Analyzing Driver Compliance to Speed Limits Using Logistic Regression
- Published: Transportation Research Board, 2015
- Authors: Gargoum, S., El-Bayouny, K.
- Date Added: 05 May 2017
- Last Update: 05 May 2017
The study examines the effects of such factors on driver compliance in the City of Edmonton, using logistic regression. Unlike previous studies, this study examines the effects of different variables on compliance, rather than collision counts or driver speed choice.
The dataset used includes vehicle spot speeds recorded at almost 700 different locations in the city. The compliance for each vehicle was used as the response variable for the regression model, which was built using data from more than 35 million cases.
The findings show that, generally, the more restricted drivers become the more likely they are to comply with speed limits; potential restrictions include street parking, bike lanes, pedestrian crossing, or the absence of shoulder lanes.
Furthermore, higher traffic activity during peak hours, and presumably on shoulder weekdays (Monday and Friday), both increase the likelihood of compliance. In contrast, as the vehicle class (length) increases, the probability of compliance decreases.
Not much can be inferred about the effects of weather on compliance to speed limits, although an interesting finding is that odds of compliance seem to drop in winter months.
Another important observation about non-compliance that is somewhat concerning is that speed limit violations are higher in residential areas relative to most of the other land uses considered.
Climate; Highway design; Land use; Logistic regression analysis; Speed limits; Spot speed; Vehicle length.
Priced: via http://amonline.trb.org