<skill_content>
<overview> Power analysis determines the sample size required to detect treatment effects of a given magnitude with specified probability. Adequate power prevents Type II errors (failing to detect real effects). Power calculations must account for clustering (design effects), non-compliance (reduces effective sample), attrition (reduces final sample), and multiple testing (inflates α). Underpowered studies waste resources and risk being misinterpreted as evidence of no effect.[Description truncada. Veja o README completo no GitHub.]