Your supervisor hands back the methodology chapter with a single comment in red: "You describe what you did, but you never explain why." This is the most common reason methodology chapters get sent back for revision. The chapter is not a lab diary. It is a written argument for why your research design, your sample, your instruments, and your analysis plan were the right choices for answering your research questions. In roughly 70% of the thesis projects we review, the first draft reads like a chronological log of activities rather than a justified set of decisions. This guide covers the six sections your methodology chapter needs, what belongs in each one, and where students lose marks.
Research Design
Open the chapter by naming your research design in precise terms. "Quantitative" alone is not enough. State whether the study is cross-sectional or longitudinal, and whether the design is experimental, quasi-experimental, or correlational. A thesis that compares test scores between a training group and a control group with random assignment is a true experimental design. A thesis that compares existing groups (for example, nurses in public hospitals versus private hospitals) without random assignment is quasi-experimental. A thesis that examines relationships between anxiety, sleep quality, and academic performance without manipulating any variable is correlational. Each label carries different implications for causal claims, and supervisors expect you to know the difference.
After naming the design, justify it. Why is a cross-sectional correlational design appropriate for your research questions? Typically because your variables cannot be experimentally manipulated (you cannot randomly assign people to have high anxiety), and because you are interested in relationships rather than causal effects. Cite a methodology textbook here. Creswell and Creswell (2018, Research Design: Qualitative, Quantitative, and Mixed Methods Approaches, 5th edition) is the standard reference, though your department may prefer Bryman (2016) or Saunders, Lewis, and Thornhill (2019). One citation is sufficient. The point is to show that your design label comes from an established framework, not from your own invention.
Population and Sampling
Define the population first, then explain how you sampled from it. The population is the full group your findings should generalize to: "undergraduate nursing students enrolled in accredited programs in the UK" or "employees of medium-sized IT companies in Germany with at least one year of tenure." Be specific. A vague population definition ("young adults") makes your sampling strategy impossible to evaluate.
Name your sampling method. Most student theses use convenience sampling or purposive sampling because the researcher distributes questionnaires to people they can access. That is fine, but you must state it openly and acknowledge the limitation. If you used stratified random sampling or systematic sampling, explain the strata or the interval. A thesis that claims "random sampling" but recruited participants through a Facebook post did not use random sampling. Supervisors catch this regularly.
Justify your sample size with a power analysis. This is where many students either skip the justification entirely or write a vague sentence about "adequate" sample sizes. Run a calculation in G*Power (Faul et al., 2009) before collecting data. For a medium effect size (f² = .15) with three predictors in a multiple regression, an alpha of .05, and power of .80, G*Power returns a minimum sample of 77 participants. For an independent-samples t-test detecting a medium effect (d = 0.50) with .80 power, you need 64 per group, or 128 total. State these numbers, report the exact G*Power parameters, and explain that your final sample of, say, 143 participants exceeds the minimum requirement. For a full walkthrough of power analysis calculations, see our sample size and power analysis guide.
Instruments and Measures
Describe every questionnaire, scale, or measurement tool you used. For each one, state who developed it, how many items it contains, what response format it uses, and what it measures. A typical entry might read: "Perceived stress was measured using the Perceived Stress Scale (PSS-10; Cohen, Kamarck, & Mermelstein, 1983). The scale contains 10 items rated on a 5-point Likert scale from 0 (never) to 4 (very often). Total scores range from 0 to 40, with higher scores indicating greater perceived stress."
Report reliability from two sources. First, the Cronbach's alpha from the original validation study or a widely cited adaptation. The PSS-10, for instance, reported α = .78 in the original 1983 study and values between .82 and .89 in subsequent validations. Second, report the alpha you obtained from your own data. If your sample produced α = .84 across the 10 items, state that explicitly. This dual reporting shows the reader that the instrument has established reliability and that it also performed reliably in your specific sample. For a detailed explanation of how to calculate and interpret this coefficient, see our Cronbach's alpha guide.
If you used a translated or adapted version of a scale, mention who translated it, whether back-translation was performed, and whether the adapted version has its own validation study. A scale validated in English does not automatically work in Turkish or Portuguese. Supervisors in non-English-speaking countries pay close attention to this point.
Data Collection Procedure
Write this section as a sequence of steps, but keep the focus on justification rather than narration. Start with ethical approval: state the name of the ethics board or institutional review board (IRB) that approved the study, the approval date, and the reference number. If your university does not have a formal IRB and instead uses a departmental ethics committee, name that committee.
Describe how you distributed the instrument. Was it an online survey via Google Forms, Qualtrics, or SurveyMonkey? Was it a paper questionnaire administered in a classroom? Did participants complete it individually or in groups? State the time period: "Data were collected between March 15 and April 28, 2025." Report the response rate if applicable. If you sent 400 email invitations and received 187 completed responses, your response rate is 46.8%. If you distributed paper questionnaires in person during a class session, your response rate is probably close to 100%, but you should still report it.
Do not write "first I contacted the dean, then I waited two weeks for approval, then I printed the surveys at the copy shop." That is a diary. Write "the researcher obtained permission from the faculty dean before distributing questionnaires" and move on. The chapter cares about what you did and why, not about the waiting, the printing logistics, or the weather on the day you collected data.
Data Analysis Plan
This section maps each research question or hypothesis to a specific statistical test and explains why that test fits. It is the part of the methodology that most directly connects to your results chapter. Write it in a structured way so the reader can trace any result back to its justification.
For each hypothesis, state the independent variable, the dependent variable, the measurement level of each, and the test you will run. A hypothesis like "There is a significant difference in academic performance between male and female students" involves one categorical independent variable (gender, two groups) and one continuous dependent variable (GPA). The appropriate test is an independent-samples t-test if the assumptions of normality and equal variances are met, or the Mann-Whitney U test if they are not. State this logic. If you are unsure which test fits your variables, our guide to choosing a statistical test walks through the decision tree. You can also run your data through Academic Stats Agent, which automatically selects the correct test based on your variable types and assumption checks.
The table below shows what a well-structured analysis plan looks like. Adapt it to match your own research questions and variables.
| Research Question | Variables | Statistical Test | Justification |
|---|---|---|---|
| RQ1: Is there a difference in job satisfaction between public and private sector employees? | IV: Sector (categorical, 2 groups); DV: Job satisfaction (continuous) | Independent-samples t-test | Compares means of one continuous DV across two independent groups |
| RQ2: Does job satisfaction differ across three educational levels? | IV: Education level (categorical, 3 groups); DV: Job satisfaction (continuous) | One-way ANOVA with Tukey HSD post hoc | Compares means across three or more independent groups; post hoc identifies which pairs differ |
| RQ3: Is there a relationship between years of experience and job satisfaction? | Variable 1: Years of experience (continuous); Variable 2: Job satisfaction (continuous) | Pearson correlation | Tests linear association between two continuous variables; assumes bivariate normality |
| RQ4: Do age, experience, and education predict job satisfaction? | Predictors: Age, experience, education (continuous/dummy); DV: Job satisfaction (continuous) | Multiple linear regression | Tests whether multiple predictors explain variance in a continuous outcome; reports R² and individual β coefficients |
State the significance level you will use (typically α = .05) and name the software. "All analyses were conducted using IBM SPSS Statistics version 29" or "using R version 4.3.2" is sufficient. Mention how you will handle assumption violations: if normality is not met, you will switch to the nonparametric equivalent; if multicollinearity appears in regression (VIF > 5), you will remove or combine the affected predictors. These conditional plans show that you thought about what could go wrong before it happened.
Ethical Considerations
This section is often the shortest, but omitting it entirely will raise questions. Cover four points. First, state that the study received approval from your institution's ethics committee or IRB, and include the approval number. Second, describe informed consent: participants received a written explanation of the study's purpose, were told that participation was voluntary, and signed (or clicked) a consent form before starting. Third, explain how anonymity or confidentiality was maintained. If you collected online data without recording IP addresses or names, say so. If the study was not anonymous (for example, you matched pre-test and post-test scores using student IDs), explain how data was de-identified after matching. Fourth, describe data storage: "Responses were stored on a password-protected university server and will be deleted after five years, in accordance with institutional policy."
If your study involved vulnerable populations (minors, patients, prisoners), describe any additional safeguards. For studies involving children, note that parental consent and child assent were both obtained. These details take only a paragraph or two, but their absence can stall your ethics approval and, later, your thesis defense.
The Biggest Mistake: Writing a Methods Diary
The single error we see most often across the 300+ projects we have reviewed is treating the methodology chapter as a timeline. Students write: "First, I reviewed the literature. Then I chose my questionnaire. Next, I went to the university office to get approval. After that, I printed 200 copies..." This reads like a personal account rather than a research argument.
The methodology chapter should read more like a legal brief than a journal entry. Every paragraph answers "why," not just "what." Why this design and not another? Why this sampling method? Why this particular instrument rather than one of the four other scales that measure the same construct? Why a sample of 150 and not 80 or 300? The student who writes "I used the Maslach Burnout Inventory because it was available online" will be asked to revise. The student who writes "The Maslach Burnout Inventory (Maslach, Jackson, & Leiter, 1996) was selected because it is the most widely validated measure of occupational burnout, with over 6,000 citations and documented reliability across healthcare populations (α = .88 to .90)" will pass review on the first attempt.
Write in third person and past tense throughout: "A cross-sectional correlational design was adopted" rather than "I decided to use a correlational design." Some universities now accept first person, but third person remains the safer default unless your supervisor explicitly approves otherwise.
Key takeaway: The methodology chapter is an argument, not a diary. Structure it in six sections: research design (named and cited), population and sampling (with a power analysis from G*Power), instruments (with dual Cronbach's alpha reporting), data collection procedure (steps with justification, not chronological narration), data analysis plan (a table mapping each research question to its variables and test), and ethical considerations (IRB approval, consent, anonymity, storage). Every decision needs a "why." If you can replace a sentence in your methodology with a sentence from someone else's thesis and it still works, you have not been specific enough.