Notice: the WebSM website has not been updated since the beginning of 2018.

Web Survey Bibliography

Title Using Mail Contact to Sample and Encourage Submission of Questionnaire Answers Over the Internet
Year 2009
Access date 30.09.2009

This presentation will summarize and integrate findings from a sequence of three experiments aimed at the development of effective internet data collection procedures for the general public.

The first experiment (Smyth, Dillman, Christian and O’Neill, Forthcoming), which was conducted on a regional sample of households, showed that it was possible to obtain a response rate of 55% from a sample of household addresses (The U.S. Postal Service Delivery Sequence File). It also showed that three-fourths of those responses came via the Internet with the remainder by mail.

The second experiment, conducted on a statewide (Washington) sample of households resulted in a response rate of 47% with two-thirds of the responses coming via the Internet (Messer, 2009; Messer and Dillman, 2009). Results of this experiment also indicate that the individuals responding to the Internet are quite different, demographically, than those who respond by mail. When the Internet and mail responses are combined together, the results are closer to representing the survey population than are either the mail or Internet proportion of the combined returns.

These experiments also provide insight into the effectiveness of token financial incentives and other data collection procedures. In addition the first experiment showed that giving people a “choice” of response mode tended to lower response rates compared to offering only a mail response option.

The third experiment (Millar and Dillman, 2009), follows-up on the issue of choice by examining the specific effect of offering people a choice of mail or web as a means of response. This experiment, conducted on a population known to have email access, provides new evidence on the effects of choice vs. the sequential offering of survey modes. It also provides insight into how mail and email contacts can be used in a supportive way to improve survey response.

Taken together these results have significant implications on alternative ways of improving the response and representativeness of survey data obtained over the Internet. Needed future research will also be discussed.

Access/Direct link

Conference homepage

Year of publication2009
Bibliographic typeConferences, workshops, tutorials, presentations
Full text availabilityAvailable on request

Web survey bibliography - Internet Survey Methodology workshop 2009 (21)