The Shewhart cycle is a repeating loop of steps used to continuously improve a process or product. Originally developed by physicist and statistician Walter Shewhart in 1939, it applies the logic of the scientific method to everyday work: form a hypothesis, test it, study the results, and adjust. You may also hear it called the PDCA cycle, the PDSA cycle, or the Deming cycle, all of which trace back to Shewhart’s original framework.
Shewhart’s Original Three Steps
Shewhart first presented the cycle as three steps arranged in a circle rather than a straight line: specification, production, and inspection. He explicitly mapped these onto the scientific method. Specification is making a hypothesis about what a product or process should look like. Production is carrying out the experiment. Inspection is testing whether the hypothesis held up.
The circular arrangement was the key insight. In a straight line, you’d set a standard, produce something, inspect it, and stop. In a circle, the results of inspection feed directly back into a revised specification, and the loop starts again. This meant improvement was built into the process rather than tacked on at the end. Shewhart framed this within his broader work on statistical process control, where he used control charts to track whether a manufacturing process was behaving predictably or drifting out of acceptable limits. The cycle gave teams a structured way to respond when the charts showed something was off.
How It Became Four Steps
W. Edwards Deming, a student and colleague of Shewhart, expanded the three-step model into the four-stage version most people recognize today. In Deming’s formulation, the stages are:
- Plan: Identify what can be improved and design a change.
- Do: Implement the change, usually on a small scale first.
- Study: Measure and analyze the results.
- Act: If the results aren’t what you hoped for, adjust and try again. If they are, standardize the change and start the next cycle.
This four-step version is where the naming gets tangled. When Deming taught in Japan in the postwar period, his students developed the term PDCA, replacing “Study” with “Check.” Deming himself later pushed back on that substitution. He felt “Check” implied simple verification, a pass/fail judgment, while “Study” captured the deeper analysis and learning he intended. He preferred PDSA (Plan-Do-Study-Act) and considered it a fundamentally different process from PDCA. PDCA, in his view, was more suited to quality control circles dealing with known faults. PDSA was about iterative learning and genuine improvement.
PDCA vs. PDSA: Why the Distinction Matters
“Check” and “Study” sound interchangeable, but the difference shapes how people use the cycle in practice. When you check something, you compare actual results against expected results. When you study something, you ask why the results turned out the way they did, what you learned, and what that learning suggests about your next move. The study phase invites curiosity; the check phase invites compliance.
In healthcare, where the cycle has been widely adopted since the 1990s, this distinction has real consequences. The earliest reported use of PDCA in healthcare was 1993, with PDSA following in 2000. Over time, PDSA use has grown while PDCA use has declined, reflecting a broader shift toward the learning-oriented version Deming advocated. Organizations that treat the third step as genuine study tend to run smaller, faster cycles and learn more from each one, rather than simply confirming whether a target was hit.
The Role of Statistical Process Control
The Shewhart cycle doesn’t operate in a vacuum. It works hand-in-hand with statistical tools, particularly the control charts Shewhart invented. These charts plot measurements over time and flag when a process has shifted beyond its normal range of variation. Without that data, the “Study” step of the cycle would be guesswork.
In a manufacturing case study involving plastic moldings, implementing Shewhart control charts allowed a production team to spot that both the width and length of their products were outside customer requirements. Over three days of iterative adjustments, following the cycle of measuring, analyzing, and fine-tuning, both parameters came into specification and the process stabilized. Production of non-conforming products dropped by 10%. That pattern, diagnose the problem with data, make a targeted change, verify the improvement with more data, is the cycle in action.
How the Cycle Works in Practice
The power of the Shewhart cycle comes from repetition. A single pass through the four steps rarely solves a complex problem. Instead, you run the cycle multiple times, each round building on what you learned in the last. The first cycle might reveal that your initial theory about the problem was wrong. The second might test a revised approach on a small group. The third might scale the change more broadly. Each round narrows the gap between where you are and where you want to be.
This iterative approach works because most real-world problems are too complex to solve in one shot. A hospital trying to reduce patient wait times, a factory trying to eliminate defects in a product line, or a software team trying to speed up its release process will all encounter unexpected variables. The cycle gives each of these teams permission to start with an educated guess, test it quickly, and learn from the results rather than trying to design a perfect solution upfront.
The key discipline is keeping each cycle small and fast. Teams that try to change too many variables at once can’t tell which change produced which result. Teams that wait too long between cycles lose momentum and context. The most effective use of the Shewhart cycle involves tight loops: a focused plan, a contained test, careful measurement, and a clear decision about what to do next.
Where It’s Used Today
The American Society for Quality (ASQ) recognizes the Shewhart cycle, the Deming cycle, PDCA, and PDSA as variations of the same foundational concept. The cycle is embedded in international quality standards, including ISO 9001, which governs quality management systems across industries worldwide. If your organization is ISO-certified, the Shewhart cycle’s logic is already woven into how you’re expected to manage and improve processes.
Beyond formal standards, the cycle shows up in lean manufacturing, Six Sigma, healthcare quality improvement, education reform, and software development. Its appeal is its simplicity. You don’t need specialized training to run a basic improvement cycle. You need a clear question, a way to measure the answer, and the willingness to act on what you find. That combination of scientific rigor and practical accessibility is exactly what Shewhart was after when he drew those three steps in a circle in 1939.