(RITA: Washington) -- The nation’s largest airlines set a single-month record in November for on-time performance for the nearly 15 years that the U.S. Department of Transportation (DOT) has collected comparable data, according to the Air Travel Consumer Report released Jan. 7 by DOT. They also set a record for the lowest rate of mishandled baggage in a single month since these data were first collected in September 1987.
|
ADVERTISEMENT |
According to information filed with the Bureau of Transportation Statistics (BTS), a part of DOT’s Research and Innovative Technology Administration (RITA), the 19 carriers reporting on-time performance recorded an overall on-time arrival rate of 88.6 percent in November, higher than both November 2008’s 83.3 percent and October 2009’s 77.3 percent. Since comparable data were first collected in January 1995, the previous high on-time mark for reporting carriers was 88.0 percent in September 2002.
…
Comments
Missing data
Unless I just overlooked it, the significant piece of data that is missing is the number of passangers that flew during the reporting periods. I would guess the numbers or flying passangers is down significantly. As a result those that did fly didn't check as many bags (due to the cost), flights were on time since it takes less time to get away from the gate. I just don't think the report (based on the article) tells us much. When the ratio of customers to employees decreases one should expect better results.
Good point
Dick. Good point. With the exception of baggage errors which were reported as the number of reports per 1000 passengers, the remainder of the numbers were raw data. It would be interesting to know how the total number of passengers (or flights) varied between the reporting periods.
Data Out Of Context
The data discussed in this article is being reported out of context. "Records" were set- So What! The data needs to be presented in the form of Process Behavior Charts so that we can determine if the so-called records represent true improvement or just normal variation. ANY data set will have a high or low "record". Far too often, government reports such data from month to month, touting improvement or deterioration of the "system" from month to month. If the numbers get better from one month to the next, someone claims to know why or take the credit. If the numbers get worse from one month to another, the panic button is pushed. This is treating natural variation (common cause, "noise", etc.) as special cause variation or assignable. When are we going to see federal agencies utilize proper "statistical thinking"?
Steven J. Moore
Director of Quality Improvement Systems
Wausau Paper Corp.
Rhinelander, WI
Add new comment