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

Web Survey Bibliography

Title The Influence of Visual Layout on Scalar Questions in Web Surveys
Year 2004
Access date 16.09.2004
Abstract This paper examines how the visual presentation of response choices influences answers to ordinal scale questions and extends previous experimentation on paper surveys to the Internet, to determine whether the results of varying scale layouts are similar or different. The web survey of a random sample of Washington State University students during Spring 2003 included two sets of experimental comparisons; 1591 completed surveys were submitted from the 3004 requested obtaining a response rate of 53%. One experiment included a response scale with all categories labeled and compared a vertical linear layout using four-five categories to three nonlinear layouts where categories were double- or triple-banked in columns across the page. A second set of experiments compared 5- and 6-point fully labeled scales to polar point scales where the verbal labels were removed for the middle three categories to an answer box format where respondents reported a number in an answer space. Multiple replications of each experiment indicate significant differences in the means between formats as well as significantly different response distributions. These findings on the web tend to confirm previous findings on paper questionnaires and suggest that respondents to both paper and web surveys are similarly affected by visual layout and presentation (Christian and Dillman, In Press, Public Opinion Quarterly, Spring 2004). Strong evidence now exists, that the visual presentation and layout of response scales influence respondent answers to self-administered questionnaires and need to be considered when designing public opinion questions and surveys that use ordinal scale questions. These construction differences appear to be important in helping surveyors understand why responses may vary across modes. The increasing use of mixed-mode surveys suggests the need to understand how the mode of communication, visual or aural, can influence how surveyors design questions and how respondents answer those questions.
Year of publication2004
Bibliographic typeConferences, workshops, tutorials, presentations