Teaching Tip: Building an SPC Chart with Airline Safety Data

A very interesting article just came out in US News and World Report (Jan.25,2011) that deals with airline safety “incidence reports”. I thought the data might make a good in class example of how to build and interpret a p-chart when you teach SPC in Supp.6. Here is the scenario US News reports:

 All the major US airlines are very, very safe, to begin. Rarely do they end with a fatal crash (the last one was Feb.12, 2009 when Continental Connection #3407 killed 50 people when it crashed in Buffalo). But safety incidents do occur. (Recall the plane that landed in the Hudson River not long ago).  Using FAA and other sources, documented incidents (such as mechanical issues) for the 8 largest carriers follow.

Jet Blue: 17 incidents per 219,000 flights in 2010. This averages to a p- value of .0000776

American Airlines: 87 per 1,241,000 or p=.0000701

United Airlines: 49 per 1,204,500 or p=.0000407

Delta Airlines: 77 per 1,994,725 or p=.0000386

Continental Airlines: 23 per 884,395 or p=.0000260

US Air: 24 per 1,131,865 or p= .0000212

Southwest Air: 23 per 1,131,500 or p=.0000203

Air Tran: 5 per 255,500 or p=.0000196

Take these 8 observations and have the class create a p-chart using these timely, real-world data.  Are any of the major airlines “out-of-control”? ( I computed that the overall p-bar =.000038 (at 95% confidence). The UCL=.000042, and the LCL=.000033. Only two airlines are “in control”, but 4 are better than the LCL. I did this by computing the total sample size to be 8,062,985 with no. incidences =305).

 Thanks to Prof. Kevin Watson at Iowa State for today’s link and idea.

Leave a Reply

Discover more from The OM Blog by Heizer, Render, & Munson

Subscribe now to keep reading and get access to the full archive.

Continue reading