Start by testing whether your research question, data and analysis fit together. A dissertation methodology chapter should explain why your choices make sense for the question and what those choices allow you to conclude. Better wording cannot repair a design that cannot produce the evidence you need.
For UK master’s students, uncertainty often concerns either the research design or its explanation. This guide helps you distinguish them, build a project-specific justification and identify decisions to discuss with your supervisor. Learn tips on the right dissertation methodology. Follow your programme handbook and approval process; no generic chapter template can validate an individual study.
First, Identify What You Are Unsure About
Write your concern as a concrete question. “Is my methodology right?” is difficult to resolve. “Can interviews answer my question about how common this experience is?” exposes a possible mismatch. “Have I explained why interviews fit an experience-focused question?” identifies a writing problem instead.
Other uncertainties concern access, participant recruitment, measurement or analysis. Separate them. A method can fit the question in principle yet be infeasible within your available access and timetable. Equally, collecting accessible data does not make it relevant to the question.
Sheffield Hallam’s methodology guidance asks both why you selected the methods and how they answer your questions. It distinguishes the overarching approach, research design and specific research tools. Use your discipline’s terminology consistently rather than treating these labels as interchangeable.
Test the Fit Between Your Question, Data and Analysis
State the answer you hope to produce, then identify what would support it. Are you interpreting experiences, describing a distribution, examining an association or assessing an intervention? These are different research purposes, even when the topic is identical.
Hypothetical workplace example: “How do new employees describe onboarding?” invites examination of participants’ accounts. “How does reported onboarding satisfaction vary across departments?” needs comparable satisfaction information and department identifiers. “Does onboarding cause better retention?” requires a design capable of addressing a causal claim; collecting opinions alone does not settle it.
For each question, write one sentence connecting the proposed material to the intended answer. If you cannot complete that connection, discuss narrowing the question or changing the design before proceeding. Do not choose a causal verb simply because it sounds more ambitious.
Qualitative, Quantitative or Mixed Methods? Start With the Question
The University of Sheffield’s research-method selection guidance starts with choosing methods that enable you to answer the research questions. Begin there rather than asking which approach is easiest or most impressive.
Table 1. Illustrative Approaches to the Same Onboarding Topic
| Purpose | Possible Approach | Decision to Justify |
|---|---|---|
| Explore participants’ experiences | Qualitative interviews with an explicitly described analytical approach. | Why these participants, questions and interpretation procedures fit the enquiry. |
| Describe or compare reported satisfaction | Quantitative analysis of suitable comparable measures. | What the measure represents, how cases were selected and what comparisons the data support. |
| Connect patterns with accounts of experience | An integrated design combining appropriate quantitative and qualitative components. | Why both are needed and how their findings will be brought together. |
These are possible starting points, not automatic assignments of a method to every question. Mixed methods are not simply adding an interview to a survey. Explain what the combination contributes and where integration occurs. Confirm that the project has the time, skills and permissions required for each component.
A desk-based project also needs a defensible design. A literature synthesis should explain how material was found, selected and examined. A secondary numerical dataset requires checks on what its variables measure and which comparisons are possible. “No primary data collection” is not a complete methodology.
Justify the Method for Your Particular Project
The University of Westminster’s methodology justification guidance recommends explaining fit to the question, discussing limitations and alternatives, and supporting choices with relevant literature. A justification should connect those considerations to your own project.
Weak rationale: “Interviews were chosen because they provide rich data and are suitable for qualitative research.”
Stronger teaching rationale: “The study asks how participants describe adjusting to university assessment. Semi-structured interviews were selected to allow follow-up questions about particular experiences. A fixed-response questionnaire would address different information needs and offer less scope for probing individual accounts. The analysis therefore concerns the participating students’ accounts, rather than estimating how common each experience is across the university [appropriate methodological references].”
The example is hypothetical and incomplete until recruitment, procedures and analysis are specified. Do not copy its rationale if you did not use follow-up questions or if your enquiry has a different purpose. Add real methodological sources that support the approach you actually adopted.
Feasibility belongs in the account, but convenience alone does not establish methodological suitability. If access restricted the setting, explain the boundary and its implications. Distinguish a deliberate scope choice from an unplanned constraint, and avoid claiming that every limitation was eliminated.
Plan the Analysis Before Listing Software
Explain what you will do with the data, why that procedure fits the question and which decisions need justification. Naming Excel, SPSS, R or NVivo does not explain an analytical method. Software may assist the procedure; it does not supply the research logic.
For quantitative work, define the outcome and relevant variables, the comparisons intended, missing-data decisions and any assumptions the proposed analysis requires. Obtain appropriate methods of advice before selecting a test. The variable names or sample count alone do not establish which test is suitable.
For qualitative work, identify the analytical approach and the actual steps used to develop interpretations. Explain your role and relevant reflexive decisions. Do not claim saturation, independent coding or participant checking unless these concepts fit your approach and the procedures genuinely occurred.
Secondary-data example: a master’s student wants to compare working hours with wellbeing. Before choosing an analysis, they check variable definitions, collection period, population, missing values and access conditions. If the dataset lacks a relevant measure, a polished chapter cannot create it. Reconsider the question with the supervisor.
Build a Chapter That Explains Your Research Decisions
Use the handbook’s structure. The planning table below identifies functions to cover, not compulsory headings or a universal word allocation. Some disciplines require a separate methods section, while others integrate methodological discussion differently.
Table 2. Chapter Components and the Decisions They Explain
| Component | Explain | Check Before Submission |
|---|---|---|
| Research purpose and design | The question, approach and design; relevant philosophical assumptions where required. | The position is coherent with the actual enquiry, not a decorative label. |
| Participants or data sources | Selection, eligibility, access and recruitment or source identification. | The account matches the material actually included. |
| Data generation or preparation | Instruments, procedures, relevant dates and processing decisions. | Enough project-specific detail is given without exposing confidential information. |
| Analysis | Steps, choices, assumptions and how the question is answered. | The explanation goes beyond software names. |
| Ethics and researcher role | Applicable approvals, consent, handling and relevant positionality. | Only actual approvals and safeguards are claimed. |
| Quality and limitations | Approach-appropriate quality considerations and bounded conclusions. | The limitations affect the claims made elsewhere. |
Discuss philosophical positioning where it helps explain the design or your course requires it. Do not attach “positivist” to every numerical study or “interpretivist” to every interview study without considering the assumptions. Your supervisor and discipline’s methods literature should guide the level of discussion.
Use the general dissertation guide for the wider project sequence, while keeping your handbook authoritative for chapter organisation.
What If You Have Already Collected the Data?
Separate what you planned from what occurred. Record recruitment changes, departures from the instrument, missing information and analytical decisions honestly. Use wording appropriate to completed versus proposed work. Do not describe a pilot, sampling procedure or quality check that never happened.
If the material can support a narrower question, discuss that option with your supervisor. If an important design or ethics concern emerges, seek guidance before further collection, reuse or changes. Do not use retrospective justification to conceal an unapproved procedure.
Ethics requirements depend on the institution and project. For example, UCL IOE’s student ethics guidance includes review of projects using human-participant data and notes that secondary analysis may be included. Secondary data is not a universal exemption: check your own approval route and any amendment requirements.
Take a Focused Decision to Your Supervisor
Bring the question, available material, proposed analysis and the claim you think they support. State the mismatch and possible responses. For example: “My interviews explore experiences, but the current question asks about prevalence. Should I narrow the question, or is a different design feasible within the remaining time and approvals?”
Ask which option is academically defensible and what must be approved or documented. Keep a record of the agreed actions. A clear methodology chapter grows from coherent research decisions; it should not make a weak design look stronger than it is.
For organising methodological literature around the decisions it informs, see our literature review synthesis guide. Avoid replacing your project rationale with a list of methods definitions.
For permitted help improving the clarity and organisation of your own chapter, explore Oxford Editors’ dissertation editing support in the UK. Research-design decisions, methodological validation and ethics approval remain with you and the appropriate university staff.
Frequently Asked Questions
Is There One Correct Methodology Chapter Structure?
No. Follow your programme handbook, assessment requirements and disciplinary conventions. A useful chapter explains the design, sources or participants, procedures, analysis and relevant ethical and quality considerations. Their order and depth vary. Use the planning table to check coverage rather than treating it as a mandatory template for every master’s dissertation.
Do Interviews Automatically Mean Qualitative Research?
The label does not determine the research approach on its own. Explain what information the interviews generate, how questions are structured and how responses are analysed. Your methodological position should fit those decisions. Avoid selecting a philosophical label merely because it is commonly associated with a particular data-collection tool.
Is Mixed Methods Stronger Than a Single Approach?
Not automatically. A combined design needs a reason for both components and an explanation of how they contribute together. Two poorly connected datasets do not resolve a research question simply by increasing the workload. Discuss whether integration is necessary, feasible and suitable for your question before committing to it.
How Do I Justify a Sample Size?
Use the requirements of your design, intended analysis, population or case boundaries, available access and relevant methodological literature. There is no universal interview count or survey threshold that validates every project. Explain the actual basis for your decision and seek methods advice where needed. Do not claim adequacy using a number alone.
Does Secondary Research Require Ethics Review?
It can. Requirements depend on the institution, the data and the proposed use. Identifiability, access restrictions, original consent and sensitive material may be relevant. Use your institution’s screening or review process and ask the ethics team or supervisor when uncertain. Public availability alone should not be treated as blanket permission for any reuse.
Can I Change My Method After Approval or Data Collection?
Discuss the proposed change with your supervisor and check whether an ethics amendment or other permission is required before implementing it. Explain what changed, why and how it affects the study. Do not assume that an earlier approval covers a different participant group, collection procedure or use of data.