Skip to main content Skip to course navigation

Factorial Experiments

Learning Objective

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

  • Explain the purpose of factorial experiments.
  • Understand main effects and interactions.
  • Distinguish between full factorial and fractional factorial designs.
  • Understand the role of replication, randomization and blocking.
  • Interpret the value of interaction plots in DOE analysis.

1. Introduction

A factorial experiment studies multiple factors simultaneously rather than examining each factor in isolation.

One of the important advantages of factorial designs is that they can reveal both:

  • Main effects — the effect associated with changing an individual factor.
  • Interactions — situations where the effect of one factor depends on the level of another factor.

This makes factorial experiments particularly useful when several process variables may work together to influence a response.

2. Main Effect

A main effect is the effect of changing one factor, averaged over the levels of the other factors.

For example, suppose a process studies:

  • Factor A — Temperature
  • Factor B — Pressure

The main effect of temperature considers how the response changes as temperature changes, while averaging across the pressure conditions included in the experiment.

The purpose is to understand the contribution of an individual factor to the process response.

3. Interaction Effect

An interaction occurs when the effect of one factor depends on the level of another factor.

For example, increasing temperature might improve the response when pressure is low but have little benefit—or a different effect—when pressure is high.

Therefore, examining only individual factor effects may not provide the complete picture.

The interaction must also be considered.

4. Full Factorial Design

A full factorial design tests all combinations of the selected factor levels.

For example, with two factors and two levels for each factor:

  • Factor A: Low / High
  • Factor B: Low / High

The experiment contains all combinations of these levels.

This allows the experiment to examine the factors together and identify possible interactions.

5. Fractional Factorial Design

A fractional factorial design tests a carefully selected subset of the combinations used in a full factorial design.

It can be useful when a full factorial experiment becomes impractical because of the number of experimental runs required.

The choice between a full and fractional design should therefore consider the experimental objective and practical requirements.

6. Interaction Plots

An interaction plot helps visualize whether the effect of one factor changes depending on another factor.

It can help the Black Belt recognize:

  • Possible interaction between factors.
  • Different response behaviour at different factor levels.
  • Conditions where the factors may need to be considered together.

Interaction plots therefore provide an important visual aid when interpreting factorial experiments.

7. Replication

Replication involves repeating experimental conditions.

Replication can provide additional information about variation in the experimental results and strengthen the basis for evaluating factor effects.

It should be planned as part of the experimental strategy rather than added without purpose.

8. Randomization

Randomization helps reduce systematic influence associated with the order in which experimental runs are conducted.

In a factorial experiment, randomized run order can help prevent time-related or uncontrolled changes from becoming confused with factor effects.

This continues the principle introduced in Lesson 18 — DOE Fundamentals & Strategy.

9. Blocking

Blocking is used to account for known sources of variation that could influence the response.

For example, if experiments must be conducted under different conditions or groups, blocking can help organize the experiment so that such variation does not obscure the factor effects being studied.

The source identifies blocking as one of the relevant techniques for factorial experiments.

10. Tools & Techniques

The key factorial experiment techniques are:

  • Full factorial
  • Fractional factorial
  • Interaction plots
  • Replication
  • Randomization
  • Blocking

11. Application Example

Paint Manufacturing

A paint manufacturer wants to study the effect of:

  • Pigment type
  • Drying time

on the performance of the finished paint.

A factorial experiment can study these factors simultaneously rather than conducting separate experiments for each factor.

The results can then be examined for:

  • Main effects.
  • Possible interaction between pigment type and drying time.

12. Case Study

Automotive Welding Process

An automotive plant studied:

  • Welding speed
  • Electrode type

using a 2 × 2 factorial experiment.

The analysis identified an interaction between the factors. After optimizing the settings, weld defects were reduced by 30%.

Black Belt Learning

The case demonstrates why interaction analysis is important.

If welding speed and electrode type influence each other, selecting the best setting for one factor without considering the other may not produce the desired result.

13. Black Belt Perspective

A Black Belt should not automatically assume that factors operate independently.

When several factors are involved, ask:

  1. What are the main effects?
  2. Could the factors interact?
  3. Should all combinations be tested?
  4. Would a fractional factorial design be appropriate?
  5. Is replication required?
  6. Should the run order be randomized?
  7. Is blocking needed to account for another known source of variation?

The objective is not simply to run more experiments. It is to extract useful information efficiently from a structured experiment.

14. Lesson Practice

  1. Why are interactions important in factorial experiments?
  2. What can an interaction plot reveal?
  3. When might a fractional factorial design be useful?
  4. What is the difference between a main effect and an interaction?
  5. Why can factorial experiments be more informative than studying factors separately?
  6. What is the purpose of replication?
  7. Why is randomization used in experimental runs?
  8. What is the purpose of blocking?

15. Key Learning Points

  • Factorial experiments study multiple factors simultaneously.
  • A main effect describes the effect of one factor averaged across other factor levels.
  • An interaction occurs when the effect of one factor depends on another factor.
  • Full factorial designs test all selected combinations.
  • Fractional factorial designs test a carefully selected subset when a full design is impractical.
  • Interaction plots help visualize interactions.
  • Replication, randomization and blocking support effective experimental design.

16. Lesson Conclusion

Factorial experiments provide a structured way to understand multiple factors and their interactions.

The key principle is:

Study Factors Together → Identify Main Effects → Examine Interactions → Use Structured Experimental Techniques → Select Effective Settings

Understanding factorial experiments prepares the Black Belt for the next stage:

Lesson 20 — DOE Analysis & Optimization