For students in Nigeria, Ghana or Kenya with a draft methodology chapter in a project or thesis. The page goes through each part of the chapter, explains what the Taro Yamane formula assumes, and shows how each part answers to Chapter One. There is no sample chapter here.
From within your field, Esmond points out where your findings claim more than your method can show and where your design choices contradict each other.
Chapter Three tells the reader how the data that answer each research question were obtained, from whom, with what instrument, and how they were analyzed. Laikipia University's graduate school guidelines put it in one sentence: “This section must clearly and logically flow from the literature and hypotheses.”
The UNIZIK Manual of Style lists the subheadings from research design to method of data analysis. Bolgatanga Technical University adds the research approach (qualitative, quantitative or mixed) and pre-testing; JKUAT's School of Public Health and Laikipia University add ethical considerations. In the 2023 Emilio Aguinaldo College manual in the Philippines, the methodology is Chapter 2 and includes the research locale.
Tense changes between proposal and project. JKUAT's School of Public Health describes the methodology chapter of the final thesis this way: “This chapter includes all what was in the proposal written in past tense”. For the general difference between method and methodology, see the thesis methodology chapter; for the other chapters, thesis chapter feedback.
A hypothetical example of a commentYour population is 1,200 staff and the Taro Yamane formula at 5% gives 300, but your data collection section says 250 questionnaires were distributed. Which number did you analyze, and how does it relate to the 300?
A first comment on one part of your own draft, right away and without an account.
Get a free previewLaikipia University's guidelines note that the wording of the objectives determines the type of research design. An objective about the "effect" of one variable on another needs a design that can support a causal claim, such as an experiment or quasi-experiment; a one-time descriptive survey shows levels and associations. Name the design and the reason it suits these objectives.
The UNIZIK manual describes the population as the totality of the people or things within the study area. State it with a number and its source, such as a staff list, and keep it consistent with the scope in Chapter One.
The formula credited to Taro Yamane (1967) gives a sample size n from a population size N and a chosen margin of error e: n = N / (1 + N × e²). With N = 1,200 and e = 0.05: 1,200 × 0.0025 = 3, so n = 1,200 / 4 = 300.
A fact sheet by Glenn D. Israel for UF/IFAS Extension at the University of Florida (US), which sets out this formula, states its assumptions: a 95% confidence level and a proportion of P = .5, the value of maximum variability. It adds that its approaches assume a simple random sample, and that more complex designs, such as stratified samples, must take into account the variances of subpopulations, strata or clusters. The formula sizes a sample for estimating a proportion within the margin of error. Whether the sample can detect a given difference or relationship is a power question, with its own calculation.
Israel adds: “Finally, the sample size formulas provide the number of responses that need to be obtained.” And: “Many researchers commonly add 10% to the sample size to compensate for persons that the researcher is unable to contact.” So if 300 responses are needed, more than 300 questionnaires go out, and the chapter reports how many came back and were usable.
The formula sets a size. It doesn't make a sample random. Three hundred respondents picked because they were available is still a convenience sample, and the chapter has to say so.
The UNIZIK manual asks for a full description of the sample and the sampling technique; Bolgatanga Technical University asks for the rationale for both the selection procedure and the sample size. A reader looks for the steps: the sampling frame, how units were selected, and how many came from each stratum. With proportional allocation, each stratum's share of the sample equals its share of the population.
In a qualitative design, participants are chosen for what they can tell you, through purposive or snowball sampling, and the number is argued from the aim and the depth of the data, for example with reference to saturation.
Describe the instrument part by part: sections, number of items per variable, response scale and how responses were weighted. Kumasi Technical University asks that questionnaire items come from the main concepts in the specific objectives and research questions.
For validity, the UNIZIK manual asks for the procedure used to validate the instrument, with a validity index where applicable. A reader then wants to know who reviewed the items, against what, and what changed. For reliability, the same manual asks for both the method and the result: “The candidate shall describe the procedure for ascertaining the reliability of the instrument as well as the index of the reliability.” Bolgatanga Technical University asks for pre-testing when the instrument is new or adapted, with the reliability coefficient where needed.
Cronbach's alpha is computed for a set of items that measure one construct, so a questionnaire with four scales has four coefficients. Report the size and source of the pilot sample. A coefficient from the study where the instrument was first used describes that study's respondents. JKUAT's School of Public Health asks students who use a standardized test to quote its existing validity and reliability levels; if you do, name the study they come from.
Bolgatanga Technical University is specific: “Indicate and justify the statistical tools used for each hypothesis/research question/objective.” The University of Ilorin's Postgraduate School thesis evaluation form for PhD and MD theses lists “Logically derived from Objectives” as a criterion for the methodology.
A sentence per research question and per hypothesis. A question on the level of a variable can be answered with mean and standard deviation; if you use a decision rule such as a mean of 2.50 on a four-point scale, state it and give its basis. A hypothesis about a difference between two groups points to a t-test, three or more groups to analysis of variance, an association between categorical variables to chi-square, and a relationship between two continuous measures to correlation or regression, provided the data meet each test's conditions. Name the significance level (for example 0.05) and the software. For qualitative data, describe the steps of coding and how themes were formed.
The procedure for data collection says who administered the instrument, when, how it was returned, and how many usable responses came back. Laikipia University's guidelines for theses and projects name what had to be in place before fieldwork: introductory letters from the graduate school, the institutional ethics review certificate, and the NACOSTI research permit. Kumasi Technical University requires ethics approval from its research ethics committee for work with human participants or animals. Report consent, anonymity and data storage, with the approval reference.
These fragments are made up. None of them comes from a real thesis.
Esmond evaluates every assignment before you submit. It never writes it.
Esmond is an AI tool that doesn't write for you. If your course doesn't allow any AI tools, don't use Esmond for that work. If you're unsure, ask your instructor or advisor.
For a methodology chapter, Esmond points out what a supervisor would question in the text you made: a design that can't carry the claims in the objectives, sample figures that differ between sections, a sample size formula with no inputs, a reliability claim with no coefficient, a research question with no analysis assigned to it. See the sample report for the format and how Esmond works for the steps.
Esmond reads what you upload. If the file holds the whole draft, it can set this chapter against the others; if it holds one chapter, the review stays within that chapter and may raise a point that another chapter already answers.
Esmond doesn't rerun your sample size calculation or your statistics, and doesn't check your ethics approval or your department's format. You decide what to change.
In the UNIZIK Manual of Style: research design, study area, population, sample and sampling technique, instrument, validity, reliability, method of data collection and method of data analysis, most of them where applicable. Other guides quoted here add the research approach, pre-testing and ethical considerations. Your department's format decides.
Both appear as n = N / (1 + N × e²), so they produce the same number. Whichever name you use, state the assumptions: an estimate of a proportion, 95% confidence, P = .5 and simple random selection.
It doesn't fit. The formula sizes a random sample for estimating a proportion. In qualitative work, participants are selected for the information they hold, and the number is argued from the aim and the depth of the data.
In a proposal the methods are still a plan; in the final project they have been carried out. JKUAT's School of Public Health asks for past tense in the final thesis.
Read on November 8, 2026.
Get your paper evaluated before you submit