Pixels or pencils? The relative effectiveness of Web-based versus paper surveys

A survey of academic reference librarians provided data for an examination of differences in survey administration on paper and the World Wide Web, with particular attention to response rates, response time/quickness, sampling bias, and differences in responses that are related to the medium of administration. Three methods of survey distribution were compared: paper survey distributed by mail, Web survey announced by mail, and Web survey announced by e-mail. Response rates for the Web survey were not as high as with traditional survey methods. Responses to the Web survey were returned more quickly, but not by a great deal. E-mail notices were more effective for promoting the Web survey than paper notices. The study detected no sampling bias or differences in the content of responses. Although the advantages of Web survey administration are alluring, researchers must use carefully designed follow-up procedures and avoid technology-related implementation problems to match the effectiveness of traditional paper surveys. (C) 2004 Elsevier Inc. All rights reserved.