DOE Terminology: Factors, Levels, Response, and Treatment - Complete Guide

DOE Terminology: A Comprehensive Guide for Six Sigma Black Belts

Why DOE Terminology is Important

Design of Experiments (DOE) is a cornerstone of the Six Sigma Black Belt's toolkit, particularly during the Improve phase. Understanding DOE terminology is critical because:

• Precision in Communication: Proper terminology ensures clear communication between team members, stakeholders, and experimenters about what is being tested and measured.
• Experimental Validity: Correct identification of factors, levels, responses, and treatments prevents experimental design flaws that could invalidate results.
• Statistical Rigor: Accurate terminology enables proper statistical analysis and meaningful interpretation of experimental data.
• Process Improvement: Correctly designed experiments lead to reliable insights that drive genuine process improvements and cost savings.
• Certification Requirement: Mastery of DOE terminology is essential for Six Sigma Black Belt certification exams.

What is DOE Terminology?

DOE terminology comprises the foundational language used to describe the components of experimental design. The four primary terms you must master are:

1. Factors

Definition: Factors are the input variables or independent variables that you intentionally change or control during an experiment.

Key Characteristics:

• They are the variables you manipulate to observe their effect on the output.
• They represent potential causes or process parameters you suspect influence your response variable.
• Examples include temperature, pressure, speed, material type, supplier, or operator skill level.
• Factors can be quantitative (continuous, like temperature measured in degrees) or qualitative (categorical, like material type: plastic vs. metal).

Example: In a manufacturing experiment, if you're testing how temperature and humidity affect product defect rates, both temperature and humidity are factors.

2. Levels

Definition: Levels are the specific values or settings that each factor can take during the experiment.

Key Characteristics:

• Each factor must have at least two levels (typically a low and high setting, or current vs. new).
• Levels can be numerical values or categorical descriptions.
• The number of levels affects the experimental design complexity (2-level designs are simpler; 3+ level designs provide more information but require more runs).
• Levels must be realistic and achievable within process constraints.

Example: If temperature is a factor, the levels might be 70°C and 90°C. If supplier is a factor, the levels might be Supplier A and Supplier B.

3. Response

Definition: The response (also called response variable or dependent variable) is the output or result you measure to assess the impact of the factors.

Key Characteristics:

• It is what you are trying to optimize, reduce, or understand.
• It should be measurable and quantifiable.
• There can be one primary response or multiple responses in an experiment.
• Responses are affected by the factors but are not directly controlled by the experimenter.
• Common responses in manufacturing include defect rate, cycle time, cost, strength, and customer satisfaction.

Example: If factors are temperature and humidity, the response might be the percentage of defective products produced.

4. Treatment

Definition: A treatment is a specific combination of factor levels tested in a single experimental run.

Key Characteristics:

• Each treatment represents one unique combination of all factor settings.
• The total number of treatments equals the product of the number of levels for each factor.
• For example, with 2 factors at 2 levels each (2²), you have 4 treatments.
• Treatments are sometimes called experimental conditions or runs.

Example: If you have two factors (Temperature: 70°C or 90°C; Humidity: 30% or 60%), the four treatments are:

• Treatment 1: Temperature 70°C, Humidity 30%
• Treatment 2: Temperature 70°C, Humidity 60%
• Treatment 3: Temperature 90°C, Humidity 30%
• Treatment 4: Temperature 90°C, Humidity 60%

How DOE Terminology Works Together

These four terms form an interconnected framework:

The Flow: You select factors you believe influence your process. For each factor, you choose levels to test. You then combine factor levels into treatments—each a complete experimental condition. You run the experiment under each treatment and measure the response to see which treatments produce the best results.

Conceptual Model:

Practical Scenario: A manufacturing team wants to improve assembly time (response). They identify two factors: worker experience (levels: novice or expert) and machine model (levels: old or new). This creates 4 treatments. They measure assembly time under each treatment combination to determine which factor has the greatest impact.

How to Answer Exam Questions on DOE Terminology

Question Type 1: Identifying Factors

What to look for: The question asks you to identify what variables are being changed or tested.

Answer Strategy:

• Factors are always input variables or independent variables.
• They are what the experimenter deliberately changes.
• Look for language like: "varying," "testing," "changing," "manipulating," or "at different levels."
• Factors are NOT the outcome or result; they are what you test.

Example Question: "In an experiment testing the effect of injection pressure and mold temperature on plastic part thickness, what are the factors?"

Correct Answer: The factors are injection pressure and mold temperature.

Question Type 2: Identifying Levels

What to look for: The question asks for the specific values or settings of factors.

Answer Strategy:

• Levels are the specific values assigned to each factor.
• Look for numerical values, ranges, or categorical descriptions.
• There must be at least two levels per factor.
• Levels should be distinct and clearly defined.

Example Question: "A study tests paint drying time at temperatures of 20°C, 30°C, and 40°C. How many levels does the temperature factor have?"

Correct Answer: Three levels (20°C, 30°C, and 40°C).

Question Type 3: Identifying the Response

What to look for: The question asks what outcome or result is being measured.

Answer Strategy:

• The response is what you measure, not what you control.
• Look for language like: "measured," "observed," "resulted in," "outcome," or "effect on."
• The response is the dependent variable—it depends on the factors.
• It should be quantifiable or observable.

Example Question: "In a study of welding temperature and welding speed on weld strength, what is the response?"

Correct Answer: The response is weld strength.

Question Type 4: Identifying Treatments

What to look for: The question asks about specific combinations of factor levels or the number of experimental runs.

Answer Strategy:

• A treatment is a unique combination of all factor levels.
• Calculate the total number of treatments by multiplying the number of levels for each factor: (Levels of Factor 1) × (Levels of Factor 2) × ... × (Levels of Factor n)
• Each treatment represents one experimental condition or run.

Example Question: "An experiment has two factors: Material Type (Aluminum or Steel) and Thickness (2mm or 5mm). How many treatments are required for a full factorial design?"

Correct Answer: 2 × 2 = 4 treatments (one for each combination of material and thickness).

Question Type 5: Scenario-Based Questions

What to look for: Complex scenarios requiring you to identify factors, levels, treatments, and responses simultaneously.

Answer Strategy:

• Break down the scenario systematically.
• Identify what is being changed (factors).
• Identify what specific values are used (levels).
• Identify what is being measured (response).
• Calculate or list all combinations (treatments).

Example Question: "A bakery experiments with three flour types and baking temperatures of 180°C, 190°C, and 200°C to optimize bread crust color. Identify the factors, levels, and number of treatments."

Answer Structure:

• Factors: Flour type and baking temperature
• Levels of Flour Type: 3 (the three flour types)
• Levels of Baking Temperature: 3 (180°C, 190°C, 200°C)
• Response: Bread crust color
• Number of Treatments: 3 × 3 = 9

Exam Tips: Answering Questions on DOE Terminology

Tip 1: Remember the Hierarchy

Keep this hierarchy in mind: Factors → Levels → Treatments → Response measured

This helps you organize your thinking when analyzing complex questions.

Tip 2: Distinguish Input from Output

Input variables (Factors and Levels): What you control and manipulate.
Output variable (Response): What you measure as a result.

This distinction is crucial for correctly identifying each component.

Tip 3: Count Treatments Mathematically

When asked about the number of treatments or runs needed, use the formula:

Total Treatments = (Levels of Factor 1) × (Levels of Factor 2) × ... × (Levels of Factor n)

For a 2² design (2 factors at 2 levels each): 2 × 2 = 4 treatments.
For a 3² design (2 factors at 3 levels each): 3 × 3 = 9 treatments.
For a 2³ design (3 factors at 2 levels each): 2 × 2 × 2 = 8 treatments.

Tip 4: Use Clear Language in Your Answers

When answering exam questions, use precise terminology:

• Say "factor" not "variable" (though technically correct, "factor" is more specific in DOE context).
• Say "levels" not "values" (though both can be correct, "levels" is standard DOE terminology).
• Say "response" not "outcome" (more technical and exam-appropriate).
• Say "treatment" or "experimental run" not just "condition."

Tip 5: Watch for Trick Questions

Common Traps:

• Confusing Response with Factors: Don't mistake the outcome (response) for the variables being tested (factors). If a question says "we tested temperature to improve product strength," temperature is the factor and strength is the response.
• Levels vs. Factors: A question might ask "how many factors" when it's really asking about levels. Read carefully.
• Counting Treatments: Ensure you multiply all factor levels, not just add them.
• Multiple Responses: Some experiments have multiple responses. Identify all of them.

Tip 6: Practice with Real Examples

Before the exam, practice with scenarios from your industry:

• Manufacturing: Speed, temperature, pressure as factors; defect rate, cycle time as responses.
• Service Industry: Process type, staff level as factors; customer satisfaction, turnaround time as responses.
• Chemistry: Reactant concentration, pH, catalyst as factors; yield, purity as responses.

Tip 7: Create a Quick Reference Table

During exam preparation, create a summary table:

Tip 8: Understand the Practical Context

Remember why DOE terminology matters in practice:

• Process Improvement: Factors are what you'll ultimately adjust in production based on experimental findings.
• Data Analysis: Correct identification ensures proper statistical tests are applied.
• Replication: Treatments must be clearly defined so others can replicate the experiment.

Tip 9: Review Full Factorial vs. Fractional Designs

While studying DOE terminology, also understand:

• Full Factorial: Tests all possible combinations of factor levels (e.g., 2³ = 8 treatments for 3 factors at 2 levels each).
• Fractional Factorial: Tests only a subset of combinations to reduce resource requirements.

This context helps you answer questions about the total number of experimental runs needed.

Tip 10: Practice Mock Exam Questions

Sample questions to practice:

1. Multiple Choice: "Which of the following is NOT a factor in this adhesive-curing experiment? A) Temperature B) Humidity C) Cure time D) Bond strength"
Answer: D (bond strength is the response, not a factor)

2. Calculation: "How many treatments are needed for a full factorial experiment with 3 factors at 2 levels and 1 factor at 3 levels?"
Answer: 2 × 2 × 2 × 3 = 24 treatments

3. Definition: "Define what a level is in the context of DOE."
Answer: A level is a specific value or setting of a factor being tested in an experiment.

4. Application: "In a study examining employee training method and work environment on productivity, identify all DOE components."
Answer: Factors = training method and work environment; Levels = specific methods and environments tested; Response = productivity; Treatments = each combination of training method and environment

Key Takeaways

• Factors are the input variables you intentionally change.
• Levels are the specific values or settings of each factor.
• Response is the output you measure to assess the experiment's results.
• Treatment is a unique combination of factor levels representing one experimental run.
• Master the mathematical relationship: Total Treatments = Product of all factor levels.
• Always distinguish between what you control (factors/levels) and what you measure (response).
• Use precise DOE terminology in exam answers for credibility and accuracy.
• Practice with real-world scenarios from your industry for context and retention.

By mastering DOE terminology, you'll be well-equipped to design rigorous experiments, analyze their results, and drive meaningful process improvements—all critical competencies for Six Sigma Black Belt success.