domains / families / research / replication-driven-research

replication-driven-research

separate

Use when starting empirical analysis, creating a data pipeline, generating results, or when data or model specifications change. Enforces end-to-end reproducibility — every number in the paper must be regenerable from raw data by a script with a fixed seed. Replaces TDD for the research domain.

When to use

• Start of a new empirical project
• First time running analysis code in a project
• Before generating any table or figure for a paper
• After any change to raw data, sample selection, or model specification
• Before declaring a result "final" or "verified"
• When the user asks "is this reproducible?"
• When reviewer comments require re-running the analysis

When not to use

• copying numbers from console output
• trusting cached outputs after input changes
• using undocumented manual steps
• using non-fixed seeds

Skill page

Source