Practice Assessment Solution

Author

Joshua Wells

Modified

April 10, 2026

# load your libraries
library(ggplot2)

Read In Your Data

anorexia <- read.delim('https://raw.githubusercontent.com/IowaBiostat/data-sets/main/anorexia/anorexia.txt')

treat_only <- subset(anorexia, Treat != "Control")

Visualization

ggplot(treat_only, aes(x = Prewt)) +
  geom_histogram(bins = 15)
ggplot(treat_only, aes(x = Postwt))+
  geom_histogram(bins = 15)

Numerical Summary

mean(treat_only$Prewt)

[1] 82.88913

mean(treat_only$Postwt)

[1] 87.46957

Conduct a test

pre <- treat_only$Prewt
post <- treat_only$Postwt
t.test(post, pre, paired = TRUE)


    Paired t-test

data:  post and pre
t = 4.16, df = 45, p-value = 0.0001413
alternative hypothesis: true mean difference is not equal to 0
95 percent confidence interval:
 2.362776 6.798094
sample estimates:
mean difference 
       4.580435

Conclusion/Interpretation

There is strong evidence that anorexic patients who received any treatment have higher post-treatment weight on average. Plausible values for the true effect size (weight gain) given by the confidence interval are between 2.4 and 6.8 pounds.