The Bell System Technical Journal 



Vol. XX January, 1^41 No. i 



Single Sampling and Double Sampling Inspection Tables 



By H. F. DODGE and H. G. ROMIG 



Introduction 



A CONSIDERABLE amount of attention has been given to the applica- 

 -^ ^ tion of statistical methods to problems of inspection with emphasis 

 on means for securing certain definite advantages such as reduction in the 

 cost of inspection, reduction in the cost of production by minimizing rejec- 

 tions, and the attainment of uniform quality of manufactured products.^- ^' ^' '' 

 This paper presents four sets of sampling inspection tables that have con- 

 tributed in a notable way to important reductions in such costs and to 

 substantial improvements in control of quality for many characteristics of 

 l)roducts used in the Bell System. 



Whether sampling may be employed to advantage in place of 100% 

 inspection usually depends, of course, on the purpose for which inspection 

 is made. The sampling tables here presented provide definite procedures 

 for conducting inspections that have certain immediate purposes which 

 are described in some detail. Through their provision for instituting a 

 "screening" inspection whenever quality falls below an acceptable level, the 

 procedures have been found in practice to enforce a program of controlling 

 quality in process as the alternative to high inspection costs. 



General Field of Application 



The sampling tables presented herewith have been developed for use in 

 consumer or producer inspections of products composed of similar individual 

 articles or pieces, where it is desired to have assurance of a definite degree 

 of conformance to specification requirements with a minimum of expense. 



The following paragraphs indicate the general conditions under which 

 the tables are applicable, as well as some of the assumptions involved in 

 their development. 



Acceptance Inspection of Lots — The tables are intended for application in 

 inspections whose immediate purpose is to determine the acceptabihty of 

 individual lots of product. 



By a lot will be meant a collection of individual pieces from a common source, 

 possessing a common set of quality characteristics, and offered as a group for in- 



1 



