(a description of the methodology (test) chosen, the motivation behind it, a discussion of alternative options that were not adopted. Either this session of the next one should contain figures as well to show the results.)
Originally, Mann Whitney U test, Chi-square test, and Fisher's Exact test were selected for test performing. The
After comparing with testing for New York City's Prison Employment Program Effectiveness (Bianco, 2018), Chi-squared test was selected as the primary option. For Chi-square test, the two variables are independent, which in this project, subscribers are independent of customers. Particularly in this project, chi-square test for equality of proportions is used (Boslaugh & Watters, 2008). Therefore, three methods were used but chi-square might be the best choice.
The Citibike project are looking at probabilities of two categorical data and therefore, Chi-square measures whether significant differences exist between the expected probability and the observed probabilities in one or more categories. As the result showing below, the Chi-Square value is way larger than 3.84 (chi-square value corresponding to 0.05 significance level). Therefore, the null hypothesis can be rejected.