A standard of 0.3 ppm has been established for
formaldehyde emission levels in wood products.
Suppose that the standard deviation of emissions
in an individual board is s = 0.10 ppm. Any lot
that contains 1% of its items above 0.3 ppm is
considered acceptable. Any lot that has 8% or
more of its items above 0.3 ppm is considered
unacceptable. Good lots are to be accepted with
probability 0.95, and bad lots are to be rejected
with probability 0.90.
(a) Derive a variables-sampling plan for this situa-
tion.
(b) Using the 1% nonconformance level as an AQL,
and assuming that lots consist of 5,000 panels,
find an appropriate set of sampling plans from
MIL STD 414, assuming s is unknown. Compare
the sample sizes and the protection that both pro-
ducer and consumer obtain from this plan with
the plan derived in part (a).
(c) Find an attributes sampling plan that has the
same OC curve as the variables sampling plan
derived in part (a). Compare the sample sizes
required for equivalent protection. Under what
circumstances would variables sampling be more
economically efficient?