Conclusions
The results of the Pearson's and Spearman's tests both show a strong correlation between birth year and ride duration. Each test shows that as birth year increases (i.e. younger riders), the average duration of CitiBike rides also increases. The Pearson's test, which tests for linear correlation, shows a very high positive correlation between the two variables with an extremely low p value bordering on 0, thus meeting my CI threshold for statistical significance (alpha = .05). The Spearman's test too meets my CI threshold with a p value of 0 and a correlation coefficient of 1.
Although the results of the Pearson's and Spearman's tests do confirm my initial theory that younger CitiBike riders tend to ride for longer, the results of the two tests cannot be taken at face value. Although the ride duration averages were calculated based on over 2 million individual CitiBike rides, each test was ultimately performed on only 32 data points. The results therefore are less powerful.
One way to strengthen the analysis would be to incorporate more data both into the initial ride duration averaging and the final analysis of average rides by rider birth year. By creating a separate data set with no filter for gender or age, the following results are produced: