TECHNOLOGY OVERVIEW
NP (Number of Defectives) Chart
BIS.Net Team
NP Charts are used for the same applications that P Charts are used i.e. to control the fraction of non-conforming products, although they can be used to also control fraction conforming (yield).
It has been argued that some non-statisticians find NP charts easier to understand.
As with P Charts Control limits are based on the binomial distribution.
Hence
LCL=Np – 3 * sqrt(NP((1-Np)/sample size)) =0 if negative
UCL= Np + 3 * sqrt(NP((1-Np)/sample size))
Control limits can be placed around a target i.e.
LCL= target – 3 * sqrt(NP((1-Np)/sample size))
UCL= target + 3 * sqrt(NP((1-Np)/sample size))
N is the sample size
P is the process proportion defectives or an estimate if not available.
But this approach implies that it is OK to target a known level of defectives. The only target should, at least philosophically be zero defectives.
If p is unknown p is replaced with an estimate of p.