"Which methodology should I use?" is one of the first questions in any thesis or research paper, and one of the most important. The methodology decides what data you collect, how you analyse it and what you can honestly claim at the end. Reviewers and doctoral committees look hard at it, because a weak methodology cannot be rescued by good writing.
The good news: you do not choose a methodology from a menu. You let your research question choose it, then you justify that choice clearly.
Start with the research question
Look at the verbs and the shape of your question:
- How much, how many, how strongly, does X affect Y? These ask for measurement and comparison, which points to a quantitative approach
- How, why, what is it like, what does it mean to people? These ask for depth and meaning, which points to a qualitative approach
- Both: for example, how common a problem is and why it happens. That points to mixed methods
If the question is vague, the method will be too. Sharpen the question first; frameworks such as PICO (for health and intervention studies) or a simple "who, what, where, when" help.
The three main approaches
Quantitative
Numbers, measurement and statistics. It suits testing hypotheses, comparing groups and measuring relationships. Common designs:
- Experimental and quasi-experimental: you change something and measure the effect, ideally with a control group
- Survey (cross-sectional or longitudinal): questionnaires to a sample, analysed statistically
- Correlational: measuring how variables move together, without claiming cause
- Secondary data analysis: using existing datasets, such as national surveys or company records
Qualitative
Words, meanings and experiences. It suits exploring new topics, understanding processes and hearing participants' own views. Common designs:
- Case study: an in-depth look at one or a few cases, such as an organisation or a programme
- Phenomenology: what an experience is like for the people who live it
- Grounded theory: building a theory from the data itself
- Ethnography: studying a group in its own setting over time
- Narrative and content analysis: stories, documents, media or interviews analysed for themes
Mixed methods
Both, combined on purpose. Three designs are widely used:
- Convergent: quantitative and qualitative data collected at the same time and compared
- Explanatory sequential: a survey first, then interviews to explain the results
- Exploratory sequential: interviews first, then a survey built from what you learned
Mixed methods are powerful but take more time and skill. Choose them because the question needs both kinds of evidence, not to look thorough.
Sampling and sample size
- Probability sampling (random, stratified, cluster) lets you generalise statistically to a population
- Non-probability sampling (purposive, convenience, snowball) is common in qualitative work and in hard-to-reach groups; be honest about its limits
- Quantitative sample size should come from a power analysis (for example with the free G*Power software), based on the effect size you expect, not from a round number
- Qualitative sample size is usually justified by depth and by saturation: the point at which new interviews add no new themes
Data collection and analysis must match
A common mistake is choosing tools that do not fit the design. A few examples of good matches:
| If you collect… | You might analyse with… |
|---|---|
| A structured questionnaire with scales | Descriptive statistics, t-tests, ANOVA, regression, factor analysis or SEM |
| Semi-structured interviews | Thematic analysis, framework analysis or grounded theory coding |
| Documents, reports or media | Content analysis, qualitative or quantitative |
| Experiment results | Comparison of groups, with effect sizes |
Our guide to choosing the right statistical test goes into the quantitative side in detail.
Ethics comes first
Research with people needs informed consent, privacy and, in most institutions, approval from an ethics committee before you collect any data. Plan the time for it. Data collected without approval may not be usable in your thesis or publishable in a journal.
How to justify your choice
In your methodology chapter or section, explain:
- Your research question and why it calls for this approach
- Your design, and why it suits the question better than the alternatives
- Your population, sampling method and sample size, with the calculation or reasoning
- Your instruments, and how you checked their validity and reliability (for example a pilot study)
- Your analysis plan, step by step
- Your ethical safeguards and the study's limitations
Cite methodology textbooks and earlier studies that used the same design. Reviewers want to see that your choice was deliberate.
A quick checklist
- Does the method answer the question exactly as it is worded?
- Is the sample large enough, and chosen in a way you can defend?
- Do the analysis methods fit the type of data?
- Is it possible within your time, budget and access?
- Do you have ethics approval?
Unsure which way to go? Our synopsis writing and thesis support services help you plan a methodology you can defend, with you as the author throughout.
