Dissertation methodology examples + how to write your methodology chapter guide
Info: 4986 words (20 pages) Study Guides
Published: 15 Oct 2026
The 60-second version
- Your methodology chapter answers two questions for every decision: what did you do, and why was that the right choice for your research questions? Markers grade the second part far more heavily than the first.
- It follows a chain: philosophy → approach → design → sampling → data collection → analysis → ethics → limitations. Each link should visibly follow from the one before.
- Write it so another researcher could repeat your study without emailing you.
- Budget roughly 15% of your total word count (see table below).
- Scroll to the annotated examples if you just want to see what “good” looks like, then come back for the structure.
UK Essays writing support
Need your dissertation methodology chapter written?
An experienced specialist in your subject will research and write your complete model methodology chapter from scratch, tailored to your research question and study design. Your chapter will explain and justify your research approach, data collection and analysis methods.
Send us your research question and, if you wish, any university guidance or marking criteria.
All our writers hold at least a 2:1 degree from a reputable university. Most hold a master’s degree or PhD, and/or have teaching experience.
What is a dissertation methodology?
The methodology is the chapter (usually Chapter 3) that explains how you conducted your research and why you conducted it that way. It sits between your literature review and your results.
It’s where your marker decides whether to trust everything that follows. The University of Southampton’s dissertation guide describes it as outlining “the research process undertaken, from selecting the method to articulating the tool or approach adopted to analyse your results,” structured in a linear way “showing how certain decisions have impacted on subsequent choices.”
Methodology vs. methods: methods are the specific tools (a survey, interviews, a regression). Methodology is the reasoning that connects those tools to your research questions. Your chapter needs both, but the reasoning is what separates a 2:1 from a First.
What markers are actually looking for
Marking criteria are remarkably consistent across institutions. Pulling from published schemes:
- Replicability. University of Nottingham’s First-class criteria require methods “comprehensive in terms of materials, methodology and protocol with appropriate references, allowing replication by others.”
- Justification. University of Malta’s dissertation template lists “justification for choice of method” as a standalone criterion, separate from description.
- Alignment. University of Calgary’s examiner rubric looks for “a relationship formulated between the problem, the questions, the method, and the patterns and significance of the data.”
- Researcher role and limitations. The same rubric asks for “the role of the researcher in the data collection procedure… along with the strengths and limitations that this role provides.”
Every section below is designed to hit one or more of these four.
Dissertation methodology structure (and word budget)
Universities commonly suggest the methodology should take around 15% of the dissertation (University of Birmingham and University of Suffolk both use 15%; Staffordshire’s health sciences handbook uses 20–25% because of expanded ethics and instrument detail; UNC Greensboro’s doctoral guidelines suggest 10–25 pages).
| Dissertation length | Methodology target | Typical level |
|---|---|---|
| 8,000 words | 1,200–1,600 | Undergraduate |
| 10,000 words | 1,500–2,000 | Undergraduate |
| 15,000 words | 2,200–3,000 | Master’s |
| 20,000 words | 3,000–4,000 | Master’s / MRes |
| 80,000+ words | 8,000–15,000 | PhD |
Here’s how a typical 2,000-word chapter breaks down. Treat the section order as a default, not a rule; check your department’s handbook and two or three past dissertations that scored well.
| Section | Share | ~Words (of 2,000) | Job it does |
|---|---|---|---|
| 3.1 Introduction | 5% | 100 | Restate RQs, map the chapter |
| 3.2 Research philosophy | 10% | 200 | Ontology/epistemology, briefly |
| 3.3 Approach & design | 20% | 400 | Qual/quant/mixed, inductive/deductive, design type, time horizon, alternatives rejected |
| 3.4 Sampling & participants | 15% | 300 | Who, how many, how chosen, why that’s enough |
| 3.5 Data collection | 15% | 300 | Instrument, procedure, pilot |
| 3.6 Data analysis | 15% | 300 | Technique, steps, software |
| 3.7 Quality (validity/reliability or trustworthiness) | 7% | 140 | How you guarded against bias |
| 3.8 Ethics | 7% | 140 | Approval, consent, anonymity, data storage |
| 3.9 Limitations | 5% | 100 | Honest, with mitigation |
| 3.10 Summary | 1% | 20 | Hand over to findings |
The single most common imbalance: too many words on research philosophy, too few on design justification and procedure.
Example 1: qualitative dissertation methodology (annotated)
The four model studies below are fictional examples. Participant details, findings and ethics reference numbers are illustrative.
Study: How do line managers in UK small and medium-sized enterprises experience the shift to hybrid working? (Master’s, business/management, ~15,000 words.) This is a full chapter compressed to ~1,000 words. Numbered markers [1]–[18] link to the annotations beneath.
3. Methodology
3.1 Introduction
This chapter sets out and justifies the research design used to address the two research questions posed in Chapter 1: (RQ1) how line managers in UK SMEs describe their day-to-day experience of managing hybrid teams, and (RQ2) what factors they perceive as enabling or constraining their effectiveness in that role. [1] It proceeds from the philosophical assumptions underpinning the study (3.2), through the research approach and design (3.3), to sampling (3.4), data collection (3.5), analysis (3.6), quality criteria (3.7), ethical considerations (3.8) and limitations (3.9). [2]
3.2 Research philosophy
This study adopts an interpretivist position. Interpretivism holds that social reality is constructed through the meanings individuals attach to their experiences, and that knowledge is therefore produced through interpretation rather than measurement (Saunders, Lewis and Thornhill, 2019). [3] Because the research questions concern how managers experience and make sense of hybrid working, rather than how frequently particular practices occur, an interpretivist stance is more appropriate than a positivist one, which would presuppose that the phenomenon could be reduced to measurable variables in advance. [4] Consistent with this position, the researcher is treated as an active participant in the construction of findings; reflexivity measures are described in 3.7. [5]
3.3 Research approach and design
The study is qualitative and inductive: themes were developed from the data rather than tested against pre-defined hypotheses, which suits an area where, as Chapter 2 showed, existing research is dominated by large-firm contexts and manager perspectives in SMEs remain under-theorised. [6] A cross-sectional, semi-structured interview design was selected. Focus groups were considered and rejected: managers in small firms frequently know one another through regional networks, and group settings risked inhibiting candid discussion of difficulties with their own teams or employers. [7] A survey was also rejected, since closed questions would have constrained participants to the researcher’s prior categories, undermining the inductive aim. [7] A longitudinal design would have allowed experiences to be tracked as hybrid arrangements matured, but was not feasible within the programme timeframe; this is acknowledged as a limitation in 3.9. [8]
3.4 Sampling and participants
Purposive sampling was used to recruit participants who could speak directly to the research questions (Patton, 2015). Inclusion criteria were: (a) currently a line manager with at least three direct reports; (b) employed in a UK organisation with 10–249 employees; (c) at least 12 months’ experience managing a team that worked a hybrid pattern. [9] Criterion (c) excluded managers whose hybrid experience was limited to emergency pandemic arrangements. Participants were recruited through two regional Chambers of Commerce and LinkedIn; 31 managers expressed interest, 19 met the criteria and 12 were interviewed (7 women, 5 men; sectors spanning professional services, manufacturing, retail and technology). [10] Twelve interviews is consistent with Guest, Bunce and Johnson’s (2006) finding that core themes in homogeneous purposive samples typically emerge within the first twelve interviews; in this study no new codes were generated after interview ten. [11]
3.5 Data collection
Interviews followed a semi-structured guide (Appendix B) of eight open questions developed from the gaps identified in Chapter 2, with probes used flexibly. [12] The guide was piloted with two managers outside the final sample; as a result, one question on “productivity” was reworded because both pilot participants interpreted it as evaluative. [13] Interviews were conducted via Microsoft Teams between February and April 2026, lasted 41–68 minutes (mean 52), were audio-recorded with consent and transcribed verbatim by the researcher. [14]
3.6 Data analysis
Transcripts were analysed using reflexive thematic analysis following Braun and Clarke’s (2006; 2022) six phases: familiarisation, initial coding, generating candidate themes, reviewing themes, defining and naming themes, and writing up. [15] Coding was inductive and semantic in the first cycle and latent in the second, and was managed in NVivo 15. Reflexive thematic analysis was preferred over content analysis because the aim was to interpret patterns of meaning rather than quantify the frequency of categories, and over interpretative phenomenological analysis because the sample size and research questions favoured cross-case patterns over idiographic depth. [15]
3.7 Trustworthiness
Following Lincoln and Guba (1985), credibility was supported through member checking (summaries returned to participants, nine of whom responded, with two minor clarifications incorporated); dependability through an audit trail of coding decisions (Appendix D); confirmability through a reflexive journal kept throughout; and transferability through the thick contextual description of participants and settings provided in Chapter 4. [16] The researcher has previously worked in an SME and held assumptions about the difficulty of hybrid supervision; these were recorded before data collection and revisited during analysis. [5]
3.8 Ethical considerations
Ethical approval was granted by the Faculty Research Ethics Committee on 14 January 2026 (ref. FREC-2026-0117). Participants received an information sheet and gave written informed consent; they were reminded they could withdraw up to two weeks after their interview. Transcripts were pseudonymised and identifying details of employers removed; recordings were stored on encrypted university OneDrive storage and will be deleted upon award of the degree, in line with the university’s data retention policy. [17]
3.9 Limitations
The sample is drawn from two English regions and over-represents professional services, so findings should be read as transferable to similar contexts rather than generalisable to all UK SMEs. Self-selection may have attracted managers with stronger views on hybrid working. The cross-sectional design captures experience at one point in time. These limitations are mitigated by the diversity of sectors represented and the transparency of the audit trail, and are revisited in Chapter 6. [18]
3.10 Summary
This chapter has justified an interpretivist, inductive, qualitative design using semi-structured interviews with twelve SME line managers, analysed through reflexive thematic analysis. Chapter 4 presents the four themes that resulted.
Annotations: why this example works
| # | What it does | Why it earns marks |
|---|---|---|
| [1] | Restates the research questions verbatim | Every later justification is tested against them; the marker never has to flip back |
| [2] | Signposts the chapter | Shows the “chain” structure markers expect |
| [3] | Defines the philosophy with a citation | Demonstrates understanding, not just name-dropping |
| [4] | Justifies philosophy against the RQs and names the rejected alternative | This is the move that separates description from justification |
| [5] | Flags researcher positionality and returns to it later | Calgary rubric: “role of the researcher… clearly described” |
| [6] | Links the inductive approach to the literature gap from Chapter 2 | Proves the chapters are connected, not written in isolation |
| [7] | Names two rejected methods with specific, context-based reasons | “Usually the most convincing sentence in the section”: shows a decision, not a default |
| [8] | Admits a practical constraint and cross-references limitations | Honesty + forward reference = maturity |
| [9] | Explicit, lettered inclusion criteria, with the reason for the non-obvious one | Replicability; pre-empts “why that cut-off?” |
| [10] | Recruitment funnel: 31 → 19 → 12, plus sample composition | Transparency about who you reached and who you didn’t |
| [11] | Justifies sample size with a source and study-specific evidence | “It’s all I could get” is not a justification; this is |
| [12] | Points to the instrument in an appendix and ties it to the lit review | Replicability without bloating the chapter |
| [13] | Reports the pilot and what changed because of it | A pilot that changed nothing looks like a box-tick |
| [14] | Platform, dates, duration range, recording, transcription | The “recipe” level of detail |
| [15] | Names the exact analytical procedure, version, software, and two rejected alternatives | Shows the label was chosen knowingly; “thematic analysis” alone is a red flag to markers |
| [16] | Uses an established quality framework with concrete actions for each criterion | Turns an abstract concept into evidence; note the “nine of whom responded” honesty |
| [17] | Approval reference, consent, withdrawal window, anonymisation, storage, retention | The complete ethics checklist in 80 words |
| [18] | Each limitation paired with a mitigation and scoped claim | “Transferable, not generalisable” is the correct qualitative framing |
Example 2: quantitative dissertation methodology (annotated excerpts)
Study: Is late-night social media use associated with poorer sleep quality in UK undergraduates? (Undergraduate psychology, 10,000 words.) The structure mirrors Example 1; below are the sections where quantitative chapters differ most.
3.3 Design
A cross-sectional correlational survey design was used to test H1 (that self-reported social media use after 22:00 would be negatively associated with sleep quality) and H2 (that this association would hold after controlling for caffeine intake and weekday/weekend differences). [A] An experimental design would permit causal inference but was neither ethically nor practically feasible, as it would require restricting participants’ device use over several weeks. [B] The design is deductive: hypotheses were derived from the displacement and arousal accounts reviewed in Chapter 2. [C]
3.4 Participants
An a priori power analysis in G*Power 3.1 indicated that detecting a small-to-medium effect (f2 = 0.08) with three predictors at α = 0.05 and power of 0.80 required a minimum of 141 participants. [D] Convenience sampling via the department’s research participation scheme and student social media groups yielded 236 responses; 22 were excluded for failing an attention check and 8 for completing in under three minutes, leaving n = 206 (71% female; mean age 20.4, SD = 2.1). [E] Convenience sampling limits generalisability beyond the host institution; this is addressed in 3.9. [F]
3.5 Measures
Sleep quality was measured using the Pittsburgh Sleep Quality Index (PSQI; Buysse et al., 1989), a 19-item instrument yielding a global score from 0–21, with higher scores indicating poorer sleep; scores above 5 indicate clinically poor sleep. The PSQI has demonstrated good internal consistency in student samples; in the present sample Cronbach’s α = 0.78. [G] Late-night social media use was operationalised as self-reported minutes of use between 22:00 and 02:00 on a typical weeknight, averaged over the past seven days. [H] Caffeine intake was recorded as number of caffeinated drinks per day. The full questionnaire is in Appendix A.
3.6 Data analysis
Data were screened for missingness (< 2% on any item; handled by pairwise deletion), normality (skewness and kurtosis within ±2) and multicollinearity (all VIF < 1.4). [I] H1 was tested with Pearson’s correlation. H2 was tested with hierarchical multiple regression, entering caffeine and weekday/weekend status in Step 1 and late-night use in Step 2, so that the unique variance explained by social media use could be isolated. [J] Analyses were conducted in SPSS 29 with two-tailed tests and α = 0.05.
Annotations
| # | Why it works |
|---|---|
| [A] | Hypotheses stated with direction; design named precisely (“cross-sectional correlational survey”) |
| [B] | Rejected alternative with ethical and practical reasons |
| [C] | Deductive approach explicitly tied to named theories from the lit review |
| [D] | Power analysis with every parameter reported, so a reader can re-run it |
| [E] | Exclusion rules stated with numbers, final n, and basic demographics |
| [F] | Limitation named at the point it arises and cross-referenced |
| [G] | Validated instrument: source, scoring, interpretation threshold, reliability in this sample |
| [H] | Operationalisation of the key variable is specific and time-bounded |
| [I] | Data screening reported with thresholds; this is routinely missing from UG dissertations and markers notice |
| [J] | Test choice justified by what it allows you to conclude, not just named |
Example 3: mixed methods dissertation methodology (annotated excerpt)
Study: Why do first-generation students disengage from university peer-mentoring schemes? (Master’s in education.) Mixed methods chapters live or die on how clearly you explain the relationship between the two strands.
3.3 Research design
An explanatory sequential mixed methods design was adopted (Creswell and Plano Clark, 2018). [i] In Phase 1, a survey of 184 mentees measured engagement levels and six candidate predictors drawn from the literature. In Phase 2, 10 participants were selected from survey respondents to explain Phase 1 patterns through semi-structured interviews. [ii] The quantitative strand therefore had priority in addressing RQ1 (which factors predict disengagement?), while the qualitative strand addressed RQ2 (how do students account for their disengagement?). [iii] Phase 2 participants were sampled purposively from the Phase 1 dataset to include both high- and low-engagement students, so that contrasting cases could be compared. [iv] Integration occurred at two points: in sampling (Phase 1 results determined Phase 2 selection and shaped the interview guide) and in interpretation, where findings were merged using a joint display (Chapter 5, Table 5.3). [v] A convergent design, collecting both strands simultaneously, was rejected because the interview guide needed to respond to survey findings that could not be anticipated. [vi] The pragmatist philosophy underpinning the design holds that the research question, rather than paradigm loyalty, should determine method (Morgan, 2014). [vii]
Annotations
| # | Why it works |
|---|---|
| [i] | Names the specific mixed design (not just “mixed methods”) with the standard citation |
| [ii] | Sequence and sample sizes for each phase stated up front |
| [iii] | Priority is explicit and mapped to specific RQs |
| [iv] | Shows how Phase 1 fed Phase 2 — the “explanatory” logic in action |
| [v] | Integration points named; “joint display” signals you know the mixed methods literature. Most weak mixed-methods chapters never explain integration at all |
| [vi] | Rejected alternative is the obvious competitor design |
| [vii] | Philosophy appears, but briefly and justified, rather than as a 500-word detour |
Example 4: secondary data / desk-based dissertation methodology (annotated excerpt)
If you’re doing a library-based dissertation (common in law, history, politics, literature, and many business degrees), you still need a methodology. The commonest mistake is writing two paragraphs saying “secondary research was conducted.” Your “data” are documents, datasets, cases or texts, and markers want to know how you selected and analysed them with the same rigour as primary data.
Study: How has climate-related disclosure in FTSE 100 annual reports changed since the TCFD recommendations? (Undergraduate accounting & finance.)
3.3 Research design and data sources
This study uses a longitudinal qualitative content analysis of corporate annual reports (Krippendorff, 2018). [a] Secondary documentary data were chosen over primary methods because the research questions concern what companies publicly disclosed and how this changed, which is directly observable in the reports themselves; interviewing preparers would capture intentions, not disclosure. [b]
3.4 Sample
The population was FTSE 100 constituents as at 31 December 2024. A stratified purposive sample of 20 companies was drawn, four from each of the five sectors with the highest Scope 1 emissions intensity, since disclosure pressure is concentrated in these sectors. [c] For each company, annual reports for FY2017 (the year the TCFD recommendations were published), FY2020 and FY2023 were retrieved from company investor-relations websites, giving 60 documents. Three companies had been acquired or delisted during the period and were replaced by the next-ranked company in the same sector. [d]
3.5 Coding framework and analysis
A coding framework was developed deductively from the eleven TCFD recommended disclosures, organised under the four pillars of governance, strategy, risk management and metrics and targets (Appendix C). [e] Each disclosure was scored 0 (absent), 1 (generic mention) or 2 (specific, company-level detail), adapting Braasch and Velte’s (2023) scoring framework to distinguish absent, generic and specific disclosure. [f] To assess reliability, a second coder independently coded 12 of the 60 reports (20%); Cohen’s κ = 0.81, indicating strong agreement, and disagreements were resolved through discussion. [g] Scores were analysed descriptively by year, sector and pillar, and illustrative extracts were selected to show how the quality of disclosure shifted, not only whether it did. [h]
3.8 Ethics
The study used publicly available documents and did not involve human participants; it was confirmed as exempt from full ethical review by the department’s ethics lead on 3 November 2025. Data were nonetheless handled in line with the university’s research integrity policy. [j]
Annotations
| # | Why it works |
|---|---|
| [a] | A named, citable method. “Desk research” is not a method; content analysis, documentary analysis, doctrinal analysis, systematic review, and critical discourse analysis are |
| [b] | Justifies secondary over primary because of the RQ, not because it was easier |
| [c] | A sampling rationale for documents, with the reason for the stratification variable |
| [d] | Dates, source, document count, and how attrition was handled: fully replicable |
| [e] | Coding framework grounded in an external standard and placed in an appendix |
| [f] | Scoring scale defined and its adaptation from prior work made explicit |
| [g] | Inter-coder reliability: proof the analysis isn’t just one person’s impression |
| [h] | Analysis described at the level of what you’ll report, previewing Chapter 4 |
| [j] | Even “no participants” studies should state their ethics status explicitly |
How to write your dissertation methodology: step by step
Step 1: open by restating your research questions
One short paragraph. Quote your RQs or hypotheses exactly as they appear in Chapter 1, then signpost the sections to come. Everything after this is justified against these questions, so put them where the marker can see them.
Step 2: state your research philosophy, briefly
Say which position you adopt (most commonly positivism, interpretivism, pragmatism, or critical realism), define it with a citation, and give one or two sentences on why it suits your questions. Keep it under 10% of the chapter. If your department doesn’t teach paradigms (common in sciences and engineering), check past dissertations; you may be able to skip this.
Step 3: justify your approach and design, and name what you rejected
Cover: qualitative/quantitative/mixed; inductive/deductive; the specific design (survey, experiment, case study, phenomenology, grounded theory, action research, content analysis, systematic review…); and time horizon (cross-sectional vs. longitudinal).
For each, use the four-part pattern that runs through every example above:
- State the choice and cite a methods source.
- Justify it against your research questions.
- Say what it means for your data and what your findings can claim.
- Name a realistic alternative and why you didn’t use it.
Weak vs. strong, side by side:
| Described (weak) | Justified (strong) |
|---|---|
| “A qualitative approach was used.” | “A qualitative approach was used because RQ1 asks how participants interpret their role, which cannot be captured through pre-defined response categories.” |
| “Interviews were conducted and analysed thematically.” | “Semi-structured interviews were chosen over focus groups because the topic (workplace conflict) risked social-desirability effects in group settings. Transcripts were analysed using Braun and Clarke’s (2006) six-phase reflexive thematic analysis…” |
| “Secondary data was used due to time constraints.” | “Documentary data were chosen because the research question concerns what organisations publicly committed to, which is directly observable in policy documents and would be filtered through recollection in interviews.” |
| “The sample size was 8.” | “Eight participants were interviewed. This is within the 6–10 range typical for interpretative phenomenological analysis (Smith, Flowers and Larkin, 2022), where depth of idiographic analysis takes priority over breadth.” |
Step 4: describe sampling so someone could repeat it
Population, sampling technique (and why), inclusion/exclusion criteria (and why), recruitment route, response/recruitment funnel, final sample, and a justification for the size. Quantitative: power analysis. Qualitative: cite the sample-size literature for your design and give study-specific evidence (e.g., no new codes after interview n).
Step 5: detail data collection as a procedure
Instrument (and where it came from), piloting (and what changed), mode, location, dates, duration, recording, and storage. If you used a validated scale, report its source, scoring and reliability in your sample. Put full instruments in appendices.
Step 6: specify your analysis technique and steps
Name the technique precisely, cite its source, walk through the steps, name the software and version, and justify the technique against an alternative. For quantitative work, report data screening (missing data, outliers, assumption checks) before you name the tests.
Step 7: address quality
Quantitative: validity (internal, external, construct) and reliability. Qualitative: Lincoln and Guba’s trustworthiness criteria (credibility, transferability, dependability, confirmability) with a concrete action for each. Include researcher reflexivity for interpretive work.
Step 8: cover ethics completely but concisely
Approval body, date and reference; informed consent; right to withdraw (with a deadline); anonymisation/pseudonymisation; data storage and retention; any special considerations (vulnerable groups, sensitive topics, deception). Secondary-data studies: state your exemption.
Step 9: own your limitations
For each: what it is, how you mitigated it, and why your findings still hold. Three or four well-handled limitations beat a defensive list of ten. Don’t hide what your marker can already see.
Step 10: summarise and hand over
Two to four sentences recapping the key decisions and pointing to the findings chapter. No new information.
10 mistakes that cost marks in methodology chapters
- Describing instead of justifying. The most common and most costly.
- Writing a methods textbook. Three pages comparing positivism and interpretivism in the abstract, then one sentence on your study. Keep theory in service of your choices.
- Using a label you don’t honour. Calling the study “phenomenological” or “grounded theory” commits you to specific sampling, analysis and reporting conventions. If you don’t follow them, the label becomes a liability.
- Misaligned RQs and methods. An RQ about “lived experience” answered with a Likert survey, or an RQ about prevalence answered with six interviews.
- No rejected alternatives. Markers read this as not having considered any.
- Unjustified sample size. Especially “it was all I could get,” with no reference to what the sample still allows you to conclude.
- Skipping data screening / reliability. For quantitative work, no assumption checks; for qualitative, no audit trail or reflexivity.
- Reporting results in the methodology. Describe how you analysed; save what you found for Chapter 4.
- Future tense in the final submission. Proposals are written in the future tense; the final dissertation describes what you did. University of Westminster’s guide is explicit on this.
- Treating a desk-based dissertation as needing no methodology. It needs one just as much; see Example 4.
Your dissertation methodology checklist
Before you submit, every line should be a yes:
- Research questions/hypotheses are restated at the start of the chapter
- Philosophy is stated, defined, cited and justified in ≤ 10% of the chapter (or omitted if your field doesn’t use it)
- Approach (qual/quant/mixed; inductive/deductive) is justified against the RQs
- Specific design is named, cited, and at least one alternative is rejected with reasons
- Time horizon is stated
- Sampling technique, criteria, recruitment funnel and final sample are reported
- Sample size is justified with a source and study-specific evidence
- Instrument source, piloting and any changes are described
- Procedure includes dates, mode, duration, recording and storage
- Analysis technique is named precisely, cited, stepped out, with software and version
- Data screening (quant) or audit trail/reflexivity (qual) is reported
- Quality criteria are addressed with concrete actions
- Ethics: approval reference, consent, withdrawal, anonymity, storage, retention
- Each limitation has a mitigation
- Past tense throughout (final submission)
- Appendices contain instruments, consent forms, ethics approval, coding frameworks
- Nothing in the chapter is a result
How the methodology changes by discipline (click to expand)
| Discipline | Usual label | Typical emphasis | Often omitted |
|---|---|---|---|
| Psychology | “Method” | Participants, Materials/Measures, Design, Procedure (APA order); power; assumption checks | Research philosophy |
| Business / Management | “Methodology” | Saunders’ research onion; philosophy; justification of design; ethics with organisational access | Extended statistics detail at UG level |
| Education / Social work / Nursing | “Methodology” | Positionality/reflexivity; trustworthiness; ethics with vulnerable groups; practitioner-researcher role | Hypothesis-testing language |
| Law (doctrinal) | “Methodology” or “Approach” | Doctrinal vs. socio-legal; sources hierarchy; jurisdictional scope; comparative method rationale | Sampling, ethics (state exemption) |
| History / Literature | “Approach” or “Sources and methods” | Archive/corpus selection; source criticism; theoretical lens | Philosophy as a separate section |
| Engineering / CS | “Methodology” or “Design” | Requirements, system design, experimental setup, metrics, datasets, baselines | Philosophy, sampling of people |
| Sciences | “Materials and Methods” | Materials, apparatus, protocol, controls, statistical analysis | Philosophy, researcher positionality |
Always defer to your handbook and to two or three recent high-scoring dissertations from your own department.
Phrase bank: sentence starters for each section (click to expand)
Justifying a choice
- “…was selected because RQ2 requires…”
- “Given that the aim is to [explore/measure/compare]…, a … design is more appropriate than …, which would…”
- “While … would have offered …, it was rejected because…”
Sampling
- “Inclusion criteria were (a)…, (b)…, (c)…; criterion (b) was included to exclude…”
- “Of the N individuals approached, n met the criteria and n participated.”
- “This sample size is consistent with … (Author, Year), and in practice no new codes emerged after…”
Analysis
- “Following Author’s (Year) n-phase procedure, …”
- “… was preferred over … because the aim was to … rather than to …”
- “Prior to analysis, data were screened for…”
Limitations
- “A limitation of this design is… This was mitigated by… and the findings should therefore be read as…”
- “A longitudinal design would have permitted …; however, …”
Tense and voice
- Final dissertation: past tense (“Participants were recruited…”; “I conducted…”)
- Proposal: future tense (“Participants will be recruited…”)
- First person is accepted in most qualitative and interpretive fields; many sciences still prefer passive voice. Pick one and be consistent.
Where to find real dissertation methodology examples from your own field (click to expand)
- Your university repository. Most institutions publish past Master’s and doctoral theses; search for your discipline and read Chapter 3 of three that passed well.
- University of Leeds dissertation examples. Full undergraduate dissertations with marker annotations explaining how marks could have been improved.
- OATD (Open Access Theses and Dissertations). Searchable across thousands of institutions worldwide.
- ProQuest Dissertations & Theses. Usually available through your library login.
- UAGC Writing Center annotated Chapter 3. A fully annotated US doctoral qualitative methodology (PDF).
When you read them, don’t copy the content; copy the moves: where they justify, where they cite, where they name alternatives, and how they scope their claims.
Frequently asked questions
A methodology example is a chapter that states the study’s research questions, then justifies the research philosophy, approach, design, sampling, data collection, analysis, quality measures, ethics and limitations, with enough detail for replication. See the four annotated examples above for qualitative, quantitative, mixed methods and secondary-data designs.
Roughly 15% of the total word count is the most common university guideline (e.g., University of Birmingham, University of Suffolk), which means about 1,500 words in a 10,000-word dissertation and 3,000 in a 20,000-word one. Health sciences and psychology often run to 20–25% because of expanded ethics and instrument detail.
Introduction, research philosophy, approach and design, sampling, data collection, data analysis, quality/trustworthiness, ethics, limitations, and summary. Not every field uses every part.
Past tense in the final dissertation (you’re reporting what you did). Future tense in a proposal. If your proposal becomes Chapter 3, convert it and update anything that changed.
In most qualitative, education, health and social science programmes, yes, and reflexive sections almost require it. Many natural sciences prefer passive voice. Check your handbook and be consistent.
Methods are the tools (survey, interviews, regression). Methodology is the reasoning that links those tools to your questions. The chapter needs both.
Yes. You still need to justify why documentary/secondary data suit your questions, explain how you selected your sources, and describe your analytical framework. See Example 4.
Cite the sample-size norms for your specific design, then give study-specific evidence (e.g., when codes stopped emerging), and scope your claims accordingly (“transferable to similar settings” rather than “generalisable”).
Yes, and it’s common. Agree the change with your supervisor, update ethics approval if participants are affected, and briefly note the change and its rationale in the final chapter.
Describe it in the chapter; put the full instrument, consent form, ethics approval letter and coding framework in appendices and cross-reference them.
Cite This Work
To export a reference to this article please select a referencing stye below:
