An internet-based survey method for college student drinking research.

The purpose of this study was to describe and assess the utility of an Internet-based survey method for characterizing the alcohol consumption of college students. After extensive pilot research, a random sample of 1910 students aged 16-29 years was invited to complete a questionnaire, consisting of a series of web-pages linked to a relational database on a secure web-site. A branch structure allowed for tailoring of survey items by age and gender. The students received up to nine contacts, including a pre-notice letter with a token gift and an e-mail invitation (Phase 1), a reminder letter and e-mail message (Phase 2), and then telephone reminders and replacement access codes (Phase 3). Non-computer-users were offered a pen-and-paper alternative, making this a mixed-mode survey. The overall response to the survey was 82% (n = 1564). The median completion time was 16.7 min. Participants' comments showed high levels of satisfaction with the survey. Comparison of web (n = 1501) versus pen-and-paper completions (n = 63) revealed no modality effects. Technical problems addressed during the course of implementation included web-browser-operating system incompatibilities, and periodic network errors, although these resulted in little lost participation. Internet-based surveys are feasible for college student research and with carefully managed recruitment, can yield a high response.