ADM 3301 Lecture Notes - Lecture 26: Denim, Standard Deviation, Statistical Process Control
Document Summary
E. g, a car parts supplier takes and tests weekly samples of 15 subassemblies for truck doors put together on an assembly line. Over 10 weeks, the inspector records the number of defective subassemblies regardless of the extent to which any one is defective. Help the inspector construct a control chart for the performance of the assembly line. size. Week 8 proportion defective is higher than upper control limit. > if yes, delete sample and recalculate control limits, if no, collect fresh data under controlled conditions. The current process is expected to have a 0 to 22% defect rate. > less than 22% defective can be attributed to common causes. > greater than 22% proportion defective is an indication of possible special. > depending on the situation consider number of samples to calibrate control chart. > when the goal is to control the number of occurrences in a process (i. e. , number of defects per unit product) a c-chart is used.