Stage 1: CLARIFY (10 minutes)
Transform vague research interests into precise, answerable questions
The Goal: Transform a vague research interest into a precise, answerable question. Vague questions lead to unfocused research where sources pile up without clear direction. Precise questions enable efficient synthesis by giving every source collection and finding extraction a clear purpose. This 10-minute investment prevents hours of wasted effort later.
Step 1.1: Brain Dump (3 minutes)
Open your AI assistant (Claude, ChatGPT, or Gemini) and use this clarification prompt to articulate your initial thinking. The AI will ask 5 clarifying questions to help you refine your focus.
Prompt Template:
I'm exploring [YOUR TOPIC]. Help me clarify my research question.
Current thinking:
- I'm interested in: [vague description]
- I want to know: [what you're curious about]
- This matters because: [why you care]
Please ask me 5 clarifying questions to help refine this into a specific, answerable research question.Example:
I'm exploring AI's impact on productivity. Help me clarify my research question.
Current thinking:
- I'm interested in: whether AI actually makes people more productive
- I want to know: if the hype matches reality
- This matters because: I'm deciding whether to invest in AI tools for my team
Please ask me 5 clarifying questions to help refine this into a specific, answerable research question.The AI will typically ask about scope (individual vs organizational), specificity (which AI tools), context (industry/work type), and timeframe (current vs predictive).
Step 1.2: Answer and Refine (5 minutes)
Answer the AI's clarifying questions honestly and specifically. Then request 3 possible research questions using this refinement prompt:
Prompt Template:
Based on my answers, formulate 3 possible research questions using this format:
"What is the effect of [specific intervention] on [specific outcome] for [specific population/context]?"
Make them:
1. Specific enough to research in 20-50 papers
2. Broad enough to have sufficient literature
3. Interesting enough to sustain my attentionThe AI will generate three distinct research questions, each hitting different angles of specificity and scope based on your answers.
Step 1.3: Choose Your Question (2 minutes)
Review the 3 options and select the question that excites you most, has practical value to you, and feels answerable (not too broad, not too narrow).
Example Refined Question:
What is the effect of large language models (like GPT-4, Claude) on individual knowledge worker productivity in software development and content creation, based on empirical studies from 2022-2025?Save this question. It becomes your North Star for the next hour—every source you collect, every finding you extract, every synthesis you make should connect back to answering this specific question.
Why This Works: Starting with AI dialogue prevents the analysis paralysis that comes from trying to formulate the perfect question alone. The iterative refinement surfaces hidden assumptions and narrows scope before wasting time collecting irrelevant sources. You're essentially using the AI as a thinking partner to externalize and sharpen your intuitions into researchable form.