Research Methodology: Research Methods, Research Design and How to Write the Methodology Chapter

§ 01

Research methodology vs research methods

What is research and methodology, put simply? Research is the process of answering a question systematically; methodology is the logic that connects your question to your design, your sample and your analysis, and explains why that combination is the right one. Research methods, by contrast, are the specific tools methodology puts to use: a questionnaire, a structured interview, a regression model. A methodology chapter that lists methods without explaining why they fit the question has described tools without ever making the argument a committee is actually reading it for.

Research methodology in psychology applies this same framework to one field's typical designs and instruments; the framework itself, methodology as the reasoning layer above methods, is not specific to psychology or to any other single discipline. Nursing asks the same questions of a clinical intervention, education asks them of a classroom program, and business asks them of a market or workplace study; the vocabulary below applies across all of them, even where this page borrows one field's example to illustrate a point.

A useful test for which word you actually need: if you are naming a specific tool or technique, you want "method." If you are explaining and defending the reasoning that led you to that tool, you want "methodology." A chapter titled "Methodology" that only lists tools has usually mistitled itself; what a committee expects under that heading is the argument, not just the inventory. The thesis chapters around Chapter 3 show where the methodology chapter sits inside a full thesis or dissertation, between the literature review that justifies it and the results chapter that depends on it.

§ 02

Types of research methods

Types of research methods split first by what kind of data they produce and second by how that data gets collected.

Scroll the table sideways to see every column.

Most studies combine two or three of these rather than relying on just one; a survey study, for example, often adds a handful of follow-up interviews to explain a pattern the numbers alone cannot.

§ 03

Research design: the main kinds

Research design is the overall plan a study follows; the kinds of research design below differ mainly in how much control the researcher has over what happens and whether the goal is to explain a pattern, test a cause, or describe an experience in depth.

Only an experimental or quasi-experimental design can support a causal claim with any confidence; every other design in the table describes, correlates or interprets, which is a legitimate and often more appropriate goal, but not the same claim.

§ 04

Qualitative research methodology

A qualitative research methodology asks what an experience means or how a process unfolds, using non-numeric data: interview transcripts, field notes, documents, open-ended survey responses. It commonly pairs with a phenomenological, grounded-theory, ethnographic or case-study design, and analysis usually proceeds through coding, tagging chunks of text with a label, then grouping codes into broader themes. Software such as NVivo or Atlas.ti manages this coding process at scale, though the analytic judgment behind the codes stays the researcher's own work regardless of the software used. A qualitative study's strength is depth and context; its tradeoff is that findings describe the specific people or setting studied and do not automatically generalize beyond them. A committee reading a qualitative methodology section is checking for one thing above all: that the coding process is systematic and documented, not a researcher reading transcripts and reporting whatever stood out. Naming your coding approach, open coding followed by axial coding, or a start list of codes drawn from existing theory, is what separates a defensible qualitative methodology from an impressionistic summary of interesting quotes.

§ 05

Quantitative research methodology

A quantitative research methodology asks whether a relationship exists, how strong it is, or whether a difference between groups is larger than chance would produce, using numeric data analyzed statistically. It commonly pairs with an experimental, quasi-experimental, correlational or survey design, and analysis ranges from simple descriptive statistics (means, frequencies) through inferential tests (t-tests, ANOVA, regression) depending on the research question. Common software includes SPSS, R and Stata. A quantitative study's strength is that its findings can generalize to a wider population when the sample is drawn well; its tradeoff is that it usually cannot explain why a pattern exists as richly as an interview-based study can. A committee reading a quantitative methodology section checks that the analysis was chosen to fit the data and the question, a regression for a continuous outcome, a chi-square test for two categorical variables, rather than chosen because it was the test the writer already knew how to run.

§ 06

Mixed methods research methodology

Mixed methods research methodology combines qualitative and quantitative data inside one study, on the logic that numbers show a pattern's size while words explain its meaning, and that neither alone tells the full story.

A mixed methods study takes longer than a single-method study of the same size, since it runs two full data-collection and analysis cycles rather than one; committees expect a clear, stated reason for that added cost, not just an assumption that combining methods is automatically stronger.

§ 07

Sampling: who or what you study

Sampling decides who or what actually ends up in your study, and the two broad families answer different questions about how far your findings can travel.

Quantitative studies generally need probability sampling to justify generalizing to a wider population; qualitative studies more often use purposive sampling on purpose, since the goal is depth from participants who fit the question closely, not statistical representativeness. A quota sample, filling fixed numbers of participants within categories such as age or role without random selection inside each category, is a non-probability method that mimics stratified sampling, and it turns up often in business and market research, where full random sampling is rarely practical.

Sample size follows the same split: a quantitative study sets its size through a power analysis before data collection, calculating how many participants a design needs to detect a real effect of a given size, while a qualitative study stops adding participants at saturation, the point where a new interview stops surfacing a new theme. Neither number should be picked first and justified afterward; a committee can usually tell when a sample size was reverse-engineered from convenience rather than from either calculation.

§ 08

Validity, reliability and trustworthiness

Quantitative research methodology judges itself on validity and reliability. Validity asks whether a study measured what it claimed to measure, split further into internal validity (did the design rule out other explanations for the result), external validity (does the finding generalize beyond this sample) and construct validity (does the instrument actually capture the concept it names). Reliability asks whether the measurement is consistent, checked through test-retest reliability (the same result on a second administration) or internal consistency (items on a scale that measure the same thing move together, commonly reported as Cronbach's alpha).

Qualitative research uses a parallel framework built for non-numeric data, most influentially Lincoln and Guba's trustworthiness criteria: credibility (do the findings ring true, often checked against participants themselves through a member check), transferability (enough detail about the context that a reader can judge whether it applies elsewhere), dependability (would the process hold up if repeated by another researcher), and confirmability (do the conclusions trace back to the data rather than the researcher's own assumptions, often checked through an audit trail of coding decisions).

The two frameworks ask parallel questions in different vocabularies; using the quantitative terms for a qualitative study, or the reverse, is a common and avoidable mismatch a careful reader notices immediately.

§ 09

Ethics review

An ethics or Institutional Review Board (IRB) review checks that a study protects its participants: informed consent, the balance of risk against benefit, extra protections for a vulnerable population, and how collected data will be stored and kept confidential. Most reviews sort a proposed study into one of three levels: exempt (minimal risk, often secondary data or anonymous surveys), expedited (minimal risk but involving identifiable people), or full board (more than minimal risk, and some studies with a vulnerable population). Which level applies is the board's decision, made after it reads your actual protocol, not something a methodology section can assume in advance. At most institutions, recruitment cannot start until the board has approved the study or confirmed that it is exempt, so build that lead time into your timeline: a full-board review can take several weeks rather than days, plus any revisions the board requests before it signs off.

§ 10

How to write a methodology section or chapter

How to write a methodology comes down to naming and justifying each decision above, in a fixed order, so a reader can follow the logic from question to design to analysis without gaps. The chapter rarely stands alone: the literature review that justifies the method is what a strong methodology chapter argues from, pointing back to a specific gap rather than asserting a design choice from nowhere. A full proposal with its method section shows the whole argument built out, question through design, inside one complete composite example. A shorter assignment needs the same logic at a smaller scale: a methods section in a shorter research paper covers how that compresses into a page or two instead of a full chapter.

A methodology section that readers of a research paper can follow states every one of those six points once, in order, without circling back to add a missing piece halfway through the results.

The research methodology example below follows that same order. It describes an invented study, not a real one.

Notice the paragraph never claims more than the design supports: a correlational design earns a claim about a relationship, never a claim about cause, and the paragraph does not overreach into one.

§ 11

A methodology checklist

Run a finished methodology section or chapter against this list before it goes to a reader.

§ 12

When Chapter 3 is the bottleneck

A methodology chapter stalls for a specific reason more often than a vague one: an unjustified design choice, a sampling plan a committee will ask about, or an ethics section that does not yet name a review level. Dissertation support by milestone covers Chapter 3 alongside every stage around it. Chapter 3 written to your design takes your research question, your field and your preferred approach and returns a methodology chapter built to defend, not just to submit. Earlier in the process, a proposal with the method already argued covers chapters 1 through 3 together, so the design is settled before you are deep into data collection.