Internet Access

Abstract

Travel surveys can reach potential survey respondents in many ways, including intercepting them at activity centers or on transit vehicles and contacting them by telephone, mail, or online. The selection of a survey recruitment approach should consider whether the population from which a sample is drawn is fully covered in order to avoid coverage error. This paper presents two case studies that review data collected in travel surveys in metropolitan areas in the United States to identify coverage error, specifically whether lack of Internet access by a segment of the population leads to coverage error. The first case study analyzes data from several road pricing surveys to quantify differences between those who have and do not have Internet access. The second case study analyzes data from two transit origin–destination surveys in which respondents were invited to provide contact details for additional market research. This paper compares the overall sample with those willing to be surveyed in the future by telephone and those willing to receive a future survey invitation by e-mail. Both case studies find that the samples with access to the Internet are similar to the larger full samples that include those without Internet access and therefore that the coverage error found in the Internet-only samples is small. The results suggest that, for surveys of general populations of drivers or transit riders, surveying only those with Internet access does not introduce significant coverage error into travel survey samples.