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

Web Survey Bibliography

Title From Telephone to the Web: The Challenge of Mode of Interview Effects In Public Opinion Polls
Year 2015
Access date 09.07.2015

As survey research rapidly evolves, two notable trends are the growth of self-administered web surveys and mixed-mode studies. After decades in which most surveys were interviewer-administered, the new world of survey self-administration poses both opportunities and challenges. Among the challenges is how to handle long-term trend data when the mode of interview shifts, and how to minimize error in merging data in multi-mode studies. Mode-of-interview effects have been the subject of many studies, but rarely has a study examined mode effects across a wide ranging set of political and social measures. We build on past research with a comprehensive, large-scale mode-of-interview experiment that randomly assigned a nationally representative panel of respondents to telephone and web modes and administered a 75-question instrument to them. Respondents are drawn from the Pew Research Center’s American Trends Panel, recruited in early 2014 in an RDD telephone survey. The experimental groups include 1,509 respondents who took the survey on the web, and 1,494 who completed a telephone survey. Each experimental group was weighted separately to national parameters for the general public. Non-internet users (12% of the panel) were interviewed by phone but are excluded from this analysis. This presentation will provide an overview of the size, direction and nature of mode differences across standard polling questions that varied in terms of topic, format, sensitivity and cognitive complexity, among other dimensions. The mean difference in topline results between web and phone across all questions was 6.4 points; median difference was 5.0 points. Differences ranged from 0 to 29 points. The analysis shows that topline comparisons can be misleading. Even on questions with small overall mode differences, sizeable mode effects may lurk beneath the surface, often appearing only among subgroups for which a particular question is sensitive or cognitively difficult.

Year of publication2015
Bibliographic typeConferences, workshops, tutorials, presentations

Web survey bibliography - Keeter, S. (14)