agridat (version 1.5)

yates.oats: Yield of oats in a split-block experiment

Description

The yield of oats from a split-plot field trial using three varieties and four levels of manurial treatment. The experiment was laid out in 6 blocks of 3 main plots, each split into 4 sub-plots. The varieties were applied to the main plots and the manurial (nitrogen) treatments to the sub-plots.

Usage

data(yates.oats)

source

Yates, Frank (1935) Complex experiments, Journal of the Royal Statistical Society Suppl. 2, 181--247.

References

Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.

See Also

This is the same as the oats data in the MASS package, but includes the coordinates of the field layout.

Examples

dat <- yates.oats
desplot(yield ~ x*y, dat, out1=block, text=gen, col=nitro, cex=1)

# Roughly linear gradient across the field.  The right-half of each
# block has lower yield.  The blocking is inadequate!
require(lattice)
xyplot(yield ~ x|factor(nitro), dat, type = c('p', 'smooth'),
       xlab='x', span=1, as.table = TRUE)

# Typical split-plot analysis
library(lme4)
m3 <- lmer(yield ~ nitro * gen + (1|block/gen), data=dat)
# Residuals still show structure
xyplot(resid(m3) ~ dat$x, xlab='x', type=c('p','smooth'))

# Add a linear trend for x
m4 <- lmer(yield ~ x + nitro * gen + (1|block/gen), data=dat)
xyplot(resid(m4) ~ dat$x, type=c('p','smooth'), xlab='x')

# Compare fits
AIC(m3,m4)