What dissertation methodology support should clarify first
A dissertation methodology should explain how the study will obtain or select evidence that can answer its research questions, and why those choices are defensible. The chapter needs a visible line from the research problem to the design, unit of analysis, population or data source, sampling or selection logic, data-collection procedure, ethical safeguards, quality criteria, and analysis plan.
Methodology support is most useful when it tests that line for contradictions. A technically impressive method is not automatically suitable. If the research question asks for one kind of evidence but the design produces another, the problem is methodological even when each individual section sounds polished.
Start with the evidence each research question requires
Research questions define what the study must be able to learn. Questions about experiences, meanings, processes, relationships, patterns, prevalence, change, comparison, explanation, or evaluation may require different evidence and different designs. The methodology should therefore begin by identifying what each question asks the evidence to establish.
This also limits the conclusions the study can later make. A design that describes a pattern should not be written as though it automatically establishes why the pattern exists. A study that explores participants' accounts can produce rich evidence about experience without being presented as a population estimate. The design and the eventual claim need to stay compatible.
Choose the research design before choosing techniques
Interviews, surveys, observations, datasets, documents, tests, and analytical software are techniques or resources. They do not by themselves define the research design. The design explains how those elements work together to answer the research questions.
A qualitative study may be appropriate when the project needs detailed accounts, meanings, experiences, processes, or context. A quantitative study may be appropriate when the questions require numerical measurement, comparison, estimation, association, prediction, or another quantitative form of evidence. Mixed methods may be appropriate when the project genuinely needs both forms of evidence and can explain how they will be integrated. Case-study, secondary-data, document-based, and other designs also need justification in relation to the problem rather than selection by label.
Do not treat qualitative, quantitative, and mixed methods as prestige levels
No one approach is inherently more advanced because it uses numbers, interviews, multiple methods, or specialised software. The better design is the one that fits the research question and can be executed rigorously within the project's constraints. Adding a second method increases complexity and should have an analytical purpose, not merely make the study appear broader.
Define the unit of analysis and evidence source
The methodology should make clear what is actually being studied. The unit of analysis might be individuals, groups, organisations, events, documents, transactions, cases, jurisdictions, time periods, or another defined entity. Confusion at this level can lead to conclusions about one unit when the evidence was collected from another.
The evidence source should then be described in relation to that unit. Primary research may involve participants, observations, measurements, or researcher-generated records. Secondary research may use an existing dataset, administrative records, documents, archives, published material, or other existing evidence. The chapter should explain why the chosen source can provide the information the research questions require.
Build sampling or case selection around the design
Sampling is not only a question of how many observations or participants will be included. It also concerns who or what can enter the study, how selection occurs, which cases are excluded, and what that selection means for the conclusions.
Probability-based, purposive, convenience, criterion-based, theoretical, case-based, and other selection approaches serve different purposes and carry different limitations. The methodology should use the terminology and justification that fit the actual design. It should not describe a convenient sample as representative without evidence, or imply that a larger sample automatically corrects a weak selection process.
Justify sample size using the logic that fits the study
There is no single sample-size rule for every dissertation. Quantitative studies may require a justification tied to the planned analysis, variability, precision, effect assumptions, available population, or discipline-specific guidance. Qualitative studies may justify sample adequacy through the study purpose, depth, diversity of relevant perspectives, information needs, design tradition, and the logic accepted for that methodology. Existing datasets may impose a fixed available sample that has to be evaluated rather than chosen.
The important question is whether the amount and type of evidence are sufficient for the analysis and claims the study intends to make. Any formal calculation or discipline-specific threshold should be grounded in the appropriate methodological source and project requirements.
Specify data collection as a reproducible research process
The reader should be able to understand what will happen from access or recruitment through to the creation of an analysis-ready body of evidence. That may include where participants or records come from, what information is collected, the sequence of procedures, who administers an instrument, how long an activity lasts where relevant, how responses or observations are recorded, and what happens to incomplete or unusable material.
Enough detail should be provided to judge the method without turning the chapter into an operational diary. The right level of detail depends on the design and programme. A complex experimental procedure may require different documentation from a document analysis or semi-structured interview study.
Make instruments and measures fit the constructs
When the study uses a survey, scale, test, interview guide, observation protocol, extraction form, coding framework, or other instrument, the methodology should explain what it is intended to capture and why it fits the research question. The name of an instrument is not a substitute for showing the connection between the construct and the evidence it produces.
For quantitative work, this may involve explaining how variables or constructs are operationalised and what measurement properties matter for the planned use. For qualitative work, it may involve showing how interview or observation prompts allow participants or cases to address the study's questions without building the preferred conclusion into the prompt. Project-specific claims about an instrument's validity, reliability, licensing, or approved use require appropriate evidence.
Use pilot or pretest work where it has a research purpose
A pilot, pretest, or trial run can be useful when the study needs to test comprehension, timing, technical procedures, recruitment, data capture, or another element before the main collection begins. It should not be added mechanically to every methodology. Where pilot work is used, the chapter should explain what is being tested and how the findings can affect the final procedure.
Address access before describing an ideal method
A design is not feasible if the required participants, records, sites, equipment, permissions, or data cannot realistically be accessed. Methodology support should distinguish the preferred theoretical design from the design the researcher can actually implement without weakening the research question beyond recognition.
Access also affects sampling and bias. If recruitment depends on one gatekeeper, one organisation, one online channel, or a limited archive, that constraint may shape who or what appears in the evidence. The methodology should describe relevant limitations rather than conceal them behind a generic sampling label.
Treat research ethics as part of the design
Ethics can influence recruitment, consent, data collection, recording, anonymity or confidentiality, data handling, retention, researcher safety, participant risk, conflicts of interest, and the kinds of questions that can responsibly be asked. The exact requirements depend on the institution, jurisdiction, population, data type, and study.
The dissertation should follow its actual ethics approval process and guidance. Methodology support can help ensure that the written procedure is consistent with those materials, but it should not claim ethics approval that has not been granted or substitute a generic process for the institution's requirements.
Choose research-quality criteria that fit the design
Methodological quality should be evaluated using concepts appropriate to the study. Some quantitative projects discuss forms of validity, reliability, measurement error, confounding, missing data, or other threats to inference. Some qualitative designs discuss credibility, dependability, confirmability, transferability, reflexivity, audit trails, or related criteria. Other designs may use different quality frameworks.
These terms should not be inserted as a checklist without showing how the actual research procedure addresses them. If the study claims reliability, credibility, or another quality property, the chapter should explain the relevant design decision or evidence rather than rely on the label alone.
Make bias and researcher influence visible where relevant
Bias can enter through sampling, measurement, data collection, missingness, researcher decisions, coding, analysis, access, or interpretation. The relevant risks depend on the design. Methodology support should identify the plausible sources of distortion and show what the design does to reduce, monitor, or acknowledge them.
In research where the researcher plays an active interpretive or relational role, reflexivity may be important. This can include considering how the researcher's position, assumptions, access, relationship to participants, or analytical choices may shape the study. Reflexivity should be specific to the project rather than a generic personal statement.
Plan the analysis before the data are interpreted
The analysis plan should state how the collected or selected evidence will be transformed into answers to the research questions. Quantitative projects may need to identify data preparation, descriptive analysis, modelling, comparison, testing, or other techniques appropriate to the variables and design. Qualitative projects may need to explain coding, categorisation, thematic, framework, narrative, content, discourse, or another analytical approach when it fits the study.
The methodology does not need to manufacture findings. It should explain the planned procedure, the logic for choosing it, and how the output will answer each question. If the analysis requires assumptions, minimum data characteristics, coding decisions, or special handling of missing or unusual observations, those issues should be addressed at the level appropriate to the project.
Software supports analysis; it does not justify it
Naming statistical, qualitative-analysis, spreadsheet, programming, or other software does not explain the method. The chapter should first justify the analytical approach, then identify the software used to implement it where that information is relevant. A method should not be selected merely because a particular tool is available.
Integrate mixed methods rather than running two separate studies
A mixed-methods design needs a reason for combining qualitative and quantitative evidence and a plan for how the strands relate. The integration may occur through design, sampling, data collection, analysis, comparison, explanation, or interpretation depending on the project. Simply collecting a survey and interviews does not by itself explain what the combination contributes.
The methodology should make the sequence or timing clear and show what one form of evidence adds to the other. If the two strands answer unrelated questions and never inform a common conclusion, the rationale for calling the study mixed methods needs reconsideration.
Evaluate secondary data before treating it as ready-made evidence
Secondary data can make a study feasible, but the researcher did not design the original collection for the current research question. The methodology should therefore examine who or what the dataset represents, how variables were defined, how the data were collected, what is missing, which time period and context apply, what access or use restrictions exist, and whether the available measures can answer the new question.
The same principle applies to documents and records. Availability does not guarantee suitability. The source must be evaluated for provenance, completeness, relevance, and the limits it places on interpretation.
Explain methodological alternatives and trade-offs
A strong methodology can acknowledge reasonable alternatives without turning the chapter into a catalogue of methods. The useful comparison is between the selected approach and the alternatives that a reader might reasonably expect for this research problem.
Trade-offs may involve depth versus breadth, control versus natural context, direct measurement versus available records, representativeness versus access, one-time measurement versus longitudinal evidence, or analytical complexity versus feasible data. The justification should show why the selected balance is appropriate for this project.
Use supervisor feedback to locate the methodological mismatch
Feedback such as “method does not answer the question,” “sampling is unclear,” “needs more justification,” “analysis is not aligned,” “ethics section is incomplete,” or “too much textbook description” points to different weaknesses. Each comment should be mapped to the design decision it challenges.
When one decision changes, dependent sections may also change. A revised research question can affect design, sample, instrument, and analysis. A change in data access can affect the claims that remain defensible. Methodology revision should follow those dependencies rather than edit each subsection in isolation.
A methodology coherence check before data collection
Before the methodology is treated as complete, a reader should be able to answer these questions without filling gaps by assumption:
- What evidence does each research question require?
- Why is the chosen design suitable for producing that evidence?
- What is the unit of analysis?
- Who, what, or which records can enter the study, and how are they selected?
- Why is the planned amount of evidence adequate for the intended analysis?
- What exactly will happen during data collection or evidence extraction?
- Do the instruments or measures capture the constructs the study needs?
- What access and ethics conditions affect the design?
- Which threats to quality, bias, or trustworthiness are relevant, and how are they addressed?
- How will the evidence be analysed to answer each research question?
- Are methodological claims supported by appropriate project or methodological sources?
- Does the chapter stop before inventing results or interpreting findings that do not yet exist?
If several answers do not connect, the study may need a design revision rather than additional methodological terminology.
What to send for dissertation methodology support
Useful materials include the approved proposal or current research questions, dissertation brief or handbook, current methodology draft, planned population or data source, sampling or selection plan, instruments or extraction forms where available, ethics guidance or approval materials, intended analysis, relevant methodological sources, supervisor or committee feedback, and the deadline. Send them through the project order page. Existing clients can continue through client login.