Confidence Intervals for Correlation using Atlanta Commutes In Exercise 3.85 on page 204, we use the standard error to construct a 95% confidence interval for the correlation between Distance (in miles) and Time (in minutes) for Atlanta commuters, based on the sample of size n = 500 in CommuteAtlanta. The file BootAtlantaCorr contains 1000 values of bootstrap correlations produced from the original sample.

(a) Describe a process that might have been used to produce the data in BootAtlantaCorr.

(b) Use the information in the 1000 bootstrap correlations (or a similar bootstrap distribution found using StatKey or other technology) to find and interpret a 99% confidence interval for the correlation in this setting.

(c) Repeat (b) for 95% and 90% confidence levels. You do not need to repeat the interpretation.

(d) Describe how the interval changes as the confidence level decreases.