Scenario: Let’s look again at some airline delays data from Chicago. This time, we will consider all flights from March 2019 and focus on departure delays for United Airlines flights. We will need to filter the data found in Airline delays 3.

Start by using the global Data Filter to include only flights where the Carrier ID is UA, the Origin is ORD, and Cancelled = 0.

a. The departure time of a flight refers to the difference, if any, between the scheduled departure and the time at which the plane rolls away from gate. Create an XBar-S chart for DepDelay grouped by Date and report on the extent to which this process seems to be under control.

b. Create a new temporary variable as follows: highlight WheelsOff and right-click.

Choose formula, and edit the formula to equal WheelsOff – Sched Dep. This will calculate the elapsed time between scheduled departure and the moment when the wheels leave the runway. Create an XBar-S chart for this variable by Date and report on the extent to which this process seems to be under control. How does it compare to the previous chart?

c. Create a Pareto Plot for the column Main Cause. Describe what the plot tells you.

Which causes combined account for more than half of all delays?

d. Create a variability plot of DepDelay by Main Cause. Discuss your findings.

e. (Challenge) Imagine that the airline would like to control departure delays with a goal of having 90% of all non-canceled weekday flights leave the gate 20 minutes or less from the scheduled time. Run a capability analysis and report on your findings.

(Hint: the column called Weekend is a dummy variable equal to 1 for Saturday and Sunday flights.)