Skip to main content
MethodScience & Data AnalysisExperiments & scientific writingIntermediate

Design an experiment from a hypothesis

A testable hypothesis, variables, controls, sample size reasoning, procedure and analysis plan.

This is a practical, copy-paste process for lab, field or business experiments.. Use it when you want a clear structure, then add your own details and the linked evidence.

Published by Deepstory Research. Reviewed Sep 2026.

How to use this method

  1. 1. Copy or downloadCopy the prompt below, or download the method file and attach it to your chat.
  2. 2. Paste into any AIChatGPT, Claude, Gemini or Copilot. Replace each [PLACEHOLDER] with your details, or let the AI ask you.
  3. 3. Add evidenceAttach the evidence files listed on this page so the AI works from sourced numbers, not guesses.

The prompt

Paste it as-is. The AI will ask for anything still in [BRACKETS].

Download .md
# Design an experiment from a hypothesis
Method by Deepstory Research: deepstoryresearch.com/method/hypothesis-experiment-design

You are a principal investigator. Follow the method below step by step. Goal: A testable hypothesis, variables, controls, sample size reasoning, procedure and analysis plan.

## My inputs
- Question: [QUESTION]  (example: does a new fertiliser raise wheat yield)
- Constraints: [CONSTRAINTS]  (example: 20 plots, one season)
- Measurement tools: [MEASUREMENT_TOOLS]  (example: what you have)

Before you start: if any input above is still in [BRACKETS] or missing, ask me for it in one short numbered list and wait. Do not guess. If I attached files (data, briefs, Deepstory evidence files), treat them as the primary source.

## Steps to follow
1. Write null and alternative hypotheses.
2. Define independent, dependent and control variables.
3. Choose design (randomised, blocked, crossover).
4. Reason about sample size and power.
5. Write the step-by-step protocol.
6. Pre-specify the analysis.

## Deliver the output in this structure
- Hypotheses
- Variables
- Design
- Sample size reasoning
- Protocol
- Analysis plan
- Threats to validity

## Quality rules
- Pre-register the primary outcome to avoid p-hacking.
- Label every number as FACT (with source and date), ESTIMATE (show the calculation) or ASSUMPTION.
- If a Deepstory evidence file is attached, use its figures as the reference values and cite its source references; tag claims that are not from it as [UNVERIFIED].
- End with: the 3 assumptions that matter most, what would change the answer, and the single best next step.

What you need to give the AI

InputPlaceholderExample
Question[QUESTION]does a new fertiliser raise wheat yield
Constraints[CONSTRAINTS]20 plots, one season
Measurement tools[MEASUREMENT_TOOLS]what you have

Use it when

  • Lab, field or business experiments.

What you get

  • Hypotheses
  • Variables
  • Design
  • Sample size reasoning
  • Protocol
  • Analysis plan
  • Threats to validity

The procedure, step by step

Read this before you run it, so you can check the AI actually followed the method.

  1. 1

    Write null and alternative hypotheses.

  2. 2

    Define independent, dependent and control variables.

  3. 3

    Choose design (randomised, blocked, crossover).

  4. 4

    Reason about sample size and power.

  5. 5

    Write the step-by-step protocol.

  6. 6

    Pre-specify the analysis.

Quality checks built into the prompt

  • • Pre-register the primary outcome to avoid p-hacking.
  • • Every number labelled FACT, ESTIMATE or ASSUMPTION; claims outside attached evidence tagged [UNVERIFIED].

Evidence to use with this method

Methods tell the AI how to think. Evidence gives it sourced numbers to think with. Download a context file and attach it with the prompt.

EvidenceGeopolitical

Internet use ranking

Which selected economies reported the highest internet-use rates? Individuals using the Internet (% of population) · common year 2024 · 20 selected economies with published values

Open evidence
EvidenceGeopolitical

Income and internet use

Did higher-income selected economies also report more internet use? GDP per capita (current US$) vs Individuals using the Internet (% of population) · common year 2024 · n=20

Open evidence

Run next

Methods that chain well after this one.

Science & Data AnalysisChoose the right statistical testThe appropriate test for your data and question, with assumptions check and interpretation template.Science & Data AnalysisAnalyse an A/B testA correct read of an A/B test: lift, confidence, sample adequacy, guardrails and decision.

Deepstory methods are process guidance for AI tools. Review AI output before relying on it; legal, tax, medical and investment decisions need a qualified professional.