | Table 4.1, 4.2 (Chapter.Position) |
| Bar / column chart | Comparing groups on one variable | Figure 4.1, 4.2 |
| Line chart | Trends over time | Figure |
| Scatterplot | Showing relationships between continuous variables | Figure |
| Thematic map | Qualitative theme hierarchy | Figure |
| Forest plot | Systematic-review effect sizes | Figure |
Every table and figure needs: (a) a number, (b) a caption above (tables) or below (figures), (c) source/note line if data is borrowed, (d) reference in body text (“Table 4.1 shows…”). Don’t include a table or figure without referencing it in prose.
9 results-chapter mistakes that lose marks
- Mixing results and discussion. The single biggest error. Results = what you found; discussion = what it means.
- Reporting only p-values. Effect sizes (Cohen’s d, eta-squared, R²) matter more than significance — required at most institutions.
- Skipping assumption checks. ANOVA without normality and homogeneity checks; regression without multicollinearity check. Examiners will ask.
- No descriptive statistics. Inferential tests come after descriptives; jumping straight to t-tests skips the foundation.
- Quote dumps. Long verbatim passages without analytical framing. Each quote should support a specific analytical claim.
- Cherry-picked themes. Presenting only themes that support your hypothesis; ignoring counter-cases is misconduct in qualitative work.
- No participant context. Quotes without demographic identifiers leave examiners unable to assess sample diversity.
- Tables not referenced in text. Every table must be introduced (“Table 4.1 presents…”) and discussed in prose.
- Inconsistent precision. Reporting some p-values to 2 decimal places, others to 3. Be consistent throughout.
Results chapter stuck? Get expert help
PhD-qualified writers in your subject. SPSS, STATA, R analysis with annotated tables and viva-defendable interpretation.
Authoritative references
- Field, A. (2024) Discovering Statistics Using IBM SPSS Statistics. 6th edn. London: Sage.
- Hayes, A. F. (2022) Introduction to Mediation, Moderation, and Conditional Process Analysis. 3rd edn. New York: Guilford.
- Braun, V. and Clarke, V. (2022) Thematic Analysis: A Practical Guide. London: Sage.
- Creswell, J. W. and Plano Clark, V. L. (2018) Designing and Conducting Mixed Methods Research. 3rd edn. Thousand Oaks, CA: Sage.
- American Psychological Association (2020) Publication Manual. 7th edn. Washington, DC: APA.
- Tabachnick, B. G. and Fidell, L. S. (2019) Using Multivariate Statistics. 7th edn. Boston: Pearson.
- UK Council for Graduate Education (2024) UK PhD Examiner Reports 2023–2024. Lichfield: UKCGE.
- Charmaz, K. (2014) Constructing Grounded Theory. 2nd edn. London: Sage.
Get your results presented at distinction-grade
PhD-qualified specialists handle stats analysis and presentation. Tables ready to drop into your dissertation; viva-defendable interpretation.
