Statistical Problem Formulation

What it does

When installed, this skill tells an AI agent to pin down a research topic before any method design, theory or report writing. The agent works through a fixed checklist: what data is observed and how it was sampled, what statistical model is assumed, what quantity is being estimated, which assumptions must hold, what claims are being tested, how success is measured, and what theoretical property should be derived. It also gives a standard YAML handoff structure and a Markdown template for the write-up, plus a quality bar that the formulation must be implementable without guessing.

Try asking

• “Formulate a research question about whether a new drug dosage changes recovery time, with data model, estimand and assumptions”Copy
• “Turn my topic on causal effects of remote work on productivity into a formal problem statement with testable hypotheses”Copy
• “Produce a YAML handoff listing claims, evaluation criteria and theory targets for my survival analysis project”Copy

Get better results

• Provide your raw data description (sample size, sampling scheme, censoring) up front so the agent can fill the observed-data section accurately.
• Review the listed assumptions yourself; the agent proposes them but a statistician must confirm they are realistic.
• Use the blocking_ambiguities field to see what still needs clarification before moving to method design.

What's inside

• SKILL.md