Afternoon Ridership Patterns — Regression Results

This analysis groups riders departing at the top nine exit-stations (where the off-peak ridership uptick was most present), verse riders traveling to the bottom nine exit-stations (where there was no sign of an off-peak bias). The time periods of interest are the four 6-minute intervals surrounding the first period of off-peak fares [7:00–7:05 entry].

In the regression equation, the exit_dummy takes on the value 1 for riders traveling to the top nine exits, and 0 otherwise. The large and significant coefficient on Time, confirms the declining trend of ridership over this afternoon period. Importantly, the interaction term (and dummy intercept) is non-significant. Meaning the pattern of ridership over this period does not change depending on which group of riders we look at, those departing at the top nine or bottom nine exit stations.