Use these guides before data collection to document the assumptions behind recruitment targets. The collection explains why calculators disagree, how repeated measurements affect power, and how to separate an analyzable target from the number that must be recruited.
A defensible sample size follows from the primary hypothesis, statistical model, effect-size definition, alpha, power, allocation, design effects, and expected loss to follow-up.
Different sample-size answers usually trace to a specific input, statistical method, or reporting convention. Learn how to compare them and choose a defensible result.
Read guide →Repeated-measures sample size depends on more than the number of groups. Correlation, sphericity, interaction targets, and dropout can materially change the result.
Read guide →Begin with the guide closest to your immediate decision. Record the assumptions, settings, and definitions you use, then open the linked DataStatPro workflow to apply the method. Educational examples should be adapted to the actual study protocol and data structure.
