Category
15 AI prompts for Research
Literature, synthesis, interviews and evidence you can cite.
Written and tested for The AI University. Each carries its variables, a usage guide and the cases where it will let you down. Free with an account.
15 prompts
- Research
Literature Synthesis With Disagreement
Synthesises sources into what is established, what is contested and what is merely asserted — keeping the disagreement rather than averaging it away.
ClaudeChatGPTGemini000Open → - Research
Interview Guide That Avoids Leading
Builds a user or customer interview guide that asks about past behaviour rather than future intentions, which is the only reliable signal.
ClaudeChatGPT000Open → - Research
Claim Verification Trail
Traces a claim back to its original source and reports where the chain breaks, what got distorted, and what the original actually said.
ClaudeChatGPTGemini000Open → - Research
Survey Design That Avoids Garbage
Designs a survey whose answers mean something — no leading questions, no scales that cannot be interpreted, and a plan for the analysis before it is sent.
ClaudeChatGPT000Open → - Research
User Research Synthesis
Turns a pile of interviews into patterns with evidence, keeping the disconfirming cases rather than smoothing them away.
ClaudeChatGPTGemini000Open → - Research
Market Sizing
Builds bottom-up with every assumption stated, since top-down sizing is unfalsifiable and therefore useless.
ClaudeChatGPT000Open → - Research
Source Credibility Assessment
Assesses funding, method, date and whether it is primary — in that order, since the first two explain most disagreements.
ClaudeChatGPTGemini000Open → - Research
Competitive Intelligence Brief
Reads hiring, pricing and changelogs, since press releases say what a company wants believed.
ClaudeChatGPT000Open → - Research
Expert Interview Guide
Asks what they have changed their mind about, which is where the non-consensus view lives.
ClaudeChatGPT000Open → - Research
Research Question Refinement
Narrows until the question is answerable, since broad questions produce work nobody can act on.
ClaudeChatGPT000Open → - Research
Desk Research Plan
Timeboxes and defines done, since desk research otherwise expands to fill the whole project.
ClaudeChatGPT000Open → - Research
Diary Study Design
Prompts at the moment rather than at the end of the day, since recall degrades within hours.
ClaudeChatGPT000Open → - Research
Usability Test Script
Gives tasks rather than instructions, since the moment you say where to click the test is over.
ClaudeChatGPT000Open → - Research
Research Repository Structure
Indexes by finding rather than by study, since a study-organised archive is one nothing is ever found in again.
ClaudeChatGPT000Open → - Research
Triangulation Plan
Uses two methods that fail differently, which beats doing one method twice as carefully.
ClaudeChatGPT000Open →
