1. An exercise for keeping in mind the true model versus the estimated model and the true error versus the estimated error (i.e., the residuals):

Simulate linear regression data in R with white noise.

Simulate the data for y = 5.0 – 4.0x + error, experiment with the variance until the plot clearly looks linear and clearly has random noise (0.75 < R2 < 0.85).

Fit a model.

Plot the data with both the true line and fitted line superimposed on the data.

Find the residuals.

Plot the residuals versus the true noise.

Comment of the two plots created.

Assess the assumptions with appropriate residual plots.