The OC Curve is used in
sampling
inspection. It plots the probability
of accepting a batch of items against the
quality level of the batch.
The figure shows an 'OC' (Operating Characteristic)
Curve for a sample of 50 items taken from
a batch of 2000 and using a critical acceptance
number 'c' of 2 (the batch will be accepted
if there are two or less defectives in the
sample). From the curve you can see that
there is about a 23% probability of accepting
a batch that contains 8% of defective items.
When designing a sampling
plan it is usual to decide on two points,
the AQL
and LQL
and the associated Producer's
Risk and Consumer's
Risk. The necessary sample
size and acceptance number for the curve
to pass through these points is then calculated
and hence the shape of the curve.
OC Curves are mainly associated
with sampling inspection but they are also
used to find the Average
Run Length in control
charts.
