The effects of the progress indicator's design on online surveys

Abstract

A progress indicator (PI) is often used to inform respondents of the task completion status on online surveys. When conducting online surveys, reducing the dropout rate and the participants' cognitive loads is important for improving survey efficiency. Although there have been many studies on the effect of applying the PI on online surveys, it is unclear which design of PIs should be used for reducing the dropout rate and the participant's perceived task load. Moreover, even in the web content accessibility guidelines (WCAG), which should be followed to improve web accessibility when building websites, design guidelines for PIs are still vague compared to other graphic elements. This study aims to provide design guidelines for PIs that can be universally applied to various online survey environments by paying attention to generic forms of PI, which are used to precisely communicate information on progress. Representative design factors that compose PIs are bar-graph and label. To select the PI's design elements to be investigated in this paper, we collected use cases of PIs applied to previous studies and various online services. Five bar-graph elements and five label elements were selected based on how the PI communicates information. Then, 25 PI types were created

5 bar-graph elements x 5 label elements. To find out whether the design factor of the PI affects the online survey's dropout rate and abandoned items ratio, and the mental load of the respondents, an online survey study was conducted. We built a web survey site with 25 PI types and collected a total of 1948 participant data through mTurk. The online survey study results showed that the PI's label design had a significant effect on the number of items answered by the participants. In addition, it was also found that the mental load of respondents significantly changed according to the combination of bar-graph and label designs of PI. This study provides strong support for designing the way PI communicates the task process without deceiving the speed of progress, which encourages respondents to answer more items and decrease their mental loads in online survey environments. Furthermore, we analyze the various online services that need to provide progress information about the task and discuss ways to communicate progress in each environment effectively.