• Federal Highway Administration >
• Publications >
• Research Publications >
• 06125 >
• Appendd.Cfm
• Pedestrian and Bicyclist Intersection Safety Indices

Pedestrian and Bicyclist Intersection Safety Indices

Final Report

PDF Version (2.25 MB)

APPENDIX D. LESSONS LEARNED ABOUT ONLINE VIDEO-BASED SURVEYS

Through the process of creating and conducting the online safety survey, the research team encountered many issues related to Web-based surveys. Researchers who intend to conduct similar surveys may benefit from the lessons learned in this study. The online format was convenient for the ability to distribute the survey widely across the United States, and even internationally if needed. However, some survey participants only had dial-up Internet access, which caused the video clips to download very slowly. Some video clips were too small or of insufficient quality to provide ideal visibility of the intersection or crosswalk. Most of the issues encountered, however, came from the decision to distribute the video clips in RealPlayer™ format (".rm" files). In order to play RealPlayer video, it is necessary to download and install the free RealPlayer program. This program does not come pre-installed on most computer systems, unlike Microsoft^® Windows Media^® Player. The process of downloading and installing RealPlayer was confusing to many survey participants. Most city and State employees also had issues with firewall restrictions that prevented them from downloading and/or installing software on their computer. The research team recommends that future researchers create their video clips in a more easily read format, such as Windows Media (.wmv).

One of the difficulties in filming video clips of intersections is determining how to get the right vantage point to provide the viewer with all necessary information. Pedestrian crosswalks are relatively easy to film since they occupy only a small space in the intersection; however, bicycle approaches can be more difficult. It is often hard to strike the balance between positioning the camera too close (good detail of the intersection, but no view of the approach) and too far away (good view of the approach, but intersection details are unclear). Although this study used a single vantage point per clip, the research team suggests that including multiple vantage points would be a better alternative. For instance, the authors suggest a video clip design that would show the intersection from two or three positions, ranging from far away from the intersection to closeup. It might also be good to include a panning shot at the intersection to give participants a feel for the quality of the sight distance at the intersection.

FHWA-HRT-06-125