An Empirical Investigation of Passenger Wait Time Perceptions Using Hazard-Based Duration Models

Waiting time in bus stops heavily affects traveler attitude towards public transportation and therefore is an important element for consideration when planning and operating a bus system. Furthermore, what passengers perceive as waiting time is often quite different from their actual waiting time at a bus stop. In this context, we present an empirical investigation of actual and perceived waiting times at bus stops for the case of a large bus network, using hazard-based duration models. The analysis is based on a questionnaire survey undertaken at bus stops of the Athens, Greece, bus network. Results indicate that age, trip purpose, and trip time period seem to have an impact on that perception, with older individuals, work, and education trips being factors that increase perceived waiting time and lead to an overestimation of actual waiting, while perceived waiting time decreases during morning time periods. Introduction Transit is an important element of a city’s transportation system; transit systems offer sustainable and equitable transportation services to all travelers at low cost. Journal of Public Transportation, Vol. 14, No. 3, 2011 110 However, despite advantages of transit usage for societies, its share in ridership is often significantly lower compared to that of private vehicles, a fact usually attributed to the reduced performance and quality of services offered by transit systems. Bus systems are the most common among transit systems in cities worldwide (Vuchic 2004); however, these systems exhibit considerably lower attraction to passengers, which is mostly a result of problematic operations and their interaction with the rest of traffic. Indeed, operational characteristics such as low travel speeds, inadequate frequencies, and lack of punctuality and schedule reliability do have a negative effect in bus transit ridership. Waiting times in bus stops are among those elements heavily affecting the attitude of passengers towards using bus transit, as well as their opinions on the quality of transit services. Riders are expected to wait at stops for buses to arrive, being exposed to adverse weather conditions, crowding, and a surrounding environment of poor quality, while being stressed by waiting anxiety. As a result, what passengers perceive that waiting time is often much larger compared to the actual waiting time imposed by bus operations, especially when no information is given to them on expected bus arrivals (Mishalani et al. 2006). In that context, this paper provides an empirical investigation of actual and perceived waiting times at bus stops in the case of a large bus network with the use of duration models. The analysis is based on a survey undertaken at over 30 bus stops of the Athens bus network, which consists of over 300 lines and serves 1.3 million passengers on a daily basis. The remainder of the paper is structured as follows: the next section provides a brief background of past work on bus waiting times. Next, a description of the dataset and the methodological aspects of the paper are given. Empirical findings are then presented and discussed. The paper concludes with major insights drawn from the preceding analysis. Background The issue of waiting times at bus stops has been a topic of interest for researchers for at least three decades, with efforts concentrating both on the actual and perceived waiting times. Jolliffe and Hutchinson (1975) offered a behavioral explanation of the relationship between bus and passenger arrivals at bus stops and their impact on waiting times, considering random and not random passenger arrivals. Turnquist (1978) identified the effects of service frequency and reliability on waiting times, and Bowman and Turnquist (1981) developed a model based on pas-