An np-chart is an attribute control chart that plots the number of defective items in successive samples of fixed size. It supports rapid visual monitoring of process stability, helping quality teams identify unusual variation, investigate causes, and apply corrective action. section 14.0 section 14.2.2
Use an np-chart when:
- Each observation contains the same number of inspected units. section 14.2
- The monitored value is D, the count of units that fail the specified quality criteria. section 14.1.7
- A direct count is easier to interpret than the defective proportion tracked by a p-chart. section 14.2 section 14.2.2
- Sample sizes vary, use a p-chart instead; when one unit can contain multiple defects, consider a c-chart. section 14.1.7 section 14.2
Orange juice cans
For the oj data frame, each batch contains 50 cans. Compute the mean defective count, np-bar, then set the demonstrated limits at two standard deviations: UCL = np-bar + 2√np-bar and LCL = np-bar − 2√np-bar. Plot D by sample, draw the mean and limits, and highlight points outside them. The example shows a generally stable process, with two points above the UCL requiring investigation for possible special-cause variation. section 14.2.2
