Frame a researchable question

Back to: Research & Decisions with AI

Turn a broad topic into a question with scope, criteria, and a usable decision at the end.

The idea

Research starts by reducing ambiguity. “Research electric vehicles” has no natural stopping point. A researchable question names the decision, population or context, comparison, criteria, time horizon, and acceptable evidence.

Write the decision sentence first: “This research will help [person] decide [choice] by [date].” Then list what the answer must compare. The criteria may include cost, performance, risk, implementation effort, equity, or reversibility.

Ask AI to challenge the frame, not answer it yet. It can surface hidden assumptions, alternative interpretations, missing stakeholders, and terms that need definitions. You decide which scope is appropriate.

Try it now

1. Choose a real but non-sensitive decision or question.
2. Write the decision sentence, including who decides and when.
3. Define the population/context, alternatives, five criteria, and time horizon.
4. Ask AI for ambiguities, assumptions, excluded viewpoints, and possible reframings.
5. Revise the question and write two explicit out-of-scope statements.

DECISION SENTENCE: This research will help [WHO] decide [WHAT] by [WHEN].\nCONTEXT/POPULATION: [scope]\nALTERNATIVES: [what is compared]\nCRITERIA: [five criteria]\nTIME HORIZON: [period]\nOUT OF SCOPE: [two boundaries]\n\nDo not answer yet. Identify ambiguous terms, hidden assumptions, missing stakeholders, and three alternative framings.

Quality check

• A useful answer would lead to a named decision or action.
• Criteria are defined before evidence is gathered.
• The time horizon and exclusions prevent endless scope growth.
• Alternative framings were considered rather than silently ignored.