The PDCA or PDSA Cycle began as a series of steps within the world of product development. It wasn’t until much later did it become a method for process improvement. Let’s go through the history together.
Below are the elements in the Toyota House, followed by the rest of the PDCA or PDSA article.
1. The Shewhart Cycle
Dr. Walter Shewhart is best known for his development of statistical control. Back in 1939, Dr. Shewhart struggled with the current view of his time regarding quality and product development. He didn’t believe it was a linear model, but rather a cycle. In his words,
These three steps must go in a circle instead of in a straight line, as shown . . . It may be helpful to think of the three steps in the mass production process as steps in the scientific method. In this sense, specification, production, and inspection correspond respectively to making a hypothesis, carrying out an experiment, and testing the hypothesis. The three steps constitute a dynamic scientific process of acquiring knowledge. 1
In his mind, it looked like this:
2. The Deming Wheel
Dr. Walter Shewhart’s student at the time was Dr. W. Edwards Deming. It was Deming who took the Shewhart Cycle and modified it. In 1950, while speaking to the Japanese Union of Scientists and Engineers (JUSE) regarding statistical quality control, Dr. Deming modified the Shewhart Cycle by adding a 4th step that he called – at the time – “redesign through marketing research”. Historians believe that it was at that meeting the term “Deming Wheel” was born 2.
At that meeting, Dr. W. Edwards Deming discussed these steps:
- Design the product
- Make it and test in lab and production line
- Launch the product
- Test it in service via market research and user interviews
- Redesign and improve the product given customer’s feedback
3. PDCA Cycle is Born
According to Misaki Imai, Japanese executives took the Deming Wheel from that 1950 seminar and called it PDCA and took Deming’s steps (1-4) and codified it as Plan-Do-Check-Act 3.
According to Imai, in the 1960’s, the what was initially the Deming Wheel was called PDCA by the Japanese. It evolved from a method to describe product development to become a method for process improvement.
4. PDSA Cycle is Reborn
In 1986, Dr. W. Edwards Deming reintroduces the Shewhart Cycle and warned his audience that Plan-Do-Check-Act (PDCA) version is not accurate because the English definition for “check” means to “hold back”. Deming continues by encouraging his audience to use Plan-Do-Study-Act (PDSA) and not the “corruption PDCA” 4.
5. PDCA or PDSA?
I’m not picking sides. I’ve always used PDCA. In fact, at Toyota PDCA is the recognized and approved framework, not PDSA. I’m not sure if that matters. I think in the grand scheme of things, in practice it really doesn’t matter. But, to theoreticians and academics, they will likely continue to quibble over PDCA versus PDSA. But for practitioners like you and me, we’ll be busy making things better.
What’s your opinion? Does it matter to you? 5
- Shewhart, W. A. 1939. Statistical Method from the Viewpoint of Quality Control. Department of Agriculture. Dover, 1986, page 45 ↩
- Deming, W.E. 1950. Elementary Principles of the Statistical Control of Quality, JUSE ↩
- Imai, M. 1986. Kaizen: The Key to Japan’s Competitive Success. New York: Random House, page 60 ↩
- Deming, W.E. Personal letter to Ron Moen on November 17, 1990. ↩
- Some parts of this article was taken from “Evolution of the PDCA Cycle” by Ronald Moen and Clifford Norman ↩
Become a Lean Six Sigma professional today!
Start your learning journey with Lean Six Sigma White Belt at NO COST
Ondiappan Arivazhagan "Ari" says
I does matter. I go with PDCA.
However in the above explanation, why 5 steps are needed to explain a 4-step CI cycle.
IMO, the step# “3.Launch the product” is redundant, hence waste (NVA), and can be removed or can be merged with step #4.
It looks like, from the above points, that the real owners of the “PDCA cycle” is the set of unknown Japanese engineers, not Deming. Is it correct.?
Deming should have explained how the word “Study” is superior to the word “Check” in this context and In 1986, the did “study” not mean “examine” also. BTW, the word “Check” also means ” examine (something) in order to determine its accuracy, quality, or condition, or to detect the presence of something”.
I am comfortable with PDCA and Deming was known for his clarity of thought and, I am afraid, not for such confusing remarks on trivial matters.
Guy Ramsey says
Reference opinion #3. The PDSA fits better with the Theory of Knowledge facet Deming’s later work. Checking is not enough, it presumes you know what is true. Deming asks us to question what is true, to consider the system, the stake holders, with an appreciation for variation as a guide. We are not merely checking.
Attila Hujak says
If the “check” is used in general terms and the whole cycle is not just about the product development cycle, the “CHECK” can mean anything that is necessary to do after doing something to evaluate the outcome of the “DO” ! And the main message of the PDCA/PDSA is, not let the loop open! Semantic exercises on the wording could corrupt this great philosophy and approach itself!
Cole D. Lehman says
PDCA does the job. The cycles are the same. And, the secondary definition of “check” as something that:
‘stops or slow down the progress of (something undesirable).”efforts were made to check the disease” ‘
Actually strengthens its relevance to process improvement, it doesn’t hinder it. (think countermeasures)
I appreciate the effort to emphasize thoroughness in that step, but arguing over the finer points of word meanings is something people mostly do to try and establish dominance in expertise of a field by making someone else look wrong.
Arguments over minutiae make for great wastes of time when people could be working on applying the concepts to reality.
Chris McNicholas says
Agree that squabbling about the terminology can be a waste when time and effort can be spent better elsewhere.
I think as long as the cycle is used continually, (i.e. to iteratively develop a change aimed at improvement over time) and other key principles such as use of data over time, making predictions and starting on a small scale, then the labeling of the stages should not matter.
Our systematic review demonstrates a tendency for those using PDCA cycles to only report a single cycle, however I would caution against concluding that this is always the case. Applying the principles of the method can be improved for both uses of terminology. These principles should be the focus to drive learning and improvement, not just capture the narrative of improvement work.
Andrew McDowell says
The differences between PDCA and PDSA are striking and vast, let me explain because there are so many who don’t get this point and it needs be understood.
Deming’s system is the planning and study of actions taken related to manufacturing processes. The study portion is there to increase knowledge about the actions taken. Deming understood that people were not perfect and neither were processes. He understood that all processes contained variation who’s cause was both Common (inherent to the process exa machine setup, employee competence) and Special (effects the process but not inherent to its daily operation exa external power outage). Ignoring this Study step of the functioning process, could result the company not observing variation which is effecting the process which was not thought out or considered during initial planning. Deming knew that every change to a process was not and could not be perfectly thought out, therefore the step of study was added to observe via measurements the end result (whether it be desirable or undesirable or a mix). The Act is in place to start the process over considering what has been learned as a result of the Study.
The PDCA cycle uses check and is all too often; simply verification that predetermined conditions have been achieved (limited focus). It does not indicate that the process needs be studied and determination made as to the results of the study. Check is verification that predetermined conditions have been met while study indicates gaining knowledge of what actually occurred as a result of planned process change. This is why Deming used the word hold back when he saw the word Check substituted for the word Study.
Marriam-Webster defines the words:
Study: Application of mental faculties to the acquisition of knowledge Careful examination or analysis of a phenomenon development or question
Check: A sudden stoppage of a foreword course of progress One that arrests, limits, Restrains – A standard for testing and evaluation – The act of testing or verifying.
Therefore study would indicate determining from analysis what conditions (efficiency, variation etc) are occurring related to a process or its relative changes. Check would indicate a verification that a process change provided some predetermined end result, regardless of the conditions.
Process changes have the ability to provide an end result while mucking up all sorts of conditions. For instance, if a process change placed an operator’s extension and flexion at their greatest lengths but achieved a conforming product, was the change effective? Or, if a process change provides conforming product but a necessity to continually adjust machines, is this desirable? The study of PDSA, would have discovered these things where the check PDCA might allow them to be ignored.
The other failure is to combine PDCA and its siblings 7D and 8D as a companies process for addressing ALL Non-Conformance. These tools are very useful when applied to their intended purpose, which was supplier processes under PPAP control. However I still see many people applying these techniques to ALL their problem solving related to discovered or observed non-conformance. I would caution that processes not under PPAP control should under go good problem solving techniques and not the Automotive PPAP 7D or 8D. The reasoning lies behind Deming’s PDSA and the knowledge of processes which is acquired from study. Further, is the opportunity presented via non-conformance for businesses to perform a Strategic Assessment to consider current state VS desired state, Risk, and Value then choose to completely replace a current process.