PDCA - the scientific method

The Scientific Method

1. Come up with a hypothesis - a theory of how something works, and how something that you might do might affect it.
2. Try an experiment - to test your theory.
3. Observe, study, analyze, and reflect on the results.
4. Do something with your newfound knowledge:
   1. If your theory seems wrong or incomplete - come up with a new hypothesis, and a new experiment to test it
   2. If your theory seems right - apply your new knowledge, and publish your results so others can benefit.

      And then choose your next puzzle to solve - in a virtuous cycle of continuous improvement.

In the world of lean process improvement, various authors and teachers have popularized their own approaches for the scientific method under the acronyms of PDCA, PDSA, SDCA, DMAIC, and many more.

In the not-too-distant future, some new author or consultant will come up with some new acronyms and buzzwords to describe the scientific method. Whatever labels come and go, you can have confidence that as long as you are using the scientific method - you are using a lean system.

---

PDCA Cycle

also known as the PDSA Cycle
because late in W. Edwards Deming's career, he started using the word "study" instead of "check"

Plan Identify problems, goals, metrics, and plans Do Try it, experiment as soon as possible, and ideally on a small scale Check / Study Study, Analyze, and Reflect on results This is by far the most important (yet most likely to be skipped) part of the cycle Act Either standardize to Anchor the change, or Adjust to respond to a failed experiment in a virtuous cycle of continuous improvement

There are 2 layers of PDCA cycles: 1) Strategic (or macro) * Strategic Planning * Policy Deployment * Value Stream Mapping 2) Tactical (or micro) * The A3 Problem Solving Method * Kaizen Events * Other Problem Solving Methods * Project Management * Metrics-based flowcharts * Standard Work * Business Process Engineering * QFD Quality Function Deployment * Six Sigma quality improvement The PDCA Cycle is foundational to every lean six sigma tool and method

SDCA Cycle

For Daily Kaizen...

(continuous improvements consistently found & implemented by Daily Accountability Teams - not kaizen events)

1. There is no planning needed
2. Adhering to Standard Work is even more critical

so the cycle emphasizes Standard Work even more - becoming abbreviated to:

• Standardize - establish work standards
• Do - apply the standards
• Check / Study - Any deviation from the standard should prompt the following 3 questions:
  1. Did the deviation occur because no standard exists?
  2. Did the deviation occur because the standard is inadequate?
  3. Did the deviation occur because the standard was ignored?
• Act - Adjust either the standard or the behavior of the worker that ignored the standard

The Shewhart Deming Cycle Edward Deming (1986) What could be the most important accomplishments of this team? What changes might be desirable? What data is available? Are new observations needed? Plan an experiment to change or test something. Make the change (preferrably on a small scale) Observe the effects Study and reflect upon the results. What can be learned? What can be predicted? Repeat the cycle - accumulating wisdom The Six Steps of Kaizen In their book Toyota Kaizen Methods, Isao Kato and Art Smalley popularized the following Six Steps to Improvement: Discover improvement potential Analyze current methods Generate original ideas Develop an implementation plan Implement the plan Evaluate the new method

---

DMAIC Cycle

Define Measure Analyze Improve Control

The DMAIC Cycle is similar to the PDCA Cycle, and is another way to apply the scientific method. At their core, every tool and method originating from Lean Six Sigma teachings is a unique way of applying the scientific method to unique circumstances that you're likely to encounter on your lean journey. The scientific method is especially obvious in Hoshin Planning Policy Deployment, 8D Problem Solving, A3 Problem Solving, Kaizen Events, and other problem solving methods.

And there are a half dozen acronyms spawned for DMAIC as it is applied to new product development and other special situations...

---

Design of Experiments

aka DOE, D of E, designed experiments

Design of Experiments is a Six Sigma method to design, conduct, and analyze scientific experiments to test assumptions and hypotheses about how different variables affect the outcomes of a process.

Designed experiments are usually conducted in two phases:

Phase 1) Screening Experiment - to identify which of many possible variables have the most significant impact.

Phase 2) Optimimzation Study - to study the few most impactful variables more attentively.

General Procedure:

1. Choose the process to study, and the purpose of the study.
2. Clarify the output measures ("response") to study.
3. Determine the appropriate measurement precision and accuracy
   (perhaps using repeatability and reproducibility studies)
4. Using tools such as flowcharts, brainstorming, root cause fishbone...
   Identify potential variables ("factors") that might affect the output.
   Variables that can be controlled (and therefore experimented with).
   Identify each factor with a letter/label, such as A, B, C, etc.
5. Choose low and high settings ("levels") for each factor,
   and identify them as A+, A-, B+, B-, C+, C-, etc.
6. Document the design of experiments, including:
   • The different combinations of levels ("treatments")
   • How many times each treatment will be performed ("replication")
   • Sequence ("randomization")
7. Identify variables that might interfere with the experiment -
   and plans for how to minimize or at least monitor them.
8. Perform the experiment - carefully - as designed.
9. Analyze the data and document conclusions
   perhaps using Pareto and other types of charts for clearer understanding.
10. Use the conclusions to improve the process, and verify that the changes are actually improvements.
11. Standardize the improved process.
12. Decide whether or what additional experiments should be done next. Do it again.

Like everything else in the realm of Six Sigma... there is more to it than this brief summary.
For a more complete undertanding - refer to the suggested reading: An Introduction to Design of Experiments, by Larry Barrentine.

---

Others

And there are other frameworks for applying a systematic approach to problem solving,
such as the 10 Step Generic Quality Improvement Process in Nancy Tague's book, The Quality Toolbox,

You get the idea... as long as you are applying the scientific method, you are practicing lean six sigma.

---

Lean Six Sigma is a Practice that requires thinking What's the difference between a scientist and a third-grade science student? They both know the definition of the scientific method... but a scientist actually practices the scientific method - with a high degree of skill - to solve endless diverse real-world problems That's your goal... to be a skilled and experienced lean six sigma practitioner... actually practicing the lean six sigma scientific methods - with a high degree of skill - to solve endless diverse real-world problems

"Don't be too timid and squeamish about your actions. All life is an experiment. The more experiments you make the better." ~ Ralph Waldo Emerson