Overview
Published in 1950, "Some Theory of Sampling" is an early technical work on statistical sampling theory. The publication date is significant — the same year Deming delivered the juse-lectures-1950 that would launch the Japanese quality revolution. This simultaneity reveals the dual track of Deming's career during the japan-and-the-quality-revolution-1947-1960s era: rigorous academic statistics published through traditional channels in America, and revolutionary management philosophy delivered through lectures in Japan.
Technical Content
At 602 pages, this is Deming's major technical work on sampling theory. The book covers probability sampling, stratification, multistage sampling, replicated designs, and — crucially — the relationship between sampling design and cost. Deming treats sampling not as a purely mathematical exercise but as a practical problem of allocating finite resources to obtain information of specified quality. This cost-consciousness reflects his years of practical experience at the u-s-census-bureau and USDA during his usda-and-census-bureau-1927-1946 period.
The book also addresses estimation, variance calculation, and the design of sampling plans for various applications. It is a contribution to the statistical literature, not a management text. The audience is professional statisticians and researchers working on surveys, quality inspection, and experimental design. The work reflects the training Deming received at Yale and his formative exposure to walter-a-shewhart's methods at Bell Labs during his education-and-early-statistical-career-1900-1927.
Intellectual Development
"Some Theory of Sampling" documents a stage in Deming's intellectual development when the statistical and philosophical strands of his thinking were still largely separate. The sampling theory here would later feed into his understanding of variation as a management concept — but in 1950, Deming had not yet articulated the management philosophy that would appear in out-of-the-crisis three decades later. The book is a snapshot of Deming as pure statistician, before the fuller synthesis emerged.
Relationship to Later Work
This book was followed a decade later by the more applied sample-design-in-business-research, showing steady development of Deming's sampling methodology. The theoretical foundations laid here also connect to on-the-distinction-between-enumerative-and-analytic-studies, the 1953 paper that would become central to Deming's epistemology. The progression from sampling theory to the distinction between enumerative and analytic studies to the full System of Profound Knowledge traces Deming's intellectual arc from technical statistician to management philosopher.
Historical Note
The fact that Deming was simultaneously publishing rigorous sampling theory in America and teaching top Japanese executives how to transform their industries underscores the remarkable compartmentalization of his career during the usda-and-census-bureau-1927-1946 period and its aftermath. The American statistical community knew him for books like this one; the Japanese industrial community knew him as the man who had shown them how to build quality into their processes. These two Demings would not be reconciled in American consciousness until the 1980 NBC documentary during the american-revival-and-legacy-1980-1993.
The book remains a foundational text in survey methodology, cited in sampling theory courses alongside Cochran and Hansen-Hurwitz-Madow. Its treatment of replicated sampling designs was particularly influential.
Source
Full text available via Archive.org controlled digital lending (borrowable copy). Published by Wiley (later Dover reprint). The Archive.org copy allows one-hour borrowing for registered users.