Here's the equation given:

P'+-3(sqrt (P'(100-P')/N)) where P' is historical value of percent defective and N is number of pieces inspected each week.

After 6 weeks, P'=12, Week 1-6 data is 11, 11, 14, 15, 10, 12. And 600 pieces were inspected each week.

I'm using :=B3+3*(SQRT((B3*(100-B3))/B4))

B3=P'=12

N=3600 (600 per week * 6 weeks)

UCL = 13.6

LCL = 10.4

With the right calculations, all of the values should fall within the control limits, but they aren't for me, so I'm doing something wrong.

(This is from the August 1978 Quality Progress, p. 22, Applications, #3.)

Regards,

leaning