Ingest: Evolution of the PDCA Cycle

Type: kb/sources/types/ingest-report.md

Source: evolution-of-the-pdca-cycle.md Captured: 2026-07-22 From: https://uploads.teachablecdn.com/attachments/aYpoH9SdSDO0k3KYVRRj_NA01MoenNormanFullpaper.pdf

Classification

Genre: scientific-paper -- a conference/journal paper by Moen and Norman that documents a lineage with dated citations and figure-by-figure attribution; it argues a historical thesis rather than reporting an experiment, but the snapshot's scientific-paper classification holds because it is a peer-facing scholarly account with a full reference apparatus. Domains: discovery, improvement-cycles, scientific-method Author: Ronald Moen and Clifford Norman, Associates in Process Improvement (API) — insider participants in the history they narrate (co-authors of the Model for Improvement and correspondents of Deming). High credibility on the API/PDSA lineage, with a corresponding stake in that lineage's framing (see Limitations).

Summary

The paper traces the plan-do-check-act (PDCA) and plan-do-study-act (PDSA) improvement cycles back to the scientific method and the philosophy of science, running from Galileo and Bacon through American pragmatism (Peirce, James, Dewey, C.I. Lewis) to Shewhart's cycle (1939), Deming's "wheel" (1950-51), the Japanese PDCA (1951), Deming's PDSA (1986-93), and finally the API Model for Improvement (1994-2009). Its load-bearing move is Shewhart's own mapping of specification-production-inspection onto hypothesis-experiment-test, from which the authors argue that industrial improvement cycles are the quality-engineering descendants of the scientific method's conjecture-test loop. Read it if you want the dated provenance of PDCA/PDSA, the "check vs study" correction Deming insisted on, or a primary-source anchor connecting improvement cycles to the theory-of-knowledge tradition.

Connections Found

Casebook notes for this material already exist, and this source's durable role is historical evidence for two of them. It corroborates definitions/discovery-lifecycle.md, whose conjecture-consequence-test core is described as adapting Peirce's abduction-deduction-induction: this paper is an independent record of that same Peircean loop being transmitted through Shewhart and Deming into a widely adopted improvement cycle, and it sits naturally beside that note's existing Peirce-SEP evidence. It also grounds a-proposal-selection-loop-requires-search-evaluation-and-retention.md: the paper's PDCA definition supplies exactly the reject-or-retain fork that note treats as definitional ("action -- back to plan if unsatisfactory, or standardization if satisfactory"), making PDSA a long-validated human instance of the note's search/evaluation/retention loop. The snapshot has two sibling sources on the same PDCA-to-PDSA history (foundation-and-history-of-the-pdsa-cycle.md and circling-back-clearing-up-myths-about-the-deming-cycle.md); the three together are candidate raw material for a synthesis note, but none yet abstracts them.

Extractable Value

  1. Shewhart's explicit hypothesis-experiment-test mapping -- Shewhart himself equates specification-production-inspection with "making a hypothesis, carrying out an experiment, and testing the hypothesis," a primary-source datum that strengthens the discovery-lifecycle note's claim that improvement cycles are instances of the scientific loop rather than a mere analogy. [quick-win]
  2. The reject/retain fork in PDCA -- the "act = standardize if satisfactory, back to plan if unsatisfactory" definition is concrete evidence for the proposal-selection-loop note's requirement of a reject-capable evaluator. [quick-win]
  3. "Check" vs "study" correction -- Deming's insistence that PDCA is a corruption because "check" means "to hold back," where PDSA's "study" names the knowledge-building step, is a reusable vocabulary distinction for how a learning loop's evaluation phase should be framed. [just-a-reference]
  4. The three improvement questions -- "What are we trying to accomplish? How will we know a change is an improvement? What changes can we make?" is a compact, reusable planning scaffold that operationalizes theory-based prediction before action. [quick-win]
  5. Candidate synthesis note -- across this and its two sibling snapshots, the transferable claim "the industrial improvement cycle is the quality-engineering instance of the conjecture-consequence-test loop" is not yet captured in kb/notes/, and would give the discovery and self-improving-systems clusters a shared historical grounding. [deep-dive]
  6. Dated provenance table -- the paper pins each transition (Shewhart 1939, Deming wheel 1950-51, Japanese PDCA 1951, PDSA 1986/1993, Model for Improvement 1994/1996/2009) to a citation, a ready reference whenever a note needs to attribute the lineage precisely. [just-a-reference]

Limitations (our opinion)

This is our editorial opinion. As a scientific paper the lens is "what was not tested," and this paper tests nothing — it is a historical-lineage argument by participants in the lineage. The authors are API principals and Deming correspondents, so the framing that the Model for Improvement is "an important evolution of the PDCA cycle" is not a neutral finding; it flatters their own contribution and should be read as advocacy, not verdict. The intellectual genealogy (Galileo → Bacon → pragmatists → Lewis → Shewhart) is argued by juxtaposition and selective quotation rather than by demonstrating that each figure actually transmitted the idea to the next; the Peirce-to-Shewhart link in particular rests on C.I. Lewis as a "bridge" without direct evidence of influence. The claim that PDSA improves on PDCA relies on a linguistic argument about the word "check" plus reported experience, not on comparative results. Treat the paper as authoritative for dated provenance and primary quotations, but as one interested school's reading when it comes to which cycle is "correct" or which improvement framework is best.

Write definitions/discovery-lifecycle.md's evidenced-by link to this snapshot (evidence: evolution-of-the-pdca-cycle.md), recording the paper as historical corroboration that the note's conjecture-consequence-test loop is the same one Shewhart mapped onto hypothesis-experiment-test — the highest-reach edge connect surfaced, and the smallest durable step that pulls this orphaned source into the discovery cluster.