Using Cumulative Logistic Regression Model for Evaluating Bicycle Facilities on Urban Arterials

This paper presents the results of a research study conducted to identify roadway characteristics that significantly influence the perception of the quality of service provided to bicyclists on urban streets. The data used in this study was collected in the National Cooperative Highway Research Program (NCHRP) 3-70 Multimodal Level of Service for Urban Streets research study. The data for the bicycle mode was gathered using the video simulation techniques from Tampa, Florida on urban streets with varying geometric and operational characteristics. Ten video clips were shown to 145 participants, ranging in age and gender, at four locations: Oakland, California; Chicago, Illinois; New Haven, Connecticut; and College Station, Texas. The participants were asked to rate these clips for the bicycle facilities on the scale of 1-6 with 6 being equal to the level of service (LOS) rating “A” and 1 being equal to LOS “F”. Taking advantage of 1,400 data points collected in the NCHRP 3-70 study, this study utilized cumulative logistics regression modeling technique to obtain the entire distribution of bicycle LOS under a given set of conditions. The model results indicate that with the absence of bicycle lanes or shoulders, the LOS rating for a bike mode reduces by 1.22 units. Additionally, on roadways with speed limit > 30 mph, the LOS rating for the bicycle mode reduces by 1.6 units and for travel lanes number greater than 1, the LOS rating reduces by 0.64 units. If there are no unsignalized conflicts per mile on the roadway, the LOS rating increases by 0.42 units. The results of the cumulative regression model agree with the perception that the presence of bicycle lanes or shoulders increases the comfort level of bicycle users, while higher auto speeds, higher number of lanes, and presence of unsignalized conflicts along roadway segment causes impedance to the bicycle mode. Based on the cumulative logistics regression model, cumulative probability and individual probability distribution for each LOS rating were plotted. The individual probability calculations reveal that the probability of rating a roadway segment without a bicycle lane as “D” or worse is 10 percent higher than a roadway segment with a bicycle lane or shoulder. Also the probability of rating a roadway segment with a bicycle lane as B or better is 14 percent higher than the probability of rating a roadway segment without a bicycle lane or shoulder.

• Corporate Authors:

  Transportation Research Board

  500 Fifth Street, NW
  Washington, DC  United States  20001
• Authors:
  • Ali, Asma T
  • Cristei, Cerasela M
  • Flannery, Aimee
• Conference:
  • Transportation Research Board 91st Annual Meeting
  • Location: Washington DC, United States
  • Date: 2012-1-22 to 2012-1-26
• Date: 2012

Language

• English

Media Info

• Media Type: Digital/other
• Features: Figures; References; Tables;
• Pagination: 16p
• Monograph Title: TRB 91st Annual Meeting Compendium of Papers DVD

Subject/Index Terms

• TRT Terms: Arterial highways; Bicycle facilities; Highway design; Level of service; Logistic regression analysis; Multimodal transportation; Urban highways
• Geographic Terms: Tampa (Florida)
• Subject Areas: Pedestrians and Bicyclists; Terminals and Facilities; I72: Traffic and Transport Planning;

Filing Info

• Accession Number: 01366661
• Record Type: Publication
• Report/Paper Numbers: 12-2518
• Files: TRIS, TRB
• Created Date: Mar 30 2012 1:09PM