“Shih and Fan [2008] conducted a meta-analysis of 39 study results published within the last ten years that directly compared web and mail survey modes. See Appendix for response rate results. They found that mail surveys have higher response rates than web surveys in general but there were two factors that significantly influenced response rates: population type and usage of follow-ups. In comparative studies that used college populations, the weighted average web survey response rate was higher than that of mail surveys by 3%. However, for other populations (e.g., professionals, employees, and general population), the weighted average web survey response rate was lower than the mail survey response rate (23%, 10%, and 13% lower, respectively)….Selm and Jankowski (2006)…also suggest several techniques they may increase response rates. First, the authors suggest that researchers undertake multiple attempts to contact potential respondents via pre-notifications, reminders, replacement surveys, and thank you notes. Second, researchers should utilize a mixed-mode strategy, including both electronic and pencil-and paper questionnaires, in order to reach respondents without access to the Internet. Third, researcher should try to ensure that the survey topic is relevant to the target group. Lastly, researchers should use respondent incentives to stimulate questionnaire completion.” [Tomaszczyk, 2008]

Alicia Tomaszczyk, “Research Note for Response Rates in Web Surveys – Summer 2008″
http://www.src.uwaterloo.ca/Services/Research%20Note%20for%20Response%20Rates%20in%20Web%20Surveys.htm