Randomista

Randomista is the research companion for anyone who designs and runs randomised controlled trials — in social science, behavioural research, education, or field experiments. It walks each study through the five stages of a defensible trial: Design → Power → Randomise → Preregister → Run. DESIGN YOUR STUDY Define arms with allocation weights and a control condition, stratification factors, and primary and secondary outcomes. Once a study has a sealed sequence or a registered preregistration, the design locks automatically — no accidental edits to a live trial, ever. CALCULATE YOUR SAMPLE SIZE A built-in power calculator for two-sample t-tests, two proportions, and one-way ANOVA — validated against R's pwr package. Cohen benchmark chips, attrition inflation, a power-versus-N curve, and one tap to adopt the result as your recruitment target. RANDOMISE — SEALED AND REPRODUCIBLE Generate simple, permuted-block, or stratified allocation sequences from a reproducible seed. Every sequence is sealed at creation: parameters frozen as canonical JSON, protected by a SHA-256 content hash that is re-verified every time you open it. The audit card — seed, PRNG identifier, engine version, hash — is designed to be screenshotted straight into your paper's appendix. Anyone with your seed and settings can reproduce the exact sequence. ALLOCATE WITH CONCEALMENT The allocation table stays concealed by default to protect against selection bias. In the field, reveal the next assignment with one large button — each reveal is stamped once and can never be edited. Export full or concealed CSVs at any time. PREREGISTER BEFORE YOU COLLECT An AsPredicted-style nine-question form that fills itself from your design: outcomes, conditions, the power-analysis sentence, and the full randomisation audit line. Export OSF-ready Markdown or a print-styled PDF. Marking the study as registered freezes the answers and outcomes permanently — amendments happen the honest way, by duplicating into a new version. STAY ON SCHEDULE Per-study checklists with reminders, an RCT starter template (ethics, pilot, prereg freeze, launch, attrition check), and a cross-project dashboard showing every active study's stage and what's due in the next seven days. BUILT FOR RESEARCHERS • iPad-first, with full keyboard shortcuts (⌘N new study, ⌘1–5 stage navigation) • Everything stays on your device — no account, no tracking, no cloud • Statistical engines tested against independently computed reference values • One-time purchase — no subscription, no in-app purchases Randomista does not replace your statistician — it makes the rigorous path the easy one.