Methodology
The research hypothesis was that males make longer rides on average than females. This was based on anecdotal evidence from observing Citibike riders and reading blogs and civic group's reports on bicycling in New York. In particular, the literature examined indicated that females often have greater safety concerns than males, which may act as a behavioral barriers to riding for long periods.
The following hypotheses were investigated:
H0: Mean ride length (males) <= mean ride length (females).
H1: Mean ride length (males) > mean ride length (females).
Significance level: alpha = 0.05.
The test adopted was a two-sample T-test. The T-test is required because the variance of the two samples is not known, therefore the sample follows a t-distribution rather than a normal distribution.
The decision rule was to reject the null hypothesis if the p-value / 2 was less than if alpha. This is because we are conducting a one-sided test, and the t-test produces a T-test produced a p-value of less than alpha.
The T-test produced output of:
- t statistic: -146.323
- p value: 0.00.
Accordingly, the null hypothesis was rejected.
The statistical test supported the data visualization shown in Figure 4, providing a robust support for the conclusion.