SAMPLING INSPECTION TABLES 



15 



process inspections of parts where the inspection lots tend to lose their 

 identity by merger in a common storeroom from which quantities are with- 

 drawn on order as needed. 



Other things being equal the average amount of inspection for double 

 sampling is less than for single sampling. Fig. 5* gives a direct comparison 

 for the lot protection tables (SL and DL). The saving obtained by using 

 double instead of single sampling is greatest for large lot sizes and low process 

 averages. Over the area of the tables found most useful in practice (per- 



^AVERAGE PER CENT INSPECTED FOR SINGLE SAMPLING 

 ( (NUMBERS IN CIRCLES GIVE AVERAGE PER CENT INSPECTED 

 \^ FOR DOUBLE SAMPLING, AT SPECIFIC POINTS) 



Fig. 5 — Relative amount of inspection, double and single sampling 



centage inspection less than 25 or 30%), the saving generally exceeds 10% 

 and may be as great as 50%. The saving that results from using the 

 double sampling instead of the single sampling AOQL tables (SA and DA) 

 is of the same order of magnitude and may be estimated roughly from Fig. 

 5 by using the associated lot tolerance values listed in the AOQL tables, 

 for a chosen set of AOQL, lot size, and process average values. While the 

 amount of inspection is a major cost item, other costs a.ssociated with double 



* The curves and figures on this chart should be regarded as approximate. The mathe- 

 matical relations involved are such that there exist unique values to be plotted on the 

 M-k plane when certain approximate probability equations, referred to in the appendix, 

 are employed in the solution, but not when exact equations are employed. 



