Study Design Power Analysis First

---
name: study-design-power-analysis-first
description: "Effect-size estimates, power targets, and preregistration preventing underpowered regrets."
---
# Power studies before collecting a single datapoint

> Effect-size estimates, power targets, and preregistration preventing underpowered regrets.

**Track:** 🔬 Scientific Research · **Domain:** Study Design · **Level:** foundation · **~30 min**

**Who this is for:** Researchers, PhD Students, Postdocs, Research Software Engineers

## When to Use This Skill

Effect-size estimates, power targets, and preregistration preventing underpowered regrets.
Use it whenever a matching task appears in conversation — the agent loads these instructions on demand.

## Steps

1. Estimate effect size from prior literature or pilot data honestly
2. Set α=0.05, power=0.80 minimum; compute required N per group
3. Account for expected attrition/contamination in recruitment numbers
4. Preregister hypotheses, exclusions, and analysis plan
5. Define stopping rules and interim analysis boundaries upfront
6. Archive design doc citable regardless of outcome

## Common Pitfalls

- N=10 convenience samples dressed as exploratory
- HARKing after unblinded looks at trends

## Commands

**Install with skills CLI**
```bash
npx skills add aniruddhaadak80/skills --skill study-design-power-analysis-first
```

**Install globally**
```bash
npx skills add aniruddhaadak80/skills --skill study-design-power-analysis-first -g
```

---

Part of [aniruddhaadak80/skills](https://github.com/aniruddhaadak80/skills) · Browse all at https://skills.sh/aniruddhaadak80/skills