Display of Major Risk Categories for Mobile Apps. (September 2016)

• Record Type: Journal Article
• Authors: Chen, Jing; Ge, Huangyi; Proctor, Robert W.; Li, Ninghui; Yang, Weining
• Is Part Of: Proceedings of the Human Factors and Ergonomics Society ... Annual Meeting. Volume 60:Part 1(2016)
• Page Start: 779
• Page End: 779
• Publication Date: 2016-09
• Subjects: Human engineering -- Congresses; 620.8205
• DOI: 10.1177/1541931213601178
• Languages: English
• ISSNs: 1541-9312

Abstract

Smart mobile devices have become indispensable, yet the associated risks are not fully understood or known by users (Felt, Greenwood, & Wagner, 2011), and an effective way of communicating the risk information to the users is needed. In addition to a summary risk/safety index that has been shown to be effective (Chen, Gates, Li, & Proctor, 2015; Gates, Chen, Li, & Proctor, 2014), prior research (Jorgensen, Chen, Gates, Li, Proctor, & Yu, 2015) proposed an intermediate-level risk display that includes selected major risk categories that are identified by individual users ( personal privacy; monetary loss; device stability ). The goal of the present study was to evaluate the various formats for displaying these major risk categories, by comparing bar graph and numerical table representations (see Tversky, 1997), as well as the horizontal and vertical display orientations in each format. These displays were tested in two user tasks: Risk rating and app selection. In Experiment 1, participants were asked to rate the potential risk of an app based on the provided risk display. The bar graphs took shorter time and elicited higher risk ratings than the numerical tables. As the risk score increased in each category, the potential risk of the app was rated higher overall, and the self-reported mostly concerned risk category was weighed more than others. In Experiment 2, participants were required to choose one app that they would install out of two alternative apps given the potential risks of the two apps. There was no significant difference in terms of response time or app-selection percentage between the display formats. The apps with lower risk scores were chosen more often, and this effect again was influenced by the self-reported risk concerns. No difference was found between horizontal and vertical displays in either experiment. These results indicate that it is potentially beneficial to include this intermediate-level risk display in the app interface, and individualized design incorporating self-reported risk concerns of the users may be more informative. Moreover, the bar graph display seems to be more efficient and effective than the table display, at least when users need to evaluate an individual app.