acplot(x, alpha=0.4, k=5, ...)
fac.aov
, typically created by fitting
a fractional factorial model using
fac.aov
.
alpha=0.4
as a good starting value. Box and Meyer (1986)
recommend
alpha=0.2
.
k=5
as a good starting value. Box and Meyer (1986) recommend
k=10
.
Posterior probabilities are calculated by the function
accalc
which takes
alpha
and
k
as parameters. The function
accalc
calls a C function which was adapted from Stephenson et al. (1989).
The basic assumption of this approach is that the effects arise
from a scale contaminated normal; namely,
alpha*N(0,sigma^2) + (1-alpha)*N(0,k^2*sigma^2)
NULL
Box and Meyer (1986), An analysis for unreplicated fractional
factorials, Technometrics, 28, 11-18.
Haaland, P. D. (1989), Experimental Design in Biotechnology,
New York: Marcel Dekker.
Stephenson, W. R., F. L. Hulting, and K. Moore (1989), Posterior
probabilities for identifying active effects in unreplicated
experiments, Journal of Quality Technology, 21, 202-212.
buffer.fac <- fac.aov(buffer.df) acplot(buffer.fac,0.2,10) acplot(buffer.fac,0.4,5)