If the airline fight that you are on is 20 minutes late departing, can you expect the pilot to make these minutes up by, say, fying faster than usual? These data summarize the status of a sample of 984 fights during August 2006. (Like the data used to illustrate Simpson’s paradox in Chapter 5, these data come from the Bureau of Transportation Statistics.)
(a) Do you expect the number of minutes that the fight is delayed departing to be associated with the arrival delay?
(b) Make a scatterplot of the arrival delay (in minutes) on the departure delay (also in minutes). Summarize the association present in the scatterplot, if any.
(c) Find the correlation between arrival delay and departure delay.
(d) How is the correlation affected by the evident outlier, a fight with very long delays?
(e) How would the correlation change if delays were measured in hours rather than minutes?