The methodology chapter explains what you did and why it was the right thing to do, in enough detail that another researcher could repeat it. Work through the eight steps below in order. The single thing examiners test hardest is internal consistency: every component must follow from your research questions and agree with every other component.
Before you start, open your faculty’s FYP or thesis guidelines. Malaysian universities differ on required subsections and headings, and the handbook overrides any general advice, including this article.
Step 1: Restate the research questions at the top
Begin the chapter by repeating your research questions verbatim from Chapter 1. This is not filler. It anchors everything that follows and lets you check each later decision against the question it serves.
If you find yourself unable to connect a section of this chapter back to one of those questions, that section does not belong.
Expected output: the questions restated word for word, unchanged from Chapter 1.
Step 2: State the research design and justify it
Name your approach and your specific design, then explain why it fits. Naming alone is not enough; the justification is what earns the marks.
Model sentence you can adapt:
“This study adopts a quantitative approach using a cross-sectional survey design. This design was selected because the research questions require measuring the strength of the relationship between service quality and customer loyalty across a defined population at a single point in time. A qualitative design was not adopted because the objective is to test hypotheses derived from existing theory rather than to explore meaning.”
Note the final sentence. Explaining why the alternative was not chosen is the element most FYP chapters omit, and adding it changes how the choice reads: as a methodological decision rather than a default.
Step 3: Define the population and the sampling
Population and sample are different things and must be written separately.
- Define the population precisely: who, where, and during what period. “Employees of manufacturing firms in Selangor” is too loose. “Full-time production staff at [Company] in 2026, totalling 320 people” is defensible.
- State where the population figure came from, including the source and its date. An unsourced population number is a standing invitation for a viva question.
- State your sample size and how it was determined, showing the calculation rather than only the result.
- Name the sampling technique and make sure it is consistent with the sample size method you used.
That last point is where chapters quietly contradict themselves. A sample size formula that assumes random selection sits badly beside a procedure describing whoever happened to be available. If circumstances force convenience sampling, say so plainly and record it as a limitation.
Step 4: Describe the instrument
Explain exactly how each variable was measured. For a questionnaire study, cover:
- The structure, section by section, and what each section measures.
- The source of each set of items, whether adapted from a published instrument or newly developed.
- The response scale used for each construct.
- Any modifications you made to adapted items, and why.
If you adapted a published instrument, cite it properly and state what you changed. Adapted items must be re-tested for validity and reliability, because results reported for the original version do not automatically transfer to your modified version or to a different population.
For a build-and-test project, this section describes the development method, the tools used, and the criteria against which the system will be evaluated.
Step 5: Set out the data collection procedure
Write this as a chronological sequence a reader could follow, including approvals.
- Ethics or institutional approval, with the process named.
- Permission obtained from the organisation, school or clinic involved.
- Pilot test, including how many respondents took part and what changed as a result.
- Main data collection: how the instrument was distributed, over what period, and how responses were returned.
- Response rate: how many were distributed, how many returned, and how many were usable.
Start the approval process early. Research involving human participants, personal data or clinical settings typically needs review at Malaysian universities, approval cannot be granted retrospectively, and waiting for it is one of the most common causes of a delayed FYP.
Step 6: State the data analysis plan
Say which analysis answers which research question. Mapping them explicitly is the fastest way to demonstrate that the chapter holds together.
Model sentence:
“Research question one is answered using descriptive statistics reporting means and standard deviations for each construct. Research question two is answered using multiple linear regression, with customer loyalty as the dependent variable and service quality and price fairness as predictors. Assumption testing for normality, linearity and multicollinearity was conducted before interpreting the regression output.”
If you are unsure which test matches your question, work through the decision guide on which statistical test to use for your FYP data before writing this section. Deciding here rather than after data collection saves a rewrite.
Step 7: Address validity and reliability
For quantitative work, report the pilot results for each construct and state the criteria you applied. Run item validity first, remove items that fail, then recompute reliability on the retained items. Computing reliability while keeping items you have already declared invalid is a frequent and easily spotted error.
For qualitative work the equivalent section addresses trustworthiness, and you should name the specific strategies you used, such as triangulation across data sources, member checking with participants, an audit trail of decisions, or peer debriefing. Naming the strategies is not enough on its own; describe how each was actually carried out in your study.
Step 8: State the limitations honestly
Every design has limits. Naming them yourself is a strength, because the alternative is an examiner naming them for you while you appear not to have noticed.
Useful limitations to consider: a sample drawn from a single organisation restricting generalisability, a cross-sectional design preventing causal claims, self-reported data carrying response bias, and a response rate below the calculated target.
Write each as a specific consequence rather than a vague apology. “Because data were collected from one company, the findings may not generalise to other firms in the sector” is useful. “There were some limitations in this study” is not.
The consistency check before you submit the chapter
Read the chapter once with a single question in mind: does every element follow from the research questions and agree with every other element? Specifically check that:
- The design matches the verbs in your research questions.
- The sampling technique matches the sample size method.
- The instrument measures every variable named in your framework, and nothing extra.
- The analysis plan covers every research question, with none left unanswered.
- The number of questionnaire items is compatible with the analysis you intend to run.
Most returned methodology chapters fail on one of these five, not on the quality of the writing. The depth of justification expected does rise with the level of your degree, as set out in the comparison of FYP, thesis and dissertation in Malaysia.
Writing a chapter that holds together
Methodology is the most structured chapter in the document, which is exactly why inconsistencies stand out so clearly. Most students know what they did; the difficulty is setting it down in an order where each part visibly follows from the last.
Tesify helps you build the chapter section by section, keeping design, population, instrument and analysis aligned as you draft, with you remaining the author and responsible for every methodological decision. More than 9,000 students have used it to produce over 15,000 chapters.
Build your methodology chapter in Tesify
Frequently asked questions
How long should the methodology chapter be?
There is no national standard, and faculties differ. Completeness matters more than length: every component present, every choice justified, and enough detail that another researcher could replicate the study. Check your handbook for any stated limit.
Should the chapter be written in past or present tense?
Write the proposal version in future or present tense, then convert to past tense once the work is done, because you are reporting what you actually did. Many faculties specify this, so check the handbook, and above all be consistent throughout.
Do I need a research philosophy section?
Some faculties require a short statement of philosophical position, others consider it unnecessary at undergraduate level. Follow your handbook. Where it is required, keep it brief and connect it to your design rather than reproducing textbook definitions.
How many respondents should the pilot test have?
A commonly used working figure in Malaysian faculties is around 30 respondents with characteristics similar to your sample but excluded from the main study. This is convention rather than a national rule, so confirm against your own guidelines.
Can I write the methodology chapter before collecting data?
Yes, and you should. It is your plan of work, so most of it can be drafted before data collection begins. That is why it is the most efficient chapter to write early while waiting for approvals or supervisor feedback on other sections.
What if my actual procedure differed from the plan?
Report what actually happened, not the plan. If the response rate fell short or the timeline shifted, describe the reality and discuss the consequences in your limitations. Examiners handle honest deviation routinely; a chapter that describes a procedure the data could not have come from is a far worse problem.
Do qualitative studies need a sample size calculation?
No. Qualitative sampling is governed by data saturation, the point at which further interviews stop producing new themes, together with clear selection criteria for participants. Explain both rather than presenting a formula.
Should I include the questionnaire in this chapter?
Describe its structure in the chapter and place the full instrument in an appendix. Most faculties require the complete instrument to be appended, so check the handbook for the expected format.
How do I justify convenience sampling?
State plainly why probability sampling was not feasible, such as no accessible sampling frame or restricted access to the population. Then acknowledge the effect on generalisability in your limitations. Honest justification is accepted far more readily than a technique described inaccurately.
What is the most common reason methodology chapters are returned?
Internal inconsistency. The research question asks about a process while the analysis compares groups, or the sampling technique contradicts the sample size formula. Running the five-point consistency check above before submitting catches nearly all of these.