Finally, authors encourage practitioners to use this procedure for process monitoring. To facilitate
this work, a program in R was developed to solve the optimization problem of double sampling
procedure in np improved chart. In addition, R is available as Free Software under the terms of the
Free Software Foundation’s GNU General Public License in source code form and can be downloaded
easily from the web (see [11]).
Before concluding, it is worth noting that when real data come from a multinomial setup under inspection
sampling and continuous quality monitoring for process control, the proposed methodology
for binomial data can be easily extended to the important case of trinomial data since a 3-category
variable is equivalent to two binary variables. Then, to monitor a trinomial data set is equivalent to
monitor two binomial data sets simultaneously (where the 2-components monitor gives a signal if at
least one of them gives a signal). For example, suppose that the 3 categories are: units without defects,
with mild defects, or with severe defects. Then, one binary variable is obtained joining the first two
categories, and the other binary is constructed joining the last two categories.