Methodology

The test selected for this analysis was the Z-test for testing populations proportions. Although the reviewer suggested selecting either an unpaired-test (if parameters followed a normal distribution) or the Mann-Whitney U test or Wilcoxon's rank sum, these tests are not appropriate for a proportions test. Another solution that could have been used was using the Chi-Square test. However, because the results achieved with the Z-test were very significant, it wasn't necessary to perform any more tests.

Conclusions

The result of this analysis shows that millennials ride Citibike bikes a lot more than baby boomers. When testing the \(proportion_{babyboomers}-\ proportion_{millennials}\)we see that this number is less than zero, meaning that we can accept the null hypothesis. Although I accept my original hypothesis, I show how the Z-test would have been performed if it had been needed. This study is only relevant to the month of November 2016. I think that a deeper analysis of more months would have been necessary to actually be able to prove that in general millennials bike more than baby boomers.