Hypothesis Generator and Tester
Generate testable hypotheses from observations and design experiments to test them.
Category: research
Difficulty: beginner
Platforms: chatgpt claude perplexity
Tags: hypothesis testing experimental design scientific method research design critical thinking
Prompt Template
You are a research scientist specializing in experimental design. Help me generate and test hypotheses.
Context:
- Domain: {{domain}}
- Observation or pattern I have noticed: {{observation}}
- Available data or resources: {{resources}}
- Constraints: {{constraints: time/budget/sample size/equipment}}
- Type of research: {{type: academic/product/business/personal curiosity}}
Provide:
1. HYPOTHESIS GENERATION:
- 5 testable hypotheses that could explain my observation
- For each: the independent variable, dependent variable, and predicted relationship
- Rank by testability and potential impact
2. NULL HYPOTHESES: Formal null hypothesis for each
3. EXPERIMENTAL DESIGN: For the top 2 hypotheses:
- Recommended study design (experiment, quasi-experiment, observational)
- Control conditions needed
- Sample size estimation with rationale
- Variables to control for (confounds)
4. MEASUREMENT PLAN: How to operationalize and measure each variable
5. ANALYSIS PLAN: Which statistical tests to use for each hypothesis
6. FALSIFICATION CRITERIA: What results would disprove each hypothesis
7. ALTERNATIVE EXPLANATIONS: Rival hypotheses to rule out
Prioritize hypotheses that are both testable with my resources and meaningful if confirmed.
Tips
- Start with the simplest hypothesis that could explain your observation
- Design the experiment to rule out the most obvious alternative explanation first
- Pre-register your hypotheses and analysis plan to avoid HARKing
- If you cannot specify what result would disprove your hypothesis it is not testable