- First, we will revisit the mtcars data set built into R. Attach the data file and review the variable names.
attach(mtcars)
names(mtcars)
We will begin by looking at some diagnostic tools, starting with the pairwise simple correlation values amongst all variables presented in the correlation matrix. To obtain this enter:
cor(mtcars)
Paste the results. Is there evidence of multicollinearity? State a few specific correlation values to support your answer.
- Now let’s look at the variance inflation factors (???)?. To do this, we must define the estimated regression function for predicting mpg based on all other variables and then find the ??? values. The function for ??? is in the car package.
Enter:
library(car) > reg = lm( mpg ~ ., data=mtcars ) > vif(reg)
Paste the results. What can you conclude about multicollinearity based on these values?
Sample Solution
