agridat (version 1.23)

woodman.pig: Weight gain in pigs for different treatments

Description

Weight gain in pigs for different treatments, with initial weight and feed eaten as covariates.

Usage

data("woodman.pig")

Format

A data frame with 30 observations on the following 7 variables.

pen

pen

treatment

diet

pig

pig number

sex

sex

weight1

initial weight in pounds, week 0

weight2

final weight in pounds, week 16

feed

feed eaten in pounds

w0

initial weight

g

average weekly gain

h

half rate of change in growth

Details

Six pigs in each of 5 pens were fed individually. From each litter there were 3 males and 3 females chosen for a pen. Three different diet treatments were used.

Note: Woodman gives the initial weights to the nearest 0.5 pounds.

The w0, g, h columns are from Wishart 1938. Wishart used the weekly weight measurements (not available) to fit quadratic growth curves for each pig and then reported the constants. These are the data that are widely used by many authors.

References

Wishart (1950) Table 2, p 17.

Bernard Ostle (1963). Statistics in Research, 2nd ed. Page 455. https://archive.org/details/secondeditionsta001000mbp

Henry Scheffe (1999). The Analysis of Variance. Page 217.

Peter H Westfall, Randall Tobias, Russell D Wolfinger (2011). Multiple Comparisons and Multiple Tests using SAS. Sec 8.3.

Examples

if (FALSE) {
  library(agridat)
  data(woodman.pig)
  dat <- woodman.pig
  
  # add day of year for each weighing
  dat <- transform(dat, date1=36, date2=148)
  plot(NA, xlim=c(31,153), ylim=c(28,214),
       xlab="day of year", ylab="weight")
  segments(dat$date1, dat$weight1, dat$date2, dat$weight2,
           col=as.numeric(as.factor(dat$treatment)))
  title("woodman.pig")

  # Average gain per week
  dat <- transform(dat, pen=factor(pen), treatment=factor(treatment),
                   sex=factor(sex))
  m1 <- lm(g ~ -1 + pen + treatment +sex + treatment:sex + w0, data=dat)
  anova(m1)
  # Compare diets. Results similar to Westfall 8.13
  libs(emmeans)
  pairs(emmeans(m1, "treatment"))
  # NOTE: Results may be misleading due to involvement in interactions
  #  contrast estimate    SE df t.ratio p.value
  #  A - B      0.4283 0.288 19 1.490   0.3179 
  #  A - C      0.5200 0.284 19 1.834   0.1857 
  #  B - C      0.0918 0.288 19 0.319   0.9456 
}