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

Web Survey Bibliography

Title Investigating Response Quality in Mobile and Desktop Surveys: A Comparison of Radio Buttons, Visual Analogue Scales and Slider Scales
Year 2014
Access date 16.08.2016
Presentation PDF (184 kB)
Mobile devices have smaller displays, touch screens and different methods of navigation compared to desktop computers. This may limit the amount of information that can be placed on a mobile phone screen and it can also affect how a survey is comprehended and completed.
The most traditional rating scales in Web surveys are made from radio buttons. Radio buttons require quite a lot of space. Only a limited number of response options can be presented simultaneously. Otherwise, respondents have to scroll to see all options which may bias ratings.
Visual Analogue Scales (VAS) are operated by point and click: respondents move the mouse arrow to any position on the line and after clicking the mouse button a marker appears. In contrast, slider scales have a handle visible directly on load of the Web page and ratings are 261|Page done by drag and drop. Both scales can either be implemented as discrete or continuous rating scales. A continuous implementation is especially valuable if respondents use mobile devices like smart phone where an efficient use of space is required.
A comparison is needed of radio buttons, VAS, and slider bars to see how they affect usability and data quality on mobile phones compared to regular desktop completion. Finger navigation on mobile phones is less precise than mouse navigation on desktops. This could result in selecting the wrong (not intended) answer option in radio buttons. Slider bars or VAS might be more efficient in selecting the intended response option. We look at response quality indicators, paradata, evaluation of the questionnaire as well as personal characteristics.The usability of question formats is conjectured to be related to the number of scale points. We use an experimental design with question format and 5, 7, 11 and continuous scales. Data are collected in a probability-based panel in the Netherlands. 
Year of publication2014
Bibliographic typeConferences, workshops, tutorials, presentations

Web survey bibliography - The American Association for Public Opinion Research (AAPOR) 69th Annual Conference, 2014 (20)