Understand both sides of a debate fairly
Get the strongest version of each side of a contested question, with the evidence, values and cruxes that divide them.
Turn a research question into keyword sets, Boolean strings and database tips for finding relevant papers.
Help me build a literature search for this research question: {{question}}
Databases I will use: {{databases}}. Time window: {{years}}. Study types I prefer: {{study_types}}.
Provide:
1. **Concept breakdown**: split the question into 3 to 4 core concepts (e.g. population, intervention or factor, outcome, context).
2. **Term table**: for each concept, list synonyms, related terms, alternative spellings, abbreviations and broader/narrower terms.
3. **Boolean search strings**: one broad string (high recall) and one narrow string (high precision), written in generic syntax with AND, OR, quotation marks and truncation (*). Note how to adapt it for the databases I named, including controlled vocabulary like MeSH where it applies.
4. **Filters and limits** I should apply, and which to avoid because they hide good results.
5. **Screening criteria**: a short inclusion and exclusion list I can use to triage titles and abstracts.
6. **Gaps**: terms or perspectives I may be missing, plus two backup strategies (citation chaining, key authors to check).
Do not list specific papers or authors unless you are certain they exist; if you mention any, tell me to verify them. If you are unsure about a fact, say so plainly. Never invent citations, quotes, statistics or URLs.Running a broad and a narrow string lets you balance coverage and precision. Test the strings, check hit counts, then ask the model to adjust them based on what you see.
Licensed CC0 (public domain) Version 1 · history
Get the strongest version of each side of a contested question, with the evidence, values and cruxes that divide them.
Get a staged learning path of book and article types, from first intro to advanced, with what to look for in each.
Break a claim into checkable parts, say what evidence would support it and rate confidence honestly.
Turn a broad topic into a focused research question with scope, variables, methods and feasibility checks.
Line up two articles or reports side by side, separate facts from interpretation and judge which is more reliable.
Summarize a paper in plain English, then assess its methods, evidence and limits without overstating anything.
Comments (0)
Sign in to join the discussion.
No comments yet. Be the first to share how it worked.