Conducting Efficient Transit Surveys of Households Surrounding Transit-Oriented Developments

This paper presents best practices for transit survey protocols targeted at small geographic areas such as those in and around transit-oriented development and for households proximate to transit stations. The widespread dissemination of cell phones and rapidly decreasing presence of in-household landlines has made telephone interviewing prohibitively expensive; moreover, the portability of cell phone numbers has confounded small-area probability sampling. A postal mail push-to-web protocol described here is a more cost-efficient approach for small-area data collection, particularly when probability samples are sought. Address files are available from postal databases, and the strategy is to send respondents details on how to access an online questionnaire; continued nonresponse is then followed up, with the final mailed contact including a paper questionnaire. Embedded in the contacting protocol is a survey research experiment in which a method for maximizing response rates, the use of web instruction cards, is examined in the context of the mail push-to-web survey protocol. Results led to four optimization techniques: first, the mail push-to-web protocol described in this paper minimizes printing and mailing costs as well as computer data entry error from paper survey instruments; second, prenotification letters also serve to reduce those costs by indicating nondeliverable addresses to be deleted from the sample frame; third, web cards—whether generic or tailored—are ineffective; and fourth, from persuasive evidence, name association (i.e., having a commercial sample firm append names to the sample frame and using those names to personalize the mailing) has a strong positive effect on response propensity.