Web Survey Bibliography

Title Why Internet Samples Are Inherently Biased
Author Neustadtl, A., Kestnbaum, M., Robinson, J. P.
Source The American Association for Public Opinion Research (AAPOR) 57th Annual Conference, 2002
Year 2002
Access date 04.05.2008
Abstract Surveys by Internet samples are becoming increasingly popular, questions arise about whether the respondents who can be reached by Internet are systematically different from non-users on variables that make generalizations from such samples highly problematic. For example, it is known that Internet users are younger, and more highly educated and white than non-users; however, one might be able to offset such biases in surveys by simply applying weighting to Internet samples to bring them into line with face-to-face or telephone samples. It became possible to test this assumption using data from the Internet module on the year 2000 General Social Survey (GSS). The GSS sample not only contains about half users to compare with half non-users, but data on more than 500 attitude and behavior variables on which to compare them - as well as extensive background factors on a sample of 2386 adults with over a 70% response rate. In brief, GSS Internet users are significantly different from non-users on well over 100 of these factors. While some of these differences can be explained by Internet users’ higher education, lower age or other background factors, most others cannot, suggesting that no simple weighting factor or adjustment strategy can satisfactorily make Internet samples comparable to samples conducted by home or phone interviews.
Year of publication2002
Bibliographic typeConferences, workshops, tutorials, presentations
Print

Web Survey Bibliography - The American Association for Public Opinion Research (AAPOR) 57th Annual Conference, 2002 (35)