The influence of campus characteristics, temporal factors, and weather events on campuses-related daily bike-share trips

Abstract

Recently, there has been an increase in bike-sharing programs on university campuses. However, there is scarce literature on the campus characteristics that are most favorable for these programs. This study evaluates the impacts of campus characteristics, temporal factors, and weather events on daily bike-sharing trips within U.S. university campuses. The study used data collected between 2014 and 2018 from 177 stations located within 25 campuses. With an average of 38 trips per day, a high variability in the number of daily trips was observed, with a maximum of 708 and a minimum of zero. Both traditional Negative Binomial (NB) and Mixed Effect Negative Binomial (MENB) models were applied, and their performances were compared based on a set of criteria. The MENB performed better than NB. The MENB model results revealed that among campus related factors, number of faculty/staffs, number of full-time students, private universities, campus size, and Latino and Asian student dominance are positively associated with an increase in trips per day, whereas number of part-time students, distance from campus to central business district, and students' ages are negatively associated. The temporal factors depicted that weekends, public holidays, time between semesters, and summer semester were associated with decreases in number of daily trips, while years in operation and fall semesters were positively associated with daily ridership. Severe weather events resulted in less bike-sharing, and average daily temperature was positively associated with number of trips. Campus bike-sharing planners can utilize these findings for estimating daily bike-sharing ridership on campuses.