Screen shot 2018 11 07 at 11 55 47 pm
This experiment was conducted by Mark Bauer and Manrique Vargas (NYU CUSP).Abstract Are New Yorkers subscribing to Citibike’s annual pass? Are there more short-term users of Citibike in New York? What are the proportions of these user types throughout the week? We analyzed these questions during this experiment, and thought it might be intriguing to learn more about these user types. Ultimately, we concluded that Subscribers bike on weekends is less than the proportion of customers biking on weekends.Introduction - CitiBike in New York CityLaunched in May 2013, Citibike is New York City’s bike share system and is the largest in the nation. Motivate, one of the leaders in bike share systems in the world, currently operates Citibike. For this experiment, we used Citibike data to perform our analysis.Key Words:Customer = 24-hour pass or 3-day pass userSubscriber = Annual Member Additional Links:Citibike: https://www.citibikenyc.com/ Data source: https://www.citibikenyc.com/system-dataMotivate: https://www.motivateco.com/ Our IdeaWe wanted to analyze the characteristics of Subscribers (annual members) and Customers (24-hour pass or 3-day pass user) throughout the week. Specifically, we had questions about the user types and their proportions during the week compared to the weekend. Additionally, we asked ourselves questions such as how are these two user type options performing? To start, we chose to analyze the month of June in 2015 (possibly to account for more tourists), and plotted the number of riders (Users) by day (Mon-Sun). This would tell us how the annual membership is performing in relation to the 24-hour or 3-day pass, and would provide us with more information about the proportion of these user types per day of the week. Enhancing our understanding of the proportions of user type by day could tell us about the distribution of these user types. A breakdown of our experiment is below:Idea: Subscribers bike on weekends is the same or higher than the proportion of customers biking on weekends Null Hypothesis: Proportions of Subscribers bike on weekends is the same or higher than the proportion of customers biking on weekends. Alternative Hypothesis: Proportions of  Subscribers bike on weekends is less than the proportion of customers biking on weekends.