Downloaded from https://www.everyspec.com AMCP 706-110 3.28, and 3-29 become 3-6 8-NORMAL PARAMETER, MEAN a p The standard deviation is presumed to be known exactly, and the mean is the param- l= "V0 eter upon which acceptance rests. The I. aO 'n1-0 I(. (3 sample plans ae well known in the quality S•n y - Rn l(3-32) control field. Seq-sample plans are feasible, and some forms of the plans are given in bi,0 . - 1, b-. 1 1 - 1/,y this paragraph, No truncation data are b, - a yreadily available, The equations are adapted from Ref. 23, Chap. XV. h. a/bt Notation followa: o (,y 1)/(Rn 7)'/ (3-33) A a actual mean at 1 W L•2).1 ") aWlo Ij .
a higher mean VA h- + 0 X (•T) 1 1 L - lower mean h h + 80 X (,\iT) (3-34) A a (AN -JAL )/2 r),•-hk+ soxЧ (X, 7) = h + #I x (XIT) a - known standard deviation It is now worthwhile tabulating b 2 , So, s, as aH, -z * probability of incorrect decision functions of X,; see Table 3-9. when A IF sH or PL Example Nos. 27 and 28 Illustrate the pro- x a random sample from s-nonnal cedure. population with mean I and standard deviation a TABLE 3-9 ZL, ZH• (X - ML )/U or (x - As)/O a (Ap)/o, discrimination ratio Sie Eqs.
3.32, 3-33, and 3-34 In the text, FACTORS FOR SEQ-SAMPLE TISTI far 1.25 0.223 1.120 0.o96 1.50 0.406 1.233 0.822 n * number of individual items 2.0 0.W 1,443 0.721 tested so far 2.5 0.910 1.637 0.655 3.0 1.099 1.820 0,607 3.5 1.253 1.990 0.570 The equations for the choose "j is Low" 4.0 1.386 2,164 0.541 or "IA is High" are (when AL is used as a 5.0 1,O6W 2.485 0,497 reference) ___ Se u, sum over all measurements so