To show the majority records, extreme values (trip durations above 3000 seconds) are removed.
Methodology
To figure out whether male and female riders have different trip durations, this research test whether the trip durations of the two genders follows the same distribution. Based on the shape of the two histograms, the two distributions do not follows normal distribution. Thus, techniques that require data follow normal distribution such as anova analysis should not be used. This research uses Mann Whitney U test, which does not need data follow normal distribution. Mann Whitney U test's null hypothesis is the two distribution follow the same distribution.