a)Without making any scatterplots, create a regression model for each of the pairs. That is, create regression model for xA v.s. yA,then for xB v.s. yB, then for xC v.s. yC, and for xD vs yD. Report R² for each. You do not have to report the complete model.

b)NOW make a scatterplot for each of the datasets (include the regression line). For this problem you do not have to do residual plots.

c)For which if any of these datasets would you trust a regression model?

d)Thinking question: What does this process tell you? i.e. What lesson do we learn from Professor Anscombe?

\f