Case 1:

Modified last: October 27

HP is America West.

Another link to an airport code site: http://www.uni-karlsruhe.de/~un9v/atm/ase.html

Questions and answers. -last modified October 27

Here is an Excel workbook that illustrates how to easily select random samples. Automatic calculation is turned off, so you have to manually recalcuate when you want to "sample."

Data:

ASQP data elements description (excel)

ASQP data for July 1997 (tab delimited text) - for flights into and out of Boston

ASQP data for February 1998 (tab delimited text) - for flights into and out of Boston

ASQP data for February 1998 with column headings

Airport codes - a list of the three letter codes

http://www.airnav.com/ - details about airports; can search for codes not in the airport codes link

http://www.fare.net/air.html - list of airline web sites, includes IATA carrier code, e.g., AA

 

  1. Look at the AA (American Airlines) flights between BOS (Boston Logan) and SJC (San Jose, California) as well as the AA flights between SJC and BOS.
    1. Graph the frequency distribution of the block times for each of the flights.
    2. Calculate the mean and standard deviation for each flight number's (e.g., 128) block times.
    3. Do you notice anything interesting? What?
  2. Assume each flight is a Bernoulli trial. Assume that success could be defined in two different ways: 1) in which sucess is defined to be "actual arrival time within 15 minutes of scheduled arrival time" and 2) in which success is defined to be "actual arrival time less than or equal to scheduled arrival time." Assume that for passengers, p (the probability of success) is the primary measure of system performance. Analyze system performance across different city-pairs (origin and destination for a given flight), different airlines and different seasons.
  3. Assume that the ASQP data includes all flights into and out of BOS (it doesn't). Estimate the maximum arrival acceptance rate and maximum departure rate at BOS. Analyze the operation of BOS relative to these maximum rates.
  4. Are the mean block times between city-pairs the same independent of direction? Are the mean block times significantly different in different months?
  5. Analyze delays (e.g., departure, arrival, total) for flights into and out of Boston. A few suggestions (don't limit yourself to just these):
    1. Develop a delay model. Estimate various parameters related to your model. Consider factors such as city-pairs, airlines, congestion, season, time of day, congestion, etc. See if there are significant differences in delay parameters across these factors.
    2. Analyze the relationship between tardiness (define tardy as: a flight is tardy if it arrives after the scheduled time) and scheduled block times. Analyze the relationship with other variables. Consider factors such as city-pairs, airlines, congestion, season, time of day, congestion, etc.