R-help to exercise 16 in BSS
# Read the data into a dataframe, give names to the variables, and inspect the data:
gun<-read.table("http://www.math.uio.no/avdc/kurs/STK4900/data/gun.dat", col.names=c("method","phys","team","rounds"))
gun
# Check that the data correspond to those given in the exercise.
# Attach the dataframe
attach(gun)
# QUESTION 1)
# Compute correlations:
cor(gun)
# How are the correlations between the covariates ("method","phys","team")?
# Can you explain the reason for this?
# How are the correlations between the covariates and "rounds"?
# QUESTIONS 2-4)
# Define the covariates as factors (categorical covariates):
method=factor(method)
phys=factor(phys)
team=factor(team)
# Use sum-contrasts (which is common for anova):
options(contrasts=c("contr.sum","contr.poly"))
# Fit a model with main effects and interactions and write the anova table:
gfit<-lm(rounds~method*phys*team)
anova(gfit)
# What does the anova table tell you? Which factors are significant?
# Look at the estimates:
summary(gfit)
# Give an interpretation of the (most important) estimates.