difference-in-difference plots regarding a new bike lane
Below are a few exploratory difference-in-difference line plots generated from a survey which questioned pedestrians about their travel and shopping behaviour before and after the implementation of a bike lane in central Toronto (Bloor Street). Plots were built using ggplot2 and Rmarkdown sometime during the summer of 2017 as part of preliminary descriptive analysis for a report on the economic impacts of the bike lane.
Click here for a detailed research report regarding this project
The plots were generated using ggplot2 and Rmarkdown. The p-value pertains to a statistical test that compares whether the changes in percents are significantly different between the two sites - Jeff Allen
Survey Summary
n table
| Bloor (case) | Danforth (control) | Total | |
|---|---|---|---|
| pre bike lane | 842 | 173 | 1015 |
| post bike lane | 1577 | 412 | 1989 |
| total | 2419 | 585 | 3004 |
Mode Share of Visitors
= Bloor 
= Danforth
percent of visitors who arrive by bike percent of visitors who arrive by car
DiD p = 0.073 DiD p = 0.402
percent of visitors who arrive by transit percent of visitors who arrive by walking
DiD p = 0.186 DiD p = 0.469
Cycling Frequency
= Bloor = Danforth
percent of visitors who cycle to the street at least one day per week
DiD p = 0.001
Cycling Safety
= Bloor = Danforth
percent of visitors who feel safe biking on the street
DiD p = 0.000
Visit Frequency
= Bloor = Danforth
Percent who visit 15 or more days per month
DiD p = 0.335
Monthly Spending:
= Bloor = Danforth
percent of all visitors who spend more than $100 per month
DiD p = 0.399
Subseting by mode:
percent of cyclists who spend more than $100 per month: percent of drivers who spend more than $100 per month:
DiD p = 0.198 DiD p = 0.415
percent of transit riders who spend more than $100 per month: percent of walkers who spend more than $100 per month:
DiD p = 0.161 DiD p = 0.354