Chapter Three is the chapter a defence panel can check. Every other chapter can be argued about; methodology is either internally consistent or it is not, and a panellist with a calculator can prove it in ninety seconds. This guide walks the nine sections most Nigerian departments expect, in order, with the actual sentences you would write in each one.
Before you start, confirm the section list your department issues. There is no national template. The National Universities Commission lays down minimum academic standards and accredits programmes, but it does not prescribe the internal structure of a methodology chapter, and Nigerian universities differ. The nine sections below cover the union of what University of Ibadan, University of Nigeria Nsukka, Ahmadu Bello University and Obafemi Awolowo University departments typically require. If your handbook lists seven, delete the two you do not need rather than inventing them.
Step 1: Write the chapter introduction (two or three sentences, no more)
Open by stating what the chapter covers, in the order it covers it. This is the shortest section in the chapter and students routinely inflate it into a page of definitions of “research”.
Worked example. “This chapter describes the methods used in carrying out the study. It covers the research design, the population and sample of the study, the sampling technique, the research instrument, the validity and reliability of the instrument, the method of data collection and the techniques used in analysing the data.”
That is the whole section. Do not define research methodology. Do not quote a dictionary.
Step 2: State the research design and justify it against your objectives
Name the design in the first sentence, then give one reason tied to your research objectives. The commonest failure here is naming a design and then justifying it with “because it is widely used”.
Worked example. “The study adopted a descriptive survey design. The design was considered appropriate because the study set out to describe the current state of a phenomenon across a defined population at a single point in time, rather than to manipulate any variable or establish causation.”
Three practical rules. First, if you administered a questionnaire once and did not manipulate anything, you ran a survey, not an experiment. Second, if you analysed documents, records or existing datasets, say so plainly and call it a documentary or secondary-data design. Third, if you combined two approaches, name the mixed-methods variant and say which strand came first, because a panel will ask.
Step 3: Define the population of the study with a number and a source
A population without a number is the single most common defect in a Nigerian Chapter Three, and it makes every subsequent section unverifiable. State who is in the population, where the boundary sits, and where the figure came from.
Worked example. “The population of the study comprised all 1,200 final year students of the Faculty of Management Sciences of the University, as obtained from the Faculty Examinations Officer in June 2026.”
Where your population is national rather than institutional, cite a published figure with its year attached. For higher education numbers, the National Universities Commission publishes its Nigerian University System Statistical Digest, and its 2019 edition is the most recent full national enrolment count the Commission itself publishes on its website. Whatever source you use, name it and date it in the sentence rather than in a footnote.
Step 4: Determine the sample size and show the arithmetic in full
Do not write “a sample of 300 was selected”. Write the formula, substitute your own numbers, and carry the arithmetic through to the answer. This is the section panels most reliably test.
The formula most commonly taught in Nigerian departments is the Taro Yamane formula, published in Yamane’s Statistics: An Introductory Analysis (1967):
n = N / (1 + N(e)2), where n is the sample size, N is the population and e is the level of precision, usually 0.05.
Worked example. “Applying the Taro Yamane formula at a 0.05 level of precision to a population of 1,200: n = 1,200 / (1 + 1,200 × 0.052) = 1,200 / (1 + 1,200 × 0.0025) = 1,200 / (1 + 3) = 1,200 / 4 = 300. A sample size of 300 respondents was therefore adopted.”
Two alternatives are equally acceptable if your supervisor prefers them. The Krejcie and Morgan (1970) table gives a determined sample size for a given population without any calculation, and is convenient where your population is awkwardly large. Cochran’s (1963) formula is used where the population is very large or effectively unknown. Whichever you choose, name it, give the year, and never present a sample size that is larger than the population you declared in Step 3.
Step 5: State the sampling technique honestly
Name the technique and describe the mechanism, in that order. The failure mode here is labelling convenience sampling as random sampling because random sounds more rigorous. A panel spots it instantly: if you cannot describe how every member of the population had a known chance of selection, it was not random.
Worked example (probability). “A simple random sampling technique was used. The names on the Faculty’s final year register were numbered serially from 1 to 1,200, and 300 numbers were drawn using a random number generator, without replacement.”
Worked example (non-probability, and perfectly defensible). “A purposive sampling technique was used. Respondents were selected on the basis of having completed at least one semester of industrial training, because the study’s second objective concerned the transfer of workplace skills.”
Where you used stratified sampling, state the strata and the proportional allocation per stratum. Where you used cluster sampling, state what the clusters were.
Step 6: Describe the instrument, section by section
State the instrument type, how many sections it has, how many items in total, and the response format of each section. If you adapted an existing instrument, say whose it was and what you changed.
Worked example. “Data were collected using a structured questionnaire titled ‘Workplace Skills Transfer Questionnaire’ (WSTQ), developed by the researcher. The instrument had three sections. Section A collected demographic data in six items. Section B contained fifteen items on skills acquisition, and Section C contained twelve items on skills application. Sections B and C used a four-point Likert scale of Strongly Agree, Agree, Disagree and Strongly Disagree.”
A four-point scale removes the neutral midpoint and is common in Nigerian departments; a five-point scale retains it. Either is defensible, but you must be able to say why you chose yours. Attach the full instrument as an appendix, because a panel that cannot see the instrument will assume it does not exist.
Step 7: Establish validity and reliability with the actual route you used
These are two different things and students frequently merge them. Validity is whether the instrument measures what it claims to measure. Reliability is whether it measures it consistently.
Worked example (validity). “The instrument was subjected to face and content validation. Copies of the draft instrument, together with the research objectives, were given to the project supervisor and two other lecturers in the Department, whose comments led to the rewording of four items and the removal of two ambiguous items.”
Worked example (reliability). “Reliability was established using the test-retest method. The instrument was administered to twenty students of a comparable faculty who were not part of the study sample, and re-administered after two weeks. The two sets of scores yielded a reliability coefficient of 0.81, which was considered adequate.”
Do not report a Cronbach’s alpha you did not compute. If you did compute one, be ready to say on what software, on how many items and on what pilot sample. A coefficient with no test behind it is the fastest route to a rewrite.
Step 8: Describe data collection, including the response rate
State how the instrument reached respondents, over what period, who administered it, and how many usable copies came back. The returned figure you state here must be the same number that appears in Chapter Four and in your abstract.
Worked example. “The researcher administered the questionnaire in person over a two-week period in June 2026, with the assistance of two trained research assistants. Of the 300 copies distributed, 284 were returned and 276 were correctly completed and usable, giving a response rate of 92 per cent.”
If you collected data online, say which platform and how the link was distributed. If you obtained permission or ethical clearance, state which body granted it and when, and put the letter in an appendix. Health, nursing and education departments increasingly require this.
Step 9: State the method of data analysis, tool by research question
Match each analytical tool to a specific research question or hypothesis rather than listing tools in the abstract.
Worked example. “Data were analysed using descriptive and inferential statistics with the aid of SPSS version 27. Research questions one and two were answered using frequencies, percentages and mean scores, with a criterion mean of 2.50 on the four-point scale. Hypotheses one and two were tested using chi-square at the 0.05 level of significance.”
Two things to get right. First, state your decision rule explicitly, meaning the criterion mean for descriptive items and the significance level for tests. Second, do not name a tool you never ran. If you eventually analysed with Excel, write Excel.
The three consistency checks to run before you submit the chapter
- Population is greater than or equal to sample. Check the two numbers against each other. This sounds obvious and it fails often.
- Sample equals distributed, and returned is less than or equal to distributed. The number you calculated in Step 4 should be the number you say you distributed in Step 8, and the returned figure must be smaller.
- Every research question has an analytical tool. If Chapter One posed four questions, Step 9 must account for four.
Run the same reconciliation across chapters as well. The sample size in Chapter Three, the number of analysed responses in Chapter Four and the figure quoted in your abstract are the three numbers a panel checks first, and they are the numbers most often found to disagree. The full inventory of what else gets tested is in the guide to the questions asked during a project defence.
How Chapter Three connects to the chapters around it
Chapter Three is not a standalone document. Its operational definitions must match the ones you set in Chapter One, and any instrument you adapted must be traceable to a study you actually reviewed in Chapter Two. Whether your department expects three, four or five chapters in total is itself an institutional decision rather than a national one, as set out in the guide to how many chapters a Nigerian final year project has.
Every named source in this chapter also has to be formatted in the style your department mandates, and Nigerian universities do not agree on one. Confirm yours against the guide to Nigerian referencing requirements before you build the reference list.
Draft Chapter Three without losing a week to the format
Tesify structures your methodology chapter section by section, keeps your citations attached to the sources you actually read, and lets you fill in your own design, population and instrument rather than starting from a blank page. Over 9,000 students have used it, and every word is still written by you.
Frequently asked questions
How long should Chapter Three be?
Most Nigerian undergraduate methodology chapters run between eight and fifteen pages. There is no national page rule; length follows from how many sections your department requires and how much your design needs explaining. A documentary study is legitimately shorter than a mixed-methods one.
Do I have to use the Taro Yamane formula?
No. It is the most commonly taught formula in Nigerian departments, but Krejcie and Morgan’s table and Cochran’s formula are equally acceptable. What is not acceptable is a sample size with no stated basis at all.
Can I write Chapter Three before collecting my data?
You can draft it, and most students do, because it forms the core of the research proposal. But rewrite it in the past tense once collection is complete, and correct any number that changed. A chapter still written in the future tense tells a panel you never went back to it.
What tense should Chapter Three be in?
Past tense throughout in the final submission, because you are reporting what you did. The proposal version is written in the future tense, and forgetting to convert it is one of the most common supervisor corrections.
What is the difference between population and target population?
The population is everyone the study concerns. The target population is the subset you can actually access and from which the sample is drawn. If your department asks for both, state the accessible boundary explicitly rather than repeating the same figure twice.
Is secondary data acceptable for a final year project?
Yes, in many departments, and it is standard in economics, accounting and some management fields. State the source, the period covered, the unit of observation and any gaps in the series, and treat the source’s own methodology limitations as your limitations.
Do I need ethical clearance?
It depends on your department and your respondents. Studies involving patients, minors, clinical records or identifiable personal data usually require clearance from an institutional committee. Ask your supervisor early, because clearance can take weeks and cannot be obtained retrospectively.
What if my response rate was poor?
Report it honestly and discuss it as a limitation. A 60 per cent response rate that is stated and discussed is far more defensible than a 100 per cent rate that a panel does not believe.
Should the instrument go in the chapter or in an appendix?
Describe it in the chapter and attach the full copy as an appendix. Reproducing all thirty-three items inside Chapter Three interrupts the argument and inflates the chapter without adding anything.
Can I change my methodology after my proposal was approved?
Yes, with your supervisor’s agreement, and it happens constantly when field access turns out to be harder than planned. Document what changed and why, and make sure every downstream section reflects the change rather than the original plan.