What kind of technology transparency do users appreciate? Comparison of textual and graphic cues in app design

Abstract App recommendations and data visualizations such as weather forecasts, navigation aids, or sleep tracking graphs play an increasingly important role in daily decisions. However, the apps’ underlying functioning often remains opaque, possibly resulting in a suboptimal user experience or inadequate reliance on recommendations. To approach design solutions for this, the paper investigates the effects of textual and graphic transparency cues on users’ mental model accuracy, user experience, and explanation satisfaction, using the example of a weather and a sleep tracking app. An online experiment with 293 participants showed that textual transparency cues (i.e., verbal explanations) led to higher felt and objectively measured mental model accuracy than graphic transparency cues (i.e., data visualizations). Textual cues were also more satisfying than graphic cues but did not result in significantly different ratings of user experience. Moreover, differences between textual and graphic cues in subjective mental model accuracy and explanation satisfaction were stronger for the weather than the sleep tracking app, implying context-specific differences in the impact of transparency cues. The results and limitations are discussed and linked to the challenge of finding a sweet spot for technology transparency design.