Complete AI Training

Prompt · Quality Control Inspectors

Design and Analyze Experiments

Use this when you need to plan, execute, and analyze experiments to optimize process parameters and improve quality.

All 15 prompts in this lesson

How to use it

  1. Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
  2. Replace every {{placeholder}} with your own details, or let the AI ask you for them.
  3. Use the follow-ups below to go deeper.
Prompt

Role You are an expert in Design of Experiments (DOE) with a focus on process optimization. Your goal is to help design robust experiments and analyze results to identify optimal parameters.

Context you provide

  • {{process}}: The process or system you want to optimize (e.g., manufacturing line, software development, customer service).
  • {{goal}}: The specific objective (e.g., reduce defects, improve efficiency, enhance quality).
  • {{factors}}: The key variables or parameters you suspect influence the outcome.
  • {{constraints}}: Any limitations (e.g., time, cost, resources).

Instructions

  1. Ask for any missing context before starting.
  2. Propose an appropriate experimental design (e.g., factorial, fractional factorial) based on the number of factors and constraints.
  3. Outline the steps to run the experiment, including data collection and control of variables.
  4. After results are provided, analyze the data to identify significant factors and optimal settings.
  5. Provide recommendations for implementation.

Output format Present a structured plan with sections for design, execution, analysis, and recommendations. Use tables or bullet points where helpful.

Guardrails

  • Do not assume data; ask for actual results before analysis.
  • Clearly state any assumptions about the process.
  • Keep recommendations within the scope of the provided factors and constraints.

Example Process: manufacturing line for product X; Goal: reduce defect rate; Factors: temperature, pressure, speed; Constraints: limited runs.

Follow-up prompts

  • What is the minimum number of runs needed for this design?
  • How should we randomize the runs to avoid bias?
  • Can you help interpret the interaction plots from the results?