research-grilllisted

# Research Grill An interview, not an evaluation. You ask; the researcher decides. The session ends when nothing about the project is left silently assumed and every decision is written down. **Related skills.** `$research-wayfinder` carries the same interview across many sessions as a ticket map; use it when a project has more open decisions than one conversation can hold, and use this skill for one sitting. `$diverge` puts several distinct options on the table when a question needs them before it can be answered. `$hypothesis-building`, `$survey-design`, `$conjoint-design`, `$list-experiment`, `$cross-national-design`, and `$pre-registration-writing` are where a settled decision goes to be worked out in full; name the relevant one in your recommended answer so the researcher can go deeper. ## Stage Infer the stage from what the researcher brings; an argument forces it. | Stage | Researcher brings | Done when | |---|---|---| | `idea` | a topic, a hunch, a puzzle, "I want to study X" | one falsifiable research question, the population it covers, the contribution claim in one sentence, and the reason an experiment or survey is the right instrument rather than existing data | | `design` | a research question | the design tree below is fully visited: estimand, identification, sample and power, measurement, analysis plan, ethics and pre-registration venue, output venue | | `defend` | a finished design, a pre-analysis plan, or a draft | every objection a careful reviewer woul