There are several types of attribute control charts when the variables are categorical. Two of them are the P and NP charts. The first are used to control proportions, and the second, to control counts, within groups of a certain size, for example, batches, orders in a day, etc.
The form of the P and NP graphs is the same, what changes is the scale: proportions and natural numbers, respectively.
But many times, we can make an P graph but not an NP. Here, we are going to explain very briefly and with a practical case when this happens.
Let’s suppose that we have data from 24 product samples produced per shift in a slaughterhouse. In this case, we have the following database:
- In the column “production” we have the total product manufactured per shift: number of package of lamb ribs.
- The column “n_defectives” indicates the number of defective items per shift (packages that do not meet the required quality standards)
- In the column “N_Defects” we have the number of defects found in the production of each shift (the weight of the tray lower than the established one, wrong labeling, incorrect placement of the pieces, …)
- The column “p” is the number of defective products between the production of that shift.
Note that for each shift the number of defects is equal to or greater than the number of defectives, since a defective item can have one or more defects.