Aaron Springer

Aaron Springer, Steve Whittaker

Extensive recent media focus has been directed towards the dark side of intelligent systems, how algorithms can influence society negatively. Often, transparency is proposed as a solution or step in the right direction. Unfortunately, research is mixed on the impact of transparency on the user experience. We examine transparency in the context an interactive system that predicts positive/negative emotion from a users' written text. We unify seemingly this contradictory research under a single model. We show that transparency can negatively affect accuracy perceptions for users whose expectations were not violated by the system's prediction; however, transparency also limits the damage done when users' expectations are violated by system predictions.

Aaron Springer, Victoria Hollis, Steve Whittaker

We demonstrate that users may be prone to place an inordinate amount of trust in black box algorithms that are framed as intelligent. We deploy an algorithm that purportedly assesses the positivity and negativity of a users' writing emotional writing. In actuality, the algorithm responds in a random fashion. We qualitatively examine the paths to trust that users followed while testing the system. In light of the ease with which users may trust systems exhibiting "intelligent behavior" we recommend corrective approaches.

Aaron Springer, Steve Whittaker

As we increasingly delegate important decisions to intelligent systems, it is essential that users understand how algorithmic decisions are made. Prior work has often taken a technocentric approach to transparency. In contrast, we explore empirical user-centric methods to better understand user reactions to transparent systems. We assess user reactions to global and incremental feedback in two studies. In Study 1, users anticipated that the more transparent incremental system would perform better, but retracted this evaluation after experience with the system. Qualitative data suggest this may arise because incremental feedback is distracting and undermines simple heuristics users form about system operation. Study 2 explored these effects in depth, suggesting that users may benefit from initially simplified feedback that hides potential system errors and assists users in building working heuristics about system operation. We use these findings to motivate new progressive disclosure principles for transparency in intelligent systems.

Aaron Springer, Steve Whittaker

The rise of machine learning has brought closer scrutiny to intelligent systems, leading to calls for greater transparency and explainable algorithms. We explore the effects of transparency on user perceptions of a working intelligent system for emotion detection. In exploratory Study 1, we observed paradoxical effects of transparency which improves perceptions of system accuracy for some participants while reducing accuracy perceptions for others. In Study 2, we test this observation using mixed methods, showing that the apparent transparency paradox can be explained by a mismatch between participant expectations and system predictions. We qualitatively examine this process, indicating that transparency can undermine user confidence by causing users to fixate on flaws when they already have a model of system operation. In contrast transparency helps if users lack such a model. Finally, we revisit the notion of transparency and suggest design considerations for building safe and successful machine learning systems based on our insights.