Weight of chickens — chickens • hsds

Data from a randomized blocked experiment to investigate the effect of a drug added to the feed of chicks in an attempt to promote growth. The comparison is between three treatments: standard feed (control), standard feed plus low dose of drug, standard feed plus high dose of drug. The experimental unit is a group of chicks, reared and fed together in the birdhouse. The experimental units are grouped three to a block, with physically adjacent units going in the same block.

Usage

chickens

Format

A data frame with 24 rows and 3 columns:

block: Block corresponding to a bird house
dose: Dose of drug added to the diet (control, low or high)
weight: Average weight per bird (in pounds) at maturity for the group of birds in each experimental unit

Source

Snee, R.D. (1985) Graphical display of results of three-treatment randomized block experiments. Applied Statistics, 34, 71-7.

Examples

library(hsds)
library(ggplot2)
library(lme4)
#> Loading required package: Matrix

# Plot of the average weight per diet, on top of a box plot
ggplot(chickens, aes(x = dose, y = weight)) +
  geom_boxplot(color = "gray") +
  geom_line(aes(color = block, group = block)) +
  geom_point(aes(color = block), size = 2) +
  theme_classic()


# Model of the weight by dose, taking into account the random variation within
# each individual unit
mod <- lmer(weight ~ dose + (1 | block), data = chickens)
summary(mod)
#> Linear mixed model fit by REML ['lmerMod']
#> Formula: weight ~ dose + (1 | block)
#>    Data: chickens
#> 
#> REML criterion at convergence: -43
#> 
#> Scaled residuals: 
#>      Min       1Q   Median       3Q      Max 
#> -1.59348 -0.62704 -0.02187  0.75035  1.42969 
#> 
#> Random effects:
#>  Groups   Name        Variance Std.Dev.
#>  block    (Intercept) 0.000994 0.03153 
#>  Residual             0.004765 0.06903 
#> Number of obs: 24, groups:  block, 8
#> 
#> Fixed effects:
#>             Estimate Std. Error t value
#> (Intercept)  3.86625    0.02683 144.100
#> doselow      0.13875    0.03451   4.020
#> dosehigh     0.17125    0.03451   4.962
#> 
#> Correlation of Fixed Effects:
#>          (Intr) doselw
#> doselow  -0.643       
#> dosehigh -0.643  0.500

# The `modelbased` package is great to estimate marginal means
modelbased::estimate_means(mod, at = c("dose"))
#> Error in loadNamespace(x): there is no package called ‘modelbased’

# Or to estimate contrasts to see if there is a significant difference in weight
# between diets
modelbased::estimate_contrasts(mod, contrast = "dose", p_adjust = "tukey")
#> Error in loadNamespace(x): there is no package called ‘modelbased’