define-a-research-question

# define-a-research-question A good question determines what evidence belongs and what can be left out. Make the boundary visible before searching. ## When to use - Use when the topic is broad, value-laden, ambiguous, or likely to expand during research. - Use for explanatory, descriptive, comparative, causal, evaluative, or decision-support questions. - Do not force an exploratory project into a hypothesis before the user understands the terrain. ## Procedure 1. State the audience, decision or assignment, deadline, and required output. 2. Write the broad topic and list the important actors, outcomes, mechanisms, places, periods, and constraints it may contain. 3. Separate terms that can be observed from terms that need an operational definition. 4. Choose the question type and write a one-sentence candidate using only the necessary concepts. 5. Define inclusions, exclusions, unit of analysis, geography, time window, comparison, and acceptable evidence. 6. Test feasibility: relevant sources exist, the scope fits the time, and the answer would change understanding or action. 7. Identify assumptions and nearby questions that are useful but out of scope. 8. Produce the final question, a short rationale, key definitions, and two or three subquestions that partition the work. ## Failure plan - If the question presumes its conclusion, rewrite it neutrally. - If evidence cannot distinguish the proposed causes, downgrade a causal question to an associational or descriptive one. - If important terms remain disputed, carry competing definitions instead of silently choosing one. - If the scope is still too large, narrow one dimension at a time and record the tradeoff. ## Done - The question is answerable with identifiable evidence and has a clear audience or decision - Scope, definitions, assumptions, exclusions, and subquestions are explicit - The wording does not assume the answer or promise a stronger conclusion than the design can support