Shewhart Walter A. Economic Control of Quality of Manufactured Product

New York, D. Van Nostrand Company, 1931. — 467 pp.This important book, written by a recognized authority in the fast developing field of mathematical statistics, comes as a welcome addition to the all too sparse collection of expository books. Those who have wished in their statistical courses for practical problems to illustrate sampling theory, the correlation surface, frequency distribution of parameters such as the mean and standard deviation, and the use of Chi Square, will be enthusiastic about the splendid collection of practical problems most of which have their origin in the telephone business.
After briefly indicating in a general way in Chapters 1, 2, 3, and 4 how it is possible to use modern statistical theory to control quality of the manufactured product, Dr. Shewhart digresses to give a practical presentation of statistical theory, especially the modern theory of sampling. Some elementary but important problems of presenting data by tables and graphs are considered in Chapters 5 and
6. Such statistical concepts as arithmetic mean, median, mode, standard deviation, skewness, kurtosis, correlation coefficient, and correlation ratio are defined and their calculation is illustrated in Chapter
7. A discussion of the relative usefulness of these statistics and a development of Tchebychefï's theorem are given in Chapter
8. A study of correlation and relationship is presented in Chapter
9. Laws basic to the control of quality, that is, the law of large numbers, the point binomial, and the meaning of statistical laws are described in Chapter
10. The development of statistical concepts will be found to be sufficiently complete for those who do not require the details of difficult proofs. Those who wish to investigate proofs will find ample references to periodical literature. An excellent bibliography is given as Appendix III. Chapters 11 to 21 can best be described by presenting outlines of some of the problems of control.
At present we are coming more and more to realize that we cannot do things just as we would like to do them. We can, however, approach a state of perfection by eliminating known causes of variation from standard. Thus, we may have control within limits. In Dr. Shewhart's words: "since we are thus willing to accept as axiomatic that we cannot do what we want to do and cannot hope to understand why we cannot, . we must also accept as axiomatic that a controlled quality will not be a constant quality." "A phenomenon will be said to be controlled when through the use of past experience...we can state, at least approximately, the probability that the observed phenomenon will fall within the given limits."