Introduction

New York City introduced the Citibike program in late 2013 as a public paid service enabling citizens to rent branded bikes for 30-minute ride sessions. Rich data exists for every historical ride in the system that includes the rider's starting and stopping station, age, gender, and duration. Like the New York City subway, the system is operational 24 hours a day, 365 days a year. I am interested in exploring whether there is a statistically significant difference in the age of Citibike riders in the daytime versus the nighttime. My intuition is that late-night riders will tend to be younger than daytime riders. This question may be a proxy for the effect of things like bedtime and sleep preferences, sense of safety late at night, use of Citibike for commuting vs. "going out", and the higher availability of alternative transportation methods for older bike riders.

Data

Citibike data is publicly available online and contains comprehensive ride records for one-month periods from July 2013 to present \cite{nyc}. I used the July 2017 Citibike data set, consisting of 1,735,599 rides. A ride record consists of the the following fields: trip duration, start time, stop start, start station ID, start station name, start station latitude, start station longitude, end station ID, end station name, end station latitude, end station longitude, bike ID, user type, user birth year, and user gender.