The survey procedure was as follows: Lawrence Berkeley National Laboratory (LBNL) security personnel directed vehicles out of the traffic stream and into a coned-off survey lane in groups of approximately six at a time, corresponding the number of survey personnel in each direction. The interview generally took about 45 seconds to complete. When all interviews in a group were complete, the six-vehicle queue was released, and another group was directed into the lane. Prior to releasing the vehicle, surveyors handed the drivers an informational form and, for those who responded that they were headed to or from the University, a mail-back survey form which contained additional questions relating to parking location and mode choice. The results of this mail-back survey are being summarized and evaluated by the UCB Campus Planning Office.
In addition to the security personnel directing traffic, two to three Berkeley Police officers were present throughout the day to ensure safety and adequate traffic control.
The survey was continuous throughout the day with the exception of 20 - 30 minute breaks during the morning commute peak (approximately 8:00 - 8:30 a.m.), the afternoon commute peak (approximately 4:30 - 5:00 p.m.), and during the lunch hour (approximately 11:45 -12:15 p.m.) Additional breaks were taken by one surveyor at a time throughout the day, so that the total number of surveyors varied between five and six per direction.
The commute period breaks served not only to relieve surveyors but also to provide a clearing period for the longer-than-typical vehicle queues which developed due to the slowing effect of the traffic direction. Ideally, such traffic direction would not increase vehicle queues at this all-way-stop controlled intersection, but in reality, additional delays were introduced by uncertainty and hesitation on the part of some drivers, drivers who stopped to argue with or ask questions of the traffic directors, and generally slower speeds near and through the intersection. The lengths of the queues relative to typical lengths during peak hours was not quantified.
The interview forms contained questions designed to determine the driver's trip origin, trip destination, trip purpose, and for southbound drivers, the local route they planned to take. The interview for northbound drivers had four questions, and the interview for southbound drivers had five questions. The surveyors recorded the time of each interview on the form, and also the number of persons in the vehicle (vehicle occupancy). The interview survey forms are included in Section VI.
For the entire survey day (approximately 7:00 a.m. to 6:45 p.m.), a total of 1,342 surveys were conducted in the northbound direction, and a total of 1,415 surveys were conducted in the southbound direction. This constitutes a sample rate of 21% in the northbound direction and 19.2% in the southbound direction, based on the total hose count volume for the survey period. Table S-1 presents the hourly sample rate for the interview survey. The rate varies from hour to hour due to variations in interview duration, drivers' willingness to participate, and surveyors' breaks. It should be noted that a few surveys (between five and twenty) were partially or wholly unusable in the data reduction process due to either missing or unintelligible answers on one or more questions. The survey proceeded smoothly throughout the day, with drivers generally cooperating and following the manual traffic direction. Most drivers appeared to realize that the survey was brief, because they could see the surveyed vehicles moving through the survey lane periodically. Some drivers refused to enter the survey lane when directed to do so, and they were released into the traffic flow with the assurance that the survey was voluntary.