Our alternative hypothesis was that, on average, the drip duration is shorter for a sample of users BORN AFTER 1960 compared to users BORN BEFORE 1960 throughout the timeframe of Jan 2016 - Feb 2016.
While thinking of our initial hypotheses, we decided that we will examine future results based on the level of significance of alpha = 0.05.
There were different options for testing if people born after 1960 take shorter trips. As we had a relatively sizable dataset (more than 100 rows), we concluded that the One-Sided Two-Sample t-test would be appropriate. Two-Sample t-test is used to test the difference between two population means.
Our reasoning for choosing this parametric test as opposed to a non-parametric test was that our dataset was large enough to make the test robust to non-normality due to the central limit theorem. We explored using Mann Whitney U test (also known as Wilcoxon-Mann Whitney), which required us to change our hypothesis to "the sample distributions are not equal". We decided not to proceed with such test since it would not necessarily provide answers to our initial question (whether younger people make shorter rides on average or not).
We used the SciPy library and it's t-test for the means of two independent samples (scipy.stats.ttest_ind) functions to do the hypothesis testing.
Conclusions