RESEARCH GUIDE / 03
The methods section explains exactly how evidence will be produced and analysed. It should be detailed enough for a reader to judge validity and, where appropriate, repeat the process without guessing what the researcher did.
Follow the reasoning
Start in order, but do not treat research as a straight line. Later evidence often sends you back to fix an earlier decision—and that is part of doing the work properly.
State the design by name, explain why it suits the objective, describe the study setting, and specify when data collection will occur.
Quality check: The setting helps readers judge whether findings may apply elsewhere.Explain the target and accessible populations, eligibility criteria, sampling frame, sampling method, sample-size calculation, and recruitment process.
Quality check: Report assumptions used in the sample-size calculation—not only the final number.Define outcomes, exposures, predictors, confounders, and derived variables. State units, categories, cut-offs, and scoring rules.
Quality check: A variable name is not an operational definition.Describe questionnaires, instruments, data sources, pilot testing, interviewer training, calibration, and the sequence used to collect data.
Quality check: Identify whether tools were adapted, translated, or previously validated.Explain coding, validation, missing-data checks, secure storage, descriptive analysis, models or tests, assumptions, effect measures, uncertainty, and software.
Quality check: Map each analysis directly to an objective and avoid choosing tests from the variable names alone.Cover ethics approval, consent, confidentiality, risks, benefits, permissions, data access, foreseeable biases, and how limitations will affect interpretation.
Quality check: Ethics is part of the method—not an administrative sentence added at the end.See it in context
Write the methods in the same order that the study will happen, then connect each objective to its variables, collection procedure, and analysis.
Analytical cross-sectional survey in named Kenyan university settings during a defined semester
Eligible registered students selected using a justified probability or clearly explained alternative method
Pilot-tested questionnaire with defined exposure questions and an appropriate sleep measure
Training, range checks, required fields, duplicate review, and a documented cleaning log
Descriptive summaries, unadjusted relationships, then a justified adjusted model with uncertainty reported
Approval, informed consent, minimal identifiers, controlled access, and a retention plan
Before you move on
Design, setting, population, and dates are unambiguous.
01Sampling and sample-size assumptions can be evaluated.
02Variables, tools, procedures, and quality controls are operationally defined.
03The analysis plan matches the objectives and reports uncertainty.
04Ethics, data protection, bias, and limitations are treated seriously.
05Good methods make research decisions visible; hidden decisions make findings difficult to trust.
Follow the next decision
A good question should shape the literature search; the evidence should sharpen the design; and the methods should make the result possible to understand and challenge.