preregistering-analysis

Enforce timestamped preregistration of hypotheses and analysis plans before results.

282|26|Updated May 28, 2026
One-click install
npx skills add https://github.com/K-Dense-AI/science-superpowers --skill preregistering-analysis
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: preregistering-analysis
Source: https://github.com/K-Dense-AI/science-superpowers/tree/main/skills/preregistering-analysis
Command: npx skills add https://github.com/K-Dense-AI/science-superpowers --skill preregistering-analysis

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Prevents post-hoc hypotheses (HARKing) and p-hacking by enforcing that hypotheses, analysis plans, and decision rules are written and frozen before observing outcomes.

Core Features & Use Cases

  • Lock: write the hypotheses, primary analysis, predictions, decision rules, stopping rules, and multiplicity plan before looking at results.
  • Freeze: commit a timestamped preregistration to the repository path (e.g., docs/science-superpowers/preregistrations/).
  • Execute & Separate: run exactly the registered analysis and clearly label any unregistered follow-ups as Exploratory.
  • Supports multiple formats and templates for preregistration as part of the research workflow.

Quick Start

Create a preregistration documenting hypotheses, the exact analysis plan, the prediction, and stopping rules, then commit it before analyzing data.

Frequently Asked Questions about preregistering-analysis

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I prevent p-hacking and HARKing in my research design?

To prevent p-hacking and HARKing, you must preregister your hypotheses, analysis plans, and decision rules before observing outcomes. This Skill enforces a LOCK, FREEZE, EXECUTE, SEPARATE protocol to freeze predictions with a timestamp before data analysis begins.

What is a preregistration protocol for confirmatory analysis?

A preregistration protocol for confirmatory analysis requires writing hypotheses, primary analysis, predictions, and stopping rules before looking at results. It commits a timestamped preregistration to your repository path and clearly labels any unregistered follow-ups as exploratory.

How do I create a timestamped preregistration for an observational study?

You create a timestamped preregistration for an observational study by documenting your hypothesis testing plan and decision rules, then committing them to a repository path like docs/science-superpowers/preregistrations/. This freezes your analysis plan before executing the study.

Can I use this preregistration workflow for any hypothesis testing project?

Yes, you can use this preregistration workflow for any confirmatory-analysis project, including experiments and observational studies. It applies across hypothesis testing, p-values, and reporting claims by enforcing explicit documentation of predictions before results.

What's the best way to separate exploratory analysis from registered predictions?

The best way to separate exploratory analysis from registered predictions is the EXECUTE and SEPARATE protocol. You run exactly the registered analysis first, then clearly label any unregistered follow-ups as Exploratory in your final reporting claims.

Do I need to define stopping rules before preregistering an analysis plan?

Yes, you need to define stopping rules and a multiplicity plan as part of the LOCK phase before preregistering your analysis plan. Writing these decision rules before observing outcomes is required to prevent post-hoc hypotheses and ensure reproducibility.