Evaluating the Impact of Uncertainty Visualization on Model Reliance

Jieqiong Zhao, Yixuan Wang, Michelle V. Mancenido, Erin K. Chiou, Ross Maciejewski

Room: 109

2023-10-26T00:09:00ZGMT-0600Change your timezone on the schedule page
Exemplar figure, described by caption below
In a human-AI teaming collaborative environment, humans have to appropriately incorporate AI/ML model predictions to fully leverage the benefits of automation. We conducted an experiment using a crowdsourcing platform to assess the impact of the presence and absence of uncertainty information on a college admission forecasting task. Furthermore, we examined how uncertainty information influences model adoption behavior using two uncertainty visualization techniques and two levels of task difficulty. Our findings indicate that ordinal forms of expressing model uncertainty are more likely to calibrate humans' adoption of model predictions.
Fast forward
Full Video

Uncertainty;model reliance;trust;human-machine collaborations


Machine learning models have gained traction as decision support tools for tasks that require processing copious amounts of data. However, to achieve the primary benefits of automating this part of decision-making, people must be able to trust the machine learning model's outputs. In order to enhance people's trust and promote appropriate reliance on the model, visualization techniques such as interactive model steering, performance analysis, model comparison, and uncertainty visualization have been proposed. In this study, we tested the effects of two uncertainty visualization techniques in a college admissions forecasting task, under two task difficulty levels, using Amazon's Mechanical Turk platform. Results show that (1) people's reliance on the model depends on the task difficulty and level of machine uncertainty and (2) ordinal forms of expressing model uncertainty are more likely to calibrate model usage behavior. These outcomes emphasize that reliance on decision support tools can depend on the cognitive accessibility of the visualization technique and perceptions of model performance and task difficulty.