How to Write the Methodology Chapter

Your methodology chapter is where your examiners decide whether to believe your findings. If your design choices look arbitrary, everything built on them looks shaky too, however strong your results are.

Having coached 10,000+ students, we’ve found the fix for a weak methodology chapter is rarely a better design. It’s a better explanation of the one you have. In this post, I’ll show you what the chapter covers, how to justify each choice, and eight real dissertation extracts that do it well.

What is the methodology chapter?

The methodology chapter sets out how you designed your study and why. It’s usually Chapter 3, after the literature review and before your results.

It matters for two reasons. It shows you understand research design, because a flawed design produces flawed findings. And it makes your study repeatable: another researcher should be able to follow your chapter, run the same study and compare their results with yours.

The choices in it aren’t a list of separate items, either. They form a chain. Your research philosophy shapes which approaches make sense, your approach narrows the research design options worth considering, and your design decides which sampling, data collection and analysis methods follow. If you’re unsure where methodology ends and methods begin, this explainer covers the difference.

How to write up the methodology chapter

The exact structure varies by field and by university, so check your program’s guidelines first and, if you can, read a few past dissertations from your department. You can also browse our collection of real dissertations and theses, which is where the eight extracts below come from. What follows is a common structure in the social and applied sciences.

Before you write a word, draft an outline. If you start writing without knowing what goes where, you’ll end up stitching disconnected pieces together later, which takes far longer than planning would have. Most methodology chapters have four parts, and the middle one does most of the work:

The middle part carries the chapter’s weight: each choice follows from the one before, and each needs its own justification.

Here’s what goes into each one.

---

1. Introduction

Remind your examiner what the study is trying to achieve, especially your research aims, objectives and questions, because every choice that follows will be justified against them. Then give a brief map of how the chapter is organized. A paragraph or two is plenty.

Here’s how one business PhD opens its methodology chapter:

John’s opening points straight back at the research questions, then tells the examiner what the chapter will justify first and where to find it. Nobody reading the rest of that chapter will get lost. Do the same in yours: anchor first, then map.

---

2. Methodological choices

This is the heart of the chapter, so get specific. It’s not a “less is more” situation. Here are the choices most dissertations need to cover, in the order they usually appear, followed by a simple way to justify each one.

Research philosophy

Research philosophy is the set of assumptions behind your study: what you believe counts as knowledge, and how it can be gathered. Three positions come up most often:

• Positivism assumes a single reality that can be observed and measured objectively, and it commonly underpins quantitative studies.
• Interpretivism, at the opposite end, assumes reality is shaped by the people experiencing it, and it commonly underpins qualitative studies.
• Pragmatism sits between the two, judging methods by whether they answer the research question, and it commonly underpins mixed methods studies.

Whichever you adopt, state it and say why it suits your research questions.

Here’s what that looks like in a public health PhD, stated in the first person:

Munn names the position and gives the reason in a sentence anyone could follow. The extract also combines positions that are often paired differently, and says so openly. That’s allowed, as long as you explain why, the way this one does.

Side note: Not every program expects a philosophy section, so check a few past dissertations from your department before writing one.

Research approach

Next, say whether your research approach is inductive or deductive. Inductive research works bottom-up, starting from data and building toward a theory, so it tends to be exploratory. Deductive research works top-down, starting from a theory or hypothesis and testing it against data, so it tends to be confirmatory.

Then say whether your study is qualitative, quantitative or mixed methods. This choice is closely tied to your philosophy, so check the two agree: a positivist study built entirely on open-ended interviews needs some explaining.

In this education PhD, the choice of a quantitative approach arrives with its reasons attached:

Huang doesn’t just declare the study quantitative. The reasons are numbered, and the first is the one that carries the weight: the thing being studied can be measured with a valid instrument. If what you’re studying can’t be measured that way, that’s your cue to reconsider the approach rather than argue harder for it.

Research design

Your research design (sometimes called your research strategy) is the overall plan for answering your questions. Common designs include experimental research, surveys, case studies, ethnography, grounded theory, action research and phenomenology.

They suit very different aims. An experiment manipulates one variable while holding others constant, so it’s well placed to establish cause and effect. Ethnography observes people in their natural setting, so it can’t establish causation but is well suited to understanding a group’s culture from the inside. The right design follows from what your research questions ask, and it’s one of the places where naming the alternative you rejected pays off most.

Time horizon

Say whether your data were collected at one point in time (cross-sectional) or at several points (longitudinal). If your research questions are about change over time, you’ll probably need a longitudinal design. Be realistic, though: most degree timelines only leave room for a cross-sectional study, and “a longitudinal design wasn’t feasible within the program” is a legitimate justification, as long as you note it as a limitation too.

Sampling strategy

Explain who you collected data from and how you chose them. Sampling methods fall into two broad families:

• Probability sampling selects participants at random from a defined population, which supports generalizing your findings to that population.
• Non-probability sampling selects participants deliberately or by convenience, which is common in qualitative work and can’t support that kind of generalization.

Then give the detail that would let someone repeat it. State your inclusion and exclusion criteria, and say why each one is there: a criterion that looks arbitrary invites the question of whether it skewed your sample. Explain how you recruited people (through an organization, a mailing list, social media or referrals from early participants), how many you approached and how many took part. If some declined or dropped out, say so.

Finally, justify the size of your sample. In a quantitative study that usually means a power analysis, showing the sample was large enough to detect the effect you were looking for. In a qualitative study it’s more often an argument that your data answered the question: themes had started repeating, or your sample is in line with other studies using the same method.

“It was all I could get” is honest, but it isn’t a justification on its own. Say what the sample still lets you conclude, and list any shortfall under your limitations.

This development sociology thesis names its sampling technique, then ties it straight back to the research question:

Look for the word “Because”. Madsen names the technique, then ties it to one specific part of the research question. That link is what turns a sampling label into a justification.

Data collection

Explain what data you collected and how. The usual methods differ by approach:

• Quantitative studies often use surveys, existing datasets or measurements from equipment.
• Qualitative studies more often use interviews, focus groups or observation.

Whichever you used, give the detail that lets someone repeat it: the instrument, where it came from, how you tested it, and how each session ran.

In this master’s thesis in education, one data collection method is weighed against another before the choice is made:

This is justification done well. Barnes names the option considered, why it was set aside, and what was chosen instead. The reason is an honest one, too: time. Practical constraints are a legitimate justification, as long as you say what they cost you.

Data analysis

Be precise about how you analyzed your data, and justify each technique. For qualitative studies, common options include content analysis, thematic analysis and discourse analysis; our overview of qualitative analysis methods compares them. Quantitative studies almost always report descriptive statistics, and most go on to inferential tests such as correlation or regression.

Also explain how you prepared the data before analyzing it: removing duplicates and incomplete responses, transcribing interviews, translating them if needed. Name any software you used.

An analysis justification doesn’t need to be long. This one, from a public health PhD, takes two sentences:

Two sentences, and both reasons point at the data rather than the method’s reputation: the tests don’t assume a normal distribution, and they cope with unequal groups, which this study had (its measurement points ran from 183 to 904 women). The thesis also admits the cost in the very next sentence: the results are harder to interpret. Point your own analysis justification at your data or your questions, and be as honest about the trade-off.

Ethics

If your study involves people, you’ll almost certainly need approval before you collect any data: from an institutional review board (IRB) in the US, or an ethics committee in most other countries. Say which body approved the study, how you obtained informed consent, how you protected participants’ identities and how you stored the data. The detail your university expects varies, so check its template. Our guide to research ethics covers the principles behind each of these.

How to justify each choice

You’ll have noticed that each extract in this section does more than name a method. That extra step is what your examiners grade, and it follows the same pattern whatever the choice. Here’s the one our coaches teach, and it works for everything from your paradigm to your statistical tests:

1. State and define the choice, citing a methods source for it.
2. Justify it against your research aims and questions. Why does this choice suit what you’re trying to find out?
3. Say what it means for your study: how it shapes your data collection, your analysis or what your findings can claim.

For your research design and your analysis methods, add a fourth step: name a realistic alternative and say why you didn’t use it. That’s usually the most convincing sentence in the section, because it shows you made a decision rather than defaulted to one.

Sampling and data collection bend the pattern slightly. There, step three becomes procedure: inclusion and exclusion criteria, how you recruited people, what you asked them, how you stored the data. Think of it as a recipe. Someone else should be able to follow it without calling you.

The second version states the choice, points to its source, ties it to the research question and names what it means for the findings. That’s the whole pattern in three sentences. For how much justification a tight word count leaves room for, see how to justify your research methodology.

---

3. Limitations

No methodology is perfect. There are always trade-offs between the ideal design and what your time, budget and access allowed. This section is where you name those trade-offs and explain why they were reasonable.

Limitations range from time and budget constraints to sampling or selection bias: a sample smaller than you’d hoped for, say, or one skewed toward a particular group. Be critical, since hiding a weakness your examiner can already see costs you more than admitting it. But don’t beat your study to death either. For each limitation, say what it was, how you reduced its impact and why the study still has value.

Here’s a nursing PhD owning a small sample without apologizing for the study:

Herron names the limitation plainly, then gives two reasons it doesn’t sink the study: small samples are normal in phenomenology, and the themes had started repeating. Limitation, context, then why the findings still stand. If you claim saturation yourself, be ready to explain how you judged it.

---

4. Summary

End the chapter with a paragraph or two recapping your key decisions. A figure can help here, especially if your university favors a particular model such as the research onion. Keep it to what you’ve already said: the summary is not the place for new information.

Here’s how the education thesis quoted under data collection closes its methodology chapter, in three sentences (EC is early childhood, the study’s setting):

It restates the chapter’s two main decisions, a qualitative case study and theoretical thematic analysis, then names the data sources and hands over to the findings. Nothing in it is new and nothing is argued again, because the chapter has already done that work. Aim for the same in yours: recap, don’t re-argue, and point to what comes next.

---

Methodology chapter example

That’s the chapter, part by part. The extracts above show how individual choices are justified; if you’d like to see a whole chapter in one place, our research methodology example walks through a complete chapter from a real dissertation, section by section.

Where methodology chapters go wrong

Most weak methodology chapters fail in one of a handful of predictable ways. Here are the ones I’d check a draft for first.

Using a label without knowing what it commits you to. Calling your study phenomenological, or grounded theory, or a case study, brings expectations about sampling, data and analysis. If the rest of the chapter doesn’t meet them, the label becomes a liability.

Describing steps instead of justifying choices. “Interviews were conducted and analyzed thematically” tells your examiner what happened. It says nothing about why, and the why is what’s being graded.

Writing it like a journal article. Published papers compress their methods for expert readers under tight word limits. A dissertation is assessing whether you understand your methods, so it needs the detail an article would leave out.

Losing the link to your research questions. If your question explores people’s lived experience and your instrument is a closed-ended survey, the mismatch will be obvious. Every choice should trace back to a question.

Treating your philosophy as a personal belief. Your paradigm is a position you adopt for this study because it suits your question. It isn’t a statement about who you are, and it doesn’t have to match how you see the world outside your research.

Leaving your choices unsupported. You’re not expected to invent new methods. You’re expected to apply established ones well, so cite the methods textbooks and studies your choices come from.

If you can read your draft and find none of these, you’re likely already in quite a good position.

Still have questions?

How long should a methodology chapter be?

Your program’s guidelines come first, since expectations vary a lot between fields and degrees. As one reference point, Kathy Hytten’s dissertation guidelines at UNC Greensboro suggest 10-25 pages for the methods chapter of a doctoral dissertation, and put the typical dissertation at 150-200 pages in total. For a master’s dissertation, read a few from your department that passed well and use their proportions.

Should the methodology be in past or future tense?

It depends on the document. A proposal describes a study you haven’t done yet, so it’s written in the future tense. The final dissertation describes what you actually did, so it’s in the past tense: the University of Westminster’s guide says so plainly. If your proposal becomes Chapter 3, update the tense and revise what you planned into what you did, including anything that changed along the way.

Can I write the methodology in the first person?

In many fields, yes. Qualitative and interpretive research often expects “I”, because your role in collecting and interpreting the data is part of the method; two of the extracts above use it. Some sciences still prefer the passive voice. Check your department’s style guide and a few recent dissertations, and whichever you choose, be consistent.

Do I need a research philosophy section?

If your program teaches research paradigms or the research onion, you almost certainly do. Many science and engineering programs don’t expect one, and some examiners find an unnecessary philosophy section padding. Past dissertations from your department, or your advisor (supervisor, in the UK), will settle it quickly.

What’s the difference between methodology and methods?

Methods are the specific tools you used, such as a survey, interviews or a regression model. Methodology is the reasoning behind them: why those tools, in that combination, suit your research questions. The chapter covers both, which is why it describes what you did and justifies it.

Can I change my methodology after my proposal?

Yes, and it’s common. Agree the change with your advisor first, and if it affects participants, check whether you need to update your ethics approval. The final chapter describes what you actually did, and it’s good practice to note significant changes from the proposal and why you made them.

Can’t find your answer here? Ask a Grad Coach directly – the initial chat is free.

Speak with a friendly coach →