eff.aovlist
Compute Efficiencies of Multistratum Analysis of Variance
Computes the efficiencies of fixedeffect terms in an analysis of variance model with multiple strata.
 Keywords
 models
Usage
eff.aovlist(aovlist)
Arguments
 aovlist

The result of a call to
aov
with anError
term.
Details
Fixedeffect terms in an analysis of variance model with multiple strata may be estimable in more than one stratum, in which case there is less than complete information in each. The efficiency for a term is the fraction of the maximum possible precision (inverse variance) obtainable by estimating in just that stratum. Under the assumption of balance, this is the same for all contrasts involving that term.
This function is used to pick strata in which to estimate terms in
model.tables.aovlist
and
se.contrast.aovlist
.
In many cases terms will only occur in one stratum, when all the efficiencies will be one: this is detected and no further calculations are done.
The calculation used requires orthogonal contrasts for each term, and will throw an error if nonorthogonal contrasts (e.g., treatment contrasts or an unbalanced design) are detected.
Value

A matrix giving for each nonpureerror stratum (row) the efficiencies
for each fixedeffect term in the model.
References
Heiberger, R. M. (1989) Computation for the Analysis of Designed Experiments. Wiley.
See Also
Examples
library(stats)
## An example from Yates (1932),
## a 2^3 design in 2 blocks replicated 4 times
Block < gl(8, 4)
A < factor(c(0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,
0,1,0,1,0,1,0,1,0,1,0,1))
B < factor(c(0,0,1,1,0,0,1,1,0,1,0,1,1,0,1,0,0,0,1,1,
0,0,1,1,0,0,1,1,0,0,1,1))
C < factor(c(0,1,1,0,1,0,0,1,0,0,1,1,0,0,1,1,0,1,0,1,
1,0,1,0,0,0,1,1,1,1,0,0))
Yield < c(101, 373, 398, 291, 312, 106, 265, 450, 106, 306, 324, 449,
272, 89, 407, 338, 87, 324, 279, 471, 323, 128, 423, 334,
131, 103, 445, 437, 324, 361, 302, 272)
aovdat < data.frame(Block, A, B, C, Yield)
old < getOption("contrasts")
options(contrasts = c("contr.helmert", "contr.poly"))
(fit < aov(Yield ~ A*B*C + Error(Block), data = aovdat))
eff.aovlist(fit)
options(contrasts = old)