agridat (version 1.8.1)

ryder.groundnut: Wet and dry groundnut yields

Description

Wet and dry groundnut yields with a possible data error.

source

K. Ryder (1981). Field plans: why the biometrician finds them useful, Experimental Agriculture, 17, 243--256.

Details

Ryder (1981) uses this data to discuss the importance of looking at the field plan for an experiment. Based on analysis of the residuals, he suggests that varieties A and B in block 3 may have had their data swapped.

Examples

dat <- ryder.groundnut

# RCB model
m1 <- lm(dry~block+gen,dat)
dat$res1 <- resid(m1)

# Table 3 of Ryder.  Scale up from kg/plot to kg/ha
round(dat$res1 * 596.6,0)

# Visually.  Note largest positive/negative residuals are adjacent
desplot(res1 ~ col + row, dat, text=gen, main="ryder.groundnut")

# Swap the dry yields for the plots and re-analyze
dat[dat$block=="B3" & dat$gen=="A", "dry"] <- 2.8
dat[dat$block=="B3" & dat$gen=="B", "dry"] <- 1.4
m2 <- lm(dry~block+gen, dat)
dat$res2 <- resid(m2)
desplot(res2 ~ col+row, dat, text=gen, main="ryder.groundnut")