In Six Sigma, an alias occurs in Design of Experiments (DOE) when two or more effects become indistinguishable from each other in a fractional factorial experiment. You cannot tell which factor, or which interaction between factors, caused the observed change in the output. This is not a general statistical correlation problem. It is specific to fractional factorial experimental designs used in the Analyze and Improve phases of DMAIC.
Table of contents
Key Takeaways
- Alias is a DOE-specific term. It does not mean general variable correlation. It describes a condition in fractional factorial experiments only.
- Full factorial designs have no aliasing. Every effect is independently estimable. Aliasing only appears when you run a fraction of the possible experimental runs.
- Two effects are aliased when their column patterns of plus and minus signs are identical. You cannot tell which one is causing the observed change in the output.
- Every fractional factorial design has an alias structure. The alias structure lists every aliased relationship. Review it before running the experiment, not after.
- Design resolution describes how severe aliasing is. Resolution III aliases main effects with two-factor interactions. Resolution IV keeps main effects clear but aliases two-factor interactions with each other. Resolution V keeps both main effects and two-factor interactions clear of each other.
- Alias appears in the IASSC Black Belt Body of Knowledge under Design of Experiments in the Improve phase.
- The fix for severe aliasing is more experimental runs. Follow-up experiments break the alias and allow separated estimation of confounded effects.
What Is an Alias in Six Sigma?
In Six Sigma, alias has a specific and technical meaning.
iSixSigma defines it precisely: “Aliasing occurs when the estimate of a factor effect is difficult to distinguish because of the impact of other factors in your experiment. The description of all the possible aliased factors is defined in an alias structure.”
Quality America, citing Paul Keller’s Six Sigma Demystified (McGraw-Hill, 2011), defines alias as a situation “when an independent estimate cannot be made for a parameter which is the alias of another parameter.”
Six Sigma Daily confirms the same concept: “An alias is when the pattern of pluses (+) and minuses (-) in columns are identical.” In fractional factorial designs, this means two effects produce the same pattern of experimental results. There is no way to separate their individual contributions to the output.
Alias is not about general correlation between variables. Alias is specifically about what happens when you run a fractional factorial DOE and do not have enough experimental runs to separate all effects from all interactions.
Public, Onsite, Virtual, and Online Six Sigma Certification Training!
- We are accredited by the IASSC.
- Live Public Training at 52 Sites.
- Live Virtual Training.
- Onsite Training (at your organization).
- Interactive Online (self-paced) training,
Why Alias Only Appears in Fractional Factorial Designs
To understand alias, you need to understand the difference between full factorial and fractional factorial designs.
Full factorial design: You test every possible combination of factors at every level. A 3-factor, 2-level full factorial requires 8 experimental runs (2³). You get complete information about all main effects and all interactions. There is no aliasing in a full factorial design.
Fractional factorial design: You test only a fraction of the possible combinations. A half-fraction of the 3-factor design uses only 4 runs instead of 8. This saves time, materials, and cost.
The trade-off is aliasing. When you run only a fraction of the possible combinations, some effects become mathematically indistinguishable from other effects or from interactions between effects.
iSixSigma explains the consequence directly: “If two factors are aliased, and the DOE shows that these factors have a significant effect on the response, there is no way to tell which factor or interaction is causing the effect.”
Six Sigma Daily states the practical implication plainly: “There really isn’t a use for aliasing. Instead, it’s just good to know that when a full factorial design is fractionated, you lose some of the power of the design.”
The Alias Structure
Every fractional factorial experiment has an alias structure.
The alias structure is a complete list of all the aliased relationships in a given design. It tells you exactly which main effects are aliased with which two-factor interactions, which two-factor interactions are aliased with each other, and so on.
Number Analytics defines the alias structure precisely: “The description of all the possible aliased factors is defined in an alias structure.”
Minitab and statistical software generate the alias structure automatically when you create a fractional factorial design. You review it before running the experiment to understand which effects you will be unable to separate.
A simple example: in a Resolution III design, the alias structure might show that main effect A is aliased with the two-factor interaction BC. If A and BC both appear to have a significant effect on your output, you cannot determine from the data alone which one is driving the result. You need additional experimental runs to separate them.
Alias vs Confounding: The Difference
Alias and confounding are related but not identical.
Six Sigma Daily explains the distinction clearly: “An alias when employing the use of a designed experiments methodology is the pattern of pluses (+) and minuses (-) into columns are identical. Confounding is similar, but it doesn’t mean 100% overlap with the pattern of pluses and minuses in the columns. Perhaps the column might be 80% confounded, or 90% confounded.”
Alias means 100% overlap. Two effects are completely indistinguishable. The overlap is total.
Confounding means partial overlap. The two effects are partially mixed together, but the degree of mixing is less than 100%. Both are problems. Both should be avoided where possible. But alias is a stricter version of confounding.
GreyCampus, in its Lean Six Sigma Black Belt course materials, confirms: “An alias occurs when two factor effects are confused or confounded with each other.”
Design Resolution and Its Connection to Alias
Design resolution is the concept that describes how severe the aliasing is in a given fractional factorial design.
Number Analytics explains the relationship: “The resolution of a design is a key concept that describes the degree of confounding. Higher resolutions imply less confounding among lower-order effects. Lower resolutions may result in crucial effects being masked by aliasing with higher-order, often negligible, interactions.”
The three most common resolution levels in Six Sigma DOE work are:
Resolution III Main effects are aliased with two-factor interactions. This is the most severe aliasing. Number Analytics warns: “Although economical, they tend to be risky when interactions exist and are unknown.” You cannot tell the difference between a main effect and a two-factor interaction.
Resolution IV Main effects are clear of two-factor interactions. However, two-factor interactions are aliased with each other. This is better than Resolution III. You can trust main effect estimates. But you cannot separate two-factor interactions from each other.
Resolution V Main effects and two-factor interactions are all clear of each other. Two-factor interactions may be aliased with three-factor interactions. In most practical applications, three-factor interactions are negligible. Resolution V designs are generally considered strong for identifying main effects and two-factor interactions cleanly.
The general rule: higher resolution means less aliasing but more experimental runs required. Lower resolution means more aliasing but fewer runs.
Texas Lean Six Sigma, an NIST MEP-approved center, notes: “Running only part of the experiment requires less resources, but some factors are confounded. Confounding or aliasing means that the effect of some factors or interactions can’t be separated.”
Where Alias Appears in the DMAIC Cycle
Design of Experiments sits in the Analyze and Improve phases of DMAIC.
Teams use DOE in the Analyze phase to identify which factors significantly affect the process output. They use it in the Improve phase to optimize those factors.
Alias becomes relevant as soon as a team chooses a fractional factorial design. The choice of design determines the alias structure. The alias structure determines what conclusions the team can draw from the results.
Texas Lean Six Sigma confirms: “Screening DOEs are typically done using a Fractional Factorial design. This approach will run a fraction of the total number of experimental runs for a Full Factorial Experiment.” (Source: Texas Lean Six Sigma, March 2024)
A Black Belt leading a DOE must review the alias structure before conducting the experiment. This review prevents drawing incorrect conclusions after the experiment is complete.
The IASSC Lean Six Sigma Black Belt Body of Knowledge includes Design of Experiments as a core topic in the Improve phase. Alias, alias structure, and design resolution are concepts that appear in Black Belt-level DOE content.
How to Manage Alias in Your DOE
Alias cannot always be eliminated, especially when experimental resources are limited. But it can be managed deliberately.
The following four approaches are used in Six Sigma project work:
1. Review the alias structure before running the experiment. Statistical software including Minitab generates the alias structure for any fractional factorial design. Review it before committing to the design. Know which effects will be aliased.
2. Choose a higher-resolution design when resources allow. A higher-resolution design aliases lower-order effects with higher-order interactions. Higher-order interactions are typically negligible. The risk of a false conclusion is lower.
3. Use prior knowledge to make safe aliasing choices. In practice, three-factor and four-factor interactions are rarely significant. Aliasing main effects with these high-order interactions is relatively safe. Aliasing main effects with two-factor interactions is riskier.
4. Run follow-up experiments to de-alias important effects. If the DOE results show a significant effect but the alias structure means you cannot tell which factor is responsible, run additional confirmation experiments. These breaks the alias and separates the effects.
iSixSigma confirms this sequential approach: full factorial experiments give complete information with no aliasing. Fractional factorial designs introduce aliasing as a trade-off for efficiency. The experimenter manages that trade-off by choosing the design carefully and planning follow-up runs when needed.
Alias in the IASSC Black Belt Exam
The IASSC Lean Six Sigma Black Belt Body of Knowledge includes Design of Experiments as a formal topic in the Improve phase.
Within DOE, the IASSC BOK covers fractional factorial designs, confounding, and design resolution. Alias is the term that describes what happens to effects in a fractional factorial design. Understanding alias and alias structure is testable content at the Black Belt level.
At SSDSI, our Black Belt training programs cover DOE in the Improve phase content. That includes fractional factorial design, alias structure, design resolution, and how to interpret DOE results when aliasing is present.
FAQ: Alias in Six Sigma
What does alias mean in Six Sigma?
In Six Sigma, alias refers to a specific Design of Experiments (DOE) problem. An alias occurs when two or more effects in a fractional factorial experiment produce identical patterns of experimental results. You cannot distinguish between them statistically. iSixSigma defines it as a situation where “the estimate of a factor effect is difficult to distinguish because of the impact of other factors in your experiment.” This is specific to fractional factorial designs, not to general statistical analysis.
What is the difference between alias and confounding in DOE?
Alias means 100% overlap between two effects. They are completely indistinguishable. Confounding means partial overlap, such as 80% or 90%. The effects are mixed but not entirely identical. Six Sigma Daily states: “An alias is when the pattern of pluses (+) and minuses (-) into columns are identical. Confounding is similar, but it doesn’t mean 100% overlap.” Both are problems to be managed in fractional factorial design.
Does aliasing occur in full factorial experiments?
No. iSixSigma confirms: “When you conduct a full factorial experiment, you will test all the possible unique combinations of your factors. This gives you a complete picture of the main effect of all your factors along with all the possible higher order interactions. There is no aliasing or confounding of factors.” Aliasing only occurs when you run a fractional factorial design.
What is an alias structure?
An alias structure is a complete list of all aliased relationships in a fractional factorial design. It identifies which main effects are aliased with which interactions, and which interactions are aliased with each other. Statistical software such as Minitab generates the alias structure automatically. You review it before running the experiment to understand the limitations of the design and what conclusions you will and will not be able to draw from the results.
What is design resolution and how does it relate to alias?
Design resolution describes how severe the aliasing is in a fractional factorial design. Resolution III designs alias main effects with two-factor interactions. Further, Resolution IV designs keep main effects clear of two-factor interactions but alias two-factor interactions with each other. Resolution V designs keep main effects and two-factor interactions clear of each other. Higher resolution means less aliasing but requires more experimental runs.
Where does alias appear in the IASSC Black Belt Body of Knowledge?
Alias appears in the Improve phase of the IASSC Lean Six Sigma Black Belt Body of Knowledge under the Design of Experiments topic. The BOK covers fractional factorial designs, confounding, and design resolution at the Black Belt level. Understanding alias and alias structure is required knowledge for the IASSC Black Belt certification exam.
How SSDSI Teaches DOE and Alias
At Six Sigma Development Solutions, our Black Belt programs cover Design of Experiments as a core Improve phase topic.
We teach full factorial versus fractional factorial designs, alias structure, design resolution (Resolution III, IV, and V), and how to use Minitab to generate and review alias structures before running experiments.
We deliver training in three formats. Onsite training brings instructors to your facility for team-based learning. Live virtual training delivers the same instructor-led content in real time online. Online self-paced training lets professionals work through the content on their own schedule.
Every format prepares you for the IASSC Lean Six Sigma Black Belt certification exam. SSDSI is an IASSC Accredited Training Organization.
Ready to master DOE, alias, and the full Black Belt body of knowledge?
Explore SSDSI’s Black Belt programs in onsite, live virtual, or online formats.
About Six Sigma Development Solutions, Inc.
Six Sigma Development Solutions, Inc. offers onsite, public, and virtual Lean Six Sigma certification training. We are an Accredited Training Organization by the IASSC (International Association of Six Sigma Certification). We offer Lean Six Sigma Green Belt, Black Belt, and Yellow Belt, as well as LEAN certifications.
Book a Call and Let us know how we can help meet your training needs.
