Airport Passenger-Related Processing Rates Guidebook (2009)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

This chapter provides guidance on managing day-to-day field operations during a data collection event. Airports are unique places for collecting data and a failure to adequately take this factor into your study will seriously compromise your results. In Chapter 7, the proposed action plan and strategies are tested through site visits and field testing. Successful implementation of the final action plan is met through team preparation. Finally, an overview of two data collection events is presented as a checklist. 7.1 Understanding the Environment In Chapter 6, the importance of coordinating research efforts with airport personnel is empha- sized. The world, however, is imperfect, and even when one does solicit and get cooperation, problems can still arise. At one airport, for example, the dominant carrier refused to permit data collectors access to its gate hold areas, essentially crippling the study. The purpose of the study had been communicated to all station managers through memos from a high-level airport administrator, and all other airlines had signed off on the study. Eventually, a face-to-face meet- ing with airline representatives, airport management, and the lead consultant was required to obtain cooperation. There are many other examples of the unique challenges to data collection in an airport. Many of these challenges intensified after the September 11, 2001, terrorist attacks, as airports and the agencies working in them become more sensitive to potential security threats. This section addresses how to better understand the dynamics at an airport that can impact the study, as well as how to reduce the likelihood of unanticipated problems. In particular, consider both the bureaucratic and the physical environment of airports. 7.1.1 The Bureaucratic and Physical Environment In many ways, an airport is like a microcosm of a small city. Its functions are myriad, and author- ity is spread across multiple agencies and staff with differing reporting relationships. As such, coordination must be maintained with those with functional responsibility for the following: • Airport management (those not directly involved in the study); • TSA; • Airport police; • Airport operations; • CBP; • Airlines; • Transit agency (if transit station is on-site); and • Ground transportation providers. 70 C H A P T E R 7 Managing and Implementing Data Collection

The amount of communication and coordination with these entities can vary significantly depending on factors such as the following: • Existing relationships among the organizations; • Security threat levels; • Their prior experience with similar data collection events; and • Local, regional, and national policies (relevant to national agencies). Largely related to security issues emerging after September 11, 2001, protocols and policies may change substantially over time at both a local and national level. For example at one large airport, prior to September 11, 2001, data collection personnel could enter security along with passengers, well-wishers, and greeters. The next year (after the terrorist attacks), the FAA/TSA required personnel to present a letter from airport management with the person’s name on it. The following year, personnel were required to have a security badge. It is important to moni- tor changing policies and practices, and to prepare for changes sufficiently in advance of data collection. (See Chapter 6 for scheduling recommendations.) 7.1.2 Site Visit and Reconnaissance As the one responsible for collecting the data, the data collection project manager should visit the functional element(s) to be studied. The purpose of this site visit is to observe processing, clarify the particular data elements to be collected, and confirm how special situations, if any, should be handled. Most importantly, these initial observations allow the researcher to see the elements in their setting. The following are some recommendations for the site visit: 1. Invite the researcher (end user) to accompany you. This will not only contribute to a com- mon understanding between the two of you but may also result in modifying elements of the project’s purpose and scope. 2. If possible, schedule the visit for a similar time period to the one anticipated for the actual data collection event. 3. Meet with airport staff to identify potential modifications to the existing layout prior to the scheduled data collection. 4. To the extent possible, pre-test the data collection protocol, data recording instruments, and so forth. Is there a situation that wasn’t initially anticipated? Do the data entry sheets/templates need to be modified? Will additional staff be needed to cover the desired locations? The next section provides a recommended field testing approach. 7.1.3 Field Testing Research projects are costly efforts. While you can try to anticipate contingencies in your plan- ning, a formal walk-through before collecting usable data will permit you to make needed mod- ifications to forms, schedules, and procedures, thereby increasing the likelihood of gathering reliable data. What has been proposed in this Guidebook is grounded in a substantial and mul- tidisciplinary body of research. The concept in its most basic form is that, before you can trust data, you need to validate that you are collecting the “right” data (validity) and doing so in the right manner (reliability). After Deming,1 the approach is sometimes referred to as Plan, Do, Study, Act which means the following in the context of this chapter: • Plan involves specifying the performance criteria for data collection methods and materials (see Chapter 3); Managing and Implementing Data Collection 71 1Deming, W. E. (1986). Out of the Crisis. MIT Center for Advanced Engineering Study.

• Do refers to implementing these methods and materials on a trial basis; • Study relates to evaluating performance in light of requirements, and; • Act is using the feedback to make changes for improvement. You might reasonably think of this as a calibration activity. The approach often also incorporates a triangulation method that is commonly used in research to enhance confidence. In both surveying and navigation, triangulation is used to deter- mine coordinates and distance to a point using two other known reference points of a triangle. In research methodology, triangu- lation implies using multiple sources of information, multiple methods, and/or multiple researchers. In an organizational setting in which one is studying employee satisfaction, for example, interviews may be collected with persons from different units and at different hierarchical levels (multiple sources). To supple- ment interviewing as a data gathering strategy, the researchers might also review employee turnover statistics as well as survey employees (multiple methods). Finally, as an illustration of multiple researchers, two or more researchers may participate in analyzing focus group data, independently reviewing the data and then meeting to identify where the researchers’ conclu- sions converge and diverge. As an analogy, if it looks like a duck, quacks like a duck, and walks like a duck, the conclusion that it is a duck is more persuasive because more than one attribute was examined. A Generic Field Testing Process Passenger check-in is used to illustrate the process, but it can be adapted to other data gather- ing foci. Schedule Field Test Session. Arrange for all persons involved in the pilot test to meet approximately 1.5 hours prior to the actual start. If possible, orientation should be done on site (i.e., at the study airport), in a conference room facility capable of accommodating the entire survey staff. An airport representative should make welcoming remarks noting the airport’s significance to the region and the importance of the par- ticular study. An example of a training outline and agenda, as well as sample orientation materials is included in Appendix C of this Guidebook. Field Test Orientation. Define the purpose of conducting a field test and refer partici- pants to that section of the handout related to field testing. Depending on the number of per- sons in the team, consider having two persons collect the same data on the same elements at the same time. In other words, both persons observe and record identical data for purposes of calibration. Review task particulars and times. Synchronize watches, and direct the team to the field-test location. Initiate Field Test. Allow approximately 15 to 20 min for data collection. While the team is collecting data, observe each person, looking specifically for any factor that might impact that person’s ability to collect those specific data for which he or she has responsibility. Note any unanticipated problems or difficulties that the staff is having. At the end of the period, meet with the team to debrief. In particular: • Ask each team member to comment on any questions, problems, etc., that arose during the trial. • Review the logic of the observation strategy for possible change. • Have each person run a quick calculation of averages, ranges, and standard deviations and have each person report the results. 72 Airport Passenger-Related Processing Rates Guidebook In research, triangulation involves use of multiple researchers, multiple methods, and multiple sources to enhance the credibility of the findings.

Review Field Test Results. Review the results in regard to the following: • Consistency in data where expected. For example, if two persons were making identical obser- vations but of different check-in lines, are the averages relatively similar? Are there relatively large standard deviations for each of these observers? If so, is there something different between the two lines that suggest that you should not treat them alike? If so, examine the ranges to see if it might be a function of one or two extreme values. • Consistency across observers. If you chose to have two persons making identical observations for part of the trial, do the observations match? While a small amount of observation is nor- mal, you should not expect to find wide variation. If there is, what were the two persons doing differently? Why are they not calibrated? • Consistency in data where not expected. If you have opted to stratify based on the assumption that groups will differ in some way (e.g., business and first class passengers will, on average, be unalike in regard to time to check-in), have these differences emerged? If the averages and standard deviations are relatively similar, should you consider ignoring the stratification strat- egy and treat them as a single group (i.e., combining business class and first class)? Spend a few minutes by yourself to consider if you want to make any changes. Then recon- vene the team, communicate any changes, and direct them to begin actual data gathering. 7.2 Team Preparation and Training To help ensure a successful data collection event, a staff training session will be necessary. It is recommended that this be broken out into three components: 1. Overview of study and role of data collection, 2. Procedural/management issues, and 3. Actual training. 7.2.1 Conduct Team Briefing While some staff preparation might have already occurred during the planning phase, it is still important to communicate to the assembled data collection team the overall purpose of the study, i.e., the actions that might arise as a result of the data. This helps to place the frequently mundane task of data collection in the greater context of the study. There is evidence from the research literature that performance is enhanced when workers understand not only how to per- form a task but why the task is being done and how it fits into a larger framework. 7.2.2 General Requirements and Logistics It is important to stress to staff that the data must be collected at the designated times (typi- cally, peak times) and the importance of meeting their agreed upon responsibilities, i.e., that if even just one person doesn’t show up, the entire event can be compromised. Once the scheduled period is gone, it’s too late, particularly if the period was a peak time. Finally, given that other data collection events may have been scheduled, it is often very difficult to make up the missed one. Regardless, even a highly motivated and well prepared group of people will, from time to time, encounter unexpected and unavoidable circumstances. Emphasize the importance of con- tacting the supervisor at the first indication of a potential conflict. Give each data collector a handout which summarizes the purpose of the study, his/her role, contact information of other team members (including key airport contacts), a detailed sched- ule, and instructions for data collection events in which he/she would be participating (see Appendix C for examples). Managing and Implementing Data Collection 73

The amount of time spent on each of these items and when they are covered will depend on the size and complexity of the event, its schedule, and experience of the personnel. For a large study, for example, with many inexperienced personnel, it may be appropriate to hold the introductory session the day before the actual data collection event. At this first session, one could outline the general study and how the collected data will be used to make a recom- mendation. One could then review procedural issues, such as when to report, how to dress, who key airport and study team members are and how to get in touch with them, and how to record data. The next item on the agenda would be a walk-through of the airport, including visits to the sites where data collection will take place. 7.2.3 Specific Data Collection Training The next item on the agenda would be the actual training. It is recommended that this occur as near as possible to the actual data collection event—preferably, immediately prior to it. Sufficient time should be allowed (at least 30 min, preferably, more) to allow staff to enter sam- ple data and iron out the exact process. During this time, staff can also be placed in position. For example, surveyors monitoring skycap transactions at the curb would be taken outside, those monitoring kiosks would be shown the best location for monitoring those transactions, and those monitoring agents would be placed in their positions. Any personnel acting as “floaters” (i.e., surveyors who will be giving breaks to the primary surveyors) would need to understand the data collection process and location of each position. 7.3 Data Collection Case Studies The previous chapters reviewed typical airport passenger-related processes, provided guid- ance on defining the research and developing sampling plans, and gave direction on determin- ing when to sample and how to staff the survey. The previous sections of this chapter focused on reconnaissance, field testing, and training. In the next section of Chapter 7, a typical data collec- tion event at an airline check-in counter is described. Finally, Sections 7.3.2 and 7.3.3 provide guidance for collecting data at bag claim and restrooms. Appendix E provides example templates for recording data using paper and pencil. 7.3.1 Collecting Data at an Airline Check-in Facility This section includes a one-day data collection event at a moderately sized airport where check-in processing rates are being estimated. Based on an understanding of the project (See Chapter 3), the end users wish to compare processing rates among three airlines (versus simply obtaining an aggregate number). In addition, a distinction is to be made between first class and coach customers as well as those using express and full service e-ticket kiosks. The peak hour, in terms of seat departures, for Airline A is between 6:30 a.m. and 7:30 a.m.; the peak hour for Airline B is between 7:00 a.m. and 8:00 a.m.; the peak hour for Airline C is between 8:30 a.m. and 9:30 a.m. Assuming passengers will reach the check-in counter about 90 min prior to their scheduled departure time (Section 5.6.3), it is estimated that the peak period of activity at the check-in counter will actually be between 5:00 a.m. and 6:00 a.m. for Airline A, 5:30 a.m. and 6:30 a.m. for Airline B, and between 7:00 a.m. and 8:00 a.m. for Airline C. Allowing for the desire to also collect demand profiles at each counter, one hour before and after each scheduled peak is applied. This means that the data collection schedule would be as follows: • Airline A from 4:00 a.m. to 7:00 a.m. • Airline B from 4:30 a.m. to 7:30 a.m. • Airline C from 5:00 a.m. to 8:00 a.m. 74 Airport Passenger-Related Processing Rates Guidebook

Sample sizes are calculated (Section 5.5), a sampling plan is developed, and a data collection technology and staffing method are selected (Section 6.1.4 and 6.3). A general, preliminary action plan is generated, and refined after a reconnaissance visit and field testing effort (Section 7.1.3). Based on the scope of the effort, the various sources for staffing (airport staff, consultant staff, temporary agency staff, or market/research staff) are reviewed and a preferred source is identified. Recognizing the early start times, the employees are asked to come the previous day for training and a walk through of the airline check-in facility (Section 7.2). On the day of the survey, the survey personnel arrive about 30 min prior to the scheduled start time. The project manager distributes survey materials; positions staff (as shown in Exhibit 7-1); and reviews procedures. Once data collection begins, the project manager allows 10–20 min to transpire and then revis- its each surveyor to review data and address any remaining questions. At the conclusion of the event, survey materials are collected, and surveyors are queried as to whether they observed anything out-of-the-ordinary that would warrant documentation. 7.3.2 Collecting Data for a Bag Transfer Timing Study This section provides guidance from measuring the interval between an aircraft’s arrival at the gate and the appearance of its bags at the bag claim device. There are two methods that can be used. The first is having each surveyor collect data for a particular flight. In this instance, the data collector would travel to the gate to record an aircraft’s arrival time; he/she would then head directly to the bag claim area to await the arrival of the first bag from that flight. A second option would be to have two data collectors—one in the concourse and one at the claim area. The two would then relay information about a particular flight via cell phone. With this setup, however, it is imperative that their watches be synchronized for accuracy. The passenger boarding time includes getting passengers onto the aircraft and seated and also having all their carry-ons stowed. Unless the survey effort is being undertaken directly by or for an airline, data collection will not occur onboard the sample aircraft, but rather at the gate; therefore, the definition of the start and finish time for boarding will, by necessity, be different. Managing and Implementing Data Collection 75 Curbside Check-in 1st Class Passengers Kiosks Express Kiosks Dom. Coach Passengers Full-service (no chkd bags) Int’l Travel Exhibit 7-1. Airline check-in staffing locations for case study.

In this instance, each discrete activity should be recorded by time, expressed in hours, minutes, and seconds. Therefore, a PDA that can “time-stamp” their occurrence would be ideal. Items to record include the following: • Time when gate door opens for boarding; • Time of pre-board announcement; • Time when each pre-board passenger enters the loading bridge; • Time of first-class boarding announcement; • Time when each first-class passenger enters the loading bridge; • Time of each successive boarding announcement (whether by row or zone); • Time when each passenger enters loading bridge; • Time of final boarding announcements; • Time when gate door closes; and • Time when aircraft door closes (or when loading bridge is pushed away from aircraft). If a PDA is not used, the time when each passenger enters the loading bridge should be noted, until the last passenger’s entrance time is recorded. Enplaning rates vary by aircraft size, composition of passengers, destination (e.g., domestic or international), and individual airline practices. It is recommended that your sampling plan take these variables into consideration, and if separate analyses of these factors are desired, the sam- pling plan should include stratification (Section 5.2.1). For drawing a sample of departing flights only, the most straightforward method is to obtain an OAG schedule pull and randomly draw flights based on the strata you have identified. For arriving flights, it is often much less efficient to identify sampled flights from an OAG due to the greater variability between a flight’s scheduled and actual arrival times, and because there often seems to be a general lack of information regarding a particular flight’s arrival gate, especially at large airports. For these reasons, it is recommended that a different sampling plan be used to obtain deplaning rates (see Section 5.2.6). 7.3.3 Collecting Restroom Use Data Based on previous restroom data collection and planning efforts, it appears that restroom facility planning (in particular determining the number of fixtures required to accommodate demand) is most closely tied to simultaneous demand (i.e., how many fixtures are occupied at any given moment). A method for collecting this demand is described below. It is recommended that collecting simultaneous demand data for restroom use be accom- plished by teaming a male and female surveyor and having this pair enter the respective rest- rooms at regular intervals to record the number of fixtures occupied at that time and the number of people in queue. (An initial inventory of fixtures—including those which are out of order—needs to be conducted prior to the survey.) It is recommended that the observations be obtained at intervals no greater than 10 min because longer periods would increase the chance of missing the peak. Determining the actual time period for the survey should be based on looking at scheduled seat arrivals and seat departures, with more weight given to the peak arrival hour, since the deplaning process results in greater surges of restroom demand. It is recommended that the event begin about 1 hour before the anticipated peak hour and con- clude 1 hour after the anticipated peak hour to minimize the chance of missing the actual peak. For example, if the scheduled peak hour is between 2:30 p.m. and 3:30 p.m., the data collec- tion period should begin at 1:30 p.m. and end at 4:30 p.m. 76 Airport Passenger-Related Processing Rates Guidebook When monitoring a flight, confirm the flight number and origin by asking a nearby gate agent or a deplaning passenger.

Exhibit 7-2 shows an actual example of simultaneous demand results at a women’s restroom at a large international airport. As shown, a capacity of 11 fixtures is clearly inadequate to accom- modate the surges of demand, which reached 29 women at about 5:30 p.m. This resulted in a queue of more than 15 women. Managing and Implementing Data Collection 77 0 5 10 15 20 25 30 35 1700 1710 1720 1730 1740 1750 1800 1810 1820 1830 1840 1850 1900 1910 1920 1930 Number of Stalls D em an d Exhibit 7-2. Example of simultaneous demand for women’s restroom facilities.