Are sliders too slick for surveys? An experiment comparing slider and radio button scales for smartphone, tablet and computer based surveys

Abstract

"The continued rise in smartphone penetration globally afford survey researchers with an unprecedented portal into personal survey data collection from respondents who could complete surveys from virtually any place at any time. While the basic research into optimizing the survey experience and data collection on mobile devices has continued to develop, there are still fundamental gaps in our knowledge of how to optimize certain types of questions in the mobile setting. In fact, survey researchers are still trying to understand which online design principles directly translate into presentation on mobile devices and which principles have to be modified to incorporate separate methods for these devices. One such area involves the use of input styles such as sliding scales that lend themselves to more touch centric input devices such as smartphones or tablets. Operationalizing these types of scales begs the question of an optimal starting position and whether these touch centric input styles are equally preferred by respondents using less touch capable devices. While an outside starting position seems optimal for slider questions completed via computer, this solution may not be optimal for completion via mobile devices as these devices are subjected to far more space and layout constraints compared to computers. This experiment moves the mixed device survey literature forward by directly comparing outcomes from respondents who completed a collection of survey scales using their smartphone, tablet or computer. Within each device, respondents were randomly assigned to complete one of 20 possible versions of scale items determined by a combination of three experimental factors including input style, length and number formatting. Results from this study suggest more weaknesses than strengths for using slider scales to collect survey data using mobile devices and also suggest that preference for these touch centric input styles varies across devices and may not be as high as the preference for the more traditional radio button style." (author's abstract)