Compare two sources and find where they disagree

Line up two articles or reports side by side, separate facts from interpretation and judge which is more reliable.

LLazyPrompt(@lazyprompt)Research & LearningAnalyzeWorks with any model
No ratings yetSign in to rate
Save 0 Fork Report
Prompt template
Compare these two sources on the topic: {{topic}}.

For the comparison, produce:
1. **Core claims** of each source, in two or three bullets each.
2. **Agreement**: points where both say essentially the same thing.
3. **Disagreement**: a table with columns: issue, Source A says, Source B says, likely reason for the difference (different data, definitions, time period, framing, or bias).
4. **Evidence quality**: for each source, what kind of evidence it gives (primary data, expert opinion, anecdote, undisclosed), and anything missing.
5. **Framing and tone**: loaded words, what is emphasized or omitted.
6. **Bottom line**: which claims I can rely on, which I should check elsewhere, and what specific evidence would settle the disagreements.

Use only the texts below. Do not guess about the publishers beyond what the text shows. If you are unsure about a fact, say so plainly. Never invent citations, quotes, statistics or URLs.

Source A ({{label_a}}):
{{source_a}}

Source B ({{label_b}}):
{{source_b}}
Example output
**Core claims.** A: productivity rose 12% after going remote, driven by fewer interruptions. B: no significant productivity change, but higher attrition among new hires. **Agreement.** Both examine remote work's effect on output and both treat interruptions and onboarding as relevant context. | Issue | Source A | Source B | Likely reason | |---|---|---|---| | Productivity | +12% | No significant change | Perception vs measured output; different workforces | | Retention | Not mentioned | +9% attrition for new hires | Omitted in A | **Evidence quality.** A relies on manager opinion (survey of 800), which reflects perception. B uses output data from 1,500 workers, but only one job type; sampling method not stated for either. **Framing.** A stresses benefits; B reports both a null result and a cost. **Bottom line.** Trust B more on measured productivity in call centers; treat A as a signal of manager perception. To settle it: objective output data across several job types, and attrition data from A's firms.

Why it works

Separating what each source claims from why they differ avoids picking a winner too early. Add a third source and ask for a three-way table to triangulate.

Licensed CC0 (public domain) Version 1 · history