How to Write Dissertation Findings Chapter (Thematic Analysis + Quotes Like a Pro)
Learning how to write dissertation findings chapter with thematic analysis? This guide gives you the exact structure, word counts, 30/70 quote rule, cross-case table, and a full thematic analysis findings chapter example — built for UK masters students writing findings right now.
If you are staring at 12 coded transcripts and wondering how to write dissertation findings chapter without just dumping quotes, you are not alone. For most UK masters students, the findings chapter is where marks are won or lost: examiners want to see clear themes, carefully chosen interview quotations with commentary, and a tight link back to your research questions — not a transcript pasted under headings.
This guide walks you through a proven thematic analysis findings chapter example structure used in distinction-level qualitative dissertations. You will learn exactly what goes where, how much of your chapter should be quotes versus your own analysis (the 30/70 rule), how to do qualitative findings cross-case comparison that impresses markers, and whether results vs discussion should be separate at your university. If you have already built themes but struggle to write them up, first read our guide on how to turn codes into themes, then come back here to present them.
- What Goes in a Findings Chapter (and What Does Not)
- How Should You Structure a Thematic Findings Chapter?
- How Much Quote vs Analysis? The 30/70 Rule + Example
- How Do You Present Cross-Case Comparison?
- Should Results and Discussion Be Separate?
- How Do You Link Findings to Objectives Without Repeating Literature?
- Findings Chapter Checklist
- FAQs
What Goes in a Dissertation Findings Chapter?
The job of your findings chapter is simple to state and hard to do: answer “What did you find?” — not yet “What does it mean?” That second question belongs to your discussion chapter. In practice, qualitative findings always contain some interpretation (choosing themes is interpretation), but the emphasis stays descriptive and evidential.
Think of the division of labour taught across UK universities from Southampton to Reading to Westminster:
- Findings chapter = present. Report themes, patterns, prevalence across participants, and illustrative verbatim quotes. You can add brief analytical commentary on what a quote shows within your dataset — for example, noting it was the most prevalent theme or appeared mainly among mature students — but do not bring in theory or prior studies yet.
- Discussion chapter = interpret. Connect those themes to literature, explain why you got these results, answer “So what?”, acknowledge limitations, and state your contribution. That is where you will need the moves covered in our dissertation discussion chapter guide.
What examiners expect to see in findings specifically for thematic analysis:
- Chapter opening (150–250 words): restate aim and research questions in one paragraph, describe sample briefly (e.g., 12 semi-structured interviews, purposive sample), and give a roadmap: “This chapter presents four themes…”. Include a theme overview table here.
- Themes as sections (80% of chapter): 3–5 main themes, each with a clear analytic heading, opening claim paragraph, 2–4 quotes from different participants, and commentary after every quote.
- Variation and negative cases: explicitly report where participants diverged. Markers trust findings that show complexity over findings where everyone magically agreed.
- Closing summary (150–250 words): recap key findings question-by-question with no new data and a handover sentence to the discussion.
What to leave out: raw data dumps, full transcripts (put those in appendices), counts dressed up as statistics (“87.5% proved…”), new literature, and evaluative language like “this proves” or “this contradicts Smith (2022)”. Save those verbs for discussion. Use past tense throughout — you have already done the analysis.
How Should You Structure a Thematic Findings Chapter?
For most masters dissertations using Braun and Clarke thematic analysis, organise theme-by-theme, not participant-by-participant. Participant-by-participant (P1 said…, P2 said…) reads like 12 mini case studies and hides patterns. Theme-by-theme shows you synthesised across the dataset, which is exactly what thematic analysis claims to do.
A distinction-level structure looks like this:
4.1 Introduction + participant table reference + theme map
4.2 Theme 1: [Analytic name, not topic label] (answers RQ1)
4.2.1 Sub-theme 1a + 4.2.2 Sub-theme 1b
4.3 Theme 2: [Analytic name] (answers RQ1/RQ2)
4.4 Theme 3: [Analytic name] (answers RQ2) + cross-case comparison
4.5 Theme 4 (if needed, often divergent/minority view)
4.6 Chapter summary linked to research questions
Name each theme as a claim, not a topic. “Lack of feedback” is a topic label any code list could give you. “Feedback was valued only when it told students what to do next” is an analytic claim — it tells the reader what you argue the data shows. Examiners decide in that first paragraph whether you did thematic analysis or just sorted quotes under headings.
Use this word-count planner. It assumes a 12,000–15,000 word dissertation where findings carry 20–30% of marks:
| Section | Words (12k diss.) | Words (15k diss.) | Purpose |
|---|---|---|---|
| 4.1 Introduction + theme map | 200–300 | 250–350 | Restate RQs, sample, roadmap, overview table |
| Theme 1 (core, RQ1) | 600–800 | 800–1,000 | Claim + 3 quotes + commentary + prevalence |
| Theme 2 (core, RQ1/RQ2) | 600–800 | 800–1,000 | Claim + 3 quotes + commentary + sub-themes |
| Theme 3 (core, RQ2 + cross-case) | 600–800 | 700–900 | Claim + matrix + variation across groups |
| Theme 4 (nuance / negative case) | 300–500 | 400–600 | Minority view, keeps credibility high |
| 4.6 Summary | 200–250 | 200–300 | Recap by RQ, no new quotes, link to discussion |
| Total | 2,500–3,450 | 3,150–4,150 | 3–4 themes ideal; 5 max |
Two alternatives: if your handbook asks for findings by research question, keep themes but nest them under RQ headings (e.g., “4.2 Findings addressing RQ1” then Theme 1 and 2 inside). If you did multiple case studies, present a one-paragraph within-case pen portrait per case, then move to cross-case themes. Do not stay in within-case description — marks come from synthesis.
Start every theme section the same way for consistency: claim → prevalence → evidence → link. Example opening: “A prominent finding was that students valued feedback most when it gave actionable next steps (RQ1). Nine of twelve participants emphasised this, particularly those returning to study after employment.” Then give quotes. End each theme with one transition sentence to the next theme so the marker never gets lost.
How Much Quote vs Analysis? The 30/70 Rule + Before/After Example
The most common feedback on failed findings chapters is over-quoting or under-interpreting. Either the chapter reads like a transcript with headings (quote after quote, no voice), or it reads like an essay with no evidence (claims with one tiny snippet). The fix used by supervisors across the UK is the 30/70 rule: no more than ~30% of your findings word count should be verbatim quotes; ~70% should be your words — introducing quotes, interpreting them, and connecting them.
In practice that means 2–4 quotes per main theme from at least 2–3 different participants, each handled with the quote sandwich:
- Introduce (1–2 sentences): tell the reader what to look for. Not “As Participant 7 said…”, but “The contingent nature of support was explicit in how junior nurses described asking for help…”
- Quote (1–3 sentences, <40 words ideally): verbatim, trimmed with […] for omissions, with consistent attribution — e.g., (Amara, P4, final-year Business). For longer passages over 40 words, use an indented block quote sparingly.
- Interpret (2–3 sentences): explain significance. What does word choice reveal? How does it build or complicate the theme? Do not just paraphrase the quote — that is description, not analysis.
For formatting and ethics detail — block vs embedded quotes, anonymisation, cleaning fillers, line referencing — see our full guide to presenting interview quotes in your dissertation.
Theme 2: Support. Participants talked about support. As P3 said, “My supervisor was supportive and helped me a lot with my work and gave me support when I needed it and it was good.” Another participant noted, “Support from peers was good, yeah, like we supported each other.” P7 also said, “I got support from the library team which was helpful.” This shows support was important.
Why it fails: topic label not claim, three quotes stacked with no interpretation, no prevalence, no variation, “shows” with no explanation of what or why.
Theme 2: Support was experienced as contingent permission, not guaranteed provision (RQ1). Seven of eleven participants described hesitating before seeking help, waiting until problems became urgent rather than asking proactively. The contingent nature of support was explicit in how several participants described post-restructure supervision:
“I didn’t want to be the person who didn’t know how uni worked, so I just tried to figure it out myself first.” (Maria, P4, first-generation, Business)
Maria’s use of “the person” is revealing: she read help-seeking as an identity risk — exposure as an outsider — rather than a routine academic act. The result was avoidance, a pattern echoed by five others who described office hours as governed by unwritten “protocol” they had never been taught.
“I thought I’d get billed for showing up too much, so I saved [tutorials] for emergencies.” (Jordan, P9, BSc Economics)
While Jordan similarly delayed contact, his reason differed — a practical misunderstanding about provision rather than social discomfort. The divergence matters: the same observable behaviour (late help-seeking) had two distinct drivers, suggesting targeted induction could address each differently. This nuance is developed further in the discussion.
Why it works: analytic claim, prevalence (7/11), two vivid quotes from different participants, word-level commentary (“the person”, “protocol”), named variation, stops before literature — saving interpretation for discussion.
Three selection rules to apply to every quote: (1) it says something a paraphrase could not — a distinctive phrase, metaphor, or reasoning step; (2) it is representative of a pattern you have already evidenced, not a one-off you find interesting; (3) across the chapter you represent breadth — if P3 appears fifteen times and P11 zero times, markers will question whether themes are truly shared. Aim for every participant quoted at least once overall, but select per theme by fit, not by rota.
Findings chapter help in 72h
Stuck with coded data and a blank Chapter 4? Send your transcripts or codebook — we structure your themes, select quotes, and draft your findings around your research questions. Plagiarism-free, 0% AI.
How Do You Present Qualitative Findings Cross-Case Comparison?
Markers award analysis marks for qualitative findings cross-case comparison — showing how a theme played out differently across groups, sites, or participant types — because it proves you looked for patterns and variation rather than smoothing data into one story. Even if you did not do a formal multiple-case design, you can compare across meaningful subgroups (e.g., placement vs non-placement students, early-career vs experienced nurses, UK vs international students).
Do it in two moves: a theme summary table at the chapter opening, then a quote matrix inside the theme where comparison matters most. Do not matrix every theme — one well-placed comparison table signals rigour without bloating word count.
Move 1 — Theme overview table (Chapter 4.1). Gives the examiner the whole analytic structure at a glance:
| Theme (answers RQ) | Sub-themes | Prevalence |
|---|---|---|
| 1. Actionable feedback drives use (RQ1) | Specific next steps; feed-forward language | 9/12; strongest among returners to study |
| 2. Support as contingent permission (RQ1) | Identity risk; practical confusion | 7/11; all first-gen participants |
| 3. Timeliness beats detail under pressure (RQ2) | 48-hour window; deadline proximity | 8/12; strongest in placement group |
Move 2 — Cross-case quote matrix (inside Theme 3). Suppose your RQ2 asks how placement experience shapes perceptions of feedback timeliness. Present the pattern, then evidence it compactly:
| Group | Pattern observed | Illustrative quote |
|---|---|---|
| Placement students (n=6) | Prioritised speed over depth; wanted 48-hour turnaround to act before next shift | “If it comes after my next set of shifts, it’s dead — I needed two lines in two days, not two pages in two weeks.” (Daniel, P6) |
| Non-placement students (n=6) | Preferred detailed written feedback even if slower; used it for redrafting | “I’d rather wait a week and get paragraph-by-paragraph notes I can work through.” (Priya, P2) |
| Divergent case | One placement student preferred depth, linked to dissertation module not placement tasks | “On placement, quick is king — but for my dissertation I wanted the long version.” (Sam, P10) |
Follow the matrix with 3–4 sentences of within-data commentary: what the contrast shows, how strong it is, and what complication the divergent case adds. Example: “The contrast suggests timeliness is task-contingent rather than a fixed preference, with placement rotas compressing the window in which feedback remains actionable. Sam’s dual preference cautions against treating group membership as deterministic.” That is enough — the why (link to literature on formative feedback timing) waits for discussion.
Practical tips: never present counts as statistics (“75% proves…”). Use cautious prevalence language — “most”, “seven of twelve”, “particularly among…”. Always attribute with pseudonyms, never real names, and strip identifying details. Reference appendices for full codebook or extended extracts (e.g., “see Appendix D, lines 142–156”) so the chapter stays selective while the audit trail stays complete.
Should Results and Discussion Be Separate or Combined?
For results vs discussion separate vs combined, the UK norm is: separate by default, combined only if your handbook or supervisor says so. Southampton, Reading, Westminster and Manchester handbooks all describe the same logic — sciences and quantitative projects almost always separate (objective results, then interpretation), while qualitative projects have more freedom to integrate because theme presentation already involves analysis.
| Option | When to use it | Risk if you choose wrong |
|---|---|---|
| Separate Chapters 4 + 5 | Default for most UK masters; required for quantitative / mixed-methods; safest for thematic analysis | None — markers can see evidence before argument |
| Single Results & Discussion | Handbook explicitly allows it; short dissertations (8,000–10,000 words); essay-style humanities structure | Looks like you cannot distinguish evidence from claim |
| Integrated theme-by-theme (present + discuss each theme) | Supervisor approves; case-study / narrative designs honouring holistic stories | Repeats literature in every theme; word-count blowout |
If you combine, keep the two jobs visually distinct inside each theme: present data first under “Findings”, then interpret under “Discussion” sub-heading, or use signpost sentences (“Having presented the pattern, I now consider what it means in light of…”). If you separate — recommended — end findings with a handover paragraph that lists the 3–4 findings the discussion will interpret, in the same order the discussion will follow. Parallel structure between Chapters 4 and 5 is one of the fastest ways to look coherent.
Still unsure? Ask your supervisor one precise question: “Do you want Chapter 4 to be findings-only with literature reserved for Chapter 5, or are you happy with integrated analysis-discussion per theme?” Their answer determines your headings. Do not guess — handbooks vary even within the same university.
How Do You Link Findings to Objectives Without Repeating the Literature Review?
Every theme must answer a research question, but without dragging whole paragraphs of theory into findings. The technique is signpost, don’t synthesise: name the RQ or objective each theme addresses in the theme heading or first sentence, note prevalence and relevance, and stop. Save authors, debates, and “consistent with / extends / contradicts” moves for discussion.
Use these sentence stems (copy-adapt them):
- Theme opening link: “This theme addresses RQ1 (How do final-year students experience academic support?) by showing…”
- Prevalence + relevance: “This was the most prevalent theme in the dataset, appearing in nine of twelve transcripts, and directly bears on Objective 2…”
- Within-data connection: “Unlike Theme 1, which concerned written feedback, this theme concerns face-to-face encounters, suggesting…”
- Handover to discussion: “The implications of this divergence for formative feedback theory are considered in Section 5.3.”
What “without repeating lit review” means in practice: in findings you may reference your conceptual framework as an organiser (“consistent with the study’s framing of feedback as dialogue…”) in one clause, but you do not cite three authors to agree or disagree. If you catch yourself writing “As argued by Nicol and Macfarlane-Dick (2006)…”, cut it and paste it into your discussion plan. Findings cites participants; discussion cites scholars. That single discipline keeps both chapters clean and stops self-plagiarism of your own literature review.
Finally, keep a tight thread to methodology: if you said you used reflexive thematic analysis of semi-structured interviews, your headings must be themes (not interview questions), your quotes must be attributed consistently, and your appendix references must let a marker trace any theme back to codes. One inconsistency — e.g., findings organised by interview question when methods promised themes — costs credibility fast.
Findings chapter help in 72h
Deadline close? Get a findings chapter drafted from your data in as little as 72 hours — themes structured, quotes sandwiched, cross-case table included. Message us now for a free outline.
Dissertation Findings Chapter Checklist (Before You Submit)
Run this checklist honestly. Every “no” is a revision task:
- Structure: 3–5 themes as analytic claims (not topic labels), each mapped to an RQ; introduction roadmap + closing summary with no new data.
- Evidence: 2–4 quotes per theme from ≥2–3 participants; every quote introduced + interpreted (quote sandwich); quotes ≤30% of words.
- Breadth: all participants represented across chapter; no single voice dominates; pseudonyms consistent; identifying details removed.
- Variation: at least one divergent / negative case reported; prevalence stated cautiously (“7 of 12”, “most”, “particularly among…”).
- Tables: theme overview table + one cross-case matrix where comparison earns marks; tables supplement, never replace, narrative.
- Separation: no literature debate in findings; past tense; no “proves / contradicts X” claims — those live in discussion (see how to write the discussion chapter).
- Traceability: appendix + line references for codebook / extended extracts; headings match methods promise (themes, not interview Qs).
- Polish: fillers trimmed with […] where meaning preserved; ellipses/brackets used correctly; transitions between themes; proofread attributions.
If you tick all eight, your findings chapter is ready to carry the discussion. If two or more are “no”, fix findings before drafting discussion — a shaky Chapter 4 guarantees a shaky Chapter 5.
Frequently Asked Questions
For a 12,000–15,000 word UK masters dissertation, aim for 2,500–4,000 words (20–30% of total). For a 10,000-word project, 2,000–3,000 words is typical. Give core themes roughly equal space unless one directly answers your main question — then give it more and say why.
Use 2–4 quotes per main theme from at least 2–3 different participants. Follow the quote sandwich every time: introduce what to look for, present the verbatim quote with pseudonym, then add 2–3 sentences of commentary. If you need formatting help, see our guide to presenting interview quotes.
Yes by default — keep them separate unless your handbook says otherwise. Findings presents what you found; discussion interprets what it means against literature. Combined chapters suit some qualitative designs, but always confirm with your supervisor first.
Report minority views in a sub-section or cross-case table. Name the majority pattern and the deviation (e.g., 9 of 12 said X, 3 said Y) and add brief within-data commentary. Showing variation increases credibility — hiding it does the opposite.
Keep literature out except for 1–2 orienting references if integrated analysis is approved. Link themes to research questions and objectives instead. Save full comparison with theory for your discussion — our discussion chapter guide shows how to do that move well.
Final Thoughts
Knowing how to write dissertation findings chapter content that earns marks comes down to three disciplines: organise by analytic themes tied to research questions, keep quotes to ~30% with a sandwich of introduction and interpretation around every extract, and show comparison and contradiction through a single well-placed cross-case table. Do that, keep literature in the discussion, and your Chapter 4 will read as rigorous rather than descriptive.
If you are writing findings now and the deadline is close, do not lose days rearranging quotes. Build your theme map, apply the checklist above, and draft one theme at a time using the before/after model. And if you want an expert second pair of eyes — or a full draft from your coded data in 72 hours — message Academia Helper on WhatsApp today.
