Abstract

Citi Bike is a bike sharing company operating in NYC that has made their bike usage data publically available. This article investigates the question of whether or not the relative usage of customers on the weekend compared to the usage during the week is higher than that for subscribers. Pearson's Chi-Squared test is used to compare the usage frequencies.

Introduction

Citi Bike is a privately-owned bike sharing company that operates in New York City and New Jersey. The service can either be used as either pay-per-ride or via a subscription service. The two types of users are classed as customers and subscribers, respectively. The research question is based on the idea that subscribers tend to use Citi Bike for commuting, where pay-per-ride customers would be used more for leisure, thus concentrating the customer usage to the weekends.

Data

The Citi Bike data is given monthly. The data can be found at https://s3.amazonaws.com/tripdata/%Y%m-citibike-tripdata.zip, replacing the date format codes for the desired month and year. For this research, May through August 2015 was used. Each row in the dataset represents a single trip taken by a user. The relevant columns in the dataset are the date that the ride was taken and the user type, either customer or subscriber. The distributions of Citi Bike usage on each day of the week for each user type is given in Figure 1 below.