Расчёт статистической мощности и размера выборки

research

power-analysis is a Claude Code skill that computes statistical power, required sample size, and minimum detectable effect (MDE) for a study design, then produces a registry-ready power section to paste directly into a preregistration. It handles two-arm RCTs with clustering (ICC, design effect), unequal allocation, and multiplicity corrections (Bonferroni, Holm), and switches to simulation-based power for non-standard designs such as DiD, staggered event-study, IV/2SLS, and panel data with serial correlation. The skill emits a reproducible script in R, Stata, or Python alongside an MDE table, power curves, and a methods paragraph formatted for AEA RCT Registry or IRB submission. Use it before launching a field or survey experiment to justify sample size, during R&R when a referee questions whether the study was adequately powered, or when /preregister needs a power section for an RCT.