Avoiding Biased Interpretation of Bicycle Surveys

Abstract

Surveys are a common technique used by transportation agencies to gather information about the socioeconomic characteristics, behavior, and preferences of the traveling public. The purpose of this paper is to help agencies and researchers understand differences in bicycle survey results when different distribution techniques are used. This study analyzes a bicycle survey that was implemented by the City of Winston-Salem, North Carolina, in early 2005. The city distributed identical survey questions in four different ways. Comparative statistics are used to show that there were significant differences in responses according to survey distribution method. Using four different methods of survey distribution allowed the City of Winston-Salem to gather feedback from a greater variety of citizens than is possible with a single survey method, target several key constituencies (people with significant interest in bicycle issues, bicyclists with less experience, and potential bicyclists), and understand potential differences in the responses to certain survey questions based on the survey method used. Practitioners can use the results of this study to help select survey distribution methods that will gather the most credible and useful data for meeting their objectives. It will also help agencies draw more informed conclusions from bicycle survey results.