A methodology chapter explains every choice you made in the study and the reason for it, in enough detail that another researcher could repeat it.
Scholars mostly forget the second part, about repeating the study. In many drafts we get, the choices are written correctly. Stratified random sampling, structured questionnaire, SPSS, all are there. The reason for each choice is not there. In the viva, examiners ask about the reason only. So this guide follows the questions you will face there. Answer them in writing now and the chapter will hold when the examiner starts asking.
“Why did you choose this design?”
Begin the chapter with your research design. Then connect it to your objectives. I will use one made-up topic to show this, adoption of drip irrigation by small farmers in Dharmapuri district.
Weak: “The present study is descriptive and analytical in nature.”
Better: “The study uses a cross-sectional survey design. The first two objectives describe the level of drip irrigation adoption and compare it across farm-size groups, which a survey of a representative sample can answer. A field experiment was not feasible, as the researcher could not assign irrigation methods to farmers.”
See what the second version is doing. The design gets a name first. After that it is linked with the objectives. And it also says why the other possible design, the experiment, was not taken. Before this you don’t have to write one full page of textbook definitions for “descriptive research”. If your guide insists on a definition, put one sentence, and then come to the reason.
| Part of the chapter | What to state | The reason an examiner wants |
|---|---|---|
| Research design | Descriptive, analytical, experimental, case study or mixed methods | How this design fits your objectives, and why you did not take the other design that looks obvious |
| Area and population | The place, the group, and the total number | Why you picked this place and this group, and the source for the population figure |
| Sample size | The final number and the working behind it | Which formula or table you used, with confidence level and margin of error |
| Sampling method | Random, stratified, multistage, purposive or snowball | Why it suits your population, and how the units were really selected in the field |
| Tool | Questionnaire, interview schedule, test, instrument or dataset | Source of the items, and the checks you did for validity and reliability |
| Pilot study | Number of respondents, month, and what you changed | One or two actual changes the pilot made in the tool |
| Data collection | Dates, mode, and the person who collected | Anything that could bias the answers, and the step you took for it |
| Analysis plan | Software and tests, written for each objective | Why that test suits that objective and that type of data |
| Ethics | Consent, approvals, anonymity and data storage | Approval number and date from the committee, where approval was needed |
“Who did you study, and how many?”
This is where examiners spend most of time in the methodology chapter. Any number you write here must have a source, and any choice must have a reason.
Write the population as a number with its source. For example, “There are 41,250 small and marginal farmers in the five blocks studied (Agriculture Department, 2024)”. That is much stronger than writing “farmers in Dharmapuri district”. Sometimes there is no official figure at all. Then write that plainly, and give the method you used to reach your own estimate.
Next comes the sample size. Writing only the final number will not do. Mention the formula or table, the confidence level and the margin of error. Say you used Cochran’s formula at 95% confidence and a 5% margin. Then show the calculation in two lines. Many times the final sample is bigger than what the formula gave. Then the reason should be written down also. In our experience it is mostly for the half-filled questionnaires that come back. Sometimes one stratum had very few cases and you took some more there.
Describe your sampling the way it actually happened in the field, and not the way the textbook writes it. If you write “multistage random sampling”, give each stage. First how the blocks were chosen, then the villages, then the households. Suppose one village refused, or a block had to be replaced after the floods. Put that in the chapter also. When an examiner reads these small field problems, he believes the rest of chapter more. A neat textbook version falls apart once they start asking one or two questions on it.
“Where did your questionnaire come from?”
When you have built the questionnaire yourself, the chapter must tell where each section came from. Mostly scholars take items from earlier researches and adapt them for their own questionnaire, and nobody objects to that. Only, name the source of every scale and mention what you changed in it.
The next paragraph is about checking the tool. Content validity is usually judged by experts. Give how many experts there were and their fields. Reliability is usually given as Cronbach’s alpha for each scale separately. This can come from the pilot data, the main data, or both. Most guides look for 0.70 or above. For a new scale with only few items, some guides will accept a little lower value.
Some scholars translate the questionnaire into Tamil, Telugu, Kannada or Malayalam. In that case write who did the translation and whether a back-translation was done. Committees in South Indian universities ask this question many times, because most field surveys here are done in the local language and not in English.
“What did the pilot study change?”
In Indian survey-based theses a pilot of around 30 respondents is common. That number is only a convention and no rule fixes it. Still, many chapters give the pilot study just one line. “A pilot study was conducted with 30 respondents and the tool was found reliable.”
Writing it like that throws away the pilot. What you should report is the change the pilot study brought. Take our drip irrigation example. In the pilot the farmers understood “subsidy” as the PM-KISAN payment only, so two items were reworded. One section was cut, and the interview time came down from 40 minutes to 25. Alpha for the “perceived cost” scale went up from 0.61 to 0.78 when one of the item was removed. Changes like these show the examiner the pilot actually did some work.
“Why this test for this objective?”
Take each objective one by one and write the analysis plan for it. Many scholars put this in a small table. Objective goes in the first column, then the variables, then the test, and in the last column the assumption checked. Later in the results chapter, every test will already have its reason sitting in the methodology chapter. If any test has no reason written here, examiner will surely ask you for one.
Not sure which test suits which objective? Our guide on choosing a statistical test matches the common types of objectives with their tests. In a qualitative study the approach has to be named. It may be thematic analysis or content analysis, or grounded theory in some cases. Mention software if you used any, for example NVivo or ATLAS.ti. And write how your codes were grouped into themes.
Write the software name with its version also, like SPSS 26, R 4.3, AMOS 24 or SmartPLS 4. It looks like a small thing. Still, one version of a software can give slightly different results from another. Anybody checking your results will look in this chapter to know which version you ran.
“Did you get permission?”
When people took part in your study, the chapter needs an ethics section, even if it is short. First mention how consent was taken from respondents, on a signed form or orally. Write how respondents were kept anonymous and where the data are kept now. Your university or hospital may have asked for Institutional Ethics Committee approval. Then give the name of the committee, the approval number and the date.
Clinical and health studies in India usually need the ethics committee to approve first, and only then the data collection can start. Examiners will ask to see the approval letter. Getting approval after the data collection is over is a serious problem. Any small doubt about approval for your study, go and ask your guide today itself. Don’t keep it for the pre-submission stage.
Gaps we see most often in draft methodology chapters
In the methodology chapters scholars send us for checking, the same gaps keep coming.
- Sample size is given, but the working is missing.
- “Simple random sampling” is written in the chapter, but the scholar had no sampling frame, so no random pick was possible.
- Reliability for the full questionnaire, with nothing for each scale.
- Some variables appear in the objectives and never appear in the questionnaire.
- Several pages copied from textbooks on what is “research” and what is “methodology”. Only after all that the actual study is mentioned.
- Future tense still sitting there from the synopsis, like “The researcher will collect data…”. In the final thesis, write what you did in past tense.
Our research methodology support goes through a draft for these same points. It helps most before data collection starts, because at that stage the fixing costs very little. If the objectives themselves are not clear, first read the guide on research objectives and hypotheses. The methodology chapter can be only as clear as the objectives behind it.
Questions scholars ask
How long is a methodology chapter in a PhD thesis?
Mostly between 15 and 30 pages. It depends on your field and how many methods you used. For example, a mixed-methods study takes more pages than one survey. Nobody in the committee is going to count your pages. What they check is whether each choice has its reason written.
Should I write “research methodology” or “research methods” as the title?
See what recent theses from your department used, or ask your guide. In Indian theses “Research Methodology” is the title you see more. A few universities fix the chapter titles in their format rules, so check those rules also.
Can I change the method after the synopsis was approved?
Small changes are usually settled with your guide, for example taking a slightly bigger sample, and you explain them in the methodology chapter. A big change is different. If you move from a survey to an experiment, the doctoral committee normally has to agree to it in writing.
Do I need to explain limitations in the methodology chapter?
Yes, but briefly, and only the limits that come from the method. A sample taken from one district, or data that respondents reported themselves, are examples. The full discussion of limitations usually goes in the discussion or conclusion chapter.