Exercise

Inference with and without outlier (t-test)

Not only can one point change the estimated regression line, but it can also change the inferential analysis.

The datasets with and without the outlier are provided as hypdata_outlier and hypdata_no_outlier respectively.

Instructions

100 XP
  • Run a linear regression of response vs. explanatory on hypdata_outlier, and extract the coefficient-level information.
  • Do the same for hypdata_no_outlier.
  • Look at the output. How do the coefficients and p-values differ?