Scientific Fact-Checking
Critical examination of claims for their scientific supportability
Prompt template
Check the following claims for their scientific accuracy and evidential basis. **Claims to Verify**: [CLAIM 1] [CLAIM 2] [CLAIM 3] **Context**: These claims come from [SOURCE, e.g., press release, social media, article, study] and concern the field of [FIELD]. For each claim, conduct the following analysis: 1. **Core Assertion**: Summarise the essential empirical claim in one sentence. 2. **Evidence Base**: Which scientific studies or meta-analyses support or refute this claim? Distinguish between strong evidence (RCTs, meta-analyses), moderate evidence (observational studies), and weak evidence (case reports, expert opinion). 3. **Accuracy Rating**: Is the claim correct, oversimplified, misleading, contested, or false? Justify your assessment. 4. **Context and Nuance**: What important caveats, contextual factors, or nuances are missing from the original statement? 5. **Correction Suggestion**: Formulate a more precise, scientifically accurate version of the claim. Rating scale: ✓ Correct | ⚠ Oversimplified | ✗ Misleading/False
Replace placeholders in [square brackets]: [CLAIM 1][CLAIM 2][CLAIM 3][SOURCE, e.g., press release, social media, article, study][FIELD]
How to use this template
Copy the prompt, replace the placeholders in square brackets with your details and paste it into the AI assistant of your choice. The more precise your details, the better the result.
Tips for better results
- Provide the source of the claim — context is essential for an accurate assessment.
- For medical or health-related claims, prefer Cochrane Reviews or guidelines from major professional societies.
- Note that AI models do not have real-time internet access and cannot verify current events.
Suitable open models
The selection names current open models as a starting point for your own tests. It is not a measurement for this template.
Keywords: #faktencheck #evidenz #wissenschaft #desinformation #kritisches-denken