Hypothesis Generation Prompt Templates

AI prompt templates for generating hypotheses. Develop testable research questions.

Overview

Coming up with good research questions is harder than it looks. These prompts help you brainstorm hypotheses, refine vague ideas into testable statements, and connect your questions to existing theory. They're especially useful early in a project when you know your general area but haven't locked down your specific research focus.

Best Practices

1

Describe what you already know about the topic so suggestions build on existing knowledge

2

Mention any theoretical frameworks you want to work within

3

Specify whether you need null and alternative hypotheses or just directional predictions

4

Include practical constraints like available data or methods so hypotheses are actually testable

5

Ask for multiple options so you can pick the most promising direction

Prompt Templates

1. Hypothesis Brainstormer

I'm researching [BROAD TOPIC AREA] and I'm interested in the relationship between [VARIABLE A] and [VARIABLE B]. Based on this context:

- What we know: [2-3 ESTABLISHED FINDINGS]
- What's unclear: [1-2 UNANSWERED QUESTIONS]
- My context: [YOUR FIELD, AVAILABLE DATA, CONSTRAINTS]

Generate 5 testable hypotheses that could contribute new knowledge. For each, explain:
1. The hypothesis statement
2. Why it's interesting/important
3. How you might test it
4. What existing theory it builds on or challenges
Topic: consumer behavior. Variable A: product sustainability labels. Variable B: purchase decisions. Known: eco-labels increase stated preference; actual behavior often differs. Unclear: when do labels change real behavior? Context: marketing PhD with access to point-of-sale data.
Hypothesis 1: Sustainability labels will have a stronger effect on purchase behavior for products with high social visibility (clothing, cars) than low visibility products (cleaning supplies). Importance: Explains inconsistent findings by introducing a moderating variable... Test: Compare label effects across product categories using your POS data...

At the beginning of a research project when you're exploring directions. Also useful when you've hit a dead end and need fresh angles.

  • •Don't commit to the first hypothesis that sounds good. Generate many, then evaluate
  • •Run this with a colleague's input to combine different perspectives

2. Hypothesis Refinement

I have a rough hypothesis but it needs work:

Draft hypothesis: [YOUR CURRENT HYPOTHESIS]

Problems I see: [ANY ISSUES YOU'VE NOTICED]

Help me improve it by:
1. Making it more specific and testable
2. Identifying the independent and dependent variables clearly
3. Suggesting appropriate operationalizations for each variable
4. Rewriting it in standard hypothesis format (If X, then Y, because Z)
5. Creating both null and alternative versions
Draft: 'Social media makes people more anxious.' Problems: too vague, no mechanism specified.
Refined hypothesis: If individuals spend more than 2 hours daily on image-based social media platforms (Instagram, TikTok), then they will report higher levels of social comparison anxiety, because repeated exposure to curated content triggers upward social comparison. IV: Daily time on image-based platforms. DV: Social comparison anxiety (measured by INCOM scale)...

When you have an idea but it's too fuzzy to test. Good for moving from 'I wonder if...' to 'I predict that...'

  • •The 'because' part of your hypothesis should connect to theory, not just common sense
  • •If you can't operationalize the variables, the hypothesis isn't testable yet

3. Theory-Based Hypothesis Generator

I want to apply [THEORY NAME] to [NEW CONTEXT OR POPULATION].

Brief theory summary: [2-3 SENTENCES ON THE THEORY'S MAIN CLAIMS]

The new context: [DESCRIBE THE SITUATION YOU WANT TO STUDY]

Generate 3 hypotheses that:
1. Directly test whether the theory applies in this context
2. Test a boundary condition of the theory
3. Extend the theory in a novel direction

For each, explain the theoretical reasoning.
Theory: Self-Determination Theory (autonomy, competence, relatedness drive motivation). Context: Gig economy workers who have autonomy but lack job security. Theory summary: People are more motivated when they feel autonomous, competent, and connected to others.
H1 (Direct test): Gig workers who report higher perceived autonomy will show greater job satisfaction than those with lower autonomy. Reasoning: SDT predicts autonomy increases motivation... H2 (Boundary condition): The positive effect of autonomy on gig worker motivation will be weakened when financial insecurity is high...

When you want to ground your research in established theory or test whether classic theories hold in new situations.

  • •Read critiques of the theory first so you can address known limitations
  • •Make sure you actually cite the theory in your paper, not just the general idea

4. Competing Hypotheses Framework

I'm studying [PHENOMENON] and there are different possible explanations:

Explanation A: [FIRST THEORY/MECHANISM]
Explanation B: [SECOND THEORY/MECHANISM]
Explanation C (optional): [THIRD THEORY/MECHANISM]

Design a study that could distinguish between these explanations:
1. What hypothesis would each explanation predict?
2. What outcome would support A but not B?
3. What outcome would support B but not A?
4. What study design could test this?
5. What findings would be inconclusive?
Phenomenon: People tip more after receiving a compliment from their server. Explanation A: Reciprocity norm (they feel obligated to return the favor). Explanation B: Mood improvement (compliment makes them happier, happy people tip more).
Hypothesis A predicts: Tip increase will be specific to the complimenter. A stranger's compliment won't affect tipping. Hypothesis B predicts: Any mood boost should increase tipping, regardless of source. Critical test: Measure tipping when compliment comes from server vs. random customer. If only server compliments increase tips, supports A...

When multiple explanations exist for your phenomenon and you want to design a decisive test. Useful for making theoretical contributions.

  • •The best research often pits theories against each other rather than just testing one
  • •Think about what would falsify your preferred explanation

Common Mistakes to Avoid

•

Generating hypotheses that sound interesting but can't actually be tested with available methods or data

•

Confusing research questions with hypotheses. Questions ask; hypotheses predict specific answers

•

Not grounding hypotheses in theory, which makes it hard to explain why the finding would matter

Frequently Asked Questions

Coming up with good research questions is harder than it looks. These prompts help you brainstorm hypotheses, refine vague ideas into testable statements, and connect your questions to existing theory. They're especially useful early in a project when you know your general area but haven't locked down your specific research focus.

Related Templates

Literature Review Prompt Templates

AI prompt templates for literature reviews. Synthesize sources and identify research gaps.

Research Summarization Prompt Templates

AI prompt templates for summarizing research. Extract key findings from academic papers.

Survey Design Prompt Templates

AI prompt templates for designing surveys. Create effective questionnaires for research.

Have your own prompt to optimize?

Try Prompt Optimizer More Research Prompts