| % File src/library/stats/man/eff.aovlist.Rd |
| % Part of the R package, https://www.R-project.org |
| % Copyright 1995-2019 R Core Team |
| % Distributed under GPL 2 or later |
| |
| \name{eff.aovlist} |
| \title{Compute Efficiencies of Multistratum Analysis of Variance} |
| \usage{ |
| eff.aovlist(aovlist) |
| } |
| \alias{eff.aovlist} |
| \arguments{ |
| \item{aovlist}{ |
| The result of a call to \code{aov} with an \code{Error} term. |
| } |
| } |
| \description{ |
| Computes the efficiencies of fixed-effect terms in an analysis of |
| variance model with multiple strata. |
| } |
| \details{ |
| Fixed-effect 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 |
| \code{\link{model.tables.aovlist}} and |
| \code{\link{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 non-orthogonal contrasts (e.g., treatment |
| contrasts or an unbalanced design) are detected. |
| } |
| \value{ |
| A matrix giving for each non-pure-error stratum (row) the efficiencies |
| for each fixed-effect term in the model. |
| } |
| \references{ |
| Heiberger, R. M. (1989) |
| \emph{Computation for the Analysis of Designed Experiments}. Wiley. |
| } |
| |
| \seealso{ |
| \code{\link{aov}}, \code{\link{model.tables.aovlist}}, |
| \code{\link{se.contrast.aovlist}} |
| } |
| |
| \examples{ |
| ## 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")) |
| ## IGNORE_RDIFF_BEGIN |
| (fit <- aov(Yield ~ A*B*C + Error(Block), data = aovdat)) |
| ## IGNORE_RDIFF_END |
| eff.aovlist(fit) |
| options(contrasts = old) |
| } |
| \keyword{models} |