Free guides
Research method guides
Plain-language guides for researchers: sample size, statistical power, design effect and ICC, choosing a study design and justifying your sample in a protocol. Every number checked.
- How to calculate the sample size for a surveyThe formula, a worked example (385 respondents for ±5 points) and how to adjust for non-response, clustering and a known prevalence.Read the guide
- Sample size for comparing two groups (proportions and means)How many participants a two-arm trial needs: worked examples for two proportions (294 per arm) and two means (142 per arm), and what changes with 90% power.Read the guide
- Statistical power explained, with a worked exampleWhat power means, why 80% is the usual minimum, and how to read it: 150 per arm gives only 52% power where 294 per arm gives 80%.Read the guide
- Design effect and the intra-cluster correlation (ICC)How clustering inflates sample size: design effect 1 + (m − 1) × ICC, with a worked trial where 294 per arm becomes 574 per arm.Read the guide
- How to choose a study designMatch your research question to a design (trial, cohort, case-control, cross-sectional, diagnostic accuracy, qualitative, review) and the reporting guideline to follow.Read the guide
- How to justify your sample size in a protocol or ethics applicationWhat reviewers expect in a sample size paragraph, with a model wording using real numbers, and the mistakes that get applications sent back.Read the guide
- Getting around ScintriaWhere everything is: the sidebar, the search shortcut, the main tools, where your saved work lives, and how to get help.Read the guide