Correlation analysis and regression of transit ridership by means of public buses – A preliminary data analysis

Abstract

Given ever-increasing private transportation ownership, a rising population, and unceasing mobility, it is crucial to ensure the usage and improvement of public transportation services. Therefore, it is important to review and understand relationships between variables affecting ridership to boost them. This paper acts as a preliminary data analysis in which correlation and multiple linear regression (MLR) analyses were constructed to examine bus transit ridership. The main goal of this study is to identify the variables that influence transit ridership and the degree to which they do so during a) peak hours and b) off-peak hours. Independent variables such as bus commercial speed, service frequency, in-vehicle travel time, bus stop distance, and rainfall were employed in this study. Results show that all the independent variables are significant and correlated with ridership. However, the regression models show that distance and travel time are not statistically significant during peak and off-peak hours, respectively. The findings of this study recommend the creation of relevant policy initiatives that can assist transit agencies in increasing ridership by analysing significant predictive factors. Additionally, the output from this study can be used in future works on designing efficient bus routes, stop locations, and even bus schedules.