Experimental science is often concerned with comparing two or more treatments. The experiment we use to compare these treatments should be designed as efficiently as possible.
Example #1 constructs confidence intervals on the mean change in murder rates between 1960 and 1970 for Southern cities. The assumptions are checked to establish the efficacy of the confidence interval.
Exercise #1 constructs a confidence interval on the mean change in the distance kicked between air and helium filled football. Assumptions need to be checked.
Exercise #2 constructs a confidence interval on the mean height difference between plant pairs containing cross- and self-fertilized seedlings.