The table shows results of whether the death penalty was imposed in murder trials in Florida between 1976 and 1987. For instance, the death penalty was given in 53 out of 467 cases in which a white defendant had a white victim.

Originally published in Florida Law Review. Michael Radelet and Glenn L. Pierce, Choosing Those Who Will Die: Race and the Death Penalty in Florida, vol. 43, Florida Law Review 1 (1991).
a. First, consider only those cases in which the victim was white. Find the conditional proportions that got the death penalty when the defendant was white and when the defendant was black. Describe the association.
b. Repeat part a for cases in which the victim was black. Interpret.
c. Now add these two tables together to get a summary contingency table that describes the association between the death penalty verdict and defendant€™s race, ignoring the information about the victim€™s race. Find the conditional proportions. Describe the association, and compare to parts a and b.
d. Explain how these data satisfy Simpson€™s paradox. How would you explain what is responsible for this result to someone who has not taken a statistics course?
e. In studying the effect of defendant€™s race on the death penalty verdict, would you call victim€™s race a confounding variable? What does this mean?

Death Penalty, by Defendant's Race and Victim's Race Victim's Defendant's Death Penalty No 414 37 16 Race White Race White Black White Black Yes 53 Total 467 48 16 Black 139 143