These classes of objects are returned from
fac.aov.
An object of class
"fac.aov" represents the fit for a fractional
factorial analysis of variance.
The class
"fac.aov" has methods for the following generic functions:
pareto
,
plot,
predict,
print,
print.summary,
qqnorm,
summary.
Class
"fac.aov" inherits from class
"aov".
The
fac.aov object is implemented as a list
with the following components in addition to an
aov object:
fcoefficients or half-effects). For unbalanced
or nonorthogonal designs, the
feffects are calculated from the
fcoefficients
using the value of
adj.coef to make all of the
standard errors equal. (See below.)
feffects. However if a
fac.aov object has been
refitted with
update, forming a non-saturated model, the saturated
effects are stored here, and the reduced model effects in
feffects
feffects =
fcoefficients*adj.coef
so that
feffects have a constant standard
error, a property that is required for the computation methods used
for
pse,
tse and
ase.
aov.