QM BenchQuality engineering tools

Six Sigma tools by DMAIC phase

Six Sigma solves problems with data in five phases: Define, Measure, Analyze, Improve and Control. Every phase has its tools, and here they are in order, free and validated, with practice tests for Yellow and Green Belts.

By Ľubomír Blažek · Process quality team leader, automotive · updated 2026-11-14

Define: what is the problem and is it worth solving?

Measure: can we trust the data and how good is the process today?

Analyze: what really causes the problem?

Improve: which settings and solutions work?

Control: how do we keep the gain?

Learn and practise

What Six Sigma is

Six Sigma is a method for improving existing processes with data. A process at six sigma level produces about 3.4 defects per million opportunities, but the name stands more for the approach than for the number: define the problem in measurable terms, make sure the data can be trusted, prove the cause before acting, and hold the gain with a control plan.

Projects are led by Green Belts (part-time) and Black Belts (full-time) and supported by a sponsor from management. Yellow Belts know the basics and work in project teams.

The DMAIC phases in one sentence each

  • Define: problem statement, goal with baseline and target, scope, team and the cost of the problem in a project charter
  • Measure: a trusted measurement system and a baseline of the process: capability, yield, defect rate
  • Analyze: possible causes collected, narrowed and verified with data and statistics
  • Improve: solutions or settings found, tested in a pilot and checked for new risks
  • Control: control plan, control charts, standards and handover to the process owner

Belts and certification

There is no single official Six Sigma standard body; ASQ, IASSC, universities and companies certify against their own bodies of knowledge, and ISO 13053 describes the DMAIC method. The practice tests on this site follow the usual topics in our own words and are for practice only, not a certification.

Coming to this section

  • · Box-Behnken and response surface designs
  • · Measurement system for destructive tests

Missing a tool you need? Write to info@qmbench.com.

Frequently asked questions

What is the difference between Six Sigma and lean?

Lean removes waste and makes the process flow; Six Sigma reduces variation and defects with statistics. Lean Six Sigma combines both and is what most companies practise today.

Do I need Minitab for a Green Belt project?

No. The tools on this site cover the usual Green Belt statistics: Gage R&R, capability, normality, t-tests, ANOVA, chi-square, regression, multi-vari and two-level factorial DOE, validated against SciPy. Larger designs and advanced models may still need a statistics package.

Are the Yellow and Green Belt tests a certification?

No. They are free practice tests with explanations and a certificate of completion for your own records. For an accredited certification, take the exam of a certification body or your company’s programme.

What does 3.4 defects per million mean?

It is the defect rate of a process whose nearest specification limit is six standard deviations from the mean, assuming the mean may shift by 1.5 standard deviations in the long term. Without the shift, six sigma would mean about 0.002 defects per million.

Which DMAIC phase takes the longest?

Usually Measure and Analyze. Making the measurement trustworthy and proving the real cause take time, but they prevent the most common failure of improvement projects: implementing a solution for a cause that was never verified.

Other areas:LeanShop floor managementQuality inspectionSupplier qualityAPQP