Skip to main content Skip to course navigation

DOE – Fundamentals and Strategy

Learning Objective

By the end of this lesson, the learner should be able to:

  • Explain the purpose of Design of Experiments (DOE).
  • Understand factors, levels and responses.
  • Recognize why DOE is preferable to trial-and-error when multiple factors are involved.
  • Understand the importance of randomization.
  • Develop a structured strategy for conducting an experiment.

1. Introduction

Design of Experiments (DOE) is a structured method for studying the effects of multiple factors on a process response.

Instead of changing one factor at a time through informal trial-and-error, DOE provides a planned approach for conducting experiments and learning how process factors affect the response.

For a Black Belt, DOE provides a disciplined way to investigate process conditions and identify settings that can improve performance.

2. Understanding the Key DOE Elements

Factor

A factor is a process variable that is deliberately studied in the experiment.

Examples may include:

  • Temperature
  • Pressure
  • Baking time
  • Ingredient ratio

Level

A level is a selected setting of a factor.

For example, if temperature is being studied, the experiment may use different temperature settings.

Response

The response is the measured outcome of the experiment.

Examples include:

  • Yield
  • Defect rate
  • Product consistency
  • Processing time

These elements establish what is being changed and what is being measured.

3. Defining the Experimental Objective

Before conducting an experiment, the Black Belt should clearly define:

  • What process response is being studied.
  • Which factors are to be investigated.
  • Which levels will be tested.
  • What the experiment is intended to learn or improve.

A clearly defined objective helps ensure that the experimental design is aligned with the project requirement.

4. Randomization

Randomization helps reduce systematic influence from the order in which experimental runs are performed and from uncontrolled conditions.

For example, if all experiments at one setting are conducted first and another setting is always tested later, changes over time could become mixed with the factor effect.

Randomization helps reduce this systematic influence.

5. Why DOE Is Preferable to Trial-and-Error

When several factors may influence a process, informal trial-and-error can make it difficult to determine which factor is responsible for an observed change.

DOE provides:

  • A planned experimental structure.
  • Defined factors and levels.
  • Controlled data collection.
  • Randomized experimental runs where appropriate.
  • A basis for statistical analysis.

This allows the Black Belt to learn systematically from the experiment.

6. Factorial Design

A factorial design studies multiple factors simultaneously.

It allows the experiment to examine how different factor settings affect the response and can reveal relationships between factors.

Factorial experiments are developed further in Lesson 19 — Factorial Experiments.

7. Tools & Techniques

The source identifies the following DOE elements and tools:

  • Factors
  • Levels
  • Responses
  • Randomized runs
  • Factorial design
  • Minitab DOE module

8. Application Example

Chemical Plant

A chemical plant wants to understand how temperature and pressure affect process yield.

The Black Belt defines:

Factors: Temperature and Pressure

Levels: Selected operating settings for each factor

Response: Process Yield

The experiment is planned systematically, with randomized runs where appropriate.

The resulting data can then be analyzed to understand the effects of the factors on yield.

9. Case Study

Food Company — Cookie Texture

A food company used DOE to study cookie texture.

The experiment considered:

  • Oven temperature
  • Baking time
  • Ingredient ratios

The analysis identified temperature as having the strongest effect, and improved settings increased consistency.

Black Belt Learning

The case demonstrates how DOE can examine several process factors systematically rather than relying on isolated trial-and-error changes.

10. Black Belt Perspective

Before starting a DOE, the Black Belt should ask:

  1. What response are we trying to improve?
  2. Which factors could influence that response?
  3. What levels should be tested?
  4. What is the objective of the experiment?
  5. How will the experimental runs be organized?
  6. How will randomization reduce systematic influence?
  7. What analysis will be used after the experiment?

A well-designed experiment begins with a clear experimental strategy, not simply with a collection of test runs.

11. Lesson Practice

  1. Why is DOE preferable to trial-and-error for multiple factors?
  2. What should be defined before running an experiment?
  3. Why is randomization important?
  4. What is a factor in DOE?
  5. What is a level?
  6. What is a response?
  7. Why is it important to define the experimental objective before conducting the runs?

12. Key Learning Points

  • DOE is a structured method for studying the effects of multiple factors on a process response.
  • A factor is a process variable being studied.
  • A level is a selected setting of a factor.
  • A response is the measured outcome.
  • Randomization helps reduce systematic influence from run order and uncontrolled conditions.
  • Factorial design allows multiple factors to be studied systematically.
  • Minitab provides a DOE module to support experimental analysis.

13. Lesson Conclusion

DOE provides the Black Belt with a structured strategy for learning how process factors influence a response.

The essential sequence is:

Define Objective → Select Factors → Select Levels → Plan Experiment → Randomize Runs → Collect Data → Analyze Results

The quality of the experiment depends heavily on the quality of the planning that occurs before the experimental runs begin.