Methodology – After categorizing the ride-times into ‘light’ or ‘dark,’ we separated ridership by gender. We plotted this to get a visual sense or ridership, and then took a ratio of men/women by which to judge how the frequency of gender in ridership. Plotting this, we were able to visually compare the light vs. dark ridership by gender ratio, and we performed a unpaired t-test to gain further statistical insight. We initially thought about performing a Mann-Whitney U test, but due to the low number of observations in our final dataset (12 items in each category), this may have introduced more noise than clarity, and so we refrained.   
Conclusions – Both the graph (Figure 4) and low p-value (0.0024 compared to our 0.05 threshold) give strong reason to reject the null hypothesis in this case. Therefore, we are able to conclude that the ratio of men to women ridership is higher than women at night. More analysis is critical before any generalized conclusions are drawn, as this data is from only New York City, and for only one month.
Figures